InSiGHT 2022 Abstract Publishing and Best Abstract Awards

OP-001: Cancer Pathway-Somatic Mutations, Pathology

The immune profile of normal colonic mucosa as a possible tumor risk modifier in Lynch syndrome

Lena Bohaumilitzky ¹, Klaus Kluck², Robert Hüneburg³, Richard Gallon⁴, Jacob Nattermann³, Martina Kirchner⁵, Glen Kristiansen⁶, Oliver Hommerding⁶, Fabian Echterdiek⁷, Nina Nelius¹, Ben Hartog⁴, Gillian Borthwick⁴, Elena Busch⁸, Georg Martin Haag⁸, Hendrik Bläker⁹, Gabriela Möslein¹⁰, Magnus Von Knebel Doeberitz¹, Toni T. Seppälä¹¹, Maarit Ahtiainen¹², Jukka Pekka Mecklin¹³, D. Timothy Bishop¹⁴, John Burn⁴, Albrecht Stenzinger², Jan Budczies², Matthias Kloor¹, Aysel Ahadova¹

¹Department of Applied Tumor Biology, Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany; Cooperation Unit Applied Tumor Biology, German Cancer Research Center (DKFZ), Heidelberg, Germany ²Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany; German Cancer Consortium (DKTK) and German Cancer Research Center (DKFZ), Heidelberg, Germany ³Department of Internal Medicine I, University Hospital Bonn, Bonn, Germany; National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Bonn, Germany ⁴Translational and Clinical Research Institute, Newcastle University, International Centre for Life, Central Parkway, Newcastle, UK ⁵Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany ⁶Institute of Pathology, University Hospital Bonn, Bonn, Germany ⁷Department of Applied Tumor Biology, Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany; Cooperation Unit Applied Tumor Biology, German Cancer Research Center (DKFZ), Heidelberg, Germany; Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany; Department of Nephrology, Klinikum Stuttgart – Katharinenhospital, Stuttgart, Germany ⁸Department of Medical Oncology, National Centre for Tumor Diseases, Heidelberg University Hospital, Heidelberg, Germany ⁹Institute of Pathology, University Hospital Leipzig, Leipzig, Germany ¹⁰Department of Surgery, Ev. Krankenhaus Bethesda Hospital, Duisburg, Germany ¹¹Department of Gastrointestinal Surgery, Helsinki University Central Hospital, Helsinki, Finland; Applied Tumor Genomics Research Program, University of Helsinki, Helsinki, Finland ¹²Department of Molecular Pathology, Central Finland Hospital Nova, Jyväskylä, Finland ¹³Faculty of Sport and Health Sciences, University of Jyväskylä, Jyväskylä, Finland; Department of Surgery, Central Finland Hospital Nova, Jyväskylä, Finland ¹⁴Division of Hematology and Immunology, Leeds Institute of Medical Research at St James's, University of Leeds, Leeds, UK

Background and aim Tumors that arise in the context of Lynch syndrome (LS), the most common inherited colorectal cancer (CRC) syndrome, are characterized by pronounced immune infiltration which is attributable to a high load of immunogenic frameshift neoantigens. Moreover, systemic frameshift neoantigen-specific immune responses have been observed in LS CRC patients and cancer-free LS carriers, the latter pointing towards a life-long immune surveillance. We asked whether evidence for immune activation could also be found in normal mucosa of individuals with LS. Hence, we assessed the immune profile in tumor-distant normal colorectal mucosa from LS CRC patients, sporadic microsatellite-unstable (MSI) and microsatellite-stable (MSS) CRC patients, and cancer-free LS carriers.

Method CD3-, CD8- and FOXP3-positive T cells were quantified immunohistochemically in a total 233 normal colonic mucosa tissue sections from 132 individuals and in 26 LS CRC. Additionally, the densities of CD3-positive T cells were determined in an independent cohort of 97 normal rectal mucosa tissue sections from LS carriers enrolled in the CAPP2 clinical trial. Gene expression analysis, covering 770 immune-relevant genes, was conducted in a subset of normal mucosa and CRC samples using the NanoString nCounter platform.

Results MSI CRC patients as well as LS carriers presented with significantly higher CD3- and FOXP3-postive T cell counts in the normal mucosa compared to the MSS patient group. The mucosal density of CD8-positive T cells was not related to tumor MSI status, but was significantly elevated in the normal mucosa of healthy LS carriers. Gene expression and cluster analysis of LS specimens revealed distinct immune profiles in the colonic mucosa of LS carriers with and without tumor manifestation. The mucosa of healthy LS carriers was particularly characterized by an overrepresentation of CD45-positive, exhausted CD8-positive, NK, mast and B cell populations. Notably, long-term follow-up of LS carriers within the CAPP2 trial showed a significant correlation between mucosal T cell infiltrate in the rectum and time to subsequent CRC occurrence.

Conclusions LS carriers present with an altered mucosal immune milieu even prior to tumor manifestation. Importantly, the immune profile in the colorectal mucosa may be a temporary or permanent tumor risk modifier in LS carriers.

Keywords Lynch syndrome, Colorectal cancer, Normal mucosa, Immune infiltration, Tumor risk.

OP-002: Cancer Pathway-Somatic Mutations, Pathology

Molecular carcinogenesis pathway of MLH1-associated Lynch syndrome colorectal cancer unraveled: “two-in-one hit” model

Aysel Ahadova ¹, Volker Endris², Jan Niklas Wittemann¹, Leonie Gerling¹, Johannes Witt¹, Robert Hüneburg³, Jacob Nattermann³, Magnus Von Knebel Doeberitz¹, Matthias Kloor¹, Hendrik Bläker⁴

¹Department of Applied Tumor Biology, Institute of Pathology, University Hospital Heidelberg, and Cooperation Unit Applied Tumor Biology, German Cancer Research Centre (DKFZ), Heidelberg, Germany. ²Department of General Pathology, Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany. ³Department of Internal Medicine I, University Hospital Bonn, and National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Bonn, Germany. ⁴Institute of Pathology, University Hospital Leipzig, Leipzig, Germany

Background and aim Lynch syndrome (LS) carriers with MLH1 or MSH2 variants display distinct clinical and molecular colorectal cancer (CRC) manifestation. Although incident CRC risk is similar in MLH1 and MSH2 carriers, MSH2 carriers present with a higher risk of colorectal adenoma compared to MLH1 carriers. On the molecular level, MSH2 LS-CRC commonly show somatic APC mutations, whereas MLH1 LS-CRCs frequently display somatic CTNNB1 mutations. Except from MLH1-LS, CTNNB1 mutations are rare in CRC, most likely because biallelic CTNNB1 mutations seem to be required for causing an oncogenic effect in the colorectum. The mechanistic reasons behind the specific association of CTNNB1 somatic mutations with MLH1 germline predisposition have been unknown. In this study, we describe a new pathway of LS CRC formation specific for MLH1-LS carriers.

Method We analyzed MLH1-LS CRC for CTNNB1 mutation status and Loss of Heterozygosity (LOH) by Sanger sequencing. Using single nucleotide polymorphisms (SNP) within the chromosomal segment between MLH1 and CTNNB1, LOH was determined in CTNNB1-mutant MLH1 LS cancers. Whole Exome Sequencing (WES) was performed to validate LOH and analyze copy number alterations.

Results CTNNB1 mutations were found in 47% (17/36) of MLH1-LS CRCs. In 15 out of 16 (93.75%; 95% CI: 69.69–99.99%) analyzable samples, evidence for LOH spanning the entire region between MLH1 and CTNNB1 was detected. WES confirmed for all analyzed samples copy number-neutral LOH (cnLOH) events of > 49 Mbp in the chromosomal region 3pter–p21 including MLH1 and CTNNB1 genes. This indicates that one single cnLOH event causes two simultaneous hits: loss of MLH1 and CTNNB1 wild type alleles, resulting in MLH1 loss-of-function and beta-catenin gain-of-function at the same time.

Conclusions CTNNB1 being a neighbor of MLH1 on chromosome 3 reduces the number of somatic events required for CRC formation in MLH1 carriers compared to MSH2 carriers. The “two-in-one hit” model provides a plausible explanation for the clinico-pathological differences between MLH1 and MSH2 carriers.

Keywords Lynch syndrome, incident colorectal cancer, MLH1, CTNNB1, carcinogenic pathway.

Graphical abstract

OP-003: Cancer Pathway-Somatic Mutations, Pathology

Mutation rate evolution drives immune escape in mismatch repair-deficient cancer

Marnix Jansen ¹, Hamzeh Kayhanian¹, Panagiotis Barmpoutis², Eszter Lakatos³, William Cross¹, Giulio Caravagna⁴, Luis Zapata⁵, Kevin Litchfield¹, Christopher Steele¹, William Waddingham¹, Dominic Patel¹, Salvatore Milite⁴, Chen Jin², Ann Marie Baker³, Khurum Khan⁶, Daniel Hochhauser⁶, Benjamin Werner³, Genomics England Research Consortium Gel³, Marjolijn J.l. Ligtenberg⁷, Iris D. Nagtegaal⁸, Andrea Sottoriva⁵, Trevor Graham³, Nischalan Pillay¹, Manuel Rodriguez Justo⁹, Kai Keen Shiu⁶

¹UCL Cancer Institute, University College London, London, UK. ²Centre for Medical Image Computing, University College London, London, UK. ³Cancer and Evolution Lab, Barts Cancer Institute, Queen Mary University of London, London, UK. ⁴Department of Mathematics and Geosciences, University of Trieste, Trieste, Italy. ⁵Evolutionary Genomics and Modelling Lab, Centre for Evolution and Cancer, The Institute of Cancer Research, London, UK. ⁶Department of Gastrointestinal Oncology, University College London Hospital, London, UK. ⁷Department of Human Genetics, Radboud University Medical Center, Nijmegen, The Netherlands. ⁸Department of Pathology, Radboud University Medical Center, Nijmegen, The Netherlands. ⁹Department of Pathology, University College London Hospital, London, UK

Background and aim Secondary mismatch repair alterations, in particular frameshift mutations in the homopolymer tracts of the minor mismatch repair (MMR) genes MSH6 and MSH3, are frequently identified in MMR-deficient tumors. Such homopolymer mutations are currently thought to constitute neutral passenger mutations because of presumed epistatic redundancy (TCGA network, Nature 2012). However, recent studies have highlighted distinct and non-overlapping functions of the MutSa/b and MutL modules in DNA repair (Fang, Science Advances 2021 and Sanders, BioRXiv 2021). These data suggest that secondary MMR alterations may modulate mutation accumulation during MMR-deficient cancer progression.

Method Here we map the clonal topography of MMR-deficient colorectal cancer to show that genomic MMR mutability co-evolves with neoantigen selection to drive intratumour diversification and immune escape. We exploit a combination of detailed molecular pathology and molecular evolution studies, bio-informatic analyses on bulk populations (Genomics England, TCGA), and mathematical modelling studies.

Results We find that microsatellite instability modulates subclonal DNA repair by toggling two hypermutable mononucleotide homopolymer runs (C8 and A8, respectively) through stochastic frameshift mutations in the MMR genes MSH6 and MSH3. This drives variation in subclonal mutation rate, mutation bias, and clonal HLA diversity during MMR-deficient cancer evolution. Combined experimental and simulation studies demonstrate that subclonal immune selection favors incremental MMR mutations. MMR-deficient cancers thus fuel intratumour heterogeneity by adapting subclonal mutation rate and mutation bias to immune selection, revealing a conserved co-evolutionary arms race between neoantigen selection and adaptive genomic mutability.

Conclusions Our data reveal layers of mutational complexity and microsatellite biology in MMR-deficient cancer evolution previously hidden in bulk analyses and demonstrate that the MMR system unravels in incremental stages during MMR-deficient carcinogenesis in Lynch syndrome and sporadic patients (Figs. 1, 2).

Keywords Mismatch repair, frameshift, reversion, microsatellite, hypermutation, evolvability.

OP-004: Genetics: Research, Germline, Genomics, Variomics

Co-evolution of mismatch repair loss and the immune response in Lynch Syndrome

Ottilie D. Swinyard ¹, Ann Marie Baker², Chris Kimberly¹, Marnix Jansen³, Kevin J. Monahan⁴, Trevor A. Graham²

¹Evolution and Cancer Lab, Centre of Genomics and Computational Biology, Barts Cancer Institute, Queen Mary University of London, London, UK. ²Genomics and Evolutionary Dynamics Lab, Division of Molecular Pathology, Institute of Cancer Research, Sutton UK. ³Research Department of Pathology, UCL Cancer Institute, University College London, London, UK. ⁴Lynch Syndrome Clinic, Centre for Familial Intestinal Cancer, St Marks Hospital & Imperial College, London, UK

Background and aim Lynch syndrome (LS) is a hereditary disease defined by hemizygous pathogenic germline mutation or hypermethylation of one of the mismatch repair (MMR) genes. Complete MMR deficiency (MMR-d) is gained through loss of the remaining functional allele, causing a loss of expression of the MMR protein. The sequence and timing of events leading to cancer development in LS are not fully understood; indeed it is still disputed whether single MMR-d crypt foci (MMR-dcf) represent cancer-initiating lesions. Furthermore, how immune surveillance responds to MMR-dcf has not been examined. We aim to investigate the earliest interactions between MMR-dcf and the immune microenvironment in the LS bowel to characterise the co-evolution of MMR-d clones and the immune system.

Method We collected fresh and archival colon tissue from 32 patients with confirmed LS. The samples were classified as normal from a LS colon free of cancer (n = 13), normal from a LS colon with a distant cancer (n = 6) or adenoma adjacent normal tissue (n = 13). Immunohistochemical staining for MMR proteins identified the frequency and location of MMR-dcf. We then performed multiplexed, whole-slide cyclic immunofluorescence (CyCIF) to characterise the immune microenvironment around MMR-dcf using a panel of 21 markers, including those identifying T cells, B cells, macrophages, neutrophils, and immune checkpoints. Cell segmentation and marker quantification were performed to reveal the immune landscape of MMR-d patches.

Results We found a significantly higher frequency of morphologically normal MMR-d crypts in adenoma adjacent areas (n = 196/5149 crypts, found in 6/13 patients) compared to areas not adjacent to adenomas (n = 1/3333 crypts, found in 1/13 patients, p = 0.0384 t test), and the patch size of these foci increased with patient age. This suggests either that MMR-dcf occur very rarely in the non-dysplastic LS bowel, or that they are quickly removed by negative selection or immune predation. The higher prevalence of MMR-dcf near polyps could indicate that they are clonal precursors of the polyp, or they could form later as a result of the immune-suppressed microenvironment surrounding a polyp. Our CyCIF analysis of the immune microenvironment revealed that MMR-d and MMR-p crypts have distinct immune neighbourhoods, with MMR-df characterised by increased infiltration of cytotoxic T cells and decreased numbers of regulatory T cells. This suggests an active inflammatory response to MMR-d, but we did not see evidence of negative selection of these populations, perhaps due to acquisition of immune evasion mechanisms.

Conclusions We present evidence that MMR-dcf are rare in the LS bowel, however patches of morphologically normal MMR-dcf are more prevalent adjacent to adenomas. These foci appear to evoke a cytotoxic immune response, although we found evidence of immune evasion mechanisms that may prevent elimination of these MMR-d lesions. Future work will use next generation sequencing to inform on the clonality of the MMR-dcf and adjacent polyps, as well as inferring genetic acquisition of immune evasion in MMR-dcf.

Keywords Lynch syndrome (LS), mismatch repair (MMR), microsatellite instability (MSI), immune landscape, clonal evolution.

OP-005: Cancer Pathway-Somatic Mutations, Pathology

Lynch syndrome-associated epithelial ovarian cancer and its immunological profile

Maria Rasmussen ¹, Kevin Lim¹, Eva Rambech², Mads Hald Andersen³, Inge Marie Svane³, Ove Andersen¹, Lars Henrik Jensen⁴, Mef Nilbert⁵, Christina Therkildsen⁶

¹Department of Clinical Research, Copenhagen University Hospital—Amager and Hvidovre, Copenhagen, Denmark. ²Institute of Clinical Sciences, Division of Oncology and Pathology, Lund University, Sweden. ³National Center for Cancer Immune Therapy, Department of Oncology, Copenhagen University Hospital—Herlev and Gentofte, Copenhagen, Denmark. ⁴Department of Oncology, University Hospital of Southern Denmark, Vejle, Denmark. ⁵Department of Clinical Research, Copenhagen University Hospital—Amager and Hvidovre, Copenhagen, Denmark; Institute of Clinical Sciences, Division of Oncology and Pathology, Lund University, Sweden; Danish Cancer Society Research Center, The Danish Cancer Society, Copenhagen, Denmark. ⁶Department of Clinical Research, Copenhagen University Hospital—Amager and Hvidovre, Copenhagen, Denmark; The Danish HNPCC Register, Gastro Unit, Copenhagen University Hospital—Amager and Hvidovre, Copenhagen, Denmark

Background and aim Lynch syndrome is a multi-tumor syndrome characterized by mismatch repair deficiency (MMR-d), microsatellite instability (MSI), and increased tumor-infiltrating lymphocytes (TILs) making these tumors potential candidates for treatment with immune checkpoint inhibitors. However, response may depend on tumor-induced immune evasion mechanisms, e.g. loss of Beta-2-Microglobulin (B2M) or upregulation of programmed death protein ligand 1 (PD-L1). We investigated the immune response and B2M and PD-L1 expression in Lynch syndrome-associated ovarian cancers.

Method We successfully analyzed 30 Lynch syndrome-associated epithelial ovarian cancers collected through the Danish Hereditary Non-Polyposis Colorectal Cancer (HNPCC) register. MMR-d, MSI, immune response (CD3, CD8, and CD68), and immune evasion mechanisms (B2M and PD-L1) were investigated. Statistical associations between these markers were evaluated in addition to survival in relation to B2M/PD-L1.

Results Of the 29 evaluable tumors, 27 were MMR-d (93.1%). Likewise of 26 evaluable tumors, 14 were MSI (53.8%). MMR-d/MMR-proficiency associated with MSI/MSS in 60.0%. Half of the ovarian tumors presented with high levels of TILs. Loss of B2M expression was observed in 46.7% of the tumors, while expression of PD-L1 was seen in 28.0% of the cases. There was no association between B2M/PD-L1 and MSI/TILs/survival. Loss of B2M was often seen in tumors with low TILs (p = 0.056 or p = 0.059 for CD3 and CD8 positive cells, respectively).

Conclusions MMR-d, MSI, and TILs are also seen in Lynch syndrome-associated ovarian cancers making these potential candidates for checkpoint-based immunotherapy. The clinical impact from immune evasion through loss of B2M needs to be investigated further in larger cohorts.

Keywords Hereditary colorectal cancer, MHC class I, HLA class I, Immunoediting.

OP-006: Cancer Pathway-Somatic Mutations, Pathology

Immunogenicity and HLA binding affinity of Lynch Syndrome associated frameshift peptide neoantigens

Alejandro Hernandez Sanchez ¹, German Litau², Katharina Urban¹, Johannes Gebert¹, Matthias Kloor¹, Magnus Von Knebel Doeberitz¹

¹Collaboration Unit Applied Tumor Biology, German Cancer Research Center (DKFZ), Heidelberg, Germany; Department of Applied Tumor Biology, Institute of Pathology, University of Heidelberg, Heidelberg, Germany. ²Department of Applied Tumor Biology, Institute of Pathology, University of Heidelberg, Heidelberg, Germany

Background and aim Lynch Syndrome (LS) is caused by inactivation of the DNA mismatch repair (MMR) system that leads to accumulation of insertion and deletion (indel) mutations. Indel mutations in coding microsatellites (cMS) result in a shift of the reading frame giving rise to frameshift peptide (FSP) neoantigens. Manifest tumors share similar cMS mutations that confer a growth and survival advantage, and at the same time result in potentially immunogenic FSPs. The shared pool of FSPs and a defined population with increased cancer risk represent a unique opportunity to develop a preventive cancer vaccine. Recently, we identified novel FSP candidates with high cMS mutation frequency in MMR-deficient cancers and high predicted immunogenicity. Here, we aimed to evaluate the immunological properties of candidate FSPs and their potential for the development of an HLA-tailored FSP-based vaccine.

Method First, IFNg ELISpot was performed with splenocytes from FSP-vaccinated HLA-A2/HLA-DR1 transgenic mice to analyze peptide-specific HLA-A2-mediated T cell responses. Secondly, we evaluated binding of predicted FSP epitopes to HLA-A2 using a competition-based binding assay to identify HLA-A2-restricted epitopes that could potentially be recognized by cytotoxic CD8 + T cells, the main mediators of antitumor immunity. Lastly, splenocytes from FSP vaccinated HLA-A2/HLA-DR1 mice were re-stimulated in vitro with the identified epitopes in an IFNg ELISpot assay to assess correct processing and presentation to CD8 + T cells, followed up by an ex vivo killing assay.

Results Most of the FSPs induced FSP-specific T cells, indicating the presence of immunogenic epitopes within the FSP. Multiple high binder HLA-A2-restriced epitopes for 3 FSPs were identified in the competition-based binding assay. IFNg ELISpot revealed multiple epitopes within 2 FSPs that induce epitope-specific CD8 + T cells, confirming correct processing and presentation.

Conclusions Our results confirm the immunogenicity and correct processing and presentation of HLA-A2 restricted epitopes that induce CD8 + T cells for two of the candidate FSPs. The number of confirmed epitopes already for one specific HLA allele highlight the great potential of FSPs for vaccination approaches to prevent LS cancers. Here, we have established a pipeline for the identification of potential neoantigen epitopes to advance clinical translation of the next-generation neoantigen-based vaccines.

Keywords Lynch Syndrome, Colorectal Cancer, Cancer Vaccine, Immune prevention, Frameshift neoantigens.

OP-007: Epidemiology and Registries

The INDICATE initiative: Is HLA Type a modulator of tumor risk in Lynch syndrome?

Aysel Ahadova¹, Richard Gallon², Johannes Witt¹, Saskia Haupt³, Lena Bohaumilitzky¹, Sanne Ten Broeke⁴, Vincent Heuveline³, Gillian M. Borthwick², John Burn², Jukka Pekka Mecklin⁵, Magnus Von Knebel Doeberitz¹, Robert Hüneburg⁶, Jacob Nattermann⁶, Toni Seppälä⁷, Matthias Kloor ¹

¹Department of Applied Tumor Biology, Institute of Pathology, University Hospital Heidelberg, and Cooperation Unit Applied Tumor Biology, German Cancer Research Centre (DKFZ), Heidelberg, Germany. ²Translational and Clinical Research Institute, Newcastle University, International Centre for Life, Central Parkway, Newcastle upon Tyne, United Kingdom. ³Engineering Mathematics and Computing Lab (EMCL), Interdisciplinary Center for Scientific Computing (IWR), Heidelberg University, and Data Mining and Uncertainty Quantification (DMQ), Heidelberg Institute for Theoretical Studies (HITS), Heidelberg, Germany. ⁴Department of Genetics, University Medical Center Groningen, Groningen, the Netherland. ⁵Faculty of Sport and Health Sciences, University of Jyväskylä, and Department of Surgery, Central Finland Hospital Nova, Jyväskylä, Finland. ⁶Department of Internal Medicine I, University Hospital Bonn, and National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Bonn, Germany. ⁷Department of Gastrointestinal Surgery, Helsinki University Central Hospital, Helsinki, Finland; Applied Tumor Genomics Research Program, University of Helsinki, Helsinki, Finland; Tampere University, Tampere, Finland

Background and aim Lynch syndrome (LS) carriers have a broad lifetime cancer risk range (30–80%). More precise individual risk estimations for LS carriers would be of high clinical value, allowing tailored cancer prevention and surveillance. LS cancers are characterized by mismatch repair deficiency, leading to an increased mutation rate and accumulation of frameshift mutations. Frameshift peptides (FSPs) generated by translation of frameshift-mutated genes are highly immunogenic, and LS tumorigenesis is under stringent surveillance by the immune system. We hypothesize that a LS carrier’s human leukocyte antigen (HLA) type may play a crucial role in the presentation of FSP antigens to the immune system, and therefore may influence the likelihood of progression from pre-cancerous lesions into cancer.

Method To evaluate this hypothesis whilst addressing diverse populations with different HLA types, we have established the INDICATE (INDIvidual CAncer risk by HLA TypE) initiative. INDICATE is an international network aiming at the analysis of HLA type as a possible cancer risk modifier in LS.

Results Tumor and peripheral blood leukocyte DNA is being collected, and initial feasibility analyses have been performed. The HLA type of the first 350 LS patients has been successfully determined and the analysis of more samples is ongoing. Case number calculations based on 80% power and two-sided 95% confidence intervals demonstrate that an increase or decrease in the lifetime cancer risk of 18% would be detectable for common HLA types, such as HLA-A*02, and of 41% for rarer HLA types, such as HLA-B*27, in a cohort of 1000 LS carriers.

Conclusions An individual’s HLA type is expected to modify the immune response to cancer development. We have shown that it is feasible to analyze HLA type as a possible cancer risk modifier in LS carriers and will pursue this in frame of the INDICATE initiative (indicate-lynch.org).

Keywords Lynch syndrome, cancer risk, HLA type, immune response, risk modifier.

OP-008: Cancer Pathway-Somatic Mutations, Pathology

MSI-H and MSI-L mark opposite ends of the tetranucleotide repeat instability phenotype spectrum

Sabine Meessen¹, Nicola Currey¹, Zeenat Jahan², Hannah W Parker², Mark A Jenkins³, Daniel D. Buchanan⁴, John L Hopper³, Eva Segelov⁵, Jane E. Dahlstrom⁶, Maija R. J. Kohonen Corish ²

¹Garvan Institute of Medical Research, Sydney, Australia. ²Woolcock Institute of Medical Research, Sydney, Australia. ³Melbourne School of Population and Global Health, University of Melbourne, Melbourne, Australia. ⁴University of Melbourne Centre for Cancer Research, Melbourne, Australia. ⁵Monash University and Monash Health, Melbourne, Australia. ⁶ACT Pathology, The Canberra Hospital and Australian National University Medical School, Canberra, Australia

Background and aim A type of DNA mismatch repair defect known as EMAST (elevated microsatellite alterations at selected tetranucleotide repeats) is found in many different cancers (1). Tetranucleotide repeat instability is caused by MSH3 gene alterations or protein dysfunction/cellular mislocalisation (2,3). It shares a molecular basis with low microsatellite instability (MSI-L) in colorectal cancer (CRC). EMAST/MSI-L is emerging as a biomarker of poor prognosis in CRC (4). MSH3 deficiency may be a suitable target for specific cancer therapies (5). MSI-H tumors can also display tetranucleotide repeat instability (EMAST/MSI-H). It is important to distinguish EMAST caused by MSH3 alteration/dysfunction from EMAST/MSI-H. Therefore, we aimed to identify subtypes of tetranucleotide repeat instability and their molecular and clinicopathological features in CRC.

Method Five tetranucleotide repeat (D9S242, D20S82, L17686, UT5320, MYCL1) and five MSI markers (1) were used to classify 100 sporadic CRC according to the number of unstable markers detected (EMAST-0/5, 1/5, 2/5 and ≥ 3/5, MSI-L and MSI-H). Promoter methylation was determined using methylation-specific PCR (CIMP panel (6), MSH3, CDKN2A/p16, MCC). MSH3 mRNA expression was analysed by qPCR. Frequency of MSH3 copy number alterations, mutations, expression and methylation status was examined in the The Cancer Genome Atlas (TCGA 2018) cohort of 524 CRC (7,8). EMAST was also analysed in 22 MMR-deficient/MSI-H CRCs related to Lynch syndrome (15 MLH1, 6 MSH2, 1 PMS2).

Results 55% of the sporadic CRC were positive for one or more EMAST markers. The EMAST-1/5 subset (26%) was associated with advanced tumour stage, MSI-L and a lower frequency of CIMP-H/L. EMAST ≥ 3/5 tumours (13%) were more likely to have a proximal colon location and be MSI-H and CIMP-H (Fig. 1). Most Lynch syndrome cancers (18/22) also showed instability with 3–5 EMAST markers, 3/22 with two markers and 1/22 with one marker.

MSI-L was found in 35% of EMAST-1/5, 12% of EMAST-2/5 but none of EMAST ≥ 3/5 and only 4% (2/45) of EMAST-0/5 sporadic tumours. 92% of MSI-L cancers were located in the distal colon and rectum while there was no significant difference in the location of EMAST-1/5 cancers along the length of the large bowel. CIMP-H was completely absent in the MSI-L group. N stage was increased in EMAST-1/5 (62%) compared to EMAST ≥ 3/5 and 0/5 (23–27%) and was most frequent in MSI-L cancers (69%). EMAST-1/5 was associated with shorter overall survival when compared with all other patients as a single group.

No CpG island hypermethylation was detected in the MSH3 gene. MSH3 mRNA expression was down-regulated more than fourfold in 13% (12/96) of the specimens and this was associated with MSI-H, CIMP-H and EMAST ≥ 3/5 in our cohort. Only two (2/524) MSH3 truncating mutations were reported in the TCGA cohort but loss of heterozygosity (LOH) was more common (24%). As tetranucleotide repeat instability marker data are not available in the TCGA 2018 cohort, it could not be determined how the EMAST subtypes are distributed among MSH3 diploid and LOH cancers.

Conclusions Tetranucleotide repeat instability is found across a spectrum of phenotypes, from Lynch syndrome cancers and sporadic MSI-H/CIMP-H tumours to the MSI-L subset with lack of CIMP-H. Cancers with only one positive EMAST marker or MSI-L are associated with an increased likelihood of lymph node metastasis. EMAST-1/5 and MSI-L are not associated with MSH3 down-regulation but may be valuable markers for MSH3 dysfunction. This study has strengthened the importance of the dinucleotide repeat markers in the reference MSI panel for the identification of EMAST/MSI-L cancers.

Funding: Cancer Council NSW RG17-05 and RG19-01, Gastroenterological Society of Australia; Australasian Colon Cancer Family Registry: U.S. NCI U01 CA167551.

References 1. Boland CR, et al. Cancer Res 1998;58:5248. 2. Haugen AC, et al. Cancer Res 2008;68:8465. 3. Tseng-Rogenski SS, et al. Gastroenterol 2015;148:579. 4. Kohonen-Corish MR, et al. J Clin Oncol 2005;23:2318. 5. Dietlein F, et al. Trends Genet 2014;30:326. 6. Weisenberger DJ, et al. Nat Genet 2006;38:787. 7. Hoadley KA, et al. Cell 2018;173:291. 8. Meessen S, et al. Cancers 2021;13:3529.

Keywords EMAST, tetranucleotide repeat instability, MSI-L, MSI-H, MSH3, CIMP.

Fig. 1.

MSI subtype, CIMP subtype, Dukes stage and N stage distribution in sporadic colorectal cancers grouped according to EMAST ≥ 3/5, 2/5, 1/5 or 0/5 status (figure modified from Meessen S, et al. Tetranucleotide and Low Microsatellite Instability Are Inversely Associated with the CpG Island Methylator Phenotype in Colorectal Cancer. Cancers 2021;13:3529.

OP-009: Genetics: Research, Germline, Genomics, Variomics

Genome-wide DNA methylation and somatic mutational landscape resolve the molecular causality of MLH1-deficient colorectal cancer and designate MLH1 epimutation carriers

Jihoon E. Joo ¹, Khalid Mahmood¹, Sharelle Joseland¹, Peter Georgeson¹, Mark Clendenning¹, Romy Walker¹, Susan G. Preston¹, Mark A. Jenkins², Finlay A. Macrae³, Christophe Rosty⁴, Ingrid M. Winship⁵, Daniel D. Buchanan⁶

¹Colorectal Oncogenomics Group, Department of Clinical Pathology, The University of Melbourne, Parkville, Victoria, Australia; University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Parkville, Victoria, Australia. ²Centre for Epidemiology and Biostatistics, The University of Melbourne, Victoria, Australia. ³Department of Medicine, University of Melbourne, The Royal Melbourne Hospital, Parkville, Victoria, Australia; Genomic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Parkville, Victoria, Australia; Colorectal Medicine and Genetics, The Royal Melbourne Hospital, Parkville, Victoria, Australia. ⁴Envoi Pathology, Brisbane, Queensland, Australia. ⁵Department of Medicine, University of Melbourne, The Royal Melbourne Hospital, Parkville, Victoria, Australia; Genomic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Parkville, Victoria, Australia. ⁶Colorectal Oncogenomics Group, Department of Clinical Pathology, The University of Melbourne, Parkville, Victoria, Australia; University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Parkville, Victoria, Australia; Genomic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Parkville, Victoria, Australia

Background and aim MLH1 epimutations are a rare form of colorectal cancer (CRC) predisposition characterised by constitutional mono-allelic methylation of the MLH1 promoter. This study aimed to i) identify unique genome-wide DNA methylation and somatic mutation profiles of MLH1 epimutation CRCs and ii) apply this to classify CRCs of uncertain aetiology including germline MLH1 promoter variants of uncertain significance (VUS) and MLH1 methylated early-onset CRCs (EOCRCs diagnosed < 45 years).

Method Genome-wide DNA methylation (HMEPIC) and somatic mutational (whole exome or targeted sequencing) profiling was performed on tumour and normal DNA of 6 reference groups: 1) primary MLH1 epimutation (n = 4), 2) MLH1 epimutation due to c.-27 germline variant (n = 2), 3) Lynch syndrome (n = 9), 4) sporadic MLH1 methylated CRCs (n = 9), 5) CRCs with double somatic mutations in a MMR gene (n = 5) and 6) sporadic MMR-proficient CRCs (n = 9). DNA methylation-based consensus clusters were determined using these 6 reference groups of CRCs and applied to 3 germline MLH1 promoter VUS carriers (2 × MLH1:c.-11 and MLH1:c.- [28,7]) and 3 MLH1 methylated EOCRCs.

Results Consensus clustering identified 4 unique clusters comprised of Lynch, double somatic MMR and MMR-proficient CRCs (clusters 1 and 2), sporadic MLH1 methylated CRCs (cluster 3) and MLH1 epimutation CRCs (cluster 4). All 3 CRCs from germline MLH1 VUS carriers and 3 MLH1 methylated EOCRCs were assigned to cluster 4, suggesting an aetiology consistent with MLH1 epimutation. APC promoter methylation (mean > 0.3) was associated with MLH1 epimutation CRCs (p < 0.01) compared with sporadic MLH1 methylated group, and also observed in all three MLH1 methylated EOCRCs and one MLH1:c.-11 CRC. Somatic second hits in MLH1 were observed in CRCs from Lynch, epimutation, both MLH1:c.-11 and all MLH1 methylated EOCRCs but were absent from sporadic MLH1 methylated CRCs. MLH1 epimutation CRCs showed frequent mutations in APC, MBD4, and KMT2C, but no BRAF V600E and infrequent RNF43 mutations compared with sporadic MLH1 methylated CRCs. The CRCs from MLH1 VUS and EOCRCs also showed a somatic profile consistent with known MLH1 epimutation CRCs.

Conclusions DNA methylation and somatic mutation profiles of MLH1:c.-11 and c. [-28,-7] carriers and MLH1 methylated EOCRCs are consistent with those of MLH1 epimutation CRCs, suggesting a constitutional disruption of MLH1. MLH1 methylated EOCRCs were molecularly distinct from late-onset sporadic MLH1 methylated CRCs. Genome-wide DNA methylation profiling may aid in classifying clinically challenging subtypes of MLH1-deficient CRC for improved cancer risk management.

Keywords MLH1 epimutation, DNA methylation, Colorectal cancer, BRAF, VUS, MMR.

OP-010: Clinical: Diagnosis, Endoscopy, Management, Implementation

An evaluation of RMH Diagnostic pathways for CRC testing considering the introduction of the NHS National Genomic Test Directory and the potential benefit of immunotherapy in a subset of MMR-d, MSI CRC patients

Andrew George ¹, Kevin Monahan², Terri Mcveigh¹

¹The Royal Marsden NHS Foundation Trust, Associate Honorary Faculty, Institute of Cancer Research. ²Faculty of Medicine, Faculty of Medicine Centre & St. Marks Hospital, London

Background and aim The introduction of the NHS National Genomic Test Directory prompts an evaluation of existing diagnostic pathways. Developments in technology and a better understanding of microsatellite instability (MSI), tumour mutational burden (TMB) and immunotherapy in cancer has opened exciting opportunities for improvement. The Royal Marsden Hospital is one of the leading cancer hospitals in the world, with the Molecular Diagnostic Laboratory designated as the lead genomic cancer testing centre for the North Thames Genomic Laboratory Hub. We propose that integration of an MSI and TMB NGS caller with MMR gene reporting in cancer diagnostics could assist clinicians in suggesting immunotherapy response and, in conjunction with MMR IHC, assist in the diagnosis of Lynch Syndrome.

Method MMR/MSI/TMB NGS was correlated with available MMR IHC and MSI NGS was validated with Promega’s OnctoMATETM MSI Dx Analysis system.

Results The MSI NGS caller showed good concordance with MSI polymerase chain reaction (PCR) methods at 94.5% (86/91). Good concordance with mismatch repair deficient (MMR-d) immunohistochemistry (IHC) at 87.5% (21/24). The TMB caller demonstrated a right shifted normal distribution of the MMR-d IHC set vs MMR proficient IHC set however significant overlap between the two sets made interpretation limited. The TMB caller detected 3 POLE hypermutator cases with > 100 mutations/Mb, the TMB scores helped classify 3 previously undescribed POLE variants as likely pathogenic. We found a hit rate of 4.4% (46/1052) of cases suspected of Lynch Syndrome when MMR and MSI NGS is combined, including one known Lynch Syndrome case confirmed by germline testing.

Conclusions We conclude that while promising, further work is required to fully validate the MSI and TMB callers as well as a wider discussion is needed before the implementation of such an extension of scope to the RMH diagnostic pathway. Caution must be exercised when interpreting such experimental results in the diagnostic settings to not confuse the referring clinician and ensure that patient benefit is paramount. The previously untapped RMH dataset presents a unique opportunity to evaluate MSI, TMB and MMR-d in a variety of cancer types, with unique insights into clinical implementation and the limitations within one of the world’s leading genomics services.

Keywords Cancer, Colorectal, Diagnostics, Sequencing, MSI, TMB.

MSI NGS caller 1

MSI NGS results—MMR-proficient IHC set vs MMR-deficient IHC set. Proposed threshold of > 3 for MSI-I/H

TMB caller 1

TMB caller results 1—MMR-proficient IHC set vs MMR-deficient IHC set. Proposed thresholds at < 10 for TMB low, 10–30 TMB Intermediate and > 30 for TMB high

MSI NGS caller 2

MSI set 1 – MMR-p IHC
No. of samples	346
Median score	0
Interquartile range	0–1
Range	0 – 4
% < 3	87.63%
MSI set 2 – MMR-d IHC
No. of samples	42
Median score	6
Interquartile range	4–7
Range	0 – 9
% ≥ 3	85.71%

MSI score metrics for the MMR-p set and MMR-d IHC set

TMB caller 2

TMB set 1 – MMR-p IHC, NGS WT
No. of samples	199
Median score	13.56
Interquartile range	9.49 – 17.63
Range	2.71 to 42.04
TMB set 2 – MMR-d IHC
No. of samples	42
Median score	27.12
Interquartile range	21.7—36.28
Range	12.21 to 54.25

TMB score metrics for the MMR-p IHC , NGS WT set and MMR-d IHC set.

OP-011: Chemoprevention

Patient-derived 3D-Spheroid Cultures: A Novel Approach to in vitro Study of Familial Adenomatous Polyposis

Alicia Adams ¹, Jennifer Devecchio¹, Jacqueline Davenport¹, Sylvain Ferrandon¹, David Liska², Matthew F. Kalady²

¹Cancer Biology Department, Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, USA. ²Cancer Biology Department, Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, USA & Department of Colorectal Surgery, Cleveland Clinic Foundation, Cleveland, USA

Background and aim Patients with familial adenomatous polyposis (FAP) develop colorectal adenomas and eventually cancer without timely surveillance and intervention. Clinical phenotypes, progression from normal to adenoma, and progression from adenoma to colorectal cancer vary by the individual. Scientific models are needed to study this process. Three-dimensional colon organoid culture systems have recently been developed to study gastrointestinal normal physiology and disease. The goal of this study was to develop a consistent and applicable in vitro system to study neoplastic progression and its prevention in FAP patients.

Method Adenomatous polyps and non-polypoid normal-appearing colonic mucosa tissues were obtained from consented patients with FAP either during colonoscopy or from colectomy specimens. Tissues were minced and dissociated in collagenase I solution, resuspended in Matrigel beads, and grown in media containing Noggin, R-spondin, and Wnt-3a. Successful establishment of an organoid line was defined by the development of sustainable spheroid cultures. Hematoxylin and eosin (H&E) stains were performed to evaluate development of a lumen and crypt-like structures. ATP-based luminescence (CellTiter-Glo 3D) and Ki-67 assays were used to assess viability and proliferation, respectively. The known chemoprevention agent erlotinib was used to treat adenoma organoid cultures.

Results Thirty FAP patients underwent tissue sampling. We successfully created normal colonic mucosa organoids from 18 of 19 (95%) patients; and adenoma organoids from 26 of 30 (87%) patients. Histology of adenoma organoids resembled that of the primary patient tissue on H&E. The organoid cultures demonstrated physiologic growth and proliferation in vitro (Fig. 1). Erlotinib caused a dose-dependent decrease in viability (Fig. 2).

Conclusions Our group established 3D-culture organoid models from both non-adenomatous and adenomatous FAP tissues that recapitulate the patient’s disease. These models are being utilized in the laboratory to study pathogenesis, progression of disease, and response to chemoprevention agents. Furthermore, these patient-derived models have the potential to develop personalized treatments.

Keywords FAP, organoids, chemoprevention.

Image

Microscopic image of an organoid grown from an FAP adenoma demonstrating a lumen and crypt-like structures (A). H&E staining (B), and Ki67 immunostaining (C) of a spheroid. All images were taken at 20 × magnification.

Graph

Sulindac causes a dose-dependent decrease in FAP organoid viability.

OP-012: Chemoprevention

Pyrvinium Inhibits Familial Adenomatous Polyposis Patient-derived 3D Adenoma Organoids

Sylvain Ferrandon¹, Alicia Adams², Rami James Aoun ¹, Jennifer Devecchio², Xiang Shao², David Liska², Matthew Kalady¹

¹Division Colon and Rectal Surgery, Department of Surgery, The Ohio State Wexner Medical Center, OH. ²Cancer Biology Department, The Lerner Research Institute, Cleveland Clinic, Cleveland, OH

Background and aim Familial Adenomatous Polyposis (FAP) is a genetic disease caused by the germline mutation of the WNT-pathway repressor, APC. Patients with FAP develop hundreds to thousands of precancerous adenomatous polyps along the inner intestinal walls at an early age that inexorably result in colorectal cancer (CRC). WNT- pathway activation is a key factor in the transition and maintenance of colorectal cancer, therefore we hypothesized that effective inhibition of WNT signaling pathway by using pyrvinium, a CK1α inhibitor, could be a viable chemotherapeutic strategy to limit the development of adenomas in FAP.

Method Adenomatous polyps were obtained from consented patients during colonoscopy of colectomy specimen. Tissues were minced and dissociated using collagen I solution and resuspended in Matrigel beads. Cells were cultivated using suitable culture media to develop a three-dimensional colon organoid culture system. FAP patient-derived organoids were subsequently tested for viability after treatment with pyrvinium. After Pyrvinium treatment, the expression of genes (AXIN, CTNNB1, CCND1, GSK3 and PYGO) involved in WNT-signaling pathway were assessed using RT-qPCR.

Results We successfully created an FAP patient-derived organoid culture system. The organoid culture demonstrated growth and proliferation in vitro. We observed a significant dose dependent decrease of viability 72 h after pyrvinium treatment (Fig. 1). Mechanistically, we highlighted a strong inhibition of the expression of genes related to WNT pathway only 4 h after pyrvinium treatment.

Conclusions We have successfully established a reliable three-dimensional FAP patient-derived organoid system that can be used in vitro to decipher FAP cells physiology. We have shown that WNT-pathway inhibition using pyrvinium has the potential to decrease the FAP cells viability in a dose dependent manner providing a potential new therapeutic strategy for patient with FAP.

Keywords FAP, organoid, chemoprevention, Pyrvinium.

OP-013: Genetics: Research, Germline, Genomics, Variomics

APC mosaicism is a relevant explanation in (mild) polyposis phenotypes and testing reveals interesting cases such as > 1 mosaic cases in a family

Diantha Terlouw ¹, Manon Suerink², Alexandra Langers³, Monique Van Leerdam³, Demi Van Egmond⁴, Dina Ruano⁴, Carli Tops², Tom Van Wezel⁴, Hans Morreau⁴, Maartje Nielsen²

¹Department of Pathology, Leiden University Medical Center, Leiden, the Netherlands; Department of Clinical Genetics, Leiden University Medical Center, Leiden, the Netherlands. ²Department of Clinical Genetics, Leiden University Medical Center, Leiden, the Netherlands. ³Department of Gastroenterology, Leiden University Medical Center, Leiden, the Netherlands. ⁴Department of Pathology, Leiden University Medical Center, Leiden, the Netherlands

Background and aim Mosaic APC mutations have been identified as a common cause (> 25%) for unexplained colonic polyposis in patients with > 20 adenomas. Nevertheless, this remains an undervalued and understudied genetic cause in diagnostics. The frequency of APC mosaicism in milder phenotypes, such as < 10 adenomas or development of adenomas above the age of 70, is still unknown.

Method We analysed APC in multiple lesions of polyposis patients using target Next Generation Sequencing.

Results The mosaicism detection rate was 10% (33/318) in the entire cohort, 3% (1/38) in patients with < 10 adenomas, 5% (6/123) in those with 10–20 adenomas and 22% (26/118) in patients with > 20 adenomas. Furthermore, mosaicism was detected in 2% (1/43) of patients aged > 70. Besides ‘true’ mosaicism cases, 22% (70/318) showed a so called hybrid mosaicism, where multiple, but not all lesions share an identical variant. Moreover, while testing for APC mosaicism we identified extraordinary mosaic cases. For instance, we found a family with two mosaic patients and a patient with a mosaic variant in 15–17% in semen DNA while this variant was detected in 6% in leukocyte DNA.

Conclusions Besides its pivotal role in unexplained polyposis patients, APC mosaicism is also relevant in milder phenotypes. Furthermore, a substantial proportion of our cohort has a hybrid mosaic pattern with varying percentages of colonic adenomas sharing an identical APC variant. Also, analyses of APC mosaicism has revealed interesting cases which might be of interest for other genetic mosaicisms or point to so far unknown genetic causes.

Keywords APC mosaicism, polyposis coli.

OP-014: Genetics: Research, Germline, Genomics, Variomics

Microbiome Profiling in Patients with Adenomatous Polyposis Compared to sporadic Subjects

Revital Kariv ¹, Naomi Fliss¹, Dana Ivankovsky², Laura Sol Grinshpon², Dror Hadas³, Shira Zelber Sagi², Elhanan Borenstein³

¹Department of Gastroenterology,Tel Aviv Sourasky Medical Center and Faculty of Medicine,Tel Aviv University. ²School of public health, Haifa university. ³Department of Clinical Microbiology and Immunology, Faculty of Medicine School of Computer Science, Faculty of Exact Sciences Edmond J. Safra Center for Bioinformatics Tel Aviv University

Background and aim Adenomatous polyposis syndromes are rare inherited conditions caused by mutations in cancer-related genes (mostly APC but also POL, MUTYH, NTHL1, MSH3, and others) with a phenotype of numerous adenomatous polyps along the gastrointestinal (GI) tract. High GI cancer risk in polyposis patients requires dedicated endoscopic surveillance, while chemoprevention and lifestyle have not been shown to have a major impact on polyp burden and cancer risk. In contrast, sporadic colorectal cancer (CRC) and its precursors have been robustly linked to lifestyle, metabolic profile, and nutrition, as well as to the gut microbiome. Both polyposis and CRC patients were shown to have distinct microbiome composition and with increase in the abundance of specific taxa. Such altered microbiome profiles may be used for early and personalized prevention, diagnosis, and therapeutic decisions.

Method To better characterize the microbiome of adenomatous polyposis patients, we aimed to conduct a comprehensive comparison of microbiome compositions between adenomatous polyposis patients and sporadic patients, using 16S rRNA metagenomic profiling. To this end, we obtained, sequenced, and analyzed samples from 19 adenomatous polyposis patients, with or without previous colonic surgery, and 49 controls. Samples were sequenced by Merck and were analyzed using QIIME2.

Results Our analysis identified significant differences between polyposis patients and controls (including subjects with and without sporadic polyps), in terms of both overall diversity and composition. Specifically, polyposis patients were associated with a significantly less diverse microbiome (Fig. 1). Further analysis, comparing polyposis patients to controls additionally identified an increase in the abundance of Blautia and Bifidobacterium and a decrease in the abundance of Bacteroides among polyposis patients.

Conclusions These initial findings suggest that adenomatous polyposis-specific microbiome patterns can be clearly detected even in a relatively small and heterogeneous cohort of adenomatous polyposis patients. Further studies and analyses to validate these findings and to pinpoint specific markers for clinically-relevant stratification are currently underway.

Keywords Microbiome, Adenomatous polyposis, Diversity, APC.

Microbiome profiling in adenomatous polyposis patients versus controls

(A) Shannon diversity and (B) composition of 19 FAP patients in comparison to 49 controls. Composition is shown at the genus level.

OP-015: Chemoprevention

Preventive Anti-inflammatory Diet to Reduce Gastrointestinal Inflammation in Familial Adenomatous Polyposis Patients: A Prospective Pilot Study

Antonino Belfiore¹, Chiara Maura Ciniselli², Stefano Signoroni³, Manuela Gariboldi⁴, Giulio Ferrero⁵, Oscar Illescas⁴, Andrea Mancini⁶, Licia Rivoltini⁷, Daniele Morelli⁸, Eleonora Bruno⁹, Alessandra Macciotta², Maria Teresa Ricci³, Clorinda Brignola³, Elena Daveri⁷, Laura Cattaneo¹, Giuliana Gargano⁹, Giovanni Apolone¹⁰, Massimo Milione¹, Paolo Verderio², Patrizia Pasanisi ⁹, Marco Vitellaro³

¹First Pathology Division, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ²Unit of Bioinformatics and Biostatistics, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ³Unit of Hereditary Digestive Tract Tumors, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁴Unit of Genetic Epidemiology and Pharmacogenomics, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁵Department of Clinical and Biological Sciences, University of Turin, Orbassano, Italy. ⁶Unit of Diagnostic and Therapeutic Endoscopy, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁷Unit of Immunotherapy of Human Tumors, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁸Laboratory Medicine Division, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁹Unit of Epidemiology and Prevention, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ¹⁰Scientific Directorate, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy

Background and aim Familial adenomatous polyposis (FAP) is an autosomal-dominant inherited condition associated with germline mutations in the adenomatous polyposis gene. Patient management involves prophylactic surgery and intensive endoscopic surveillance throughout life. Diet is a major concern for patients with FAP, who are generally symptom-free before surgery but tend to have problems related to bowel function postoperatively. We hypothesized that a low-inflammatory diet based on the principles and recipes of the Mediterranean diet would reduce markers of local and systemic inflammation.

Method Twenty-eight patients with FAP older than 18 years, who underwent rectum-sparing prophylactic colectomy and were included in our surveillance program, participated in a pilot dietary intervention study. Blood and stool samples were collected at baseline (T0), at the end of the dietary intervention (T1, three months), and at the end of the study (T2, six months after T0). Markers of gastrointestinal inflammation including fecal calprotectin, cyclooxygenase-2, and 15-hydroxyprostaglandin dehydrogenase were assessed. RNA-sequencing analysis of healthy tissue/polyps removed at T0 and T2 was also conducted. Serum calprotectin, insulin, insulin-like growth factor-1, C-reactive protein, and glycated hemoglobin were also assessed.

Results Significant changes in serum calprotectin, insulin, and insulin-like growth factor-1 levels occurred over time. Significant borderline changes were observed in the neutrophil-to-lymphocyte ratio. These changes were evident immediately at the end of the 3-month active dietary intervention (T1). A significant increase in 15-hydroxyprostaglandin dehydrogenase expression in normal crypts of matched samples was also observed between T0 and T2. A total of 80 genes, mainly involved in inflammation and immune response, were found modulated by the diet.

Conclusions This pilot study supports the hypothesis that a low-inflammatory diet may modulate gastrointestinal markers of inflammation in individuals with FAP.

Keywords Familial Adenomatous Polyposis; Anti-infllammatory diet; gastrointestinal inflammation; total colectomy.

OP-016: Clinical: Diagnosis, Endoscopy, Management, Implementation

The opinions of adolescents and young people with familial adenomatous polyposis about timing of surgery: a service evaluation

Jacqueline M. Hawkins ¹, Roshani Patel¹, Alison Leary², Susan K Clark¹

¹The Polyposis Registry, St Mark’s Hospital, London, UK. ²London South Bank University, London, UK

Background and aim Children with familial adenomatous polyposis (FAP) have adenomas from 12–14 years of age; prophylactic surgery is usually needed to prevent cancer. In our institution this surgery is usually performed in the summer holidays after public examinations, when they are 16–18 years of age. Many centres advocate operating later.

The aim of this study was to gain the opinions of adolescents and young people with FAP about our current service and the timing of their surgery.

Method A previously validated questionnaire was adapted, and a prospective, descriptive cross-sectional survey was used to collect qualitative data. Patients who had undergone prophylactic colectomy with ileorectal anastomosis between 01/06/2014 and 30/09/2019 and were aged 14–20 years at surgery, were included. Patients were recruited when they attended their routine follow-up. Demographic data and descriptive statistics were collected anonymously, and thematic analysis was used to develop themes. Table 1 shows the demographic details of the participants.

Table 1 Causes of death following duodenectomy

Results Forty-three patients were eligible, 31 had scheduled appointments during the 2-month recruitment period, and 25 attended and were recruited; 24 (96%) returned completed questionnaires.

Fifteen participants (62%) were between 16—18 years when they had their operation. Five (21%) were younger (13—15 years) and four (17%) were over 18 years. Time since surgery ranged from less than 1 year to 5 years.

Twenty-two participants (92%) said that they had their “operation at the right time” and 19 (79%) said they “preferred to get their operation over with”.

Nineteen (79%) disagreed / strongly disagreed with the comment that they would have “preferred to delay their operation” and one (4%) agreed.

Two (8%) preferred to have surgery when younger, and 18 (75%) disagreed / strongly disagreed with this statement.

The following themes emerged:

clinical information giving.

health awareness.

timing of surgery.

anxiety.

Five (21%) participants stated that they were not anxious about their admission to hospital for surgery and five (21%) were very anxious.

Conclusions This group of young patients “preferred to get their operations over with” and felt that they had their operation at the right time, after public examinations, between the age of 16–18 years.

Keywords adolescents, young people, FAP, surgery, evaluation.

Results table

Results from participant questions about their thoughts about timing of their surgery.

Table 1

Participant demographic details

OP-017: Clinical: Diagnosis, Endoscopy, Management, Implementation

Superior Rectal Artery sparing in total colectomies with ileorectal anastomosis to reduce anastomotic leakage in patients with Familial Polyposis

Gaia Colletti¹, Chiara Maura Ciniselli², Stefano Signoroni³, Emanuele Rausa³, Maria Teresa Ricci³, Federica Cavalcoli⁴, Andrea Magarotto⁴, Clorinda Brignola³, Daniela Zaffaroni³, Paolo Verderio², Marco Vitellaro ³

¹Colorectal Surgery Division, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ²Unit of Bioinformatics and Biostatistics, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ³Unit of Hereditary Digestive Tract Tumors, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁴Unit of Diagnostic and Therapeutic Endoscopy, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy

Background and aim Preventive surgery has the role to reduce the risk of cancer in patients with Familial Adenomatous Polyposis (FAP). Total colectomy with rectal sparing is the surgical procedure that better meets the needs of minimal morbidity and the best quality of life. The aim of this study is to evaluate the effectiveness of Superior Rectal Artery (SRA) preserving in reduction of the anastomotic leakage rate.

Method Retrospective study based on a prospective data collection in the Hereditary Digestive Tract Tumours Unit database of the National Cancer Institute of Milan from 2000 to 2020. Inclusion criteria: all patients treated with total colectomy and ileorectal anastomosis (IRA) for FAP condition. All patients’ characteristics, surgical (laparoscopic or open) and pathological data were collected. Exclusion criteria: patients treated with proctocolectomy with ileopouch anastomosis. The patient population included 196 patients: 68 colectomies with preservation of the SRA (Group A), 128 without preservation of the SRA, with the Inferior Mesenteric Artery (IMA) (Group B).

Results A total of 97 female and 99 male patients were included in the study; median age 30yrs (IQR range, 20–49 yrs), while median BMI 23 (IQR range, 20–26). In group A, no anastomotic dehiscence cases were observed; in group B, 10 cases (7,8%) of anastomotic leakage occurred (Fisher p-value = 0.02). We report postoperative anastomosis bleeding in 24 patients of group A (35%), all treated with endoscopic revision. Differences in time of surgery were relevant (Kruskal–Wallis p-value < 0.01): 344 min (IQR: 300–387) and 240 min (IQR: 180–300) in Group A and B, respectively.

Conclusions This study shows a significant decrease of postoperative anastomotic leakage rate in the SRA preservation group compared to control group. The SRA preservation adds a further technical surgical step to total colectomy with rectal sparing for FAP patients but these results could give a substantial advantages to this cohort of patients.

Keywords FAP; Superior Rectal Artery preserving; Preventive surgery; Anastomotic leakage.

OP-018: Clinical: Diagnosis, Endoscopy, Management, Implementation

Prevalence and Management of Cancer of the Rectal Stump after Total Colectomy and Rectal Sparing in Patients with Familial Polyposis: Results from a Registry-Based Study

Gaia Colletti¹, Chiara Maura Ciniselli², Stefano Signoroni³, Ivana Maria Francesca Cocco⁴, Andrea Magarotto⁵, Maria Teresa Ricci³, Clorinda Brignola³, Clara Bagatin², Laura Cattaneo⁶, Andrea Mancini⁵, Federica Cavalcoli⁵, Massimo Milione⁶, Paolo Verderio², Marco Vitellaro ¹

¹Colorectal Surgery Division, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ²Unit of Bioinformatics and Biostatistics, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ³Unit of Hereditary Digestive Tract Tumors, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁴Department of General Surgery, Whipps Cross University Hospital, London E11 1 NR, UK. ⁵Unit of Diagnostic and Therapeutic Endoscopy, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁶First Pathology Division, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy

Background and aim The balance between quality of life and colorectal cancer risk in familial adenomatous polyposis (FAP) patients is of primary importance. A cut-off of less than 30 polyps under 1 cm of diameter in the rectum has been used as an indication for performing ileo-rectal anastomosis (IRA) in terms of lower rectal cancer risk. This study aimed to assess clinical and surgical features of FAP patients who developed cancer of the rectal stump.

Method This retrospective study included all FAP patients who underwent total colectomy/IRA from 1977 to 2021 and developed subsequent rectal cancer. Patients' features were reported using descriptive statistics by considering the overall case series and within pre-specified classes of age (< 20, 20–30, and > 30 years) at first surgery.

Results Among the 715 FAP patients, 47 (6.57%, 95% confidence interval: 4.87; 8.65) developed cancer in the rectal stump during follow-up. In total, 57.45% of the population were male and 38.30% were proband. The median interval between surgery and the occurrence of rectal cancer was 13 years. This interval was wider in the youngest group (p-value: 0.012) than the oldest ones. Twelve patients (25.53%) received an endoscopic or minimally invasive resection. Amongst them, 61.70% were Dukes stage A cancers.

Conclusions There is a definite risk of rectal cancer after total colectomy/IRA; however, the time interval from the index procedure to cancer developing is long. Minimally invasive and endoscopic treatments should be the procedures of choice in patients with early stage cancers.

Keywords FAP; colorectal surgery; hereditary syndrome; ileo-rectal anastomosis; prevention; rectal stump cancer; total colectomy.

OP-019: Genetics: Research, Germline, Genomics, Variomics

Evaluation of novel approaches to support variant classification in DNA mismatch repair genes

Romy Walker ¹, Mark Clendenning¹, Eric J. Joo¹, Khalid Mahmood¹, Peter Georgeson¹, Sharelle Joseland¹, Julia Como¹, Sue Preston¹, Mark A. Jenkins², Finlay A. Macrae³, Christophe Rosty¹, Ingrid M. Winship⁴, Daniel D. Buchanan⁴

¹Colorectal Oncogenomics Group, Department of Clinical Pathology, Victorian Comprehensive Cancer Centre, The University of Melbourne, Melbourne, Australia; University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Melbourne, Australia. ²University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Melbourne, Australia; Centre for Epidemiology and Biostatistics, School of Population and Global Health, The University of Melbourne, Melbourne, Australia. ³Genomic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Melbourne, Australia; Colorectal Medicine and Genetics, The Royal Melbourne Hospital, Melbourne, Australia. ⁴Colorectal Oncogenomics Group, Department of Clinical Pathology, Victorian Comprehensive Cancer Centre, The University of Melbourne, Melbourne, Australia; University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Melbourne, Australia; Genomic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Melbourne, Australia

Background and aim Germline testing of the DNA mismatch repair (MMR) genes often results in the detection of variants of uncertain significance (VUS). Here, we assessed features commonly observed in carriers of germline MMR pathogenic variants to determine their utility in variant classification. We evaluated targeted tumour sequencing (TTS) data for 1) microsatellite instability (MSI-H), 2) presence of a second somatic hit in the same MMR gene as the germline VUS and 3) detection of MMR-deficient (dMMR) crypts in normal colonic mucosa as part of the multifactorial approach to MMR variant classification.

Method Germline MMR VUS (n = 17) were identified by clinical testing from 9 colorectal (CRC) and 8 endometrial (EC) cancer-affected participants of the ANGELS study. The variants comprised of 6 missense, 3 in-frame deletions, 5 intronic/splice, 2 duplications and one 3’ UTR. Data from co-segregation variant testing and RNA splicing analysis were collected as part of the canonical multifactorial approach for variant classification. Data on MSI status (MsMuTect) and presence of somatic MMR mutation or LOH (second hit) from TTS and the presence of a dMMR crypt from the patients’ normal colon tissue were collected where possible. A reference group of known MMR pathogenic variant carriers from the Australasian Colon-CFR also underwent TTS (13 CRCs and 6 ECs) and dMMR crypt screening (n = 5).

Results In our reference group, TTS identified 100% (19/19) as MSI-H and a second hit was observed in 84.2% (16/19) of the tumours while screening of normal mucosa identified at least one dMMR crypt in 60% (3/5) of those tested. The multifactorial model enabled the reclassification of 5/17 VUS (29.4%) to likely pathogenic/pathogenic. For each of these 5 reclassified variants, TTS confirmed MSI-H and identified a second hit and for 4 out of the 5 cases tested, we identified two with dMMR crypts (50%). For the remaining 12 VUS (70.6%) whose classification remained unchanged, 10 VUS tumours (83.3%) were MSI-H and harboured a second hit. Out of six normal mucosa that were available for screening, we found one dMMR crypt (16.7%). The remaining two VUS tumours were not MSI-H and absent of a second hit or a dMMR crypt.

Conclusions Our study demonstrated that confirmation of MSI-H and identification of a second somatic hit by TTS and the presence of dMMR crypts were consistent with pathogenic MMR variant classification. These approaches may add to the multifactorial approach to MMR variant classification, although further validation is needed.

Keywords Colorectal cancer, DNA mismatch repair deficient crypts, endometrial cancer, multifactorial model, targeted tumour sequencing, VUS.

OP-020: Genetics: Research, Germline, Genomics, Variomics

A fully calibrated and validated in vitro functional assay for the diagnosis of Variants of Uncertain Significance in Lynch syndrome

Emily Rayner ¹, Mark Drost¹, Yvonne Tiersma¹, Sandrine Van Hees Stuivenberg¹, Bryony A. Thompson², Cristina Fortuno³, Amanda B. Spurdle³, David E. Goldgar⁴, Marc S. Greenblatt⁵, Sean V. Tavtigian⁴, Niels De Wind¹

¹Leiden University Medical Center, Leiden, Netherlands. ²Department of Pathology, Royal Melbourne Hospital, Parkville, VIC, Australia. ³QIMR Berghofer Medical Research Institute, Brisbane, Australia. ⁴University of Utah, Salt Lake City, UT, USA. ⁵University of Vermont Medical Centre, Burlington, VT, USA

Background and aim The classification of Variants of Uncertain Significance (VUS), identified in individuals suspected of Lynch syndrome, is hindered in the absence of sufficient clinical or phenotypic data. Since a functional defect in DNA mismatch repair (MMR) is causative of the cancer predisposition, calibrated and validated functional MMR assays are useful tools for VUS classification. We have developed, calibrated and validated the Complete In vitro MMR Activity (CIMRA) assay. The assay requires only sequence information on the VUS as input.

Method We assayed the MMR function of a calibration set comprised of variants in MSH2, MLH1, MSH6 and PMS2, that were classified by the InSiGHT Variant Interpretation Committee as either pathogenic (cancer-predisposing) or benign, without using functional assay data. In case the number of available classified pathogenic variants was insufficient for calibration, we added phenotypically inactivating (i.e. proxy-pathogenic) variants, generated using a genetic screen. We then performed regression analysis of the CIMRA assay activities against their independent classifications. This yielded a regression formula for each MMR gene, which converts the CIMRA assay values into an Odds of Pathogenicity (OddsPath).

Results Validation of the regression formulae revealed that the CIMRA assay-derived OddsPath provides strong evidence in favour or against cancer predisposition for 96% of variants tested, of which 3% were discordant with their independent clinical classification. Thus, the CIMRA assay has high predictive values for the classification of VUS in all MMR genes.

Conclusions The CIMRA assay requires no patient material, uses common laboratory equipment, is ‘kittable’, scalable, predictive and reproducible. The CIMRA OddsPath can be combined with clinical and with additional quantitative data, such as in silico predictions, to further increase its predictive values. We envisage that the calibrated and validated CIMRA assay, that is embedded in the upcoming ACMG/AMP guidelines as a strong evidence type for (PS3) or against (BS3) pathogenicity, will set a paradigm for the translation of personalized cancer genetics into personalized healthcare.

Drost, M et al. (2019). A functional assay–based procedure to classify mismatch repair gene variants in Lynch syndrome. Gen Med, 21, 1486.

Drost, M et al. (2020). Two integrated and highly predictive functional analysis-based procedures for the classification of MSH6 variants in Lynch syndrome. Gen Med, 22, 847.

Keywords Lynch Syndrome, Variants of Uncertain Significance, Functional analysis-based classification, Assay calibration and validation.

Graphical Abstract

Outline of the calibrated and validated CIMRA assay to assess variant causativity of Lynch Syndrome

OP-021: Genetics: Research, Germline, Genomics, Variomics

Providing More Answers for Patients with Supplemental RNA Analysis of Colorectal Cancer Associated Genes

Daniel E. Pineda Alvarez, Laure Fresard, Victoria Carlton, Kate Krempely, Hio Chung Kang, Nick Kampshughes, Keith Nykamp, Brandie Leach

Invitae, San Francisco, CA, USA

Background and aim Identification of pathogenic/likely pathogenic (P/LP) germline variants is important to clarify a patient’s cancer risk; guide screening, determine risk reducing interventions, and treatment; and inform cascade testing. The introduction of next generation sequencing panels has improved P/LP variant rates while simultaneously increasing variant of uncertain significance (VUS) rates, relative to standard Sanger sequencing. RNA sequencing can help to resolve the clinical significance of variants predicted to alter RNA splicing and identify splice-altering variants in regions outside of the standard reportable range (± 20 bp of intronic sequence adjacent to the exons) of routine DNA sequencing. The aim of this study was to report the VUS resolution and variant discovery from RNA sequencing performed at a single commercial laboratory.

Method An RNA-sequencing assay was run on whole blood for 63 transcripts from an 84 multi-cancer gene panel. Analysis was limited to 20 genes (APC, AXIN2, BMPR1A, CDH1, CHEK2, EPCAM, GREM1, MLH1, MSH2, MSH3, MSH6, MUTYH, NTHL1, PMS2, POLD1, POLE, PTEN, SMAD4, STK11, TP53) related to hereditary colorectal cancer (CRC) risk. Data were captured to identify eligible variants reclassified to benign/likely benign (B/LB) or P/LP as well as variants discovered outside of the DNA sequencing range as a result of RNA sequencing.

Results 16,390 unique samples were run on the DNA/RNA assay. A total of 1,581 patients had eligible variants for RNA analysis. Based on RNA analysis, VUSs were downgraded to B/LB for 677 (4.1%) patients and upgraded to P/LP for 24 (0.1%) patients. 8 (0.05%) patients had deep intronic variants discovered in APC, MSH2, MSH3, NTHL1, and POLE. Of note, two of these variants were classified as LP, while the remainder were VUS.

Conclusions RNA analysis helped to reclassify eligible splice-altering variants in genes associated with hereditary CRC in 4.2%, with the majority being downgraded to B/LB. The rate of variant discovery was very low.

Keywords hereditary colorectal cancer, genes, RNA, variant interpretation.

OP-022: Genetics: Research, Germline, Genomics, Variomics

Mismatch Repair variant classification with ACMG/AMP criteria and comparison with Bayesian probability framework

John Paul Plazzer¹, Finlay Macre ¹, Bryony Thompson², Andreas Laner³, Elke Holinski Feder³

¹Department of Colorectal Medicine and Genetics, The Royal Melbourne Hospital, Parkville, Australia. ²Department of Pathology, The Royal Melbourne Hospital, Parkville, Australia. ³Medical Genetics Center Munich, MGZ, Munich, Germany

Background and aim The InSiGHT Hereditary Colorectal Cancer / Polyposis Variant Curation Expert Panel (VCEP) modified the ACMG/AMP classification guidelines for the MMR genes for the ClinGen VCEP approval process. These classification criteria incorporate both guidance from ClinGen SVI Working Group and the expertise of the previously established InSiGHT Variant Interpretation Committee, and incorporate published Bayesian parameters for Lynch Syndrome tumor characteristics, segregation and functional assays results. A pilot batch of 62 variants were classified using the new criteria.

Method Variants were selected based on new or existing evidence meeting ACMG/AMP criteria which may alter their previous InSiGHT classification or lead to a new classification. Variants were assigned to multiple reviewers and classification outcomes were discussed during teleconferences. The criteria were refined resulting in the ACMG/AMP classification which were compared to published Bayesian probabilities for 48 variants.

Results After reaching VCEP consensus classification with ACMG/AMP criteria, 22 (35%) variants were Pathogenic/Likely Pathogenic, 23 (37%) were VUS and 17 (27%) were Benign/Likely Benign. The results show concordance of 32/48 (67%) variants between the ACMG/AMP and the established InSiGHT Bayesian classifications. Categories of discordant variants include 4 out of 48 (8%) classifications that could only be achieved with ACMG/AMP criteria, and another 4 out of 48 (8%) achieved using only the Bayesian approach. An additional 8 variant classifications were discordant but it was determined that their benign Bayesian classification was not accurate due to insufficient clinical evidence. Overall, 92% of ACMG/AMP classifications were concordant or clinically valid, improving upon the 75% for Bayesian classifications alone.

Conclusions The ACMG/AMP criteria cover evidence types such as variant allele frequency, variant types and splicing data which are not in the current Bayesian framework. The Bayesian calculations were able to classify variants where evidence is marginal (e.g. tumor or co-segregation data is limited and may not meet required ACMG/AMP levels of strength). Overall, these new criteria complement the Bayesian approach and supersede previous InSiGHT classification criteria for MMR genes due to their flexibility and potential for clinical adoption. The current version of the MMR criteria, set after this pilot project, requires endorsement and publication by the ClinGen SVI.

Keywords genetic variant classification criteria.

OP-023: Genetics: Research, Germline, Genomics, Variomics

APC-specific ACMG/AMP variant classification guideline alleviates the burden of variants of uncertain significance in ClinVar and locus-specific databases

Xiaoyu Yin¹, Isabel Spier², Marcy Richardson³, Marta Pineda⁴, Deborah Ritter⁵, Tina Pesaran³, Gabriel Capellá⁶, Sean V. Tavtigian⁷, John Paul Plazzer⁸, Andreas Laner⁹, Julie Boyle¹⁰, Pilar Mur⁴, Thomas V. O. Hansen¹¹, Xuemei Shi¹², Khalid Mahmood¹³, Andrew Latchford¹⁴, Ian M. Frayling¹⁵, Sharon Plon⁵, Marc Greenblatt¹⁶, Finlay A. Macrae¹⁷, Stefan Aretz ²

¹Department of Colorectal Medicine and Genetics, Royal Melbourne Hospital, Parkville, Australia; Department of Medicine, University of Melbourne, Parkville, Australia; Institute of Human Genetics, Medical Faculty, University of Bonn, Bonn, Germany. ²Institute of Human Genetics, Medical Faculty, University of Bonn, Bonn, Germany; National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Bonn, Germany; European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS) – Project ID No 739547. ³Ambry Genetics, Aliso Viejo, California, USA. ⁴Hereditary Cancer Program, Catalan Institute of Oncology – IDIBELL, Barcelona, Spain; Centro de Investigación Biomédica en Red de Cáncer (CIBERONC), Instituto Salud Carlos III, Madrid, Spain. ⁵Baylor College of Medicine, Houston, Texas, USA; Texas Children's Cancer Center, Baylor College of Medicine, Houston, Texas, USA. ⁶European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS) – Project ID No 739547; Hereditary Cancer Program, Catalan Institute of Oncology – IDIBELL, Barcelona, Spain; Centro de Investigación Biomédica en Red de Cáncer (CIBERONC), Instituto Salud Carlos III, Madrid, Spain. ⁷Department of Oncological Sciences, School of Medicine, University of Utah, Salt Lake City, Utah, USA; Huntsman Cancer Institute, University of Utah, Salt Lake City, Utah, USA. ⁸Department of Colorectal Medicine and Genetics, Royal Melbourne Hospital, Parkville, Australia. ⁹Department of Genomics, Medical Genetics Center Munich, Munich, Germany. ¹⁰Huntsman Cancer Institute, University of Utah, Salt Lake City, Utah, USA. ¹¹Department of Clinical Genetics, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark. ¹²Greenwood Genetic Center, Greenwood, South Carolina, USA. ¹³Colorectal Oncogenomics Group, Department of Clinical Pathology, University of Melbourne, Parkville, Australia. ¹⁴Polyposis Registry, St Mark’s Hospital, London, UK. ¹⁵Institute of Medical Genetics, University Hospital of Wales, Cardiff, UK. ¹⁶Vermont Cancer Center, University of Vermont College of Medicine, Burlington, Vermont, USA. ¹⁷Department of Colorectal Medicine and Genetics, Royal Melbourne Hospital, Parkville, Australia; Department of Medicine, University of Melbourne, Parkville, Australia

Background and aim As one of the first projects to spearhead the development of the Clinical Genome Resource (ClinGen) – InSiGHT Hereditary Colon Cancer/ Polyposis Variant Curation Expert Panel (VCEP), the APC subcommittee established gene-specific variant classification criteria based on the existing ACMG/AMP framework. The criteria has been reviewed by the Sequence Variant Interpretation Working Group and is currently undergoing further refinement through pilot testing of a balanced selection of APC variants. To lay the groundwork for prospective expert panel approval for the substantial number of variants of uncertain significant (VUS) and conflicting germline APC variants in ClinVar, this study adopted a large-scale classification approach using the preliminary APC-specific classification guidelines.

Method Building off the ClinGen VCEP approval process, a streamlined algorithm using the current VCEP gene-specific classification codes was developed consisting of population (BA1, BS1, PM2_supporting), variant type (PVS1, BP1, BP7), and splice prediction (BP4, PP3) codes and applied to comprehensively assess all APC variants in ClinVar to arrive at an initial pathogenicity class. The results were compared to the ClinVar classifications to assess the efficacy of the preliminary classification criteria.

Results There were 9121 APC variants on ClinVar, which consisted of 1788 (Likely) Benign (20%), 1259 (Likely) Pathogenic (14%), and 5798 VUS (64%) and 276 conflicting variants (3%). Using the VCEP defined cut-offs, 7045 variants were classified as (Likely) Benign (77%) and 1171 variants were classified as (Likely) Pathogenic (13%), which included 5378 and 34 VUS, respectively. A prioritised list of promising causative APC variants that remained at VUS by the algorithm alone were identified from the remaining 1485 variants (16%), which will be subjected to a data mining-driven work-up to collect clinical and experimental evidence.

Conclusions The application of the preliminary APC-specific classification criteria substantially reduced the number of VUS on ClinVar. This study highlights the potential importance to clinical classification of a systematic approach built into the current APC-specific guidelines. Once the APC criteria are approved, a list of prioritised APC variants will be classified by the VCEP regularly and made available to the public through ClinVar, with detailed evidence housed in the ClinGen Evidence Repository.

Keywords adenomatous polyposis coli (APC), familial adenomatous polyposis (FAP), variants of uncertain significance (VUS), Clinical Genome Resource (ClinGen), ClinVar, ACMG/AMP guidelines.

OP-024: Genetics: Research, Germline, Genomics, Variomics

Best Oral Award:

Clinically relevant combined effect of polygenic background, rare pathogenic germline variants, and family history on colorectal cancer incidence

Stefan Aretz ¹, Emadeldin Hassanin², Hannah Klinkhammer³, Friederike David⁴, Dheeraj Bobbili⁵, Rana Aldisi², Nuria Dueñas⁶, Claudia Perne¹, Markus M. Nöthen⁴, Robert Hüneburg⁷, Peter Krawitz², Joan Brunet⁸, Gabriel Capella⁹, Patrick May⁵, Andreas Forstner¹⁰, Andreas Mayr¹¹, Isabel Spier¹, Carlo Maj¹²

¹Institute of Human Genetics, Medical Faculty, University of Bonn, Germany; National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Germany; European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS). ²Institute for Genomic Statistics and Bioinformatics, University Hospital Bonn, Germany. ³Institute for Genomic Statistics and Bioinformatics, University Hospital Bonn, Germany; Institute for Medical Biometry, Informatics and Epidemiology, Medical Faculty, University Bonn, Germany. ⁴Institute of Human Genetics, Medical Faculty, University of Bonn, Germany. ⁵Luxembourg Centre for Systems Biomedicine, University of Luxembourg, Esch-sur-Alzette, Luxembourg. ⁶Catalan Institute of Oncology-IDIBELL, ONCOBELL, Hospitalet de Llobregat, Barcelona, Spain; Biomedical Research Centre Network for Oncology (CIBERONC), Instituto Salud Carlos III, Madrid, Spain. ⁷Department of Internal Medicine I, University Hospital Bonn, Germany; National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Germany; European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS). ⁸Catalan Institute of Oncology-IDIBELL, ONCOBELL, Hospitalet de Llobregat, Barcelona, Spain; Biomedical Research Centre Network for Oncology (CIBERONC), Instituto Salud Carlos III, Madrid, Spain; Hereditary Cancer Program, Catalan Institute of Oncology-IDBIGI, Girona, Spain; European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS). ⁹Catalan Institute of Oncology-IDIBELL, ONCOBELL, Hospitalet de Llobregat, Barcelona, Spain; Biomedical Research Centre Network for Oncology (CIBERONC), Instituto Salud Carlos III, Madrid, Spain; European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS). ¹⁰Institute of Human Genetics, Medical Faculty, University of Bonn, Germany; Centre for Human Genetics, University of Marburg, Germany; Institute of Neuroscience and Medicine, Research Center Jülich, Germany. ¹¹Institute for Medical Biometry, Informatics and Epidemiology, Medical Faculty, University Bonn, Germany. ¹²Institute for Genomic Statistics and Bioinformatics, University Hospital Bonn, Germany; Centre for Human Genetics, University of Marburg, Germany

Background and aim Summarized in polygenic risk scores (PRS), the combined effect of common genetic variants for colorectal cancer (CRC) can be used to stratify individuals for CRC risk both in the general population and among individuals predisposed for hereditary CRC. The study aims to investigate to which extent PRS, high-impact monogenic variants for Lynch syndrome and polyposis, and family history (FH) affect CRC risk by assessing cancer prevalence and cancer cumulative lifetime incidence using European population-based data.

Method 163,516 individuals from the UK Biobank were stratified according to: 1. carrier status for germline pathogenic variants (PV) in CRC susceptibility genes (APC, MLH1, MSH2, MSH6, PMS2); 2. low (< 20%), intermediate (20–80%), or high (> 80%) PRS; and 3. family history (FH) of CRC. Multivariable logistic regression and Cox proportional hazards models were applied to compare odds ratios (OR) and to compute lifetime incidences, respectively.

Results Taking non-carriers with intermediate PRS as reference, we show that PV carriers with high PRS had four times higher OR than carriers with low PRS (OR = 17.5 and 3.9). CRC cumulative lifetime incidence by age 75 years for carriers of PV with low PRS is 40% and reaches 74% for carriers with high PRS, compared to 6% and 22% for non-carriers, respectively. A positive FH is associated with a further increase of the cumulative incidence reaching 98% for carriers and 26% for non-carriers. In non-carriers with a negative FH, but a high PRS, the CRC risk was doubled, whereas a low PRS even in the context of a FH resulted in a decreased risk. The full model including PRS, carrier status, and FH improved the area under the curve (AUC) for risk prediction (0.704). The PRS and FH modifies the relative risk across all five genes, however, the effect of PRS and FH is conversely related to the penetrance of the gene with the smallest effects in MLH1 PV carriers.

Conclusions Irrespective of a sporadic or monogenic background, the CRC risk is strongly influenced by the PRS. A high PRS and positive FH in non-carriers confers a CRC risk similar to a low PRS and negative FH in carriers of PV in moderate penetrance genes. FH, monogenic variants, and PRS contribute to CRC risk. The implementation of a PRS in routine patient care will likely improve individualised risk stratification for sporadic and monogenic CRC, which will guide tailored preventive strategies in high, moderate, and low risk groups.

Keywords Polygenic Risk – Family History – Risk Stratification.

OP-025: Genetics: Research, Germline, Genomics, Variomics

Assessment of the ability of the polygenic background to refine colorectal cancer risk in Lynch syndrome

Nuria Dueñas¹, Hannah Klinkhammer², Nuria Bonifaci³, Isabel Spier⁴, Andreas Mayr⁵, Emadeldin Hassanin⁶, Anna Díez Villanueva⁷, Víctor Moreno⁷, Marta Pineda¹, Carlo Maj⁶, Gabriel Capella⁸, Stefan Aretz⁴, Joan Brunet ⁹

¹Hereditary Cancer Program, Catalan Institute of Oncology-IDIBELL, ONCOBELL, Hospitalet de Llobregat, Barcelona, Spain; Biomedical Research Centre Network for Oncology (CIBERONC), Instituto Salud Carlos III, Madrid, Spain. ²Institute for Genomic Statistics and Bioinformatics, University Hospital Bonn, Germany; Institute for Medical Biometry, Informatics and Epidemiology, Medical Faculty, University Bonn, Germany. ³Hereditary Cancer Program, Catalan Institute of Oncology-IDIBELL, ONCOBELL, Hospitalet de Llobregat, Barcelona, Spain. ⁴Institute of Human Genetics, Medical Faculty, University of Bonn, Germany; National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Germany; European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS). ⁵Institute for Medical Biometry, Informatics and Epidemiology, Medical Faculty, University Bonn, Germany. ⁶Institute for Genomic Statistics and Bioinformatics, University Hospital Bonn, Germany. ⁷Unit of Biomarkers and Susceptibility, Oncology Data Analytics Program (ODAP), Catalan Institute of Oncology IDIBELL, Hospitalet de Llobregat, Barcelona, Spain; Consortium for Biomedical Research in Epidemiology and Public Health (CIBERESP), Instituto Salud Carlos III, Madrid, Spain. ⁸Hereditary Cancer Program, Catalan Institute of Oncology-IDIBELL, ONCOBELL, Hospitalet de Llobregat, Barcelona, Spain; Biomedical Research Centre Network for Oncology (CIBERONC), Instituto Salud Carlos III, Madrid, Spain; European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS). ⁹Hereditary Cancer Program, Catalan Institute of Oncology-IDIBELL, ONCOBELL, Hospitalet de Llobregat, Barcelona, Spain; Biomedical Research Centre Network for Oncology (CIBERONC), Instituto Salud Carlos III, Madrid, Spain; European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS); Hereditary Cancer Program, Catalan Institute of Oncology-IDBIGI, Girona, Spain

Background and aim Lynch syndrome (LS) is the main cause of hereditary colorectal cancer (CRC). LS-CRC incidences do not only show significant variability according to gene, sex and continent but also within the same family. The combined effect of single nucleotide polymorphisms (SNPs) as polygenic risk score (PRS) can stratify CRC risk in the general population, whereas its role in carriers of high-risk variants such as LS is still unclear. Our objective is to assess the ability of PRS to refine CRC risk prediction in European-descendant LS individuals.

Method A total of 1,465 European-descendant LS individuals (557 MLH1, 517 MSH2, 299 MSH6 and 92 PMS2) and 5,656 population controls from two independent cohorts were included. Two events were considered for analysis in the LS cohort: CRC (712 cases and 753 CRC-free LS individuals; mean age of 56.4 and 47.1, respectively) and CRC or high-grade adenoma (hgAd) (744 cases and 721 CRC or hgAd-free LS individuals, mean age of 56.2 and 46.9, respectively). A 91-SNP weighted PRS was applied. A Cox proportional hazard regression model with “family” as a random variable (Frailty model) was performed including polypectomy as time-dependant variable. A meta-analysis combining both cohorts was conducted. All p-values resulting from subgroup analyses were corrected for multiple testing by using False Discovery Rate (FDR) correction.

Results Higher PRS tended to be associated with slightly higher CRC risk in the whole LS cohort, in MSH6 mutation carriers, and in LS individuals diagnosed with early-onset CRC (< 50 years) when considering CRC as event (HR: 1.016 (1.003–1.030); p_uncorrected = 0.019, p_FDR_corrected = 0.298; HR: 1.052 (1.012–1.092); p_uncorrected = 0.010, p_FDR_corrected = 0.204; HR: 1.022 (1.007–1.038); p_uncorrected = 0.004, p_FDR_corrected = 0.204, respectively), as well as when considering CRC or hgAd as event (HR: 1.019 (1.005–1.032); p_uncorrected = 0.005, p_FDR_corrected = 0.202; HR: 1.026 (1.004–1.049); p_uncorrected = 0.022, p_FDR_corrected = 0.202; HR: 1.016 (1.002–1.030); p_uncorrected = 0.027, p_FDR_corrected = 0.202, respectively). In addition, PRS showed slightly higher CRC risk in MSH2 mutation carriers when considering CRC or hgAd as event (HR: 1.016 (1.001–1.031); p_uncorrected = 0.037, p_FDR_corrected = 0.202).

Conclusions We found indications that PRS might be able to refine CRC risk in European-descendant LS population. However, a higher sample size is needed to elucidate its usefulness to improve individualised risk assessment in LS.

Keywords Lynch syndrome, colorectal neoplasms, polygenic inheritance, penetrance, risk factor, multifactorial inheritance.

OP-026: Clinical: Diagnosis, Endoscopy, Management, Implementation

Risk of malignant and benign thyroid disease in patients with familial adenomatous polyposis – a prospective surveillance study

Robert Hüneburg ¹, Barbara Kreppel², Annika Menke¹, Tim Marwitz¹, Christian P. Strassburg¹, Markus Essler², Jacob Nattermann¹

¹National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Bonn, Germany; Department of Internal Medicine I, University Hospital Bonn, Bonn, Germany. ²Department of Nuclear Medicine, University Hospital Bonn, Bonn, Germany

Background and aim Patients with familial adenomatous polyposis (FAP) are at high risk to develop colorectal, ampullary or duodenal cancer. FAP is furthermore associated with a substantial lifetime risk for thyroid cancer (TC) and benign thyroid disease. Screening and surveillance guidelines differ substantially between countries. The aim of our study was to assess the prevalence of TC and benign thyroid disease and possible surveillance intervals in a large European center.

Method We identified consecutive patients with FAP at a single center from 2010 till September 2020. Inclusion criteria were a known pathogenic variant in the APC gene and age > 18 years. All patients were enrolled in our prospective hereditary cancer registry and signed an informed consent. Thyroid surveillance consisted of ultrasound and endocrinologic testing for thyroid disease. We collected demographics, ultrasound, laboratory and histopathological results in case of biopsy or surgery.

Results A total number of 230 patients with FAP were part of our prospective registry. We had to exclude 15 patients (< 18 y), 28 patients refused to undergo thyroid examination.

In 187 patients (100 female; mean age 39 (± 15 y); 187 (100%) pathogenic variant in the APC gene) at least one thyroid examination was performed at our center. Endocrinologic testing for thyroid disease revealed abnormal results in 92/187 (49%) with hypothyreodism diagnosed in 32/92 (35%). In 85/186 patients (46%) thyroid nodules were detected with a significant female predominance (55/99 vs 31/87; p = 0.01). In 16/187 patients (mean age 41 (± 13y) thyroid surgery was performed revealing thyroid cancers in seven patients, adenoma in one patient and C cell hyperplasia in additional one (a). No patient reported any symptoms prior surgery.

Autoimmune thyreoiditis (AIT) was diagnosed in 28/187 (15%) patients with a female predominance (21/99 vs. 7/87; p = 0.012), including one patient with thyroid cancer.

In 101 patients, a surveillance examination was done with an interval between 1–3 years. In patients with a normal index examination, no cancer or large thyroid nodule (> 5 mm) was observed.

Conclusions Patients with FAP are at increased risk for developing benign and malignant diseases of the thyroid gland irrespective to gender. Our data indicate that if the index examination is inconspicuous, the examination interval may be extended.

Keywords FAP, thyroid, thyroid cancer.

Histopathological results of surgical thyroid specimen

Patient	Histopathological subtype	tumor stage	sex	age	autoimmune thyreoditis	previous colectomy
1	multifocal papillary	pT2 mN1a (1/10) RX	w	31	n	n
2	Cribiform-morular papillary	pT1 N0 M0 L0 V0 R0	m	23	n	y
3	papillary	pT1a (m), pNx, R0, L0, V0	w	19	n	n
4	papillary	pT1b L0 V0 R0	w	48	y	y
5	papillary	pT1a R0 V0 L0	m	49	n	y
6	papillary	pT3a(m) pN0 L0 V0 R0	w	21	n	n
7	multifocal papillary	pT1b, pNX, L0, V0, Pn0, R0	w	18	n	y
8	adenoma		m	44	n	y

OP-027: Clinical: Diagnosis, Endoscopy, Management, Implementation

Outcomes following duodenectomy in patients with familial adenomatous polyposis

Isabel Martin ¹, Arthur S. Aelvoet³, Cherryl Cabalit¹, James Cockburn¹, Victoria Cuthill¹, Duncan Spalding⁴, Oliver R. Busch³, Barbara A. J. Bastiaansen³, Susan K. Clark², Evelien Dekker³, Andrew Latchford²

¹St Mark's Hospital Polyposis Registry, London, UK. ²St Mark's Hospital Polyposis Registry & Imperial College London, UK. ³Amsterdam UMC, University of Amsterdam, Cancer Center Amsterdam, Netherlands. ⁴Department of Surgery & Cancer, Imperial College London, UK

Background and aim Duodenectomy may be required for advanced duodenal/ampullary disease in familial adenomatous polyposis (FAP). Follow-up upper gastrointestinal endoscopic surveillance is recommended, but surveillance long-term outcome data are lacking. We reviewed outcomes in patients after duodenectomy.

Method Patients were identified from prospectively maintained databases (1995–2022) at two centers. Medical records were reviewed, and details of surgery, survival and endoscopic surveillance were collated.

Results One hundred and nineteen (55 female) patients were identified, with median follow-up of 114 months (range 1–341).

In total, 97 (82%) patients underwent prophylactic surgery for advanced benign disease at a median age of 50 years (range 31–71).

Twenty-one (18%) patients underwent duodenectomy for cancer (either known (n = 18) or diagnosed on post-operative histology (n = 3)), at a median age of 54 years (range 35–77). Of these 6/21 (29%) had duodenal cancer, median age 49 years (range 40–64). Twelve (57%) patients had ampullary carcinoma, median age of 63 years (range 41–66). Three (14%) patients had gastric cancer with synchronous extensive duodenal polyposis and underwent duodenectomy and total gastrectomy.

Thirty-two (27%) patients died following duodenectomy (Table 1).

Details of endoscopic surveillance were available for 84/119 (71%) patients. Twenty-six patients (33%) developed 65 gastric adenomas (GA). The median time from surgery to initial GA-diagnosis was 97 months (range 22–291). Three patients had GAs with high grade dysplasia.

During follow-up, six (8%) patients developed gastric cancer, at a median of 176 months (range 66–276) after surgery and died.

Thirty-five (44%) patients developed a median of 10 (range 1–80) jejunal adenomas in the reconstructed limbs. Jejunal adenomas developed in 25/34 patients after pancreas preserving total duodenectomy and 11/50 after pancreaticoduodenectomy (p =  < 0.001). Five patients underwent surgical resection for extensive jejunal polyposis (n = 4) or jejunal cancer (n = 1; diagnosed 14 years after duodenectomy).

Conclusions Patients diagnosed with gastric adenoma (33%), gastric cancer (8%), jejunal adenomas (44%) and jejunal cancer (1%) following duodenectomy highlights the need for endoscopic surveillance in this patient group.

Jejunal adenomas develop more commonly after PSD.

Survival following cancer diagnosis is poor, prophylactic surgery is recommended, although determining timing of surgery remains a challenge.

Keywords Familial adenomatous polyposis, FAP, Duodenectomy.

OP-028: Clinical: Diagnosis, Endoscopy, Management, Implementation

Development of ileal adenomas after ileal pouch-anal anastomosis versus end ileostomy in patients with familial adenomatous polyposis

Arthur S. Aelvoet ¹, Victorine H. Roos¹, Barbara. A. Bastiaansen¹, Roel Hompes², Willem A. Bemelman², Cora M. Aalfs⁴, Patrick M. Bossuyt³, Evelien Dekker¹

¹Amsterdam UMC location University of Amsterdam, Department of Gastroenterology and Hepatology, Amsterdam Gastroenterology Endocrinology Gastroenterology, Cancer Center Amsterdam, Amsterdam, the Netherlands. ²Amsterdam UMC location University of Amsterdam, Department of Surgery, Cancer Center Amsterdam, Amsterdam, the NetherlandsAmsterdam, the Netherlands. ³Amsterdam UMC location University of Amsterdam, Department of Epidemiology and Data Science, Amsterdam, the Netherlands. ⁴UMC Utrecht, Utrecht University, Department of Clinical Genetics, Utrecht, the Netherlands

Background and aim Patients with familial adenomatous polyposis (FAP) undergo (procto)colectomy to prevent colorectal cancer. Over time, most patients with ileal pouch-anal anastomosis (IPAA) develop adenomas in the ileum, where no adenomas were detected before surgery. Whether this is also the case in patients with end ileostomy, is not well described. We aimed to compare ileal adenoma development in patients with IPAA versus end ileostomy.

Method This historical cohort study included FAP patients with IPAA or end ileostomy who underwent surveillance endoscopies between 2001 and 2021. Primary outcomes were the proportion of patients with ileal adenomas, location of adenomas, and proportion of patients undergoing excision of the pouch/end ileostomy.

Results Overall, 144 patients with IPAA (n = 111) and end ileostomy (n = 33) with a median endoscopic follow-up of 152 and 174 months, respectively, were included. Although age at primary colectomy did not differ between the groups, the median age at construction of the end ileostomy was higher than the median age at IPAA construction (44 versus 27, p < 0.01).

Five years after surgery, 15% of patients with IPAA had developed ileal adenomas versus 4% after end ileostomy. At 10 years, these estimates were 48% versus 9%; at 20 years 85% versus 43% (log-rank p < 0.01). Moreover, patients with an IPAA developed significantly more ileal adenomas than patients with end ileostomy (Table 1). Within the IPAA group, adenomas more often developed in the pouch body (95%) than in pre-pouch ileum (5%). Within the end ileostomy group, adenomas more often developed at the everted site of the ileostomy (77%) than in terminal ileum (23%).

Within the IPAA group, two patients (2%) developed high-grade dysplasia and three patients (3%) adenocarcinoma in the ileal pouch. None of the patients with end ileostomy developed high-grade dysplasia or cancer.

Nine patients (8%) underwent pouch excision due to extensive polyposis (n = 6) or cancer (n = 3) compared to three patients (9%) who underwent end ileostomy excision due to extensive polyposis. Taking into account potential confounders in multivariable cox regression analysis, having an IPAA was still associated with ileal adenoma development.

Conclusions FAP patients with IPAA more often developed ileal adenomas than patients with end ileostomy which potentially could affect long-term management. Patients with end ileostomy might benefit from less frequent endoscopic surveillance than patients with IPAA (Fig. 1).

Keywords Familial adenomatous polyposis, endoscopic surveillance, ileal adenoma, ileal pouch-anal anastomosis, end ileostomy.

Table 1. Cumulative number of ileal adenomas per time-period following IPAA or end ileostomy

OP-029: Clinical: Diagnosis, Endoscopy, Management, Implementation

Defining the risk of desmoid disease in a high risk familial adenomatous polyposis (FAP) cohort

Massih Bahar, Susan Clark, Andrew Latchford.

Polyposis Registry, St Mark's Hospital, UK

Background and aim Desmoid disease is reported to occur in 10–20% of patients with FAP. These tumours are usually precipitated by large bowel surgery. It has previously been established that patients with FAP and a constitutional 3’ APC constitutional pathogenic variant (PV) are at high risk of desmoid disease.

There are few data to quantity the absolute risk of desmoid formation in this group, which is essential for pre-operative counselling, surgical decision making and development of a desmoid prevention strategy. We aimed to address this question.

Method We identified all patients with an APC PV 3’ of codon 1399 on a prospectively maintained registry. Registry and medical records were reviewed, and data collated. This work received local research departmental approval.

Results The cohort comprised 148 patients from 78 kindreds with a 3’ PV. In total 87/148 (59%) developed desmoid.

Desmoid was diagnosed either before surgical intervention or at the time of surgery in in 40/148 (27%).

No surgery had been performed in 50/148 (34%), who still had an intact colon. Of the 98 who had undergone surgery, 51/98 (52%) underwent total colectomy and 41/98 (42%) proctocolectomy; segmental resection or surgery details unavailable accounted for six patients.

Excluding those in whom a desmoid had been identified before/at time of surgery and the patient where large bowel surgical details were not available, post-operative desmoid developed in 46/70 (66%). A desmoid developed in 24/38 (63%) of those undergoing total colectomy and 22/32 (69%) of those undergoing proctocolectomy (p = 0.6).

Conclusions In total 60% of patients developed a desmoid.

The risk of desmoid is high even in the surgical naïve. Cross sectional imaging should be performed routinely as part of surgical planning.

Post-operative desmoid developed in 66% patients, with no significant difference in risk between total colectomy and proctocolectomy, but comparison was limited by small sample size.

The magnitude of risk we describe is such that a formal randomised post-operative chemoprevention study should be considered.

Keywords FAP, 3' variants, desmoid.

OP-030: Clinical: Diagnosis, Endoscopy, Management, Implementation

Intestinal Transplant for Desmoid Disease in FAP: A Single Institution Experience

Nicholas Smith ¹, Anil Vaidya³, Joshua Sommovilla⁴, Carole Macaron², Lisa Laguardia⁴, Margaret O'malley⁴, Susan Milicia⁴, Carol A. Burke², David Liska¹

¹Department of Colorectal Surgery, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, USA. ²Department of Gastroenterology, Hepatology, Nutrition, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, USA. ³Center for Gut Rehabilitation and Transplantation, Transplant Center, Cleveland Clinic, Cleveland, USA. ⁴Sanford R. Weiss, MD Center of Hereditary Colorectal Neoplasia, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, USA

Background and aim Up to 1/3 of patients with Familial Adenomatous Polyposis (FAP) will develop desmoid disease, and half will be intra-abdominal. Desmoid disease is associated with significant morbidity and is a major cause of death in FAP patients. Some of these patients can develop complete intestinal failure as a consequence of surgical resection, or alternatively be deemed unresectable, due to potential loss of the entire small bowel. In these patients, small intestinal and multi-visceral transplantation offer a potential treatment option. This study investigates the short- and long-term outcomes for patients with FAP undergoing small bowel transplant for desmoid disease at a major tertiary center.

Method This is a retrospective case series of a single institution’s experience of intestinal transplant for desmoid disease in FAP patients. Perioperative morbidity and mortality were recorded, as well as long-term outcomes including graft survival, TPN-dependence, desmoid recurrence, and overall survival.

Results A total of 9 patients underwent either multi-visceral transplantation (MVT; n = 5) or isolated small bowel transplantation (ISBT; n = 4), between 2014–2021. Six patients were male; 3 were female and median age was 46 years (range 34–63). Preoperatively all patients were TPN-dependent. Seven patients (78%) underwent desmoid resection at time of transplant while 5 patients (56%) had undergone total enterectomy > 12 months prior to transplant. Mean follow-up was 11.6 months (range 1.5–33.5). No patients were lost to follow-up. Mean operative time was 12.7 h (13.5–15.0), EBL 1.5L (0.15–5.0) and hospital length of stay 72 days (24–220). Seven patients (78%) incurred postoperative complications including 2 mortalities occurring > 30 days after surgery, related to opportunistic infection and immunocompromise (Clavien Dindo: 0 I; II 2; III 4; V 2). Immunosuppression-related complications occurred in 5 patients (56%), 2 of which occurred following the perioperative period. Perioperative morbidity was greater in MVT vs. ISB transplants. At follow-up, 86% of surviving patients (n = 6) were TPN-independent and without desmoid recurrence.

Conclusions Small bowel intestinal transplant is a potential salvage option for patients with FAP and desmoid disease associated with complete intestinal failure. Isolated small bowel transplant, when feasible, is associated with less morbidity compared to multi-visceral transplantation.

Keywords FAP, Desmoids, Intestinal Transplant, Intestinal Failure, Surgery.

Table 1. Outcomes of Intestinal Transplant for Desmoid Resection

OP-031: Genetics: Research, Germline, Genomics, Variomics

Characterizing POLE and POLD1 germline and somatic aetiology in colorectal cancers demonstrating the defective DNA polymerase proofreading-related SBS10 tumour mutational signatures

Khalid Mahmood ^1,2,4, Tabitha A. Harrison³, Peter Georgeson^1,2, Bernard J. Pope^1,2,4, Syed H. Zaidi⁵, Conghui Qu³, Robert S. Steinfelder³, Yi Lin³, Jihoon E. Joo^1,2, Mark Clendenning^1,2, Romy Walker^1,2, Efrat L. Amitay⁶, Sonja I. Berndt⁷, Hermann Brenner^6,8,9, Peter T. Campbell¹⁰, Yin Cao^11,12,13, Andrew T. Chan^{14,15,16,17,18,19}, Jenny Chang-Claude^20,21, Kimberly F. Doheny²², David A. Drew^14,15, Jane C. Figueiredo^23,24, Amy J. French²⁵, Steven Gallinger^26,5, Marios Giannakis^17,27,28, Graham G. Giles^29,30,31, Andrea Gsur³², Marc J. Gunter³³, Michael Hoffmeister⁶, Li Hsu^3,34, Wen-Yi Huang⁷, Paul Limburg³⁵, JoAnn E. Manson^18,28, Victor Moreno^36,37,38,39, Rami Nassir⁴⁰, Jonathan A. Nowak⁴¹, Mireia Obón-Santacana^36,37, Shuji Ogino^17,18,41,42, Amanda I. Phipps^3,43, John D. Potter^3,44, Robert E. Schoen⁴⁵, Wei Sun³, Amanda E. Toland⁴⁶, Quang M. Trinh⁵, Tomotaka Ugai^18,41, Finlay A. Macrae^47,48,49, Christophe Rosty^1,2,50,51, Thomas J. Hudson⁵, Mark A. Jenkins^2,30, Stephen N. Thibodeau²⁵, Ingrid M. Winship^49,52, Ulrike Peters^3,43, Daniel D. Buchanan^1,2,49

¹Colorectal Oncogenomics Group, Department of Clinical Pathology, The University of Melbourne, Parkville, Victoria 3010 Australia; ²University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Parkville, Victoria 3010 Australia; ³Public Health Sciences Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, USA; ⁴Melbourne Bioinformatics, The University of Melbourne, Carlton, Australia; ⁵Ontario Institute for Cancer Research, Toronto, Ontario, Canada; ⁶Division of Clinical Epidemiology and Aging Research, German Cancer Research Center (DKFZ), Heidelberg, Germany; ⁷Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA; ⁸Division of Preventive Oncology, German Cancer Research Center (DKFZ) and National Center for Tumor Diseases (NCT), Heidelberg, Germany; ⁹German Cancer Consortium (DKTK), German Cancer Research Center(DKFZ), Heidelberg, Germany; ¹⁰Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx, NY, USA; ¹¹Division of Public Health Sciences, Department of Surgery, Washington University School of Medicine, St Louis, Missouri, USA; ¹²Alvin J. Siteman Cancer Center at Barnes-Jewish Hospital and Washington University School of Medicine, St. Louis, Missouri, USA; ¹³Division of Gastroenterology, Department of Medicine, Washington University School of Medicine, St. Louis, Missouri, USA; ¹⁴Clinical and Translational Epidemiology Unit, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA; ¹⁵Division of Gastroenterology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA; ¹⁶Channing Division of Network Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts, USA; ¹⁷Broad Institute of MIT and Harvard, Cambridge, Massachusetts, USA; ¹⁸Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA; ¹⁹Department of Immunology and Infectious Diseases, Harvard T.H. Chan School of Public Health, Harvard University, Boston, Massachusetts, USA; ²⁰Division of Cancer Epidemiology, German Cancer Research Center (DKFZ), Heidelberg, Germany; ²¹University Medical Centre Hamburg-Eppendorf, University Cancer Centre Hamburg (UCCH), Hamburg, Germany; ²²Center for Inherited Disease Research (CIDR), Department of Genetic Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA; ²³Department of Medicine, Samuel Oschin Comprehensive Cancer Institute, Cedars-Sinai Medical Center, Los Angeles, CA, USA; ²⁴Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, California, USA; ²⁵Division of Laboratory Genetics, Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, Minnesota, USA; ²⁶Lunenfeld Tanenbaum Research Institute, Mount Sinai Hospital, University of Toronto, Toronto, Ontario, Canada; ²⁷Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts, USA; ²⁸Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA; ²⁹Cancer Epidemiology Division, Cancer Council Victoria, Melbourne, Victoria, Australia; ³⁰Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Victoria, Australia; ³¹Precision Medicine, School of Clinical Sciences at Monash Health, Monash University, Clayton, Victoria, Australia; ³²Institute of Cancer Research, Department of Medicine I, Medical University Vienna, Vienna, Austria; ³³Nutrition and Metabolism Branch, International Agency for Research on Cancer, World Health Organization, Lyon, France; ³⁴Department of Biostatistics, University of Washington, Seattle, Washington, USA; ³⁵Division of Gastroenterology & Hepatology, Mayo Clinic, Rochester, Minnesota, USA; ³⁶Oncology Data Analytics Program, Catalan Institute of Oncology-IDIBELL, L'Hospitalet de Llobregat, Barcelona, Spain; ³⁷CIBER Epidemiología y Salud Pública (CIBERESP), Madrid, Spain; ³⁸Department of Clinical Sciences, Faculty of Medicine, University of Barcelona, Barcelona, Spain; ³⁹ONCOBEL Program, Bellvitge Biomedical Research Institute (IDIBELL), L'Hospitalet de Llobregat, Barcelona, Spain; ⁴⁰Department of Pathology, College of Medicine, Umm Al-Qura University, Saudi Arabia; ⁴¹Program in MPE Molecular Pathological Epidemiology, Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA; ⁴²Cancer Immunology Program, Dana-Farber Harvard Cancer Center, Boston, Massachusetts, USA; ⁴³Department of Epidemiology, University of Washington, Seattle, Washington, USA; ⁴⁴Research Centre for Hauora and Health, Massey University, Wellington, New Zealand; ⁴⁵Department of Medicine and Epidemiology, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, USA; ⁴⁶Departments of Cancer Biology and Genetics and Internal Medicine, Comprehensive Cancer Center, The Ohio State University, Columbus, Ohio, USA ⁴⁷Parkville Familial Cancer Centre, and Dept of Colorectal Medicine and Genetics The Royal Melbourne Hospital; ⁴⁸Colorectal Medicine and Genetics, Royal Melbourne Hospital, Parkville, Australia; ⁴⁹Genomic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Parkville, Australia; ⁵⁰Envoi Specialist Pathologists, Brisbane, Australia; ⁵¹University of Queensland, Brisbane, Australia; ⁵²Department of Medicine, The University of Melbourne, Parkville, Australia

Background and aim Somatic exonuclease domain mutations (EDm) in the POLE and POLD1 genes are associated with increased tumor mutation burden (TMB) and a mutational signature related to single base substitutions SBS10a/b/c/d. Currently, germline ED variants in POLE and POLD1 are difficult to classify and often remain as variants of uncertain clinical significance (VUS). To date, the utility of the SBS10 signatures in assessing ED variants in POLE or POLD1 has not been determined. The aim of this study was to characterise the germline and somatic aetiology of colorectal cancers (CRCs) demonstrating the SBS10 mutational signatures to better understand the POLE and POLD1 genotype–phenotype associations.

Method Targeted multigene sequencing of tumor and matched germline DNA from 5547 CRC-affected participants of CCFR and GECCO was used to identify rare germline (gnomAD MAF < 0.005%) and somatic EDms of POLE and POLD1, estimate TMB (mutations/Mb) and calculate a set of 18 CRC-related SBS COSMIC v3.2 signatures. A CRC was considered SBS10 positive when the proportion of SBS10a/b/c/d combined was > 10%.

Results A total of 2110/5547 CRCs (38%) were positive for SBS10 signature of which 45 POLE and 25 POLD1 CRCs had an ED variant. For POLE, 56% had somatic hotspot EDms, 20% had a non-hotspot somatic EDm, and 24% had a germline ED variant. The SBS10 proportions between these three POLE aetiologies were different (p < 0.0001, 0.67 ± 0.1, 0.61 ± 0.2, 0.27 ± 0.2) as was the mean TMB (p < 0.0001, 176 ± 106, 131 ± 115 & 6.5 ± 2.6). For POLD1, 20% had somatic hotspot EDms, 52% had a non-hotspot somatic EDm, and 28% had a germline ED variant. The SBS10 proportions between these three POLD1 aetiologies were not significantly different (0.37 ± 0.3, 0.32 ± 0.2, 0.27 ± 0.1) but the mean TMB is higher in the EDm group (p < 0.05, 112 ± 88, 147 ± 134, 5.4 ± 1.4). For the remaining 2040 SBS10 positive CRCs without an EDm, the SBS10 proportion was 0.26 ± 0.15. Furthermore, 9 POLE and 28 POLD1 CRCs with somatic EDms were SBS10 negative, however, among these 3 and 10 CRCs had TMB > 100 mutations/Mb, respectively. The germline POLE or POLD1 ED variants were all classified as VUS in ClinVar and none of the somatic EDms overlapped with the germline variants.

Conclusions Higher proportions of SBS10 observed in CRC is indicative of somatic EDm for POLE, but SBS10 may have a weaker association with somatic POLD1 EDm. Further investigation of the role of SBS10 and TMB in CRC for germline ED variant classification in POLE or POLD1 is needed.

Keywords POLE, POLD1, mutational signature, variant classification, germline, somatic mutation.

OP-032: Genetics: Research, Germline, Genomics, Variomics

Identifying colorectal cancer caused by biallelic MUTYH pathogenic variants using tumor mutational signatures

Peter Georgeson ¹, Tabitha A Harrison², Bernard J. Pope³, Syed H Zaidi⁴, Conghui Qu², Robert S. Steinfelder², Yi Lin², Jihoon E. Joo¹, Khalid Mahmood³, Mark Clendenning¹, Romy Walker¹, Efrat L. Amitay⁵, Sonja I. Berndt⁶, Hermann Brenner⁷, Peter T. Campbell⁸, Yin Cao⁹, Andrew T. Chan¹⁰, Jenny Chang-Claude¹¹, Kimberly F. Doheny¹², David A. Drew¹³, Jane C. Figueiredo¹⁴, Amy J. French¹⁵, Steven Gallinger¹⁶, Marios Giannakis¹⁷, Gragam G Giles¹⁸, Andrea Gsur¹⁹, Marc J Gunter²⁰, Michael Hoffmeister⁵, Li Hsu²¹, Wen Yi Huang⁶, Paul Limburg²², JoAnn E. Manson²³, Victor Moreno²⁴, Rami Nassir²⁵, Jonathan A. Nowak²⁶, Mireia Obón-Santacana²⁷, Shuji Ogino²⁸, Amanda I. Phipps²⁹, John D. Potter³⁰, Robert E. Schoen³¹, Wei Sun², Amanda E. Toland³², Quang M. Trinh⁴, Tomotaka Ugai³³, Finlay A. Macrae³⁴, Christophe Rosty³⁵, Thomas J. Hudson⁴, Mark A. Jenkins³⁶, Stephen N. Thibodeau¹⁵, Ingrid M. Winship³⁷, Ulrike Peters²⁹, Daniel D. Buchanan³⁸

¹Colorectal Oncogenomics Group, Department of Clinical Pathology, The University of Melbourne, Parkville, Victoria 3010 Australia; University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Parkville, Victoria 3010 Australia. ²Public Health Sciences Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, USA. ³Colorectal Oncogenomics Group, Department of Clinical Pathology, The University of Melbourne, Parkville, Victoria 3010 Australia; University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Parkville, Victoria 3010 Australia; Melbourne Bioinformatics, The University of Melbourne, Carlton, Australia. ⁴Ontario Institute for Cancer Research, Toronto, Ontario, Canada. ⁵Division of Clinical Epidemiology and Aging Research, German Cancer Research Center (DKFZ), Heidelberg, Germany. ⁶Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA. ⁷Division of Clinical Epidemiology and Aging Research, German Cancer Research Center (DKFZ), Heidelberg, Germany; Division of Preventive Oncology, German Cancer Research Center (DKFZ) and National Center for Tumor Diseases (NCT), Heidelberg, Germany; German Cancer Consortium (DKTK), German Cancer Research Center(DKFZ), Heidelberg, Germany. ⁸Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx, NY, USA. ⁹Division of Public Health Sciences, Department of Surgery, Washington University School of Medicine, St Louis, Missouri, USA; Alvin J. Siteman Cancer Center at Barnes-Jewish Hospital and Washington University School of Medicine, St. Louis, Missouri, USA; Division of Gastroenterology, Department of Medicine, Washington University School of Medicine, St. Louis, Missouri, USA. ¹⁰Clinical and Translational Epidemiology Unit, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA; Division of Gastroenterology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA; Channing Division of Network Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts, USA; Broad Institute of MIT and Harvard, Cambridge, Massachusetts, USA; Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA; Department of Immunology and Infectious Diseases, Harvard T.H. Chan School of Public Health, Harvard University, Boston, Massachusetts, USA. ¹¹Division of Cancer Epidemiology, German Cancer Research Center (DKFZ), Heidelberg, Germany; University Medical Centre Hamburg-Eppendorf, University Cancer Centre Hamburg (UCCH), Hamburg, Germany. ¹²Center for Inherited Disease Research (CIDR), Department of Genetic Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA. ¹³Clinical and Translational Epidemiology Unit, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA; Division of Gastroenterology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, USA. ¹⁴Department of Medicine, Samuel Oschin Comprehensive Cancer Institute, Cedars-Sinai Medical Center, Los Angeles, CA, USA; Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, California, USA. ¹⁵Division of Laboratory Genetics, Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, Minnesota, USA. ¹⁶Lunenfeld Tanenbaum Research Institute, Mount Sinai Hospital, University of Toronto, Toronto, Ontario, Canada; Ontario Institute for Cancer Research, Toronto, Ontario, Canada. ¹⁷Broad Institute of MIT and Harvard, Cambridge, Massachusetts, USA; Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts, USA; Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA. ¹⁸Cancer Epidemiology Division, Cancer Council Victoria, Melbourne, Victoria, Australia; Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Victoria, Australia; Precision Medicine, School of Clinical Sciences at Monash Health, Monash University, Clayton, Victoria, Australia. ¹⁹Institute of Cancer Research, Department of Medicine I, Medical University Vienna, Vienna, Austria. ²⁰Nutrition and Metabolism Branch, International Agency for Research on Cancer, World Health Organization, Lyon, France. ²¹Public Health Sciences Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, USA; Department of Biostatistics, University of Washington, Seattle, Washington, USA. ²²Division of Gastroenterology & Hepatology, Mayo Clinic, Rochester, Minnesota, USA. ²³Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA; Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA. ²⁴Oncology Data Analytics Program, Catalan Institute of Oncology-IDIBELL, L'Hospitalet de Llobregat, Barcelona, Spain; CIBER Epidemiología y Salud Pública (CIBERESP), Madrid, Spain; Department of Clinical Sciences, Faculty of Medicine, University of Barcelona, Barcelona, Spain; ONCOBEL Program, Bellvitge Biomedical Research Institute (IDIBELL), L'Hospitalet de Llobregat, Barcelona, Spain. ²⁵Department of Pathology, College of Medicine, Umm Al-Qura University, Saudi Arabia. ²⁶Program in MPE Molecular Pathological Epidemiology, Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA. ²⁷Oncology Data Analytics Program, Catalan Institute of Oncology-IDIBELL, L'Hospitalet de Llobregat, Barcelona, Spain; CIBER Epidemiología y Salud Pública (CIBERESP), Madrid, Spain. ²⁸Broad Institute of MIT and Harvard, Cambridge, Massachusetts, USA; Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA; Program in MPE Molecular Pathological Epidemiology, Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA; Cancer Immunology Program, Dana-Farber Harvard Cancer Center, Boston, Massachusetts, USA. ²⁹Public Health Sciences Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, USA; Department of Epidemiology, University of Washington, Seattle, Washington, USA. ³⁰Public Health Sciences Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, USA; Research Centre for Hauora and Health, Massey University, Wellington, New Zealand. ³¹Department of Medicine and Epidemiology, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, USA. ³²Departments of Cancer Biology and Genetics and Internal Medicine, Comprehensive Cancer Center, The Ohio State University, Columbus, Ohio, USA. ³³Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA; Program in MPE Molecular Pathological Epidemiology, Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA. ³⁴Parkville Familial Cancer Centre, and Dept of Colorectal Medicine and Genetics The Royal Melbourne Hospital; Colorectal Medicine and Genetics, Royal Melbourne Hospital, Parkville, Australia; Genomic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Parkville, Australia. ³⁵Colorectal Oncogenomics Group, Department of Clinical Pathology, The University of Melbourne, Parkville, Victoria 3010 Australia; University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Parkville, Victoria 3010 Australia; Envoi Specialist Pathologists, Brisbane, Australia; University of Queensland, Brisbane, Australia. ³⁶University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Parkville, Victoria 3010 Australia; Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Victoria, Australia. ³⁷Genomic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Parkville, Australia; Department of Medicine, The University of Melbourne, Parkville, Australia. ³⁸Colorectal Oncogenomics Group, Department of Clinical Pathology, The University of Melbourne, Parkville, Victoria 3010 Australia; University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Parkville, Victoria 3010 Australia; Genomic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Parkville, Australia

Background and aim Carriers of biallelic pathogenic variants in the base excision repair gene MUTYH have a high risk of colorectal cancer (CRC), but their identification is hampered by a heterogeneous phenotype – carriers can develop polyposis in the absence of CRC or CRC in the absence of synchronous polyposis. We evaluated the capacity of tumor mutational signatures (TMS), generated from tumors and polyps, to identify biallelic carriers and classify variants of uncertain significance (VUS) from panel-sequenced data.

Method A training set of 102 CRCs, including eight biallelic MUTYH carriers, were used to determine the optimal combined SBS18 and SBS36 TMS threshold for identifying CRCs from known biallelic carriers. The classifier was refined to include somatic mutation count and TMS reconstruction error on a validation set of 2,528 panel-sequenced CRCs, and then evaluated on a test set of 3,019 panel-sequenced CRCs. The feasibility of detecting biallelic MUTYH carriers from polyp sequencing data was assessed with 50 adenomas consisting of 4 adenomas from carriers and 46 adenomas from non-carriers.

Results Our classifier correctly identified all five known carriers in the CRC test set, with no false positives. All monoallelic carriers were classified as negatives. We found evidence for a pathogenic classification for two VUS (c.1141G > T p.G381W and c.577-5A > G) and a benign classification for five VUS (c.1276C > T p.R426C, c.912C > G p.S304R, c.821G > A p.R274Q, c.925C > T p.R309C and c.1431G > C p.T477T). Somatic hotspot mutations were found to be associated with CRCs from biallelic carriers: we confirmed the previously reported association with KRAS p.G12C (p = 2 × 10–23) and found a novel association with PIK3CA p.Q546K (p = 6 × 10–11). TMS calculated on adenoma-derived sequencing data suggested that this approach can identify biallelic MUTYH carriers from adenoma data.

Conclusions Our TMS algorithm identified all cases with germline biallelic pathogenic variants with no false positives and provided evidence for or against pathogenicity for seven germline VUS. As tumor sequencing for precision oncology becomes increasingly employed, this algorithm could be implemented cost-effectively, enabling identification of carriers from both tumor-derived and adenoma-derived sequencing data, prevention of CRC through targeted screening, classification of VUS, and the investigation of biallelic MUTYH-related hotspot mutations as potential biomarkers for targeted therapy.

Keywords colorectal cancer, tumor mutational signatures, genomic sequencing data, MUTYH-associated polyposis, VUS classification, colorectal adenoma.

OP-033: Genetics: Research, Germline, Genomics, Variomics

Expanding the phenotype of NTHL1 tumor syndrome: results from a commercial laboratory

Daniel E. Pineda Alvarez, Elaine Chen, Ester Borras, Scott T. Michalski, Brandie Leach

Invitae, San Francisco, CA, USA

Background and aim NTHL1-tumor syndrome (NTS) is a rare autosomal recessive condition that predisposes to colorectal polyposis and cancer and possibly extracolonic cancers such as breast, brain, and others. Given the relative recent discovery of this syndrome and its rarity, much remains to be understood about the risks associated with this syndrome. In this study we aim to describe the phenotypes of a series of patients with biallelic NTHL1 pathogenic variants (PVs) who were tested through a single commercial laboratory.

Method Patients identified to have biallelic NTHL1 PVs were queried. Demographics and personal and family history were collected from test requisition forms completed by healthcare providers. Descriptive statistics were utilized.

Results 42 patients were identified with NTS. As outlined in Table 1, the cohort was predominantly female (n = 28) and White (n = 24). Average age at testing was 51.2 years (range 31–76). The most common genetic finding (n = 28) was p.Gln90* in homozygous. Six patients also had PVs in another gene associated with hereditary cancer risk (RAD50, BRCA1, APC I1307K, monoallelic MUTYH, BARD1 and NF1). Phenotypic data were unavailable for 4 patients. Polyposis was reported in 11 patients (26.2%) and 52 cancers in 32 patients (76%). 14 cancer types were reported with breast (n = 15) and colorectal (n = 13) being the most common. Table 2 shows the self-reported cancer and polyp personal history by patient genotype. Any history of cancer in a first or second degree relative was reported for 19 individuals.

Conclusions This is one of the largest series of NTS patients. These data add to the growing knowledge of NTS and support findings from previous studies identifying multiple tumor types among patients with NTS.

Keywords NTHL1-tumor syndrome, polyposis.

Table 1. Patient demographics

Table 2. Phenotypes by genotype

OP-034: Clinical: Diagnosis, Endoscopy, Management, Implementation

Surveillance outcomes in Hereditary Mixed Polyposis Syndrome

Sari Lieberman ¹, Menna Hawkins², Kalaikshiga Kengadaran², Menachem Schechter³, Naim Abu Freha⁴, Ido Laish⁵, Alon Basevitch⁶, Rinat Bernstein Molho⁷, Revital Bruchim⁶, Rakefet Chen Shtoyerman⁸, Elizabeth E. Half⁹, Lior H. Katz¹⁰, Sofia Naftaly Nathan¹¹, Gili Reznick Levi¹², Nadra G. Samra¹³, Sharon Simchoni¹⁴, Chana Vinkler¹⁵, Amit Wainstein¹⁴, Rachel Gingold Belfer¹⁶, Zohar Levi¹⁶, Kevin Monahan¹⁷, Ian Tomlinson¹⁸, Huw Thomas¹⁷, Yael Goldberg¹⁹, Andrew Latchford¹⁷

¹Medical Genetics Institute, Shaare Zedek Medical Center, Jerusalem, Israel. ²St Mark’s Centre for Familial Intestinal Cancer, St Mark’s Hospital, London, UK. ³Gastroenterology Institute, Shaare Zedek Medical Center, Jerusalem, Israel. ⁴The Institute of Gastroenterology and Hepatology, Soroka University Medical Center, Beer Sheva, Israel. ⁵Gastroenterology Institute, Chaim Sheba Medical Center, Tel Hashomer, Israel; Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel. ⁶Naot Clinic, Maccabi Health Care Organization, Tel-Aviv, Israel. ⁷Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel; Susanne Levy Gertner Oncogenetics Unit, The Danek Gertner Institute of Human Genetics, Chaim Sheba Medical Center, Tel-Hashomer, Israel. ⁸The Genetic Institute, Kaplan Medical Center, Rehovot, Israel; The Biology Department, Ariel University, Ariel, Israel. ⁹Department of Gastroenterology, Rambam Health Care Campus, Haifa, Israel. ¹⁰Department of Gastroenterology and Hepatology, Hadassah-Hebrew University Medical Center, Ein Kerem, 91,120, Jerusalem, Israel; Faculty of Medicine, Hebrew University of Jerusalem, Jerusalem, Israel. ¹¹Rabin Medical Center, Raphael Recanati Genetic Institute, Petach Tikva, Israel. ¹²The Genetics Institute, Rambam Health Care Campus, Haifa, Israel. ¹³Genetic Unit. Ziv Health Centre, Safed, Israel. ¹⁴Genetic Institute, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel. ¹⁵Onco Genetis Unit, Institute of Medical Genetics, Wolfson Medical Center, Holon, Israel. ¹⁶Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel; Division of Gastroenterology, Rabin Medical Center—Beilinson Hospital Petah Tikva, Israel. ¹⁷St Mark’s Centre for Familial Intestinal Cancer, St Mark’s Hospital, London, UK; Imperial College London, London, UK. ¹⁸University of Edinburgh, Edinburgh, Scotland. ¹⁹Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel; Rabin Medical Center, Raphael Recanati Genetic Institute, Petach Tikva, Israel

Background and aim Hereditary Mixed Polyposis Syndrome (HMPS) is caused by a 40 kb duplication upstream to the GREM1 gene, which predisposes to colorectal cancer (CRC). Surveillance of the large bowel is recommended but there are few data on effectiveness and long-term outcomes. We reviewed these outcomes in our HMPS cohort.

Method Patients were identified from ten medical centres in Israel and one centre in the UK. Clinical data were retrospectively collected. Details of dysplastic polyps (adenomas and serrated or juvenile polyps with dysplasia), advanced adenomas (AAs) (≥ 10 mm and/or high-grade dysplasia) and cancer diagnoses were collated from medical notes and registry family files. Research departmental approval was gained.

Results 56 patients were identified, median age at genetic diagnosis 42 years (range 10–95). Colonoscopy surveillance data were available for 55/56, comprising 491 patient year follow up.

Median age at first colonoscopy was 31 years (range 10–90). In total, 292 surveillance colonoscopies were performed, median per patient 5 (range 1–23) and median per patient surveillance interval was 2 years (range 0.5–6).

Six CRCs arose in five patients before surveillance was started (median age 42 years (36–50)). Two patients with CRC were diagnosed at index colonoscopy (age 41 and 42). One patient developed two CRC on surveillance (age 56 and 62 years, stage 3 and 4 respectively). Where data are available, 4/4 CRC were microsatellite stable (MSS).

Median age at first AAs was 38.5 years (range 25–64). Fifteen patients developed 22 AAs (21 ≥ 10 mm, 5 contained high dysplasia). Of the AAs, 5/22 were found at index surveillance colonoscopy and 17/22 arose on surveillance. AAs were seen in 5/8 of patients with CRC and 10/47 without CRC (p = 0.02).

Median age at first dysplastic polyp detected on surveillance was 35 (range 12–93). A median of 1 dysplastic polyp was found per procedure (range 0–15).

Three patients had extra-colonic cancers: basal cell carcinoma × 1, melanoma × 1, duodenal and prostate cancer in the same patient.

Conclusions CRC in HMPS appears to be MSS. There is no excess risk of extra-colonic tumours.

AAs or cancer arising on surveillance are not common but there appears to be a high-risk group which develop CRC and AAs.

Although NCCN guidelines of starting colonoscopy at age 25—30 years seems reasonable, there appears to be a spectrum of risk, so personalizing surveillance onset and intervals should be considered.

Keywords HMPS, colonoscopy, adenomas, CRC, surveillance, intervals.

OP-035: Clinical: Diagnosis, Endoscopy, Management, Implementation

Colorectal cancer risk and surveillance yield in patients with PTEN Hamartoma Tumour Syndrome (PHTS, Cowden syndrome)

Nicoline Hoogerbrugge, Linda Hendricks, Dorien Van Der Biessen Van Beek, Mariet Van Kouwen, Chella Van Der Post, Tanya Bisseling, European Pten Study Group, Meggie Drissen, Janet Vos

Radboud university medical center, Nijmegen, The Netherlands

Background and aim Patients with PTEN Hamartoma Tumour Syndrome (PHTS) are at high hereditary risk of various cancers and advised colorectal cancer (CRC) surveillance with colonoscopy every 5 years from age 40. Current PHTS guidelines are expert opinion-based and cancer risks are likely overestimated by ascertainment bias. We aimed to assess the CRC risk and surveillance yield in adult PHTS patients.

Method To assess cumulative lifetime risks (CLTRs) with Kaplan–Meier analyses and standardized incidence ratios (SIR), an adult PHTS cohort study from 20 European expert centers with data from medical files, registries and/or questionnaires. Ascertainment bias was addressed by excluding cancer indexes and by incident case analyses, and surveillance bias by assessing both adenomatous polyps and carcinomas. To assess the yield of surveillance data were retrospectively derived from medical files of patients who started CRC surveillance between 2011–2022 at the Radboudumc, Nijmegen, the Netherlands.

Results CRC risk: 455 patients were included, 50% indexes, 372 with prospective follow-up years (median 6, IQR:3–10)). Four (1%) patients were diagnosed with CRC at median age of 59, and the youngest age at CRC diagnosis was 38 years. CLTR at age 60 ranged from 2% in the cohort excluding index cases to 5% in the prospective cohort (95%CI-widths: 13% and 27%). Including modelled polyp progression, CRC risks were 3–6% and 7–10%, respectively. The SIR for CRC was not significantly increased when excluding ca-index (SIR = 2.2, 95%CI 0.4–6.5) and only slightly increased for the prospective cohort (SIR = 5.6, 95%CI 1.5–14.2).

CRC Surveillance: 38 patients, median age at first examination 45 years, had 64 surveillance colonoscopies with intervals ranging between 0.4—6 years. CRC was diagnosed in 1 patient (age 41) prior to the PHTS diagnosis. During surveillance, no CRCs were found. Adenomas were observed in 14/38 (37%) patients of whom one with advanced adenoma. Besides adenomas, hamartomas (16/38) and ganglioneuromas (15/38) were most common.

Conclusions Our unique data show that CRC risk in PHTS is not, or at most slightly, increased. The main findings at colonoscopy were benign lesions that may not give an increased cancer risk. We suggest to adapt CRC surveillance to one baseline colonoscopy around age 45 and use follow-up intervals based on findings at baseline colonoscopy and CRC family history.

Keywords herediatry colorectal cancer, COWDEN syndrome, PTEN, cancer prevention, surveillance, cancer risk.

OP-036: Clinical: Diagnosis, Endoscopy, Management, Implementation

Results of in international survey of breast cancer in Peutz-Jeghers syndrome

Andrew Latchford ¹, Anne Goverde², Massih Bahar¹, Rubi Ramzan³, Anne Marie Jelsig⁴, John Karstensen⁴, Finlay Macrae³, Susan Clark¹, Anja Wagner²

¹Polyposis Registry, St Mark's Hospital, UK. ²Erasmus MC Cancer Institute, Netherlands. ³Royal Melbourne Hospital, Australia. ⁴Copenhagen University Hospital Hvidovre

Background and aim An increased risk of breast cancer (BrCa) in Peutz-Jeghers syndrome (PJS) is reported and surveillance has been recommended for female patients with PJS. Data regarding histological classification, receptor status, and management outcomes are lacking. We aimed to gather such data, to facilitate future guideline development.

Method An international survey of members of InSiGHT was performed. Retrospective anonymised data were collated, centralised and analysed. The survey received research ethics approval at the host institution; each local centre contributing data obtained appropriate clearance according to local protocols.

Results From the InSiGHT members contacted, 23 responses were received. One responder provided Dutch national collaborative data and two other centres had cases of BrCa and provided their local data. In total the cohort comprised 25 patients with 31 breast cancers (five bilateral and one metachronous unilateral tumours). All patients were female. The median age at first cancer diagnosis was 45 years (range 26–67).

Most cancers were ductal carcinoma, either invasive (13) or in situ (DCIS; 14). The others comprised invasive carcinoma unspecified (2), mixed ductal and mucinous (1) and intra-cystic papillary carcinoma. TNM staging for invasive cancer was available in eight cases as follows: T1N0M0 (7), T1N1M0 (1).

Receptor status, where assessed, was as follows: oestrogen receptor positive 15/16, progesterone receptor positive 7/14 and HER2 positive 4/13.

Data are available for follow-up in 16 patients and comprise 143 patient years. Two patients have died at 3 and 9 years respectively after BrCa diagnosis. Of those with 5 year follow-up the survival rate is 85%.

Conclusions Given the age profiles, BrCa surveillance should be initiated at 25 years, which may be a younger age than is currently recommended in some national policies.

The predominance of DCIS and ductal carcinoma may inform the optimal modality of surveillance.

“Triple negative” BrCa is rare.

Cancers are usually early stage with corresponding good survival.

Keywords Peutz-Jeghers syndrome, breast cancer.

OP-037: Genetics: Research, Germline, Genomics, Variomics

LynchRisk: a pedigree-based model for computing risks of Lynch syndrome and future cancer risks conditional on a family history of Lynch-associated cancers and clinical data

Alexandra Lefebvre ¹, Patrick Benusiglio², Erell Guillerm², Florence Coulet², Alex Duval³, Grégory Nuel¹

¹LPSM, CNRS 8001, Sorbonne Université, Paris, France. ²CRSA, INSERM UMR 938, Paris, France; UF d'Oncogénétique et Angiogénétique Moléculaire, département de génétique, Hôpital Universitaire la Pitié-Sapêtrière, APHP Sorbonne Université, Paris, France. ³CRSA, INSERM UMR 938, Paris, France

Background and aim Screening for Lynch Syndrome (LS) is now universal in patients with colorectal and endometrial cancer (MMR IHC and/or MSI analysis). MMR-deficiency with no MLH1 promoter hypermethylation justifies dowstream germline testing. This consensual strategy is very sensitive for the diagnosis of LS but does not provide an answer for all cases. Uncertainty remains for example for patients with a Lynch-like diagnosis. In addition, one might wonder whether all MMR-d cases should have germline testing considering limited resources and as the probability of LS is sometimes extremely low. Novel mathematical models would be helpful in this context. Existing models are divided into logistic regressions (PREMM5) and pedigree-based (PB) models (MMRpro). However, they have major limitations, i.e. they do not take into account tumour testing, cancer spectrum, parameter estimation, and family history as a whole, and/or have not been updated for over a decade. In this context we have developed a new PB model which overcomes the aforementioned limitations.

Method LynchRisk (LR) is a PB model (hence a Bayesian Network) in order to exhaustively consider FH. Its parameters are extracted or estimated from recent literature (Assasi et al. 2016, Win et al. 2017, Defossez et al. 2019, Dominguez et al. 2020). We include exhaustive CD with MSI screening, IHC per protein or dimer, hypermethylation of MLH1 promoter, BRAF V600E screening. Personal histories of cancer are modeled with multi-state survival for taking into account the seven main localizations and up to two diagnoses per individuals.

Results We compare LR, MMRpro and PREMM5 (if applicable) in their LS probability and future cancer risks computations conditional on a wide variety of simulated data and families, for instance, among several other results, various personal histories and CD (Fig. 1) or entire FHs (Fig. 2). Figure 1 shows the importance of separating colon and rectum as well as considering separately MSI screening and IHC per protein/dimer. Figure 1 shows that enlarging the Lynch spectrum, considering multiple cancers and/or the whole FH greatly improves the results.

Conclusions First results show the interest and the granularity of LR when compared to other current models, the first very next step being its clinical validation on a real dataset. We expect a clinical validation in a near future on the extensive databased of Saint-Antoine Hospital (Paris) with about 2000 patients including 300 confirmed LS and an estimate of 80 Lynch-like patients.

Keywords germline, genetic predisposition, pedigree-based model, risk assessment, risk prediction, Bayesian networks

Fig. 1

Subclonal MSH6F1088fs and MSH3K383fs homopolymer frameshift mutations drive MMRd mutation burden. (A) Multiple linear regression showing relationship between homopolymer frameshifts in MSI target genes and increased or decreased mutation burden in the Genomics England MSI CRC cohort. (B) Pie charts showing mutation categories for MSH3 (top) and MSH6 (bottom). (C) Cases with MSH6F1088fs and/or MSH3K383fs homopolymer frameshifts (in red) and cases without such mutations (in blue) in the Genomics England MSI CRC cohort (n = 217) ranked by mutation burden. Clonal alterations in MMR genes MLH1, PMS2, MSH2 and MSH6, as well as subclonal MSH6F1088fs and MSH3K383fs frameshift status is indicated in the panels below. Insets show MSH6F1088fs and MSH3K383fs mutation variant allele fraction. (D-F) SNV, indel, and total mutation burden according to MSH6F1088fs and MSH3K383fs mutation status. Median values are represented by horizontal black lines

Fig. 2

Sequential expansion and contraction of the hypermutable MSH6 coding homopolymer during tumour growth. (A) Example H&E and MSH6 immunohistochemistry for a case displaying subclonal loss of MSH6 expression. Inset shows segmentation for MSH6 labelling (in red) and MSH6 loss (in blue). Dashed areas indicate nested subclones showing restoration of MSH6 labelling. Detailed photomicrographs show normal crypts (i), focal loss of MSH6 labelling (ii), and nested subclone with restoration of MSH6 labelling (iii). (B) Laser capture microdissection (LCM) of tumour background, MSH6-negative and MSH6-positive nested area followed by Sanger sequencing of the MSH6 C8 coding microsatellite. (C) Sequential expansion and contraction of the hypermutable coding homopolymers in MSH6 and MSH3 unmasks these sites as cryptic stochastic switches in microsatellite unstable tumours.

Fig. 1

Dose-dependent decrease of viability after treatment of patient-derived FAP organoid line with Pyrvinium. Organoids were treated for three days using various concentration of Pyrvinium and assessed for ATP content using CellTiter Glow®.

Fig. 1

Proportion of patients developing ileal adenomas (a) and requiring excision of pouch or end ileostomy due to polyposis or cancer (b)

Fig. 1

Individual posterior probabilities conditional on various personal histories of cancer and/or clinical data. Posterior LS probabilities, in percent, for a male for overall genes (left columns) and the two most penetrant genes for the localization, MLH1 (middle columns) and MSH2 (right columns) conditional on a diagnosed colon cancer CC or rectal cancer RC (per column) at 30, 40 or 50 years of age (per line) computed with LynchRisk (LR) and MMrpro (Mp). Note that MMRpro aggregates CC and RC into colorectal cancer (CRC). PH means personal history of cancer, hence CC, RC or CRC at indicated age. Additional conditioning on biological testing results are integrated (per sets of three rows) where ``hyper.’’ stands for hypermethylation of MLH1 promoter. Values are grayed if repeated, i.e. equal to quantities computed conditional on other data. Note that PREMM5 is not designed for computing posterior probabilities in the absence of familial data.

Fig. 2

Individual posterior LS probabilities conditional on various family histories. Evolution (per line) of individual posterior LS probabilities (on the right) conditional on various FHs (on the left) computed with LynchRisk (red), MMRpro (black) and PREMM5 (cyan). UN, CC, EC and GIC respectively stands for unaffected, colon cancer, endometrial cancer and gastro-intestinal cancer

OP-038: Epidemiology and Registries

Prevalence of Lynch syndrome in a multi-ethnic Asian population

Sock Hoai Chan¹, Weng Khong Lim², Joanne Ngeow ³

¹National Cancer Centre Singapore. ²Duke NUS Medical School Singapore. ³Lee Kong Chian School Of Medicine, Nanyang Technological University Singapore

Background and aim Lynch syndrome (LS) is an inherited autosomal dominant disorder associated with genetic predisposition to a spectrum of malignancies including cancers of the colon, rectum, stomach, liver, kidney, brain and skin. Women with Lynch syndrome are also at increased risk of endometrial and ovarian cancers. Population prevalence of LS is estimated 1:370 in European-ancestry groups. In Asia, a recent study in the Japanese population estimated LS prevalence of 1:578; however, estimates are less characterized for other Asian populations. In this study, we aim to characterize the population prevalence of LS in Singapore, a Southeast Asian city-state comprising multi-ethnic Asian populations of East Asian, South Asian and Southeast Asian ancestries.

Method The dataset for analysis is derived from SG10K_Health programme, a cross-sectional population cohort of 9051 individuals unrelated to the second degree, comprising of 5502 Chinese, 1941 Indian and 1608 Malay individuals. Whole genome sequencing was performed on DNA from blood, jointly processed and annotated on a standardized bioinformatic pipeline. Identified variants in LS genes (MLH1, MSH2, MSH6, PMS2, EPCAM) were curated using the American College of Medical Genetics and Genomics (ACMG) guidelines and carrier frequencies for pathogenic variants were normalized to the size of each ancestry group.

Results We identified 17 (17/9051) carriers of pathogenic germline variants in LS genes, translating to a prevalence of 1:532 in Singaporeans. Prevalence in Chinese was highest with 13 carriers (1:423) compared with Indian (1:970, 2 carriers) and Malay (1:804, 2 carriers). Over 88% (15/17) carriers harbored a pathogenic variant in MSH6 (8/17) or PMS2 (7/17), 2 in MLH1 and none observed for MSH2 and EPCAM. One variant, MSH6 c.3226C > T (Arg1076Cys), was observed recurrently in 0.05% (3/5502) of Singaporean Chinese and has been previously reported in association with colon cancer, Danish LS families as well as constitutional mismatch repair deficiency syndrome.

Conclusions Our findings indicate that LS prevalence in Asians is not lower than populations of European ancestry. However, our study demonstrated that Asians are a diverse population with variation in LS prevalence and genetic spectrum of LS genes among different Asian ancestry groups, hence clinical professionals serving multi-ethnic populations should be cognizant of ancestry-specific variation in their genetic risk assessment for LS.

Keywords Lynch syndrome, germline, Asian.

OP-039: Genetics: Research, Germline, Genomics, Variomics

Saturation-scale functional evidence supports clinical variant interpretation in Lynch Syndrome

Anthony Scott ¹, Felicia Hernandez², Carrie Horton², Adam Chamberlin², Xiaoyan Jia¹, Sajini Jayakody¹, Bala Burugula¹, Tina Pesaran², Elizabeth Chao², Rachid Karam², Jacob Kitzman¹

¹Dept. of Human Genetics, Univ. of Michigan, Ann Arbor, MI. ²Ambry Genetics, Aliso Viejo, CA

Background and aim Lynch Syndrome (LS) is a cancer predisposition syndrome affecting more than 1 in every 300 individuals worldwide. Tumorigenesis in LS is driven by germline loss-of-function variants in DNA mismatch repair (MMR) genes. Clinical genetic testing for LS can be life-saving but is complicated by a large burden of variants of uncertain significance (VUS), especially missense changes. We previously applied deep mutational scanning (DMS) to measure functional effects for > 94% of the 17,746 possible missense variants in the key LS gene MSH2 (Jia et al., AJHG, 2021). Therefore, we aim to improve upon clinical variant interpretation using these DMS data.

Method DMS data were overlaid on clinical databases comprising > 15,000 individuals with MMR gene variants detected by paired tumor and germline testing at a clinical genetic testing laboratory. To gauge the strength of evidence provided by the DMS, we curated a list of MSH2 germline missense variants previously classified as Pathogenic (N = 23) or Benign (N = 26) as controls. To avoid redundant application of evidence, these validation variants’ classifications were derived without use of any functional data. Subsequently, we applied the DMS to 720 standing missense VUS, obtained from a distinct cohort from the original validation set.

Results Our functional DMS measurements agreed with the clinical interpretation for all 49 variants. Following recommendations for application of the functional evidence criterion using the ACMG/AMP variant interpretation framework (Brnich et al., Genome Med, 2019), our MSH2 function map can be used with PS3/BS3 evidence codes given its observed strong concordance with these prior clinical interpretations. For the 720 VUS, while most of these are predicted to have a neutral impact on gene function, 29 (4.0%) scored as deleterious in our function map, consistent with previously published rates among other cancer predisposition genes. We are pursuing reclassification for these variants, combining functional evidence with family history and tumor characteristics consistent with Lynch Syndrome. Additionally, in addition to retrospective VUS reclassification, to date these functional data have enabled resolution of ten newly detected missense MSH2 VUSs.

Conclusions High-throughput assays for mismatch repair loss of function provide a scalable method for VUS resolution and serve as strong evidence criteria for variant classification.

Keywords Cancer syndromes, Computational tools, Bioinformatics.

Validating MSH2 DMS using known pathogenic and benign variants.

Reclassification of clinical MSH2 VUS using DMS data.

OP-040: Cancer Pathway-Somatic Mutations, Pathology

Characterization of Blood-Derived Exosomal hTERT mRNA as a Biomarker for Colon Cancer and Lynch Syndrome

Ido Laish ¹, Zohar Levi², Hussein Mahajna¹, Hadar Goldvaser³, Orit Uziel⁴

¹Institute of Gastroenterology, Chaim Sheba Medical Center, Tel Hashomer, Israel. ²Institute of Gastroenterology, Rabin Medical Center, Petah Tikva, Israel. ³Institute of Oncology, Shaare – Zedek Medical Center, Hebrew University Faculty of Medicine, Jerusalem, Israel. ⁴The Felsenstein Medical Research Center, Rabin Medical Center, Petah Tikva, Israel

Background and aim Human telomerase reverse transcriptase (hTERT)- mRNA was shown to be elevated in exosomes derived from the sera of a variety of hematological and solid cancer patients. We aimed to evaluate its role as a diagnostic marker in patients with newly diagnosed colon cancer and in hereditary syndromes with predisposition to colon cancer.

Method hTERT -mRNA levels were determined in serum-derived exosomes from.

88 patients with colon cancer, 71 Lynch-syndrome carriers with unknown active malignancies and 50 healthy controls. Data, including demographics, background diseases, clinical data regarding tumor characteristics and genetic data, were retrieved data from medical files.

Results Patients with colon cancer had both higher exosomal hTERT mRNA levels and a higher proportion of patients with positive exosomal hTERT mRNA than controls (29.5% vs. 4%, respectively, P values < 0.001). Within the cancer group, patients with a metastatic disease had higher levels of telomerase mRNA than non-metastatic disease patients, and these levels correlated with CEA levels. Likewise, Lynch syndrome carriers had a higher proportion of positive exosomal hTERT mRNA than controls (21.1% vs. 4%, respectively, P value 0.008) but only a trend towards higher exosomal hTERT mRNA levels. Higher telomerase mRNA levels were not correlated with the mutated gene.

Conclusions Exosomal serum hTERT –mRNA levels are associated with metastatic colon cancer and were also demonstrated in a subset of Lynch syndrome carriers. Its significance as a biomarker for developing malignancy should be elucidated.

Keywords Lynch syndrome, telomerase, colon cancer.

Percent of patients with positive exosomal hTERT mRNA values [Relative quantification (RQ) > 1.2] in Lynch syndrome carriers according to genetic mutation

Percent of patients with positive exosomal hTERT mRNA values [Relative quantification (RQ) > 1.2] in the study groups

OP-041: Chemoprevention

Cancer Prevention with Resistant Starch in Lynch syndrome patients in the CAPP2 Randomised Placebo Controlled Trial: planned 10-year follow-up

John C. Mathers¹, Faye Elliott², Finlay Macrae³, Jukka Pekka Mecklin⁴, Gabriela Möslein⁵, Fiona E. Mcronald⁶, Lucio Bertario⁷, Gareth Evans⁸, Anne Marie Gerdes⁹, Judy W.c Ho¹⁰, Annika Lindblom¹¹, Patrick J. Morrison¹², Jem Rashbass⁶, Raj Ramesar¹³, Toni T. Seppälä¹⁴, Huw J. W. Thomas¹⁵, Harsh Sheth¹⁶, Kirsi Pylvänäinen⁴, Lynn Reed¹⁶, Gillian M. Borthwick¹⁶, D. Timothy Bishop², John Burn ¹⁶

¹Human Nutrition Research Centre, Population Heath Sciences Institute, Newcastle University, Newcastle upon Tyne, UK. ²Division of Haematology and Immunology, Leeds Institute of Medical Research at St James’s, University of Leeds, Leeds, UK. ³Division Colorectal Medicine and Genetics, Royal Melbourne Hospital, Melbourne, Australia. ⁴Department of Education & Research, Jyväskylä Central Hospital, Jyväskylä, Finland. ⁵Center for Hereditary Tumors, Ev. BEHESDA Khs. zu Duisburg GmbH, Germany. ⁶National Cancer Registration and Analysis Service, Public Health England. ⁷Instituto Nazionale per lo Studio e, la Cura dei Tumori, Milan, Italy. ⁸Division of Evolution and Genomic Medicine, University of Manchester, St Mary’s Hospital, Manchester Universities Foundation Trust, Manchester, UK. ⁹Medical Genetics Clinic, ICMM; Clinical Genetics, Rigshospital, Copenhagen, Denmark. ¹⁰Hereditary GI Cancer Registry, Department of Surgery, Queen Mary Hospital, Hong Kong, China. ¹¹Department of Molecular Medicine & Surgery, Karolinska Institutet, Stockholm, Sweden. ¹²The Department of Medical Genetics, Queens University Belfast, Belfast City Hospital HSC Trust, Belfast, UK. ¹³MRC Genomic and Precision Medicine Research Unit, Division of Human Genetics, Institute for Infectious Diseases and Molecular Medicine, University of Cape Town, Observatory, South Africa. ¹⁴Department of Gastrointestinal Surgery, Helsinki University Hospital, Helsinki. ¹⁵St Mark’s Hospital, London & Faculty of Medicine, Imperial College London. ¹⁶Newcastle University Translational and Clinical Research Institute, Newcastle upon Tyne, UK

Background and aim The CAPP2 Trial investigated, in factorial design, the effects of aspirin and/or resistant starch on cancer incidence in Lynch syndrome (LS) patients. Resistant Starch is also known as Fermentable Fiber and is found in a range of fruit and vegetable dietary constituents such as green tipped bananas and pulses. It escapes digestion in the upper gastrointestinal tract and is fermented by gut bacteria to short chain fatty acids, such as butyrate which has anti – neoplastic effects.

Method Participants with LS were randomised double-blind to 30 g resistant starch (RS) daily or cornstarch placebo for up to 4 years. Overall, 463 participants received RS and 455 cornstarch placebo. A sachet of unflavoured powdered starch was added daily to food or drink. Compliance was measured by counting remaining sachets. Follow up was planned to ten years. For centres in England, Wales and Finland established national registries were used to identify all cancers up to the end of the second decade.

Results After up to 20 years follow-up, planned analysis revealed no difference in colorectal cancer (CRC) incidence (52 among those randomised to RS versus 53 on placebo) but significantly fewer participants had non-colorectal LS cancers, the secondary clinical endpoint, in those randomised to RS (n = 27) compared with placebo (n = 48); ITT analysis (HR 0.54 (95%CI 0.33–0.86), p = 0.010). We accounted for multiple primary cancer diagnoses among participants by calculating Incidence Rate Ratios (IRR) which showed that the protective effect of RS against non-CRC LS cancers was evident in ITT (IRR 0.52 (95%CI 0.32- 0.84), p = 0.0075) analysis. These effects are particularly pronounced for cancers of the upper GI tract; 5 diagnoses in those randomised to RS versus 21 diagnoses in the placebo group. The reduction in non-CRC LS cancers was detectable in the first 10 years, the planned endpoint, and continued in the national registry data through the next decade. There was no interaction between aspirin and RS treatments.

Conclusions 30 g daily RS appears to have a substantial protective effect against non-CRC cancers for LS patients. This may represent a systemic effect of a long term impact on the gut microbiome. One possible link to the upper GI tract may be through an impact on carcinogenic secondary bile acids which are returned via the portal venous system to the biliary system.

Keywords Resistant Starch, aspirin, lynch syndrome, microbiome, colorectal cancer.

Time to first colorectal cancer and time to first non-CRC Lynch syndrome cancer in all CAPP2 study participants followed up for 10 years and for 20 years in England, Finland, and Wales

OP-042: Chemoprevention

Genomic responses to aspirin in colonic organoids from diverse individuals with and without Lynch syndrome

Hina Usman, David Witonsky, Margaret Bielski, Sonia S. Kupfer

Department of Medicine, University of Chicago, Chicago, IL

Background and aim Aspirin (ASA) is a proven chemopreventive agent for colorectal cancer especially in Lynch syndrome (LS), though mechanisms underlying these effects are incompletely understood. Human organoids are an ideal system to study genomic host-environment interactions across individuals. Here, we utilized colonic organoids from diverse individuals with and without LS to measure transcriptional responses and to perform eQTL mapping of ASA colonic responses.

Method Human colonic organoids from 67 healthy subjects (33 African- & 34 European-Americans) were cultured, differentiated and treated with 3 mM ASA or vehicle control (DMSO) for 24 h. Gene expression was measured using RNA-seq and differential expression (DE) analyzed in dream. Gene set enrichment analysis (GSEA) was performed by SetRank. Top DE genes were validated in organoid lines from LS patients by qPCR. Expression quantitative trait loci (eQTL) mapping of ASA responses was performed using BRIdGE.

Results In total, 8,490 DE genes in response to ASA at false discovery rate (FDR) < 5% were found of which 3,859 and 4,631 were up- and down-regulated, respectively (Fig. 1A). GSEA of DE genes showed significant enrichment for fatty acid oxidation and PPAR signaling pathways among others (Fig. 1B). A significant difference between absolute values of transcriptional response to ASA across ancestry was observed compared to balance permutations (p = 0.03) (Fig. 1C). Top DE upregulated genes including ACSL1, FABP1, HMGCS2 and SCD1 were validated in LS organoid lines (Fig. 1D). eQTL mapping identified 32 genetic variants associated with treatment responses such as TMEM19 (ASA only) and PLIN2 (DMSO only) (Fig. 2A & B).

Conclusions Assessment of genomic responses to ASA in colonic organoids from diverse individuals yielded new observations about ASA effects on pathways involved in fatty acid oxidation among others that were validated in LS organoids. Notably, overall ASA transcriptional responses differed by ancestry and several genetic variants associated with ASA responses were identified. Based on these findings, work is ongoing to dissect the genetic, transcriptional and cellular ASA effects in the context of colorectal cancer chemoprevention in diverse individuals with and without LS.

Keywords aspirin, organoids, chemoprevention, transcriptional responses, eQTL mapping, Lynch syndrome.

Fig. 1

Fig. 2

OP-043: Clinical: Diagnosis, Endoscopy, Management, Implementation

Metachronous Colorectal Cancer Risk in Lynch syndrome: is Extensive Colectomy Necessary for all Carriers?

Ellis Eikenboom ¹, Sarah Moen², Ernst Kuipers², Monique Van Leerdam³, Anja Wagner⁴, Manon Spaander²

¹Erasmus MC Cancer Institute, University Medical Center Rotterdam, The Netherlands, Departments of Clinical Genetics and Gastroenterology and Hepatology. ²Erasmus MC Cancer Institute, University Medical Center Rotterdam, Department of Gastroenterology and Hepatology. ³Department of Gastroenterology and Hepatology, Leiden University Medical Center; Department of Gastrointestinal Oncology, Netherlands Cancer Institute, Amsterdam. ⁴Erasmus MC Cancer Institute, University Medical Center Rotterdam, Department of Clinical Genetics

Background and aim Treatment guidelines recommend extensive colectomy (total or subtotal colectomy) for colorectal cancer (CRC) in all Lynch syndrome (LS) carriers. However, CRC risks differ per mismatch repair gene involved: MSH6 and PMS2 have a lower risk than MLH1 and MSH2. Consequently, MSH6 and PMS2 patients might benefit from a partial colectomy, instead of extensive colectomy, to prevent metachronous CRC (mCRC) development.

Method Data from all LS carriers with CRC registered in the National prospective LS database were evaluated. Data from LS carriers with CRC were linked to the national pathology registry to assess mCRC risk after partial colectomy in low-risk LS (MSH6/PMS2) and after extensive colectomy in high-risk LS carriers (MLH1/MSH2/EpCAM), using time-to-event analysis. Patients were censored at time of event, death or assembly of database (September 13th, 2021). Ten year incidence was calculated using Kaplan–Meier survival analyses. mCRC was defined as a second CRC diagnosed at least six months after primary CRC.

Results Of the 1908 LS carriers, 531 LS carriers (median age 49 years, range 15–82; 48.6% female, n = 258) underwent surgery for primary CRC. A total of 124 mCRCs occurred after a median follow-up of 132 months (range 7- 412). mCRC risk after partial colectomy in the low-risk group did not significantly differ from the mCRC risk after extensive colectomy in the high-risk group (Hazard Ratio, HR = 1.36, 95% CI 0.68–2.72; p = 0.39). Ten year incidence of mCRC was 5.1% for the high-risk group after extensive colectomy; 9.0% for the low-risk group after partial colectomy. Within the high-risk group, partial colectomy resulted in a significantly higher mCRC risk compared to extensive colectomy (HR = 2.46, 95% CI 1.34–4.51; p = 0.004). Ten year incidence of mCRC was 16.9% for the high-risk group after partial colectomy.

Conclusions mCRC risks after partial colectomy in low-risk LS carriers did not significantly differ from the risk after extensive colectomy in high-risk LS carriers. This suggests that partial colectomy, followed by regular endoscopic surveillance, could be safe and effective in treatment of primary CRC in MSH6 and PMS2 carriers.

We would like to present our results at the InSIGHT 2022 meeting.

Keywords Lynch syndrome, colectomy, metachronous CRC, risk.

OP-044: Epidemiology and Registries

10-year Results from Universal Mismatch-Repair (dMMR) Diagnostics in Colorectal Cancer; experiences from Denmark

Jon Ambæk Durhuus ¹, Christina Therkildsen², Michael Galanakis³, Helle Gybel Juul Larsen¹, Mef Christina Nilbert⁴

¹Department of Clinical Research, Copenhagen University Hospital—Amager and Hvidovre, Copenhagen, Denmark. ²The Danish HNPCC Register, Gastro Unit, Copenhagen University Hospital—Amager and Hvidovre, Copenhagen, Denmark. Department of Clinical Research, Copenhagen University Hospital—Amager and Hvidovre, Copenhagen, Denmark. ³The Danish Cancer Society Research Center, Statistics and Data Analysis, Copenhagen, Denmark. ⁴Institute of Clinical Sciences, Division of Oncology and Pathology, Lund, Sweden. The Danish Cancer Society Research Center, Copenhagen, Denmark. Department of Clinical Research, Copenhagen University Hospital—Amager and Hvidovre, Copenhagen, Denmark

Background and aim Recognition of tumors with deficient mismatch repair (dMMR) has diagnostic, prognostic, and treatment-predictive implications. Reflex testing for dMMR was initiated in 2009 in Denmark, with the primary aim to identify individuals with Lynch syndrome. We evaluated uptake and outcome of the Danish program during a 10-year period.

Method Population-based registries were used to identify colorectal cancer diagnoses, which were linked to pathology data and health care data to assess uptake of reflex testing, define MMR status, describe dMMR patterns and map referral to genetic counseling.

Results A population of 52,480 patients diagnosed with their first colorectal cancer from 2009 to 2018 was identified. 36,644 (70%) of the patients had undergone reflex testing for MMR during the 10-year period and the uptake rate increased from 51.6% in year 2008 to 93.7% in year 2018. The probability of reflex testing was associated with age, tumor location, stage, and geographic region. In general, effect sizes were small, and we found the largest differences in the start of the study period, and small or no differences by the end of the study period. In total, 6,056 (16.5%) of the diagnosed cancers showed loss of at least one MMR protein and was considered dMMR. Diagnoses with loss of MLH1/PMS2 and a diagnosed BRAF mutation were regarded sporadic dMMR and were moved to the control group (n = 2091). We defined a potentially hereditary dMMR Lynch syndrome-suspected group of 3,965 tumors. The Lynch syndrome-suspected group was further analyzed regarding probability of referral to genetic counselling.

Conclusions Our evaluation of the Danish program for reflex MMR diagnostics, reveals that complete uptake takes a several years to achieve and displays disparities between patient groups. Further initiatives may be needed to ensure that individuals with dMMR tumors signifying potential Lynch syndrome cases are indeed referred to genetic counseling.

Keywords Colorectal cancer, Lynch syndrome, mismatch repair.

OP-045: Clinical: Diagnosis, Endoscopy, Management, Implementation

Efficacy of routine microsatellite instability test in colorectal cancer for screening Lynch syndrome

Hong-min Ahn ¹, Duck Woo Kim¹, Hyeon Jeong Oh², Hyung Kyung Kim², Hye Seung Lee³, Hye Rim Shin¹, Jeehye Lee¹, Heung Kwon Oh¹, Sung Bum Kang¹

¹Department of Surgery, Seoul National University Bundang Hospital, Seongnam, Korea. ²Department of Pathology, Seoul National University Bundang Hospital, Seongnam, Korea. ³Department of Pathology, Seoul National University Hospital, Seoul, Korea

Background and aim Lynch syndrome (LS) is a heritable genetic condition caused by mutation in DNA mismatch repair genes; MLH1, MSH2, MSH6, PMS2, and EPCAM. Universal screening for LS refers as routine microsatellite instability (MSI) test among all colorectal cancer (CRC) patients. However, the efficacy of routine MSI screening for LS in all age is still questionable. The aim of this study is to understand the efficacy of routine MSI screening for LS in CRC among different age groups.

Method CRC patients underwent surgical intervention from 2007 to 2021 were reviewed, and all patients went through the routine tumor MSI testing for LS during the study period. Information from family medical history of patients were gathered to evaluate the hereditary nonpolyposis colorectal cancer (HNPCC) by using Amsterdam II criteria. Among MSI-low and -high, the patients were divided into 7 groups by diagnosed age of CRC: 30 or less, 31 to 40, 41 to 50, 61 to 70, 71 to 80, and above 81. The yield of genetic testing for LS with or without pedigree-based diagnosis was compared.

Results Among the total 6,068 colorectal cancer patients, microsatellite stable (MSS) reported in 5,101 patients (84.1%), MSI-low in 514 (8.5%), and MSI-high in 453 (7.5%). The family histories were obtained from 5,421 patients (89.3%) and genetic testing was performed in 103 (10.7%) among MSI-low and -high patients. In age of 30 or less, four (36.4%) of 11 patients were diagnosed with LS by both genetic testing and pedigree-based diagnosis. From the age of 31 to 70, the yields of genetic testing for LS were gradually decreased as the age advanced (31 to 40; 10.9%, 41 to 50; 7.8%, 51 to 60; 4.0%, and 61 to 70; 1.2%, respectively). Meanwhile, with the combination of genetic testing and pedigree as diagnostic modality, the yields for LS increased (31 to 40; 17.4%, 41 to 50; 16.5%, 51 to 60; 5.1%, and 61 to 70; 3.1%, respectively). Above 71, the yield for LS is low as 0.4% and 0.8%, respectively, without or with the pedigree-based diagnosis.

Conclusions The routine MSI testing may be useful to early onset CRC patients under 30 but have barely effective in detecting LS in elderly patients over 60, although combination with pedigree analysis slightly improved the yield of diagnosis. MSI screening for LS could be tailored according to the age of CRC patients to avoid unnecessary testing.

Keywords Hereditary nonpolyposis colorectal cancer, Lynch syndrome, Microsatellite instability, Pedigree.

The yield of Lynch syndrome in colorectal cancer with microsatellite instability

The yields of genetic testing for LS were gradually decreased as the age advanced, and with the combination of genetic testing and pedigree as diagnostic modality, the yields for LS increased.

OP-046: Epidemiology and Registries

Using Structured Family History Data in the Electronic Health Record to Identify Primary Care Patients Eligible for Genetic Testing for Hereditary Gastrointestinal and Other Cancer Syndromes

Wendy Kohlmann ¹, Richard Bradshaw², Rachelle Chambers⁶, Ambreen Khan¹, Amanda Gammon¹, Joanne Jeter⁸, Cecilia Kessler⁵, Sara Low⁵, Rachel Monahan⁶, Rachel Hess⁴, Michael Flynn⁹, Kensaku Kawamoto², Guilherme Del Fiol², Melody Goodman⁷, Ophira Ginsburg⁷, Saundra Buys⁸, Kimberly Kaphingst³

¹Huntsman Cancer Institute and Department of Population Sciences, University of Utah, Salt Lake City, UT, USA. ²Department of Biomedical Informatics, University of Utah and University of Utah Health, Salt Lake City, UT, USA. ³Huntsman Cancer Institute and Department of Communication, University of Utah, Salt Lake City, UT, USA. ⁴University of Utah Health, Department of Population Health Sciences, Department of Internal Medicine, University of Utah, Salt Lake City, USA. ⁵Huntsman Cancer Institute, University of Utah, Salt Lake City, UT, USA. ⁶Perlmutter Cancer Center, New York University Langone Health, New York City, NY, USA. ⁷School of Global and Public Health, New York University, New York City, NY, USA. ⁸Huntsman Cancer Institute and Department of Internal Medicine, University of Utah, Salt Lake City, UT, USA. ⁹University of Utah Health, Department of Internal Medicine, University of Utah, Salt Lake City, UT USA

Background and aim Genetic testing for inherited predisposition to gastrointestinal cancers continues to be underutilized despite the availability of screening and management guidelines for individuals with these syndromes. Collection of structured family history (FHx) data within the electronic health record (EHR) holds promise for systematically identifying patients eligible for testing. However, data is lacking about whether FHx in the EHR is adequate for determining eligibility for genetic testing.

Method We examined structured FHx in the EHR among 144,484 primary care patients at the University of Utah (UHealth) and 645,264 primary care patients at New York University Langone Health System (NYU). Patients were between the ages of 25–60 and had a primary care visit in the past 3 years. A computerized algorithm identified patients who met criteria for genetic testing based on National Comprehensive Cancer Network criteria for Lynch syndrome and hereditary breast/ovarian/pancreas/prostate cancer (2018 versions).

Results 8110 (5.6%) and 14,200 (2%) of patients had EHR FHx meeting algorithm criteria for genetic testing at UHealth and NYU. The most common indications for meeting criteria were having a first- or second-degree relative with ovarian cancer (UHealth 39%, NYU 37%) or pancreatic cancer (UHealth 24%, NYU 33%). Having a relative with breast cancer diagnosed at age < 45 was the next most common indication (UHealth 14%, NYU 16%). At both centers, only 6% and 1% of patients met criteria due to a FHx of colorectal cancer or endometrial cancer diagnosed age < 50. Criteria that required documentation of > 3 relatives with specific cancers was an uncommon reason for being identified by the algorithm. Significant differences were noted in availability and completeness of FHx based on race, ethnicity, gender, and language preference. Patients from racial and ethnic minority groups, men, and Spanish-speaking patients were less likely to have cancer FHx documented in the EHR.(Table 1).

Conclusions This analysis found that FHx is not collected equitably across patient populations, and caution should be taken before developing patient identification programs based solely on EHR data in order to avoid exacerbating disparities in access to genetic services. Fewer patients met criteria based on family history of colorectal or endometrial cancer than breast cancer. Patients with a family history of ovarian or pancreatic cancer were overrepresented among those meeting criteria. This is likely due to these diagnoses being sufficient for meeting criteria while other cancer types require documentation of additional factors. These findings indicate the need for EHR tools that facilitate the collection of FHx relevant for assessing testing eligibility. Consideration should also be given to simplifying testing guidelines to better reflect FHx information that can be collected in primary care settings.

Keywords family history, primary care, genetic testing.

Table 1. Availability of Cancer Family History

OP-047: Genetics: Research, Germline, Genomics, Variomics

Population-based study shows constitutional MLH1 methylation (“epimutation”) underlies a significant fraction of incident mismatch repair-deficient, MLH1-hypermethylated colorectal cancer aged up to 55 years, but not endometrial cancers

Megan P. Hitchins ¹, Estela Dámaso², Yajing Hu³, Rocio Alvarez¹, Marta Pineda², Gabriel Capella², Marcio A. Diniz⁴, Rachel Pearlman⁵, Heather Hampel⁶

¹Department of Biomedical Sciences, Cedars-Sinai Cancer, Cedars-Sinai Medical Center, Los Angeles, California, USA. ²Hereditary Cancer Program, Catalan Institute of Oncology, Institut d'Investigació Biomèdica de Bellvitge (IDIBELL), ONCOBELL Program, Av. Gran Via de l'Hospitalet, 199–203, 08,908 L' Hospitalet de Llobregat, Barcelona, Spain; Consortium for Biomedical Research in Cancer – CIBERONC. Carlos III Institute of Health. Av. De Monforte de Lemos 5, 28,029, Madrid, Spain. ³Department of Medicine (Oncology), Stanford University, Stanford, California, USA. ⁴Department of Medicine, Cedars-Sinai Medical Center, Los Angeles, California, USA. ⁵Department of Internal Medicine and the Comprehensive Cancer Center, Ohio State University, Columbus, Ohio, USA. ⁶Department of Medical Oncology and Therapeutics Research, City of Hope National Medical Center, Duarte, California, USA

Background and aim Universal screening algorithms for the detection of Lynch syndrome (LS) among incident mismatch repair-deficient (MMRd) colorectal (CRC) and endometrial (EC) cancers use MLH1 methylation or surrogate BRAF mutation (CRC only) tumor testing to exclude sporadic cases from germline genetic testing. However, these algorithms dismiss rare cases of LS caused by constitutional MLH1 methylation (epimutation) as sporadic/non-Lynch, which may result in under-diagnosis and lack of risk-appropriate clinical management. We aimed to assess the frequency and age distribution of constitutional MLH1 methylation among incident CRC and EC cases with a MMRd (herein microsatellite unstable and/or immunohistochemical loss of MLH1 expression) and MLH1-methylated tumor in a population-based study unbiased with respect to age, family history, prior genetic testing, or BRAF status.

Method In a retrospective nested study of the “Columbus-area HNPCC Study” cohort, we tested all incident cases of MMRd, MLH1-methylated CRC (n = 95, age 34–98 years) or EC (n = 68, age 38–88 years) in available blood DNA. Informed by these findings, in the “Ohio Colorectal Cancer Prevention Initiative (OCCPI) Study” cohort, we limited testing to incident MMRd, MLH1-methylated CRC (n = 31) or EC (n = 24) cases under 60 years of age. MLH1 methylation testing was conducted on peripheral blood DNA by quantitative CpG pyrosequencing and methylation-specific real-time PCR, with positive samples confirmed by allelic bisulfite sequencing.

Results Columbus-area cohort: 4/95 (4.2%) CRC cases aged 34, 38, 52, and 74 years were positive for constitutional MLH1 methylation, yielding the highest positive detection rate of 75% (3/4) for those aged ≤ 55 years with one case missed, or 6.9% (4/58) for those aged < 75 years with all cases detected. OCCPI cohort: 4/31 (12.9%) CRC cases < 60 years were positive, aged 20, 34, 50, and 55 years, yielding the highest positive detection rate of 23.5% (4/17) ≤ 55 years. Methylation levels in blood ranged from 3% (mosaic) to 50% (hemiallelic) in positive cases. All EC cases in both cohorts were negative.

Conclusions Albeit rare overall, constitutional MLH1 methylation was a significant contributor to early-onset MMRd, MLH1-methylated CRC, but in contrast, appears to be negligible in MMRd, MLH1-methylated EC, even at young age. Our findings suggest routine testing for high-risk constitutional MLH1 methylation is warranted in CRC patients at first presentation with a MMRd, MLH1-methylated tumor ≤ 55 years for a timely and accurate molecular diagnosis at minimal cost. Where BRAF tumor mutation testing is performed instead, CRC cases with immunohistochemical loss of MLH1 expression, wildtype BRAF, and a negative/uninformative germline genetic test result should also be considered for testing for constitutional MLH1 methylation, as previously recommended.

Keywords MLH1, constitutional methylation, epimutation, Lynch syndrome,

OP-048: Clinical: Diagnosis, Endoscopy, Management, Implementation

DiagMMR, a novel carrier test validated for Lynch syndrome detection

Minttu Kansikas ¹, Laura Vähätalo¹, Jaana Putula¹, Niklas Lahti¹, Philippe Arnez¹, Sami Kilpinen², Minna Pöyhönen³, Päivi Peltomäki⁴, Heikki J. Järvinen⁵, Toni Seppälä⁶, Laura Renkonen Sinisalo⁷, Anna Lepistö⁷, Jukka Pekka Mecklin⁸, Minna Nyström²

¹LS CancerDiag Ltd, Helsinki, Finland; Faculty of Biological and Environmental Sciences, University of Helsinki, Helsinki, Finland. ²Faculty of Biological and Environmental Sciences, University of Helsinki, Helsinki, Finland. ³Department of Genetics, HUSLAB, Helsinki University Hospital Diagnostic Center, Helsinki, Finland; Department of Medical and Clinical Genetics, University of Helsinki, Helsinki, Finland; Applied Tumor Genomics, Research Programs Unit, University of Helsinki, Helsinki, Finland. ⁴Department of Medical and Clinical Genetics, University of Helsinki, Helsinki, Finland. ⁵Department of Surgery, Helsinki University Hospital, Helsinki, Finland. ⁶Department of Surgery, Helsinki University Hospital, Helsinki, Finland; Faculty of Medicine and Medical Technology, University of Tampere, Tampere, Finland; Department of Gastroenterology and Alimentary Tract Surgery, Tampere University Hospital, Tampere, Finland; Applied Tumor Genomics, Research Programs Unit, University of Helsinki, Helsinki, Finland. ⁷Department of Surgery, Helsinki University Hospital, Helsinki, Finland; Applied Tumor Genomics, Research Programs Unit, University of Helsinki, Helsinki, Finland. ⁸Nova Hospital of Central Finland, Jyväskylä, Finland; Faculty of Sports and Health Sciences, University of Jyväskylä, Jyväskylä, Finland

Background and aim Lynch syndrome (LS) is the most common hereditary cancer syndrome with a prevalence of up to 1:125. Although it is still quite underdiagnosed, well-organized cancer preventative surveillance methods exist, including preventive health care interventions to improve an individual´s prognosis. The problem is, however, in finding and diagnosing the genetic condition. The current workup is complex and involves an array of tests that combines family cancer history and clinical phenotypes with tumor characteristics and sequencing data, followed by a challenging task to interpret the found variant(s) as pathogenic. Based on the knowledge that an inherited mismatch repair (MMR) deficiency is a hallmark of LS, we have developed and validated a novel and unique functional MMR test that detects an inherited MMR deficiency from healthy tissue and thus helps to make a diagnosis without verified tumor and variant information.

Method The DiagMMR test carried out on proteins extracted from primary fibroblasts and a mismatch containing DNA substrate molecule detects the inherited MMR deficiency caused by the mismatch recognizing and binding proteins MSH2 and MSH6. The MMR efficiency is easily interpreted demonstrating either MMR proficiency or MMR deficiency (i.e. LS). Here, the DiagMMR results (normal or reduced MMR) of 117 samples were compared and validated against diagnosis based on sequencing. These samples represented LS carriers (n = 27) with a wide range of MSH2 and MSH6 pathogenic variants and non-pathogenic controls (n = 90). Following the validation, a small clinical pilot study was performed on LS suspected but not confirmed individuals. The sample collection was organized at the Helsinki University Hospital and Jyväskylä Central Hospital, Finland.

Results The method was shown to have exceptional specificity (100%) with high sensitivity (89%) compared to the reference standard (germline NGS). The ability to efficiently distinguish LS carriers from controls was further shown with a high AUROC value (0.97).

Conclusions DiagMMR offers an excellent tool for detecting inherited MMR deficiency linked to MSH2 and MSH6 and permits LS detection prior to and independent of tumor formation and sequencing data. It may be used alone, prior to and independent of tumor formation, or to complement an array of cancer reactive tests, to detect LS.

Keywords Lynch syndrome, carrier testing, MSH2, MSH6.

OP-049: Genetics: Research, Germline, Genomics, Variomics

A calibrated cell-based functional assay to aide classification of MLH1 DNA mismatch repair gene variants

Abhijit Rath ¹, Alexander A Radecki¹, Kaussar Rahman¹, Rachel B Gilmore², Jonathan R. Hudson², Matthew Cenci¹, Sean V. Tavtigian³, James P Grady⁴, Christopher D. Heinen¹

¹Center for Molecular Oncology, UConn Health, Farmington, CT, USA. ²Department of Genetics and Genome Sciences, UConn Health, Farmington, CT, USA. ³Department of Oncological Sciences, Huntsman Cancer Institute, University of Utah School of Medicine, Salt Lake City, UT, USA. ⁴Connecticut Institute for Clinical and Translational Science, UConn Health, Farmington, CT, USA

Background and aim Lynch syndrome (LS) predisposes patients to cancer of multiple tissue types and is caused by loss-of-function germline mutations in the DNA mismatch repair (MMR) genes. Identification of a pathogenic germline variant resulting in impairment of the encoded protein is confirmatory for LS diagnosis. However, discovery of a missense variant is often inconclusive, as the effects of these variants of uncertain significance (VUS) on disease pathogenesis are unclear. Importantly, clarifying the pathogenic significance of these VUS is necessary to direct further clinical management of suspected Lynch patients and their relatives. To this end, often laboratory generated functional data provide key insights regarding a variant’s ability to affect protein function, which is critical for appropriate classification of the VUS. Hence, a multi-dimensional assay system capable of informing MMR proficiency is highly desirable. We sought to develop a cell-based approach for testing the function of a panel of VUS in the MLH1 gene. We also adapted a quantitative framework for facile integration of the variant specific functional data into the existing InSiGHT classification scheme.

Method We developed an in cellulo functional assay in which we generated a panel of variant expressing cell lines by knocking-in individual MLH1 VUS using Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) gene editing in human embryonic stem cells. We examined their impact at the RNA and protein level. Further, we characterized the impact of the variants on cellular MMR functions including DNA damage response signaling and the repair of DNA microsatellites. We calibrated these assays by testing well-established pathogenic and benign control variants.

Results Out of 21 tested MLH1 VUS, five resulted in functionally abnormal protein, 15 VUS resulted in functionally normal protein, and one VUS showed mixed results. Furthermore, we converted the functional outputs into a single odds in favor of pathogenicity score for each VUS.

Conclusions Our CRISPR-based functional assay successfully models phenotypes observed in patients within a cellular context. Using this approach, we generated evidence for or against pathogenicity for utilization by the InSiGHT variant interpretation committee in the classification of these variants. Ultimately, this information will assist in proper diagnosis and disease management for suspected Lynch syndrome patients.

Keywords CRISPR-Cas, DNA mismatch repair, Lynch syndrome, MLH1, Human Embryonic Stem Cells, Variants of Uncertain Significance.

Summary of Functional Studies Outcomes and Predicted Classification for VUS

OP-050: Genetics: Research, Germline, Genomics, Variomics

A collaborative multicenter approach for the classification of mismatch repair gene variants in Spain: results of a pilot study

Víctor Lorca¹, Alysson Turcios², Anael López-Novo³, Maribel González-Acosta², Elisabet Munté², Zoila Caaveiro³, Daniel Rueda⁴, Montserrat de Miguel⁴, José Luis Soto⁵, Miguel Urioste⁶, Judith Balmaña⁷, Gemma Llort⁸, Francesc Balaguer⁹, Teresa Ramón y Cajal¹⁰, Carmen Guillén-Ponce¹¹, Núria Rodríguez Salas¹², Conxi Lázaro², Gabriel Capellá², Clara Ruiz-Ponte³, Pilar Garre¹, Trinidad Caldés¹, Marta Pineda ²

¹Hospital Clínico San Carlos, Madrid, Spain. IdISSC. ²Instituto Catalán de Oncologia, IDIBELL, Barcelona, Spain. CIBERONC. ³Fundación Pública Galega Medicina Xenomica, Santiago de Compostela, Spain. CIBERER. ⁴Hospital Universitario 12 de Octubre, Madrid, Spain. ⁵Hospital Universitario de Elche, Elche, Spain. ⁶Centro Nacional de Investigaciones Oncológicas, Madrid, Spain. ⁷Hospital Universitari Vall d’Hebron, Barcelona, Spain. ⁸Corporació Sanitaria Parc Taulí, Sabadell, Spain. ⁹Hospital Clínic, Barcelona, Spain. CIBEREHD. ¹⁰Hospital de la Santa Creu i Sant Pau, Barcelona, Spain. ¹¹Hospital Universitario Ramón y Cajal, Madrid, Spain. ¹²Hospital Universitario La Paz, Madrid, Spain

Background and aim Identification of germline pathogenic variants in mismatch repair (MMR) genes allows diagnosis of DNA MMR deficiency-associated syndromes. The identification of variants of unknown significance (VUS) precludes diagnosis. Recently, classification guidelines have improved reproducibility and transparency in variant classification. Our main aim is to accelerate and improve MMR variant classification based on the creation of a national registry of germline variants in Spain.

Method Twelve Spanish centers were invited to collaborate. Clinical and molecular data from MMR VUS carriers were submitted on a consensus template form. Collected data were curated. Monthly team conferences were held to update and review variant classification according to Insight v2.4 and Insight-ACMG v1 (pending to be approved by ClinGen) guidelines.

Results A total of 255 VUS present in 324 index cases were collected. Fifty-two of them were identified in 2 or more individuals. Classification was revisited in a subset of 161 VUS (enriched for recurrent, silent and splicing suspected variants). Update of available information allowed the reclassification of 20 variants according to Insight guidelines (8 class 1–2; 12 class 4–5). When using Insight-ACMG criteria 60 VUS were reassigned (43 class 1–2; 17 class 4–5), increasing the reclassification yield from 12 to 37%.

Conclusions The Spanish variant registry has accelerated the exchange of revised data, essential to improve classification. The use of Insight-ACMG MMR-specific guidelines further increases reclassification yield, mainly of silent/intronic VUS to likely neutral variants.

Keywords variant of unknown significance, mismatch repair, variant classification, Lynch syndrome, Constitutional Mismatch Repair Deficiency, ACMG.

OP-051: Epidemiology and Registries

Clinical description and first estimates of age-associated cancer risk and survival in the context of constitutional mismatch repair system deficiency (CMMRD syndrome): report from the European database of the European consortium C4CMMRD

Julie Robbe¹, Youenn Drouet², Lea Guerrini Rousseau³, Stéphanie Baert Desurmont⁴, Delphine Bonnet⁵, Franck Bourdeaut⁶, Karin Dahan⁷, Francoise Desseigne⁸, Christine Devalck⁹, Natacha Entz Werle¹⁰, Laurence Faivre¹¹, Maurizio Genuardi¹², Yael Goldberg¹³, Danuta Januszkiewicz Lewandowski¹⁴, Edita Kabickova¹⁵, Michaela Kuhlen¹⁶, Vanessa Perez Alonso¹⁷, Daniel Rueda¹⁸, Clara Ruiz Ponte¹⁹, Julie Tinat²⁰, Sheila Unger²¹, Anne Uyttebroeck²², Anja Wagner²³, Christine Lasset², Katharina Wimmer²⁴, Laurence Brugières²⁵, Chrystelle Colas ²⁶

¹Institut Curie, Départment de Génétique, Paris, France. ²Centre Léon Bérard, Département Prévention et Santé Publique, Lyon, F-69008, France; Université de Lyon, CNRS UMR 5558 LBBE, Villeurbanne, F-69622 France; Department of Pediatric and Adolescents Oncology, Gustave Roussy Cancer Center, Paris-Saclay University, Villejuif, France. ³Department of Pediatric and Adolescents Oncology, Gustave Roussy Cancer Center, Paris-Saclay University, Villejuif, France; Gustave Roussy Cancer Center, Unite Mixte de Recherche 8203, Centre National de la Recherche Scientifique, Paris-Saclay University, Villejuif, France. ⁴CHU Charles Nicolle, Rouen, France. ⁵CHU Purpan, Toulouse, France. ⁶Curie Institute, SIREDO Cancer Center (Care, innovation and research in pediatric, adolescents and young adults oncology), Paris, France. ⁷Hôpital Universitaire Reine Fabiola (HUDERF), Genetic department, Université Libre de Belgique (ULB), Brussels, Belgium. ⁸Centre Léon Bérard, Département Prévention et Santé Publique, Lyon, F-69008. ⁹Department of Hemato-Oncology, Hôpital Universitaire Reine Fabiola (HUDERF), Université Libre de Belgique (ULB). Brussels—Belgium. ¹⁰Pédiatrie Onco-Hématologie-Pédiatrie III-CHRU Hautepierre UdS-EA 3430, Strasbourg, France. ¹¹CHU Dijon Bourgogne, Dijon, France. ¹²Fondazione Policlinico Universitario A. Gemelli IRCCS, UOC Genetica Medica, Rome, Italy Instituto di Medicina Genomica, Università Cattolica del Sacro Cuore, Rome, Italy. ¹³Raphael Recanati Genetics Institute, Beilinson Hospital, Rabin Medical Center, Petach Tikva, Israel. ¹⁴Poznań University of Medical Sciences, Poznań, Poland. ¹⁵University Hospital, Prague, Czech Republic. ¹⁶Department of Pediatric Oncology, University Children´s Hospital, Hematology and Clinical Immunology Duesseldorf, Germany. ¹⁷Hospital Universitario Doce de Octubre, Unidad de Oncología Pediátrica, Madrid, Spain. ¹⁸Hospital 12 de Octubre, Madrid, Spain. ¹⁹Fundación Pública Galega de Medicina Xenómica (FPGMX) SERGAS, Grupo de Medicina Xenómica, IDIS, Centro de Investigación Biomédica en Red de Enfermedades Raras (CIBERer), Santiago de Compostela, Spain. ²⁰CHU Hôpitaux de Bordeaux, Bordeaux, France. ²¹Centre Hospitalier Universitaire Vaudois, Division of Genetic Medicine, University of Lausanne, Lausanne Switzerland. ²²Department of Pediatric Oncology, UZ Leuven, Leuven, Belgium; Department of Oncology, KU Leuven, Leuven, Belgium. ²³Department of Clinical Genetics, Erasmus MC Cancer Institute, University Medical Center Rotterdam, 3000CA Rotterdam, The Netherlands. ²⁴Institute of Human Genetics, Medical University Innsbruck, Innsbruck, Austria. ²⁵Department of Pediatric and Adolescents Oncology, Gustave Roussy Cancer Center, Paris-Saclay University, Villejuif, France. ²⁶Curie Institute, Genetic Department, Paris, France, INSERM U830, Paris, France / Université de Paris, France

Background and aim Constitutional Mismatch Repair Deficiency (CMMRD) is a rare childhood cancer predisposition syndrome due to biallelic germline pathogenic variants in one of the MMR genes. Patients affected with this syndrome have a high risk of developing early onset tumors such as hematological malignancies, brain tumors or Lynch syndrome associated malignancies. Although improvement of the clinical description of CMMRD, due to the scarcity of the disease, age-associated cancer risk and survival remain largely unknown. The aim of this study is to estimate the tumor risks in the context of CMMRD syndrome and to explorer genotype–phenotype correlations.

Method Data of 100 CMMRD patients were extracted from the C4CMMRD European consortium database. Overall survival and cumulative risks (CR) of cancers were estimated using Kaplan–Meier method. Subgroup analyses were conducted according to the mutated gene and the variant type.

Results These 100 patients developed 178 tumors of which 40% were brain tumors, 31% hematological malignancies and 26% Lynch syndrome related tumors. PMS2 is the main gene involved followed by MSH6, MSH2 and MLH1. The tumor spectrum varies according to the mutated gene. The median age of onset of the first tumor is 7 years (range: 0.1- 33.6) and is significatively different according to the germline mutated gene (p < 0.001) as is the cumulative risk of any tumor (p < 0.001). The highest risk is observed for biallelic MLH1 patients with 75% CR at 5 years of age. The median overall survival was 13.2 years with 71% of the patients deceased at last news. Survival was highly correlated to the mutated gene (p < 0,001): median age at death was 6y for MLH1, 7y for MSH2, 13y for MSH6 and 20y for PMS2 biallelic carriers.

Conclusions This is the largest reported series of CMMRD patients, providing the first estimates of age-associated cancer risk and survival to date. Although these estimates are likely to be upward-biased due to the selection of patients based on their personal and familial history, these data provide evidence-based arguments for a more accurate management of CMMRD patients according to their genotype. International collaboration remains crucial to obtained unbiased estimates using methods as the Genotype Restricted Likelihood and improve knowledge on this rare and severe disease.

Keywords CMMRD, tumor risk, penetrance, survival, MMR genes, C4CMMRD.

OP-052: Genetics: Research, Germline, Genomics, Variomics

The burden of microsatellite instability in the blood of constitutional mismatch repair deficiency syndrome is associated with patient genotype but not age of tumour onset

Richard Gallon ¹, Christine Hayes¹, Rachel Phelps¹, Laurence Brugieres², Lea Guerrini Rousseau², Chrystelle Colas³, Martine Muleris⁴, Neil Ryan⁵, D. Gareth Evans⁶, Katharina Wimmer⁷, John Burn¹, Michael S. Jackson¹, Mauro Santibanez Koref¹

¹Faculty of Medical Sciences, Newcastle University, Newcastle upon Tyne, UK. ²Child and Adolescent Cancer Department, Gustave Roussy Cancer Campus, Villejuif, France. ³Department of Genetics, Institut Curie, Université de Recherche Paris Sciences et Lettres, Paris, France. ⁴Sorbonne Université, Inserm, Unité Mixte de Recherche Scientifique 938, Equipe Instabilité Des Microsatellites et Cancer Centre de Recherche Saint-Antoine, CRSA, Paris, France. ⁵The Academic Women’s Health Unit, Translational Health Sciences, Bristol Medical School, University of Bristol, Bristol, UK. ⁶Division of Evolution, Infection and Genomics, University of Manchester, Manchester, UK. ⁷Division of Human Genetics, Medical University of Innsbruck, Innsbruck, Austria

Background and aim Constitutional mismatch repair deficiency (CMMRD) is a very rare recessive childhood cancer syndrome caused by inactivating germline variants in mismatch repair (MMR) genes. Microsatellite instability (MSI) in non-neoplastic tissues is a CMMRD diagnostic hallmark. Investigations of associations between constitutional MSI-burden and a patient’s genotype and phenotype have been limited by methodology and cohort size. We explored these associations by quantifying blood-MSI in a relatively large series of CMMRD patients, using novel MSI markers selected for instability in blood rather than tumours.

Method Three CMMRD, one Lynch syndrome, and two control bloods were genome sequenced to > 120 × depth. Microsatellite amplicon sequencing to 5000 × depth was used to analyse a pilot cohort of eight CMMRD and 38 control bloods, a blinded cohort of 56 CMMRD and 43 control bloods, and 80 reference control bloods. For each sample, microsatellite reference allele frequencies (RAFs) were compared to RAFs of the 80 reference controls. A blood-MSI score was calculated using –log10 of the combined RAF p-value across the MSI markers. A higher blood-MSI score indicates increased MSI relative to the controls.

Results Thirty-two novel mononucleotide repeats were selected for their instability in blood using genome sequence and pilot amplicon sequence data. MSI analysis of the blinded cohort with these novel blood-MSI markers, plus 80 reference controls for MSI scoring, gave complete discrimination between individuals with and without CMMRD, with a larger MSI score separation than with previously used tumour-derived markers. Blood-MSI scores were reproducible (R = 0.994, p < 10^-15). Lower scores were associated with MSH6 variants (p = 0.001) and variants classified as VUS by InSiGHT and/or ClinVar databases (p = 0.0024), but did not correlate with age at first tumour (p = 0.287) or age at sample collection (p = 0.100).

Conclusions Constitutional MSI-burden is associated with CMMRD patients’ genotypes, but no phenotype association was found. The new blood-MSI markers enhance microsatellite sequencing as a diagnostic assay for CMMRD. The method is cheap and scalable, costing less than $100 per sample and allowing 3–80 samples per run depending on MiSeq kit used.

Keywords Constitutional mismatch repair deficiency; Paediatric cancer; Diagnosis; Genotype phenotype correlation.

OP-053: Early Onset Colorectal Cance

Genetic Anticipation in MLH1-associated Lynch syndrome

Arti Sharma Pandey ¹, Christine Drogan², Dezheng Huo³, Kristen Postula¹, Shreshtha Madaan Garg¹, Sonia S. Kupfer²

¹Feinberg School of Medicine, Northwestern University, Chicago, Illinois,USA. ²Gastroenterology, Hepatology and Nutrition, University of Chicago, Illinois, USA. ³Department of Public Health Sciences, University of Chicago, Illinois, USA

Background and aim Genetic anticipation, defined as earlier age of onset of an inherited disease and/or an increase in severity in successive generations, is controversial in Lynch syndrome. Most studies have reported an anticipation of 2 to 11 years, while a few have found no differences in age of cancer diagnosis in subsequent generations. Statistical estimation of anticipation effects is subject to biases such as ascertainment bias, testing bias, right truncation effects and secular effects associated with birth year. The aim of this study was to determine anticipation effects in MLH1-associated Lynch syndrome controlling for several biases.

Method Pedigree data from families with MLH1 pathogenic variants and at least 2 first degree relatives with cancer seen at an academic institution between 1992–2021 were collected. A Cox-proportional hazard regression model to determine hazard ratios for colorectal or endometrial cancer between generations using clinical data on families with (MLH1 pathogenic variants. Crude anticipation effects with Wilcox signed rank test using parent–child-pairs were also determined.

Results A total of 703 members from 31 MLH1 families including 62 parent–child pairs were included. Four generations were defined from the data with respect to the probands. At a given age, the hazard ratio of cancer diagnosis the grandparents, parents and probands were lower by 81%, 67% and 65% respectively, as compared to the children’s generation after correction of biases and adjusting for birth-cohort effects (Fig. 1). Wilcox test showed a decrease of 5.5 years in ages of diagnosis of colorectal and endometrial cancers in children compared to parents, which stayed significant for parent-children-pairs where children were born after 1945 (Table 1). Hazard ratio for colorectal or endometrial cancers was found to be 35% lower for mutations that retained mismatch repair activity (Fig. 1) compared to those that resulted in an inactivation of the Mlh1 protein.

Conclusions Anticipation in families with MLH1-associated Lynch syndrome were observed after correcting for ascertainment bias, testing bias, right truncation, and birth-cohort effects. Validation of these findings with larger sample sizes is needed to determine implications for genetic counseling and management, though importance of screening initiation in MLH1 carriers in young adulthood per current guidelines is underscored. The effect of mutation type on mismatch repair activity may be associated with anticipation and warrants further investigation.

Keywords genetic anticipation, Lynch syndrome, MLH1.

Fig. 1

Forest plot of Hazard ratios with confidence intervals from Cox-proportional hazards regression analysis

Table 1

Parent–Child-Pairs	N	Median difference (95% CI)	p-value
All parent–child-pairs	62	5.499 (1.000,10.000)	0.013
Children born before 1945	26	1.499 (-4.499,7.500)	0.663
Children born after 1945	36	8.500 (2.499,16.500)	0.008

Wilcoxon paired signed rank test for differences in age of diagnosis of affected parents and their affected children

OP-054: Clinical: Diagnosis, Endoscopy, Management, Implementation

Endoscopic surveillance in Lynch syndrome, a prospective study of interval cancer and individual risk factors

Nigin Jamizadeh, Sophie Walton Bernstedt, Adrianna Haxhijaj, Anna Andreasson, Anna Forsberg, Ann Sofie Backman

Department of Medicine, Karolinska Institute, Stockholm, Sweden

Background and aim A hereditary component is involved in 20–25% of all colorectal cancers (CRC). Lynch syndrome (LS) is a genetic condition that increases the risk of early cancer, CRC being the most common type. Regular colonoscopies are recommended from age 25 in LS to detect CRC early. The optimal length of the interval between colonoscopies is debated. In Stockholm, the recommendation has been 12 months, but ESGE now recommends 24 months. This study aimed to describe numbers needed to scope to find a CRC, and to identify risk factors for CRC amongst LS patients.

Method Clinical data and colonoscopy findings on patients with LS in Stockholm area from June 1994 to April 2021 were registered prospectively in a structured protocol. Associations between clinical data and finding of CRC was analysed with logistic regression.

Results A total of 1831 endoscopies amongst 325 patients were analysed. 99 patients had at least one CRC during the study period. During surveillance, 28 patients were diagnosed with CRC. The median age of patients with CRC detected before surveillance (symptom-initiated colonoscopy, n = 71) was 45 years (23–75), and 49 years (22–79) if diagnosed under surveillance/index (n = 28).

A CRC was detected in 2/100 LS surveillance colonoscopies. In 57%. the CRC was detected within 18 months and 76% within 24 months. The mean interval for detecting a CRC was 24.34 months (3.07–70.83) (Fig. 1). Smoking, both ongoing (P = 0.016), and previous (P = 0.004) were associated with a higher risk of CRC. Higher BMI (P = 0.007) and being male (P = 0.014) were associated with CRC.

Conclusions Patients included in surveillance were diagnosed with CRC at an older age than those detected by symptoms. Individual risk factors for developing CRC were smoking, high BMI and male sex. The majority of CRC cases were detected within 24 months from a normal colonoscopy. Surveillance seems to delay CRC, but it is unclear whether the recommended interval by ESGE is enough to prevent CRC.

Keywords Lynch, surveillance, colorectal cancer.

Fig. 1

Histogram of surveillance interval to detection of one CRC. Intervals since a clean colonoscopy is shown in the figure. Out of 28 CRC cases diagnosed during surveillance, 17 CRC were found during surveillance. 9 were found at index and 2 had an interval < 3 months

OP-055: Clinical: Diagnosis, Endoscopy, Management, Implementation

Real-time use of artificial intelligence in colorectal cancer surveillance of patients with Lynch syndrome – a randomized controlled trial

Robert Hüneburg ¹, Karolin Bucksch², Friederike Schmeißer¹, Dominik Heling¹, Tim Marwitz¹, Dominik J. Kaczmarek¹, Stefan Aretz³, Oliver Hommerding⁴, Glen Kristiansen⁴, Christian P Strassburg¹, Christoph Engel², Jacob Nattermann¹

¹National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Bonn, Germany; Department of Internal Medicine I, University Hospital Bonn, Bonn, Germany. ²Institute for Medical Informatics, Statistics and Epidemiology, University of Leipzig, Leipzig, Germany. ³Institute of Human Genetics, University of Bonn, Bonn, Germany. ⁴Institute of Pathology, University Hospital Bonn, Bonn, Germany

Background and aim Lynch syndrome (LS), an autosomal dominant disorder caused by pathogenic germline variants in DNA mismatch repair (MMR) genes, represents the most common hereditary colorectal cancer (CRC) syndrome. LS patients are at high risk of CRC despite regular endoscopic surveillance. Quality of colonoscopy depends on the endoscopist, procedural factors and the lesion itself. Thus, a significant number of adenomas are still missed during colonoscopy. Artificial intelligence (AI)-assisted colonoscopy may help to increase adenoma detection rate (ADR). We here assessed for the first time the diagnostic performance of AI-assisted colonoscopy in LS in comparison to HD white-light endoscopy (WLE). The study was approved by the local ethics committee.

488/20] and registered on the German Clinical Trials Register (DRKS00023157).

Method Patients ≥ 18 years, with pathogenic germline variant (MLH1, MHS2, MSH6), and at least one previous colonoscopy (interval 10–36 months) were eligible for this prospective randomized controlled trial [DRKS00023157]. Exclusion criteria were (sub-)total colectomy and incomplete procedure. Patients were stratified by previous colorectal surgery and affected MMR gene with a 1:1 allocation ratio (AI-assisted vs. standard WLE). The primary objective was to explore ADR. Secondary objectives included number of detected adenomas, detection of sessile serrated lesions (SSL) and withdrawal time. High-definition technology was used for all examinations with an activated AI system for polyp real-time detection (ELUXEO 7000 system; EW10-EC02 CAD-EYE; Fujifilm, Japan) in the AI arm.

Results Between Dec-2021 and Dec-2022, 101 patients were enrolled. After excluding five patients due to insufficient bowel preparation, the final analysis included 96 patients (46 WLE/ 50 AI WLE). The cohorts did not differ significantly (Table 1). In the WLE arm, adenomas were detected in 12/46 patients compared to 18/50 in the AI arm (26.1% [95% CI 14.3–41.1] vs. 36.0% [22.9–50.8]; p = 0.379) with a higher mean number of adenomas identified (0.43 vs 0.6) using AI (Table 2). The increase in ADR was due to improved identification of flat adenomas (Fig. 1). By WLE, 4/20 flat adenomas compared to 17/30 in the AI arm (p = 0.018) were detected, the number of examinations with detection of flat adenomas was substantially higher in the AI arm (3/46 [6.5%] vs. 10/50 [20%]; p = 0.07). Among the flat adenomas found in AI, 3 were advanced adenomas. Total number of detected SSLs (5 vs. 8) and the number of examinations with SSL (4/46 [8.7%] vs 7/50 [14%]; p = 0.5) tended to be higher in the AI group. Two CRCs were detected in one patient (AI group).

Conclusions We here present first data suggesting that real-time AI-assisted colonoscopy is a promising approach to optimize endoscopic surveillance of Lynch patients, in particular to improve the detection of flat adenomas.

Keywords Lynch syndrome, artificial intelligence, endoscopy, adenoma detection rate.

Fig. 1

Endoscopic appearance of detected flat adenoma using artificial intelligence.

Table 1 Baseline characteristics

OP-056: Epidemiology and Registries

Risk-stratified FIT for urgent colonoscopy in Lynch Syndrome: A national clinical service throughout the COVID-19 pandemic

Anne Lincoln¹, Sally Benton², Peter Sasieni¹, Kevin Monahan ³

¹Comprehensive Cancer Centre, King's College London, London, United Kingdom. ²Bowel Cancer Screening South of England Hub, Royal Surrey County Hospital NHS Foundation Trust, Guildford, United Kingdom. ³The Lynch Syndrome Family Cancer Clinic, St Mark’s Hospital and Academic Institute, Harrow, United Kingdom; Department of Surgery & Cancer, Imperial College London, London, United Kingdom

Background and aim Lynch syndrome (LS) is an inherited disorder characterised by pathogenic variants within mismatch repair genes resulting in an increased risk of colorectal cancer (CRC). In England the faecal immunochemical test for haemoglobin (FIT) is used to guide colonoscopy in non-LS symptomatic and screening populations. We report results from an emergency clinical service implemented during the COVID-19 pandemic peak, with the use of FIT to prioritise colonoscopy in LS patients while endoscopy services were severely limited.

Method NHS services across England were invited to participate. Patient eligibility was determined by 1) Genetic diagnosis of Lynch Syndrome 2) Planned colonoscopic surveillance between 1 March 2020 and 31 March 2021. Requests for FIT testing from participating NHS Trusts were sent to the NHS Bowel Cancer Screening South of England Hub’s Research Laboratory. The Hub sent patients a FIT kit (OC-Sensor™ (Eiken, Japan)), FIT instructions, a questionnaire, and a pre-paid return envelope. Lab reports with faecal haemoglobin (f-Hb) results were returned electronically for clinical action. LS patients were risk-stratified for colonoscopy based upon the following f-Hb thresholds: (1) f-Hb ≥ 10 µg of Haemoglobin (Hb)/g (µg/g) faeces: triaged for colonoscopy via an urgent pathway, (2) f-Hb ≤ 10 µg/g: patients scheduled for colonoscopy within 6–12 weeks.

Results Fifteen English centres participated in the clinical service from 9 June 2020 to 31 March 2021. Uptake was 64% (375/588 invites). In 354 analysed cases, 269 (76%) had a f-Hb of < 6 µg/g. 6% (n = 23) of patients had a f-Hb between the limit of detection (≥ 6 µg/g) and < 10 µg/g, and 18% (n = 62) with FIT ≥ 10 µg/g and met criteria for urgent colonoscopy (table). Overall adenoma detection rate (ADR) was 46% (108/233) and there was effective escalation of colonoscopy in LS patients with a FIT above threshold (median 35 vs 107 days).

Conclusions The national Covid-19 response service demonstrated clinical value in improving access to colonoscopy for LS patients at highest prevalent CRC risk, Further longitudinal investigation on the efficacy of FIT in people with LS is required and will be examined as part of the study “FIT for Lynch Syndrome” (ISRCTN15740250) which is in active recruitment.

Keywords Lynch Syndrome, Bowel Cancer Surveillance, Colorectal Cancer Surveillance, Mismatch Repair, Cancer Epidemiology.

Outcomes of colonoscopy by f-Hb thresholds

Cut-off (µg/g)	Adenomas	Advanced Adenomas	CRC	No Abnormalities Detected (NAD)	Pending or missing colonoscopy	Total (%)
< 6	60	15	2	85	107	269(76)
6–9.9	4	1	0	10	8	23(6)
≥ 10	22	6	4	24	6	62(18)

OP-057: Early Onset Colorectal Cance

Sex Disparities in Metabolic Transcriptomic Features Identify Novel Biomarkers of Early-Onset Colorectal Cancer

Jose Perea ¹, Lara P Fernández², Gonzalo Colmenarejo³, Abel Martel¹, Marc Marti⁴, Rosario Vidal⁵, Fernando Jiménez⁶, Marta Jiménez Toscano⁷, Araceli Ballestero⁸, Elena Hurtado⁹, Francesc Balaguer¹⁰, Gonzalo Sanz¹¹, Edurne Álvaro¹², Alfredo Vivas¹³, Damián García Olmo¹⁴, Rogelio González Sarmiento¹, Ana Ramírez De Molina²

¹Department of Molecular Medicine, Institute of Biomedical Research of Salamanca, Salamanca, Spain. ²Molecular Oncology Group, IMDEA Food Institute, CEI UAM + CSIC, Madrid, Spain. ³Biostatistics and Bioinformatics Unit, IMDEA-Food Institute, CEI UAM + CSIC, Madrid, Spain. ⁴Department of Surgery. Vall d´Hebron University Hospital, Barcelona, Spain. ⁵Department of Oncology. Universtiy Hospital of Salamanca, Salamanca, Spain. ⁶Department of Surgery. Hospital de Galdacano-Usansolo. Galdacano, Spain.

⁷Department of Surgery. Hospital del Mar, Barcelona, Spain. ⁸Department of Surgery. Ramon y Cajal University Hospital, Madrid, Spain. ⁹Department of Surgery. Gregorio Marañon University Hospital, Madrid, Spain. ¹⁰Department of Gastroenterology. Hospital Clinic of Barcelona, Barcelona, Spain. ¹¹Department of Surgery. San Carlos University Hospital. Madrid, Spain. ¹²Department of Surgery. Infanta Leonor University Hospital. Madrid, Spain. ¹³Department of Surgery. 12 de Octubre University Hospital. Madird, Spain. ¹⁴Department of Surgery. Fundación Jiménez Díaz University Hospital. Madrid, Spain

Background and aim Early-onset colorectal cancer (EOCRC) incidence rates are increasing nowadays. Although little is known about the biological mechanisms responsible of EOCRC, it has recently been observed putative disparities by sex what suggests that metabolic dysregulation, which also debuts with particularities among men and women, may contribute to the development and particularities of EOCRC. We aimed to identify metabolic sex disparities in EOCRC, easily detectable as molecular biomarkers of risk, so we conducted a plasma expression profile analysis of metabolic health related genes.

Method Fifty EOCRC patients from the Spanish Early-onset Colorectal cancer Cohort Consortium (SECOC) were included in the study using blood, collected previously to any treatment, and 30 controls of cancer-free individuals. We analyzed peripheral blood expression of 23 genes belonging to five main metabolic health related pathways: Lipid metabolism (LM), energetic metabolism (EM), immune system and inflammation (IS), insulin responses (IR) and gene-nutrient-environmental interactions (GNE). Then, we searched for sex-specific non-invasive biomarkers of EOCRC risk, and metabolic genetic risk scores (GRS) based on gene expression were developed. The models were validated through bootstrap resampling (500 resamples) and using the optimism-corrected c-index as measure of predictive capacity.

Results A total of 17 genes were significantly upregulated among men with EOCRC: ABCA1, CD36, CLOCK, CPT1A, FASN, FLCN, FNIP2, INFg, IL6, INSR, JAK1, LDLR, MTHFD, PPARg, TFAM, TNFa and LRP1B. Conversely, the expression of two genes was exclusively associated to EOCRC in women: Prostaglandin-endoperoxide synthase 2 (PTGS2, also known as COX2), that belongs to IS pathway (beta = -1.11 95%CI (-2.28, -0.452), p = 0.001) and glucose transporter SLC2A4 also known as GLUT4 (IR pathway) (beta = 2.37 95%CI (0.687, 5.16), p = 0.01). PTGS2 is the unique significant gene whose expression is downregulated among cases. Only one gene, low-density lipoprotein receptor-related protein 1B (LRP1B) was upregulated in both men and women (beta = 1.62 95%CI (0.674, 2.99), p = 0.003) and beta = 2.3 95%CI (0.837, 4.91), p = 0.002), respectively). EOCRC was associated with an upregulation of all metabolic health related pathways in men patients and among them, a GRS considering 2 genes: LDLR and CPT1A predicted CRC development with a c-index of 0.991. By contrast, women with EOCRC characteristically exhibited upregulation of IR routes and downregulation of immune responses. In women, PTGS2 expression was sufficient to predict EOCRC development (c-index = 0.896).

Conclusions In summary, we described by first time clearly distinct metabolic scenarios according to sex influencing EOCRC development. While EOCRC in men is characterized with a total deregulation of metabolic landscape, EOCRC in women is focused on immune system and insulin responses routes.

Keywords Early-onset colorectal cancer, sex disparities, metabolic pathways, transcriptomic.

OP-058: Early Onset Colorectal Cance

Risk factors associated with young onset advanced neoplasia at baseline and on follow up colonoscopy

Nataly Torrejon, Ruishen Lyu, David Liska, Carol Burke, Carole Macaron.

Cleveland Clinic, Cleveland, OH

Background and aim Risk factors associated with young onset advanced colorectal neoplasia remain unclear. Understanding the drivers for early onset neoplasia in this age group may assist with targeted screening approaches.

AIM: to identify risk factors associated with advanced neoplasia in young adults at baseline colonoscopy and on subsequent exams.

Method Patients ages 18 to 49 who underwent a baseline colonoscopy with polypectomy at Cleveland Clinic between 2011 and 2021 were included in the analysis. Patients with a history of inflammatory bowel disease and inherited colon cancer syndromes were excluded. The primary outcome was defined as the presence of advanced neoplasia (tubular adenomas > 9 mm, adenoma with any villous histology or high-grade dysplasia, sessile serrated polyp (SSP) > 9 mm or SSP with dysplasia, traditional serrated adenomas) at baseline colonoscopy. Rectal bleeding or rectal mass, iron deficiency anemia, weight loss, family history of colorectal cancer (first and second degree relative), family history of polyp, positive stool-based screening test and abnormal imaging were considered high risk indications for colonoscopy. We examined the following risk factors: age, gender, race, BMI, alcohol and tobacco use, history of diabetes or hyperlipidemia, NSAIDs and statin use, and family history of colon cancer. A multivariate binary logistic regression model assessed the association between risk factors and advanced neoplasia at baseline. To identify factors associated with metachronous advanced neoplasia, unadjusted Cox PH regression models were performed.

Results Of the 3022 patients included in the study, 714 (23.6%) patients had advanced neoplasia at baseline. Young adults with advanced neoplasia were older (median age 45.5 vs 44.4, p = 0.001), had tobacco exposure (49.8% vs 45.2%, p = 0.034), and had high risk indications for colonoscopy (63% vs 54.6%, p < 0.001), other demographics are shown in Table 2. 145 patients had at least 1 subsequent follow up colonoscopy. On univariate analysis, only a history of prior advanced neoplasia was a predictor for recurrent advanced neoplasia (HR = 2.09, 95% CI 1.05–4.1, p = 0.034).

Table 2 Colonoscopy characteristics and results

Table 1 Summary of PDAC cases (n = 36) diagnosed in 31 patients while participating in pancreatic surveillance

Conclusions Increasing age, exposure to tobacco, high-risk indications for colonoscopy, but not BMI, diabetes or hyperlipidemia, were associated with advanced neoplasia on baseline colonoscopy in young adults. The results of this study are consistent with current guidelines recommending initiating screening at an earlier age and call for particular attention to young adults exposed to tobacco.

Keywords young onset adenoma, risk factors, colonoscopy.

Table 1

Table 2

OP-059: Early Onset Colorectal Cance

Familial Component of Early-Onset Colorectal Cancer: An Opportunity for Prevention

Maria De Los Angeles Daca Alvarez¹, Marc Martí², Antonino Spinelli³, Noel De Miranada⁴, Claire Palles⁵, Alfredo Vivas⁶, Andrew Lachtford⁷, Kevin Monahan⁸, Marek Szczepkowski⁹, Wieslaw Tarnowski¹⁰, Silviu Tiberiu Makkai Popa¹¹, Rosario Vidal Tocino¹², Irene Lopez¹³, Elena Hurtado¹⁴, Fernando Jiménez¹⁵, Marta Jiménez Toscano¹⁶, Araceli Ballestero¹⁷, Teresa Ocaña¹, Lorena Moreno¹, Rebeca Moreira¹, Sabela Carballal¹, Leticia Moreira¹, Maria Pellisé¹, Rogelio Gonzalez Sarmiento¹⁸, Andreana N. Holowatyj¹⁹, José Perea¹⁸, Francesc Balaguer ¹

¹Department of Gastroenterology, Hospital Clínic de Barcelona,Barcelona, Spain; Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain; Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBEREHD), University of Barcelona, Barcelona, Spain. ²Vall d'Hebron University Hospital, Barcelona, Spain. ³Humanitas Research Hospital. Humanitas University, Rozzano, Lombardia, Italy. ⁴Department of Pathology. Leiden University Medical Center, Leiden, The Netherlands. ⁵Institute of Cancer and Genomic Sciences, University of Birmingham, Birmingham, UK. ⁶Hospital Universitario 12 de Octubre, Madrid, Spain. ⁷St. Mark's Hospital, London, UK. ⁸St. Mark's Hospital, London, UK.; West Middlesex University Hospital, London, UK. ⁹Clinical Department of Colorectal, General and Oncological Surgery, Centre of Postgraduate Medical Education, Bielanski Hospital, Warsaw, Poland. ¹⁰Centre of Postgraduate Medical Education, Orlowski Hospital, Warsaw, Poland. ¹¹Centre Hospitalier de Luxembourg, Luxembourg. ¹²Hospital Universitario de Salamanca, Salamanca, Spain. ¹³Hospital MD Anderson, Madrid, Spain. ¹⁴Hospital Universitario Gregorio Marañon, Madrid, Spain. ¹⁵Hospital Galdakao-Usansolo, Vizcaya, Spain. ¹⁶Hospital del Mar, Barcelona, Spain. ¹⁷Hospital Universitario Ramon y Cajal, Madrid, Spain. ¹⁸Molecular Medicine Unit- Department of Medicine. Biomedical Research Institute of Salamanca (IBSAL). Institute of Molecular and Cellular Biology of Cancer (IBMCC), University of Salamanca-SACYL-CSIC, Salamanca, Spain. Salamanca, Spain. ¹⁹Vanderbilt University Medical Center, Nashville, Tennessee, USA

Background and aim The incidence of colorectal cancer (CRC) in people under 50 years old is increasing without a clear established cause. Individuals with a non-syndromic family history of CRC are known to have an increased risk of this cancer and most guidelines recommend starting surveillance at age 40 with a 5-year interval. However, there is variability in the definition of familial risk. According to the European Society of Gastrointestinal Endoscopy (ESGE), this risk is considered to be clinically meaningful in those with at least one first-degree relative (FDR) diagnosed before the age of 50, and in cases with two or more FDRs with CRC. On the other hand, the U.S. Multi-Society Task Force (USMSTF) guideline defines familial risk as the presence of at least one FDR with CRC or advanced adenoma diagnosed before the age of 60, and in cases with two or more FDRs. Therefore, there is an opportunity to prevent early-onset CRC (EOCRC) in these cases. The component potentially preventable by early screening for FH in early-onset CRC is still uncertain, and a recent study conducted in the US suggests that up to 16% of cases could be potentially prevented if current screening guidelines were followed.. Therefore, there is an opportunity to prevent early-onset CRC (EOCRC) in this population. Our aim was to explore the proportion of preventable EOCRC due to family history of CRC.

Method A retrospective, descriptive, multicenter European study with patients enrolled in two different consortiums focused on EOCRC: GEOCODE (Global Early-Onset Colorectal Cancer Database) and SECOC (Spanish Early-onset Colorectal Cancer Cohort). The study population consisted of patients with non-polyposis CRC diagnosed before the age of 50. Exclusion criteria were: 1) known or suspected cancer predisposition syndromes; 2) inflammatory bowel disease; and 3) other histological diagnoses different from adenocarcinoma. A database was developed with patients diagnosed with EOCRC from January 2010 to December 2020. Baseline patient characteristics, tumor characteristics, and family history of CRC with age at diagnosis [from proband and FDR (children, siblings, parents)] and the degree of kinship (first, second and third degree) were collected. The prevalence of family history of CRC in the EOCRC population was analyzed by subgroups. We compared the impact on prevention and early-diagnosis of the European Society of Gastrointestinal Endoscopy (ESGE) and U.S. Multi-Society Task Force (USMSTF) guidelines. We defined a potentially preventable CRC if surveillance colonoscopy would have been performed > 5 years before the age of diagnosis of CRC, and early diagnosis if performed between 1 to 4 years before. A descriptive analysis was firstly carried out. Chi square test (Independence test) and Fisher's test (when necessary) were used considering a p-value less than 0.05 as statistically significant. An analysis of variance (ANOVA or Kruskall Wallis) was performed when the variables were quantitative. All analyses were performed using R statistical software.

Results A total of 903 EOCRC patients were included. All the characteristics of the cohort are found in Fig. 1. The mean age at diagnosis was 41.8 ± 6.5 years, with similar distribution by sex. Regarding tumor location, the majority of cancers were located in the rectum (39%) followed by the left colon (36%) and the right colon (25%), and 63% were diagnosed at an advanced stage (stages III and IV). Family history of CRC was noted in 275 (30.4%) cases. Criteria for familial CRC risk for ESGE and USMSTF guidelines were met in 6.3% and 9.4% of patients, respectively. Following ESGE guidelines CRC could have been potentially prevented in 41.1% and diagnosed earlier in 10.7% of cases. If surveillance had started 10 years before the youngest relative, we observed a significant increase in prevention (41.1% vs 55.36%; p = 0.01). Following USMSTF guidelines, 55.42% of cases could have been prevented and 14.46% early detected (Fig. 2).

Conclusions 6–9% of EOCRC patients fulfilled the familial CRC risk criteria. In this group, approximately 50% CRC could be prevented if early surveillance was initiated. Early and accurate identification of familial CRC risk and increase in the uptake of early colonoscopy are key to decrease familial EOCRC.

Keywords early-onset colorectal cancer, family history, surveillance, prevention.

Characteristics of the Cohort

Impact of the application of the ESGE and USMSTF guidelines on CRC prevention and early detection

OP-060: Clinical: Diagnosis, Endoscopy, Management, Implementation

Diagnostic Yield of Endoscopic Screening for Identification of Signet Ring Cell in Carriers of a Pathogenic Variant in CDH1: Single Center Experience

Margaret Omalley ¹, Lady Katherine Mejia Perez², Arjun Chatterjee⁴, Lisa Laguardia¹, Susan Milicia¹, Sarah Mcgee⁵, Joshua Sommovilla¹, Nicholas Smith¹, David Liska¹, Carole Macaron², Matthew Walsh³, Carol A Burke²

¹Department of Colorectal Surgery, Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, USA. ²Departments of Gastroenterology, Hepatology, Nutrition, Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, USA. ³Department of General Surgery, Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, USA. ⁴Department of Internal Medicine, Community Care Institute, Cleveland Clinic, Cleveland, USA. ⁵Department of Genetics, Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Genomic Medicine Institute, Cleveland Clinc, Cleveland, USA

Background and aim Pathogenic variants in CDH1 are associated with hereditary diffuse gastric cancer. Recommendations for prophylactic total gastrectomy (PTG) are established for carriers of pathogenic variants in CDH1. Pre-operative upper endoscopy (EGD) to detect overt cancer or occult signet ring cells (SRC) for all carriers and annual surveillance endoscopy in individuals deferring PTG is suggested. The yield of EGD detection of SRC varies from 9–60% even in expert hands. The Cambridge Protocol includes targeted biopsy of pale patches, other abnormalities, and a > 30 random biopsies in 5 gastric locations. A “Bethesda” protocol (targeted biopsy and 4 random biopsies from 22 locations) was compared to Cambridge protocol. SRC were detected preoperatively in 15% versus 36%, (p < 0.05) of patients respectively. Highly accurate detection of occult, early stage, SRC cancer is necessary for EGD surveillance to be a viable alternative for patients declining PTG. Our aim was to describe the diagnostic yield of EGD surveillance in our center for detection of SRC foci in asymptomatic CDH1 carriers using gastrectomy as the gold standard.

Method IRB-approved Cologene TM database was used to identify carriers of CDH1 pathogenic variants enrolled in the Jagelman Registry. Included patients were those with ≥ 1 EGDs. Demographics, personal and family history of cancer, endoscopic findings, and surgical histology was reviewed. Our EGD biopsy protocol evolved over time to include targeted biopsy and 7 random biopsies from 4 quadrants in 11 areas (cardia, pre-pylorus, antrum (distal, proximal), transition zone, body (proximal, mid, distal), fundus (proximal, mid, distal) under high-definition white light endoscopy. Preoperative detection of SRC was compared with gastrectomy specimens, when available, to calculate the sensitivity of the Weiss Center endoscopy protocol.

Results 61 carriers of CDH1 from 39 families were identified. 12 were excluded (presented with cancer, not in surveillance program). 49 patients with a mean age of 44.7-years, 68% female with 101 EGDs were included [Table 1]. 24 (49%) of patients underwent gastrectomy 3 without SRC on preoperative biopsy or gastrectomy specimen. Of the remaining 21, 16 had SRC detected preoperatively. The preoperative detection of SRC in the Weiss Center protocol was 81%. [Table Tab2]. 3/24 did not have SRC on the surgical specimen. 16/24 patients had SRC on pre-operative EGD. The SRC were located in the fundus, antrum, body and cardia equally.

Conclusions The Weiss Center protocol for hereditary diffuse gastric cancer surveillance for asymptomatic patients with CDH1 pathogenic variants shows a high yield for preoperative detection of SRC. The SRC are located throughout the stomach equally.

Keywords CDH1, EGD, Gastric Cancer, Gastrectomy, Signet Ring Cell.

Table 1. Demographic characteristics

Table 2. Gastrectomy characteristics

OP-061: Clinical: Diagnosis, Endoscopy, Management, Implementation

Endoscopic diagnosis of early signet ring cell cancer in Hereditary diffuse Gastric Cancer Syndrome patients: a 16-year longitudinal prospective study

Massimiliano Di Pietro ¹, Colin Yc Lee², Adriaan Olivier³, Anne Marie Lydon¹, Susan Richardson¹, Rebecca C. Fitzgerald¹

¹Department of Oncology, University of Cambridge, Cambridge, UK. ²Molecular Immunity Unit, Department of Medicine, University of Cambridge, Cambridge, UK. ³School of Clinical Medicine, University of Cambridge, Cambridge, UK

Background and aim Hereditary diffuse gastric cancer (HDGC) is a rare cancer syndrome with early-onset signet ring cell carcinoma (SRCC) linked to germline CDH1 pathogenetic variants. Prophylactic total gastrectomy (PTG) is the only definitive treatment, however, endoscopic surveillance has a role to inform the timing of PTGs via detection of microscopic SRCC foci. There is still debate on the optimal endoscopic protocol to perfom endoscopic surveillance, with particular concern to the usefulness of random sampling and advanced imaging modality. The aims of this study were: i. to assess the diagnostic value of random and targeted biopsies for detection of early SRCC; ii. to identify endoscopic criteria to predict histologic outcome on targeted biopsies.

Method We prospectively recruited 145 patients with a clinical diagnosis of HDGC, between 2005 and 2021. Patients underwent surveillance endoscopy with white light imaging and narrow band imaging (NBI) with optical magnification (85x). The procedure included at least 20 min of inspection time to assess mucosa thoroughly on different modalities and the degree of insufflation. Mucosal lesions, including pale areas, were assessed in detail using NBI magnification. The degree of suspiciousness of pale areas was scored (high/low) based on the presence of 5 endoscopic criteria, as follows: i. round shape; ii. well-demarcated borders; iii. irregular vasculature; iv., irregular micro-surface pattern; v. reproducibility of the lesion over dynamic visualization. After inspection, targeted biopsies were taken from any lesion identified, followed by Cambridge protocol biopsies including 2 biopsies in the pre-pyloric area, 4 in the antrum, 4 in the transitional zone, 6 in the corpus, 4 in the fundus and 4 in the cardia. We included in our analysis 512 surveillance endoscopies. We compared detection rate in random and targeted biopsies using PTG specimens as reference, where applicable. The accuracy of endoscopic diagnosis for prediction of SRCC findings was calculated. For comparing discrete variables of demographic characteristics, chi-square tests were applied to contingency tables where all observations were greater than 5, and Fischer’s exact test applied otherwise. For continuous variables, Shapiro–Wilk test was applied to assess for normality, followed by independent Mann–Whitney U test for non-parametric distributions. Kaplan–Meier analysis with log-rank test was applied to compare progression to SRCC in CDH1 + and CDH1– groups. P < 0.05 was considered statistically significant, for all comparisons.

Results 101 individuals (69.7%) had CDH1 pathogenic variants (CDH1 +) detected, and 44 had CDH1-variants of unknown significance or no pathogenic variant detected (CDH1-). Over a median number of 3 endoscopies, 58/145 (40%) patients were diagnosed with SRCC of which 91.4% were CDH1 + . CDH1 pathogenic variant was the only clinical parameter associated with risk of SRCC (P < 0.001). CDH1 + individuals were diagnosed with SRCC at a significantly higher rate, whereby 50% of diagnoses were made by 24 months since index endoscopy. The proportion of targeted biopsies with SRCC was significantly higher than that of random biopsies (8.67% vs 0.72%). Of the 53 patients diagnosed with SRCC and no macroscopic tumor at index endoscopy, 29 patients were first diagnosed from random biopsies (54.7%), 15 from targeted biopsies only (28.3%), and 6 had findings on both random and targeted biopsies (11.3%). Finally, 3 patients received SRCC diagnoses after undergoing PTG despite no previous positive endoscopic findings. The anatomical distribution of SRCC foci found by random biopsies more accurately reflected the distribution found on PTGs, due to low sensitivity in detecting suspicious lesions in the proximal stomach (Fig. 1). Of the 65 targeted biopsies with positive SRCC findings, 90.7% derived from pale areas. Applying the 5 endoscopic criteria for prediction of histology, the endoscopic diagnosis had a sensitivity of 67% and specificity of 90% for SRCC.

Conclusions Random biopsies promote early detection of cancerous lesions in HDGC and should be performed routinely during each surveillance endoscopy. Omitting random biopsies would lead to underdiagnosis of more than 50% of patients with early SRCC. We have identified 5 endoscopic criteria than have moderate sensitivity and high specificity for SRCC and can be used to discount lesions with low probability of cancer yield.

Keywords HDGC, signet ring cell cancer, endoscopy, pale area, CDH1, surveillance.

Fig. 1

Regional distribution of SRCC-positive biopsies or SRCC foci found on PTG specimens. For targeted biopsies (TBx) and random biopsies (RBx), each biopsy yielding SRCC foci was taken as a single positive finding (i.e. 68 and 86 positive biopsies respectively). For PTG, each gastric region where SRCC foci were found was taken as a single positive finding, to account for differences in pathology reporting.

OP-062: Epidemiology and Registries

First estimates of diffuse gastric cancer risks for carriers of CTNNA1 germline pathogenic variants

Marie Coudert¹, Youenn Drouet², Hélène Delhomelle³, Magali Svrcek⁴, Patrick Benusiglio⁵, Florence Coulet⁵, Dana Farengo Clark⁶, Bryson W Katona⁷, Liselotte P. Van Hest⁸, Lizet E. Van Der Kolk⁹, Annemieke Cats¹⁰, Jolanda M. Van Dieren¹⁰, Bita Nehoray¹², Thomas Slavin¹¹, Isabel Spier¹³, Robert Huneburg¹⁴, Silvana Lobo¹⁵, Carla Oliveira¹⁶, Lise Boussemart¹⁷, Jean Chiesa¹⁸, Mathias Swartz³, Lisa Golmard³, Anne M. Bouvier¹⁹, Valérie Bonadona²⁰, Dominique Stoppa Lyonnet²¹, Christine Lasset², Chrystelle Colas ²¹

¹Département de Génétique, Institut Curie, Paris, France; Département de prévention, ICO Paul Papin, Angers, France. ²Unité de Prévention et Epidémiologie génétique, Centre Léon Bérard, F-69008, Lyon, France; Université de Lyon, CNRS UMR 5558 LBBE, F-69622, Villeurbanne, France. ³Département de Génétique, Institut Curie, Paris, France; Full Member of the European Reference Network on Genetic Tumor Risk Syndromes (ERN GENTURIS—Project ID No 739547). ⁴Département de Pathologie, Sorbonne Université, AP-HP, Hôpital Saint-Antoine, Paris, France. ⁵Département de Génétique Médicale, Sorbonne Université, AP-HP, Hôpital Pitié-Salpêtrière, F-75013, Paris, France. ⁶Department of Medicine, Division of Hematology and Oncology, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA, USA. ⁷Department of Medicine, Division of Gastroenterology and Hepatology, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA, USA. ⁸Department of Clinical Genetics, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands. ⁹Family Cancer Clinic, The Netherlands Cancer Institute, Amsterdam, The Netherlands. ¹⁰Department of Gastrointestinal oncology, The Netherlands Cancer Institute, Amsterdam, The Netherlands. ¹¹Division of Clinical Cancer Genomics, City of Hope, Duarte, CA, USA; Departments of Medical Oncology and Population Sciences, City of Hope, Duarte, CA, USA. ¹²Departments of Medical Oncology and Population Sciences, City of Hope, Duarte, CA, USA. ¹³Institute of Human Genetics, National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Bonn, Germany; Full Member of the European Reference Network on Genetic Tumor Risk Syndromes (ERN GENTURIS—Project ID No 739547). ¹⁴Department of Internal Medicine, National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Bonn, Germany; Full Member of the European Reference Network on Genetic Tumor Risk Syndromes (ERN GENTURIS—Project ID No 739547). ¹⁵i3S—Instituto de Investigação e Inovação em Saúde, Porto, Portugal; IPATIMUP—Institute of Molecular Pathology and Immunology of the University of Porto, Porto, Portugal; International Doctoral Programme in Molecular and Cellular Biotechnology Applied to Health Sciences (BiotechHealth) from Institute of Biomedical Sciences Abel Salazar (ICBAS), University of Porto, Porto, Portugal. ¹⁶i3S—Instituto de Investigação e Inovação em Saúde, Porto, Portugal; IPATIMUP—Institute of Molecular Pathology and Immunology of the University of Porto, Porto, Portugal; FMUP—Faculty of Medicine of the University of Porto, Porto, Portugal;Full Member of the European Reference Network on Genetic Tumor Risk Syndromes (ERN GENTURIS—Project ID No 739547). ¹⁷Département de Dermatologie, Université de Nantes, Nantes, France. ¹⁸Département de Génétique, CHRU Hôpital Caremeau, Nîmes, France. ¹⁹Digestive Cancer Registry of Burgundy, University Hospital of Dijon, INSERM U1231, French Network of Cancer Registries (FRANCIM), Dijon, France. ²⁰Université de Lyon, CNRS UMR 5558 LBBE, F-69622, Villeurbanne, France; Unité Clinique d'Oncologie Génétique, Centre Léon-Bérard, F-69008, Lyon, France. ²¹Département de Génétique, Institut Curie, Paris, France; INSERM U830, Université de Paris, Paris, France; Full Member of the European Reference Network on Genetic Tumor Risk Syndromes (ERN GENTURIS—Project ID No 739547)

Background and aim Pathogenic variants (PV) of CTNNA1 are found in families evocative of hereditary diffuse gastric cancer (DGC) but no risk estimates were available until now. The aim of this study is to evaluate DGC risks for carriers of germline CTNNA1 PV.

Method Data from published CTNNA1 families were updated and new families were identified through international collaborations. The cumulative risk of DGC by age for PV carriers was estimated with the Genotype Restricted Likelihood (GRL) method, taking into account non-genotyped individuals and conditioning on all observed phenotypes and genotype of the index case to obtain unbiased estimates. A non-parametric and the Weibull functions were used to model the shape of penetrance function with the GRL. Kaplan–Meier incidence curve and Standardized Incidence Ratios (SIR) were also computed. A “leave-one-out” strategy was used to evaluate estimate uncertainty.

Results Thirteen families with 46 carriers of PV were included. The cumulative risks of DGC at 80 years for carriers of CTNNA1 PV are 49%, 57% and 77% respectively with the Weibull GRL, NP GRL and Kaplan Meier methods. Risk ratios to population incidence reach particularly high values at early ages and decrease with age. At 40 years, they are equal to 65, 833 and 21,574 respectively with the Weibull GRL, NP GRL and SIR methods.

Conclusions This is the largest series of CTNNA1 families that provides the first risk estimates of GC. These data will help to improve management and surveillance for these patients and support inclusion of CTNNA1 in germline testing panels.

Keywords gastric cancer, genetic, HDGC, CTNNA1, risks.

OP-063: Genetics: Research, Germline, Genomics, Variomics

Outcomes of universal genetic testing in diverse pancreatic ductal adenocarcinoma patients

Christine M. Drogan ¹, Hedy L. Kindler¹, Guimin Gao², Sonia S. Kupfer¹

¹Department of Medicine, University of Chicago, Chicago, United States. ²Department of Public Health Sciences, University of Chicago, Chicago, United States

Background and aim Guidelines recommend all pancreatic ductal adenocarcinoma (PDAC) patients undergo germline genetic testing. Rates of recommendation and completion of genetic testing among diverse PDAC patients are not known. The aim was to determine rates of recommendation and completion of genetic testing in diverse PDAC patients.

Method A retrospective review of PDAC patients seen at an academic center between 4/2019 and 12/2020 was performed. Recommendation, completion and results of genetic testing as well as demographic and clinical factors were recorded. Univariate and multivariate analyses of genetic testing were performed using chi-square and logistic regression.

Results In total, 580 PDAC patients were included. Median age at diagnosis was 67 years; 52% were male; 63% non-Hispanic Whites (NHW) and 20% African Americans (AA). Genetic testing was performed in 216 (37%) patients. Of those tested, 47 (22%) had a pathogenic/likely pathogenic variant identified of which 25 (12%) were in PDAC-associated genes. On multivariate analysis, age, NHW race, personal and family cancer history and medical oncology visit were independent predictors of genetic testing completion (Table 1). AA patients had significantly lower rates of recommendation and completion of genetic testing compared to NHW patients (Fig. 1).

Conclusions Genetic testing in PDAC patients is unacceptably low, especially among AA patients. Testing disparity might be due to lack of provider recommendation more than patient uptake. Lack of testing leads to missed opportunities for potential targeted therapies, improved outcomes and identification of at-risk family members who could potentially benefit from surveillance.

Keywords pancreatic ductal adenocarcinoma, genetic testing, diversity, disparity.

Fig. 1

Rates of (A) genetic testing, (B) genetic testing completion rate when recommended, and (C) documented provider recommendation in AA and NHW PDAC patients. *p < 0.0001, **p = 0.08

Table 1

Data are median (IQR) or n (%) Proportion of primary PDAC cases was calculated from the total study population (n = 347) Proportions of secondary PDAC cases were calculated from the total number of primary PDAC cases (n = 31) Proportion of male was calculated from the total study population (n = 347).

OP-064: Epidemiology and Registries

Pancreatic Cancer Surveillance in CDKN2A Mutation Carriers: Yield and Outcomes of 20 Years Prospective Follow-up

Derk Klatte¹, Bas Boekestijn², Martin Wasser², Shirin Shahbazi Feshtali², Isaura Ibrahim¹, Sven Mieog³, Saskia Luelmo⁴, Hans Morreau⁵, Thomas Potjer⁶, Akin Inderson¹, Jurjen Boonstra¹, Friedo Dekker⁷, Hans Vasen¹, Jeanin Van Hooft¹, Bert Bonsing³, Monique Van Leerdam ⁸

¹Department of Gastroenterology and Hepatology, Leiden University Medical Center, Leiden, The Netherlands. ²Department of Radiology, Leiden University Medical Center, Leiden, The Netherlands. ³Department of Surgery, Leiden University Medical Center, Leiden, The Netherlands. ⁴Department of Oncology, Leiden University Medical Center, Leiden, The Netherlands. ⁵Department of Pathology, Leiden University Medical Center, Leiden, The Netherlands. ⁶Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands. ⁷Department of Clinical Epidemiology, Leiden University Medical Center, The Netherlands. ⁸Department of Gastroenterology and Hepatology, Leiden University Medical Center, Leiden, The Netherlands & Department of Gastrointestinal Oncology, Netherlands Cancer Institute, Amsterdam, The Netherlands

Background and aim Pancreatic surveillance in high-risk individuals may lead to detection of pancreatic ductal adenocarcinoma (PDAC) at an earlier stage with improved survival. In this study, we evaluated the yield and outcomes of 20-years prospective surveillance of a large cohort of CDKN2A mutation carriers.

Method Since 2000, mutation carriers were enrolled at age of 45 or 10 years prior to the youngest age of familial onset. Surveillance consisted of annual MRI/MRCP and optional EUS. Upon detection of suspicious lesions, histological sampling was attempted and/or the surveillance interval was shortened. Suspected cases of PDAC were discussed in a multidisciplinary team in which surgical resection was considered. Kaplan–Meier analysis was used to estimate cumulative incidence and survival.

Results 347 CDKN2A mutation carriers participated in surveillance and were followed for a median of 5.6 (IQR 2.3–9.9) years. A total of 36 cases of PDAC were diagnosed in 31 (8.9%) patients at a median age of 60.4 (IQR 51.3–64.1) years. The cumulative incidence of primary PDAC was 21% by the age of 70. Five carriers (16%) were diagnosed with a second primary PDAC. Twenty-nine of 36 PDACs (81%) were radiologically resectable at the time of diagnosis. Twelve cases (33%) presented with stage I disease. The median survival after diagnosis of primary PDAC was 26.8 months, and the 5-year survival rate was 32% (95% CI, 19%–55%). Individuals with primary PDAC who underwent resection (n = 22) had an overall 5-year survival rate of 44% (95% CI, 27%–71%). Nine individuals underwent surgery for a suspected malignant lesion which proved to not be PDAC, this included five lesions with low-grade dysplasia.

Conclusions This long-term surveillance study demonstrates a high incidence of PDAC in CDKN2A mutation carriers. We provide evidence that surveillance in such a high-risk population leads to detection of early stage PDAC with improved resectability and survival (Fig. 1, 2, 1).

Keywords Pancreatic surveillance, Screening, High-risk individuals, Hereditary pancreatic cancer, CDKN2A.

Fig. 1

Cumulative incidence of primary PDAC (n = 31) in CDKN2A mutation carriers (n = 347) participating in surveillance

OP-065: Clinical: Diagnosis, Endoscopy, Management, Implementation

Surveillance Outcome and Genetic Findings in Pancreatic Cancer High-Risk Population

Guy Rosner ¹, Erez Scapa¹, Tomer Ziv², Merav Ben Yehoyada¹

¹Department of Gastroenterology and Liver Diseases, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel and Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel. ²Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel

Background and aim Pancreatic ductal adenocarcinoma (PC) has a lifetime risk of 1.3% and a poor 5-year survival rate. PC surveillance in high-risk individuals was recommended by the International Cancer of the Pancreas Screening Consortium. Our aim was to present "real world" clinical and genetic data from a referral PC high-risk center.

Method 239 high-risk PC individuals (202 families) had PC surveillance for 5.7 years and were tested genetically.

Results The cohort was divided into 3 groups: classical Familial Pancreatic Cancer (FPC) (70 individuals, 54 families), non-classical FPC (81 individuals, 73 families), and hereditary PC (88 individuals, 75 families). Five early-onset PC individuals lacking inclusion criteria were tested by whole exome sequencing (WES) but were not included in the study cohort and served as the control group. Of the 239 individuals, 11 (4.6%) had PC (most detected at an early stage), 4 (1.67%) had pancreatic neuroendocrine tumor, 6 (2.5%) had main duct intraductal papillary neoplasm (IPMN), and 41 (17.15%) had side branch IPMNs. Pathogenic variants (PVs) were found in 36.1% of the families and BRCA1,2 were the most prevalent mutated genes. PV were detected in 11.1% of both the classical and the non-classical FPC groups. WES detected PVs in 22.8% of the families tested.

Conclusions Identifying PC high-risk individuals and admitting them to a strict surveillance program may help to improve their survival. Significant number of PC high-risk individuals in our cohort carried a mutation associated with PC risk. WES improves the genetic evaluation in PC high-risk individuals.

Keywords Pancreatic cancer surveillance; Familial pancreatic cancer; BRCA1,2;

PP-001: Cancer Pathway-Somatic Mutations, Pathology

Colorectal cancer in patients with serrated polyposis syndrome: what is the contribution of the adenoma carcinoma pathway?

David Van Toledo ¹, Joep Ijspeert¹, Hannah Boersma¹, Arne Bleijenberg¹, Alex Musler², Evelien Dekker¹, Carel Van Noesel²

¹Amsterdam University Medical Centers, University of Amsterdam, Department of Gastroenterology and Hepatology, Amsterdam, the Netherlands; Amsterdam Gastroenterology Endocrinology and Metabolism, Amsterdam, The Netherlands; Cancer Center Amsterdam, Amsterdam, the Netherlands. ²Amsterdam University Medical Centers, University of Amsterdam, Department of Pathology, the Netherlands

Background and aim Patients with serrated polyposis syndrome (SPS) have an increased risk to develop colorectal cancer (CRC). Because of the large numbers of serrated polyps (SPs) in these patients, CRC is assumed to arise mainly through the serrated neoplasia pathway rather than the adenoma-carcinoma pathway. SPS patients, however, also harbour significant numbers of adenomas, and about half of all CRCs occur in the distal colon, the preferential location of adenomas. This suggests that adenomas may also play a role in CRC development in these patients. We aimed to study the contribution of the adenoma-carcinoma pathway to CRC development in SPS patients.

Method We collected a cohort of SPS patients with CRCs at our expertise center. Data regarding polyps and CRCs were retrieved from endoscopy and pathology reports. All CRCs were immunohistochemically stained for expression of MLH1, BRAFV600E and p53. CRCs were divided into ‘serrated-CRC’ and ‘adenoma-CRC’ based on the presence or absence of BRAFV600E expression. CRC characteristics (location, MLH1 expression) were compared between serrated-CRCs and adenoma-CRCs. The proximal colon was defined as CRC located proximal from the descending colon. Primary outcome measure was the proportion of serrated-CRC vs. adenoma-CRC.

Results A total of 43 CRCs from 35 SPS patients were included in the study. Patients had a median age at time of diagnosis of 63 (IQR, 59–69) and 68.6% (24/35) were female. Twenty-one (48.8%) CRCs were serrated-CRCs and 22 (51.2%) were adenoma-CRCs. Loss of MLH1 expression was observed in 10/21 serrated-CRCs and in none of adenoma-CRCs, and 17/21 of serrated-CRCs and 5/22 of adenoma-CRCs were located in the proximal colon (p < 0.001).

Conclusions Although SPS patients have an abundance of serrated polyps, still half of their CRCs evolve via the non-serrated pathway, i.e. from conventional adenomas. This finding is important in understanding carcinogenesis in SPS patients and underlines the clinical importance of conventional adenomas in these patients.

Keywords serrated polyposis, colorectal cancer.

Patient characteristics of all included patients comparing serrated-CRCs and adenoma-CRCs.

	Total	Serrated-CRC	Adenoma-CRC	Serrated-CRC & adenoma-CRC	p-value*
Total patients; n (%)	35 (100)	15 (42.9)	18 (51.4)	2 (5.7)	-
Female gender; n (%)	24 (68.6)	12 (80.0)	11 (61.1)	1 (50.0)	0.710
Patient age of SPS diagnosis; median (IQR)	63 (58–69)	63 (60–64)	64 (58–69.25)	63 (54-NA)	0.690
WHO 2019 criteria SPS; n (%)					0.719
Type 1	13 (37.1)	6 (40.0)	7 (38.9)	-	0.379
Type 2	6 (17.1)	1 (6.7)	5 (27.8)	-	0.273
Type 1 and 2	16 (45.7)	8 (53.3)	6 (33.3)	2 (100)	-
Patient with second primary CRC; n (%)	8 (22.9)	3 (20.0)	3 (16.7)	2 (100)	0.805

*Comparing serrated-CRC and adenoma-CRC with Chi-square statistics for categorical variables and Mann–Whitney U for continuous variables

Colorectal cancer characteristics comparing serrated-CRCs and adenoma-CRCs.

	Total	Serrated-CRC	Adenoma-CRC	p-value*
Total cohort size; n (%)	43 (100)	21 (48.8)	22 (51.2)
CRC location; n (%)				< 0.001
Proximal	22 (51.2)	17 (81.0)	5 (22.7)
Distal	21 (48.8)	4 (19.0)	17 (77.3)
Median CRC diameter; n (%)	32.5 (16.8–51.2)	35 (15.5–55)	25 (20–50)	0.820
First primary CRC TNM stage; n (%)				0.586
Non-advanced	29 (67.4)	15 (71.4)	14 (63.6)
Advanced	14 (32.6)	6 (28.6)	8 (36.4)
Loss of MLH1 expression	10 (23.2)	10 (47.6)	0	< 0.001

* Comparing serrated-CRC and adenoma-CRC with Chi-square statistics for categorical variables and Mann–Whitney U for continuous variables

PP-002: Cancer Pathway-Somatic Mutations, Pathology

Faecal metagenomics and Whole Genome Sequencing of patients with somatic APC variants suiting the mutation signature caused by colibactin

Diantha Terlouw ¹, Arnoud Boot², Quinten Ducarmon³, Sam De Nooij³, Manon Suerink⁴, Alexandra Langers⁵, Romy Zwittink³, Tom Van Wezel⁶, Maartje Nielsen⁴, Hans Morreau⁶

¹Department of Pathology, Leiden University Medical Center, Leiden, the Netherlands; Department of Clinical Genetics, Leiden University Medical Center, Leiden, the Netherlands. ²Department of Cancer and Stem Cell Biology, Duke-NUS Medical School, Singapore. ³Department of Medical Microbiology, Leiden University Medical Center, Leiden, the Netherlands. ⁴Department of Clinical Genetics, Leiden University Medical Center, Leiden, the Netherlands. ⁵Department of Gastroenterology and Hepatology, Leiden University Medical Center, Leiden, the Netherlands. ⁶Department of Pathology, Leiden University Medical Center, Leiden, the Netherlands

Background and aim Our APC mosaicism analyses have revealed a possible additional explanation for the development of adenomas, namely colibactin produced by pks + E.coli bacteria amongst others. More research into the impact of colibactin on colorectal mucosa and presence of putative carcinogenic bacteria in feces samples is required.

Method Twenty-one patients were selected for faecal shotgun metagenomics and somatic Whole Genome Sequencing (WGS). Two of these patients were negative controls, as Next Generation Sequencing did not show an APC variant suiting the colibactin mutational signatures (COSMIC: SBS88 or ID18) in any of their adenomas. The other patients included had at least one adenoma with an APC variant suiting SBS88 or ID18. WGS was mainly performed on adenomas without a colibactin APC variant while at least one other adenoma harbours such an APC variant.

Results Fecal metagenomics shows genes involved in colibactin production (pks island) in 60% of patients with ≥ 1 adenoma with a colibactin APC variant. Although WGS of DNA isolated from Formalin-Fixed Paraffin-Embedded material of adenomas is challenging, the mutational signature (SBS88) was slightly enriched in patients with pks (4/8; 50%) compared to patients without pks in their feces samples (2/7; 29%). The negative controls did not show pks in their feces samples, nor the mutational signatures in their adenomas.

Conclusions These results show subtle additional evidence for the carcinogenic influence of colibactin in patients with multiple adenomas. However, to draw firmer conclusions, additional negative controls will be analysed.

Keywords Colibactin, E.Coli, Mutational signatures, Fecal metagenomics.

PP-003: Cancer Pathway-Somatic Mutations, Pathology

A 10-year review of the quality and quantity of mismatch repair immunohistochemistry submissions to the UK National External Quality Assessment Scheme for Immunocytochemistry and In-situ Hybridisation

Luke A. R. Farmkiss ¹, Andrew Dodson², Ian M. Frayling³, Mark J. Arends⁴

¹Histopathology Department, Derriford Hospital, Plymouth, U.K. ²UK National External Quality Assessment Scheme ICC & ISH, London, U.K. ³St Mark's Hospital, London, U.K. ⁴Edinburgh Pathology, University of Edinburgh, Edinburgh, U.K

Background and aim The National Institute for Health and Care Excellence (NICE) recommends Lynch syndrome screening in all people initially diagnosed with colorectal cancer, either by mismatch repair (MMR) immunohistochemistry (IHC) or microsatellite instability testing.

The UK National External Quality Assessment Scheme for Immunocytochemistry and In-situ Hybridisation (UK NEQAS) has quality assessed MMR IHC since 2011.

Here we explore the quantity and quality of submissions to the scheme (2011–2021).

Method Participants submitted MMR IHC using routine clinical protocols on supplied tissue. Assessments ran quarterly, alternating between paired proteins (MLH1/PMS2 and MSH2/MSH6). Submissions were scored by four independent assessors according to pre-defined standards. Final aggregated scores (FAS) were categorised: unacceptable (score: 4–9); borderline acceptable (score: 10–12) and acceptable (score: 13–20).

FAS were collated with data for primary antibody clone, IHC detection system and automated staining platform. Data was curated and analysed in Excel (Microsoft).

Results Between 2011–2021 there were 38 assessment runs with 7038 submissions from 238 unique participants. A 2.8-fold increase was observed between runs 95 (2011, n = 114) and 135 (2021, n = 318).

The percentage of borderline or better submissions improved by 9.9% between runs 95 (81.6%) and 135 (91.5%). The average FAS (AvFAS) across all antibodies improved from 13.9 to 15.2.

Detection system data (n = 6966) showed popular systems were: Ventana Optiview (n = 3093, 44.4%, AvFAS 14.8), Leica BondMax Refine kit (n = 1504, 21.6%, AvFAS 14.1), and Dako FLEX + kit (n = 724, 10.4%, AvFAS 15.6).

Automation platform data (n = 6950) showed popular platforms were: Ventana Benchmark ULTRA (n = 2743, 39.5%, AvFAS 14.7); Leica Bond-III (n = 1460, 21%, AvFAS 14.3); Ventana Benchmark XT (n = 805, 11.6%, AvFAS 13.6).

Primary antibody clone data (n = 6967) demonstrated an AvFAS in the acceptable range for all three commonest MLH1, PMS2 and MSH2 clones and two of the three commonest MSH6 clones. Clone 44 (MSH6) (n = 514) showed an AvFAS of 11.8 (borderline acceptable).

Conclusions The observed increase in scheme participation may be linked to the introduction of NICE guidance for Lynch syndrome screening and the clinical demand for testing used to inform immunotherapy.

All four MMR antibodies showed improvement in FAS between 2011–2021. This likely reflects increased participant proficiency and a refinement of the techniques used.

Keywords Mismatch repair, Immunohistochemistry, Quality, Performance, Antibody.

Submissions scoring borderline acceptable or better.

MMR Antibody	Overall (2011–21) (%)	First Run (2011) (%)	Last Run (2021) (%)
MLH1	89.1	74.6	90.1
PMS2	87.5	81.8	86.1
MSH2	89.1	87.9	93.8
MSH6	84.6	82.1	96.3
Total	87.5	81.6	91.1

The percentage of submissions which had an average final aggregated score of equal to or greater than 10 (borderline acceptable or better).

PP-004: Cancer Pathway-Somatic Mutations, Pathology

Preliminary evaluation of highly sensitive assessment of microsatellite instability in endometrial aspirates as a tool for cancer risk individualization in Lynch syndrome

Júlia Canet Hermida¹, Fátima Marín¹, Eduard Dorca², Nuria Dueñas¹, Laura Costas³, Mònica Salinas¹, Àngela Velasco⁴, Paula Peremiquel Trillas³, Sònia Paytubi³, Jordi Ponce⁵, Laura Cárdenas⁶, Laia Alemany³, August Vidal², Eugeni López Bonet⁷, Gabriel Capellá¹, Joan Brunet⁸, Xavier Matias Guiu², Marta Pineda ¹

¹Hereditary Cancer Program, Catalan Institute of Oncology, Institut d’Investigació Biomèdica de Bellvitge (IDIBELL), ONCOBELL Program, L’Hospitalet de Llobregat, Barcelona, Spain. CIBERONC. ²Pathology Department, Hospital Universitari de Bellvitge, Institut d’Investigació Biomèdica de Bellvitge (IDIBELL), L’Hospitalet de Llobregat, Barcelona, Spain. ³Cancer Epidemiology Research Program, Catalan Institute of Oncology, Institut d’Investigació Biomèdica de Bellvitge (IDIBELL), L’Hospitalet de Llobregat, Barcelona, Spain. ⁴Hereditary Cancer Program, Catalan Institute of Oncology, Institut d’Investigació Biomèdica de Girona (IDIBGI), Girona, Spain. ⁵Department of Gynecology and Obstetrics, Hospital Universitari de Bellvitge, Institut d’Investigació Biomèdica de Bellvitge (IDIBELL), L’Hospitalet de Llobregat, Barcelona, Spain. ⁶Department of Gynecology and Obstetrics, Hospital Universitari Josep Trueta, Institut d’Investigació Biomèdica de Girona (IDIBGI), Girona, Spain. ⁷Pathology Department, Hospital Universitari Josep Trueta, Institut d’Investigació Biomèdica de Girona (IDIBGI), Girona, Spain. ⁸Hereditary Cancer Program, Catalan Institute of Oncology, Institut d’Investigació Biomèdica de Bellvitge (IDIBELL), ONCOBELL Program, L’Hospitalet de Llobregat, Barcelona, Spain. CIBERONC and Hereditary Cancer Program, Catalan Institute of Oncology, Institut d’Investigació Biomèdica de Girona (IDIBGI), Girona, Spain

Background and aim Lynch syndrome (LS) women are at increased risk of endometrial cancer (EC), characterized by mismatch repair deficiency (MMRd) and microsatellite instability (MSI). While risk-reducing gynaecological surgeries are effective there is a need to empower carriers in their decision regarding surgery timing. We aim at exploring the usefulness of highly sensitive-MSI (hs-MSI) assessment in endometrial aspirates for the individualization of gynaecological surveillance in LS carriers.

Method Hs-MSI was assessed in prospectively collected endometrial aspirates from 67 LS carriers, 30 controls with benign lesions and 26 sporadic-EC cases. MMR, PTEN and ARID1A expression patterns were evaluated in LS samples.

Results High hs-MSI levels were detected in 20 aspirates from MMRd EC cases (4 of 4 LS and 16 of 16 sporadic) being negative in aspirates from controls and MMR-proficient EC cases. Interestingly, elevated hs-MSI scores were also detected in aspirates from LS women with complex hyperplasia (3 of 3). In addition, high hs-MSI was present in 9 of 49 aspirates from LS carriers showing histologically normal endometrium. In normal endometrium hs-MSI score positively correlated with density of MMRd-glands, presence of MMRd-clusters and PTEN and ARID1A loss. Hs-MSI levels increased in follow-up aspirates from 8 LS carriers.

Conclusions Elevated hs-MSI scores were detected in aspirates from pre/malignant lesions and normal endometrium in LS carriers, correlating with MMR protein loss. Further analyses of sequential aspirates are needed to elucidate the predictive value of hs-MSI in the identification of LS carriers at higher risk of developing EC.

Keywords Lynch syndrome, Endometrial cancer, Microsatellite instability, Gynaecological surveillance.

PP-005: Cancer Pathway-Somatic Mutations, Pathology

The molecular profile of MSH6-associated colorectal carcinomas from patients with Lynch syndrome

Noah Cornelis Helderman ¹, Anne-Sophie Van Der Werf—'t Lam¹, Diantha Terlouw², Sanne Willy Bajwa—Ten Broeke³, Mar Rodríguez-Girondo⁴, Demi van Egmond², Monique van Leerdam⁵, Manon Suerink¹, Carli Tops¹, MSH6 Tumor Group¹, Hans Morreau², Arnoud Boot⁶, Tom van Wezel², Maartje Nielsen¹

¹Department of Clinical Genetics, Leiden University Medical Center, Leiden, the Netherlands. ²Department of Pathology, Leiden University Medical Center, Leiden, the Netherlands. ³Department of Clinical Genetics, University Medical Center Groningen, Groningen, the Netherlands. ⁴Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, the Netherlands. ⁵Department of Gastroenterology and Hepatology, Leiden University Medical Center, Leiden, the Netherlands; Department of Gastrointestinal Oncology, Netherlands Cancer Institute, Amsterdam, the Netherlands. ⁶GenomeScan, Leiden, the Netherlands

Background and aim Lynch syndrome (LS) patients carry a germline variant in one of the mismatch repair (MMR) genes (MLH1, MSH2, MSH6 and PMS2) and have a high risk to develop colorectal cancer (CRC). In contrast to other LS CRCs, the molecular profile of MSH6-associated CRCs has not been studied on a larger scale. Therefore, this study aims to examine the molecular profile of CRCs from MSH6 variant carriers in order to gain more insights in the cancer development in this group of patients.

Method The DNA of 14 MSH6-associated CRCs was analyzed using whole-exome-sequencing and was subsequently compared to that of the other LS subgroups. The sequence of mutational events was studied by classifying variants either as being related to MMR-deficiency or not, using COSMIC signature 6.

Results Variants of CTNNB1 were more frequently found in MSH6-associated CRCs (21%) than in MSH2- (6%; P = 0.295) and PMS2-associated CRCs (0%; P = 0.061). APC variants were found in 36% of the MSH6-associated CRCs and frequently co-occurred with mutations in TCF7L2 and RNF43. KRAS and TP53 variants were prevalent in all four LS subgroups, though the contribution of signature 6 to these variants was lower in the PMS2-associated CRCs (30% & 17%, respectively) than in the MSH6- (57% & 50%), MLH1- (50% & 63%), and MSH2-associated CRCs (88% & 63%).

Conclusions Our findings suggest that MSH6-associated CRCs can develop via three distinct molecular pathways yet mainly originate from MMR-deficient crypts with an intervening adenoma stage, and could influence future guidelines for screening and treatment.

All results discussed in this abstract are preliminary results, as the analysis of 10–15 additional MSH6-associated CRCs is currently pending. The final results will be available for presentation in September.

Keywords Lynch syndrome, MSH6, Colorectal cancer, Molecular profile, Mutational signature.

Visual abstract—The molecular profile of MSH6-associated colorectal carcinomas from patients with Lynch syndrome

PP-006: Cancer Pathway-Somatic Mutations, Pathology

Loss of Mismatch Repair Protein Expression in Nonneoplastic Crypts/Glands is Common in Lynch Syndrome, but None/Rare with MMR Double Somatic Mutations in Colorectal/Endometrial Cancers

Wei Chen ¹, Cody Eric Freitag¹, Rachel Sarah Pearlman², Heather Doll Hampel³, Peter Paul Stanich⁴, Casey Mcdonald Cosgrove⁵, Wendy Lynne Frankel¹

¹Department of Pathology, The Ohio State University Wexner Medical Center, Columbus, Ohio, USA. ²Department of Internal Medicine, Clinical Cancer Genetics Program, The Ohio State University Wexner Medical Center, Columbus, Ohio, USA. ³Division of Genetics and Genetic Counseling, City of Hope, Duarte, California, USA. ⁴Division of Gastroenterology, Hepatology and Nutrition, The Ohio State University Wexner Medical Center, Columbus, Ohio, USA. ⁵Division of Gynecologic Oncology, The Ohio State University Comprehensive Cancer Center, Columbus, Ohio, USA

Background and aim Mismatch repair protein-deficient (dMMR) nonneoplastic glands (NNG) have been reported as a unique marker of Lynch syndrome (LS), as it was not observed in MLH1-methylated or MMR-proficient colorectal cancers (CRC) and endometrial cancers (EC). However, it is unknown if CRC and EC with double somatic (DS) mutations harbor dMMR NNG. Carcinoma of the lower uterine segment (LUS) has been associated with LS, and it is unclear if dMMR NNG are enriched in LUS. We aim to study the frequency of dMMR NNG in CRC and EC patients with LS and DS, and if dMMR NNG is seen more frequently in LUS.

Method The study was approved by an institutional IRB. For CRC, 23 LS and 18 DS colectomy specimens were identified. 4 blocks with normal mucosa (nontumor blocks, bowel margins and/or random normal mucosa) were selected from most cases. If less than 4 blocks were available, additional sections of selected block(s) were used. For EC, 9 LS and 11 DS hysterectomy specimens were identified. 4 blocks with benign endometrium (nontumor blocks, 2 from LUS and 2 from the wall of uterine corpus) were selected from most cases. 3 cases had only 1 section of LUS available. MMR immunohistochemistry specific for the known mutation from each case was performed and reviewed for dMMR NNG. Statistical analysis was performed using Fisher’s exact test and Student’s t-test.

Results All CRC cases were adenocarcinoma. dMMR NNG were found in 65% (15/23) of LS and 0% (0/18) of DS CRC cases (p < 0.05; Table 1). There was no significant age difference between LS and DS for CRC (47 vs 53; p = 0.08). All EC cases were endometrioid carcinoma. dMMR NNG were found in 67% (6/9) of LS and 9% (1/11) of DS EC cases (p = 0.02; Table Tab2). Most (5/6) dMMR NNG in LS were found in the uterine corpus. dMMR NNG were found in the LUS of 1 LS case and 1 DS case. There was a significant age difference between LS and DS (48 vs 62; p = 0.0003), but no difference in BMI (29 vs 35; p = 0.16). The DS case with dMMR NNG was from a 58-year-old with negative LS germline testing, although there was unspecified cancer history reported in her mother, maternal grandmother, and paternal grandfather and uncle. For both types of cancers in LS patients, there was no significant age difference between those with dMMR NNG and those without.

Conclusions dMMR NNG were common in tumors from LS, but rare in DS. Taken together with previously reported lack of dMMR NNG with MLH1 methylation, the presence of such glands should raise a high suspicion for LS. Although LS-associated EC tends to involve the LUS, dMMR NNG were more common in the uterine corpus than LUS.

Keywords MMR deficient crypts, MMR deficient glands, Lynch syndrome, double somatic mutations, colorectal cancer, endometrial cancer.

MMR-deficient Nonneoplastic Crypts in Lynch Syndrome Patients with Colorectal Cancer

Case	Age	Genetic Testing Results	MMR IHC Performed	dMMR NNGs
20	54	MLH1 pathogenic mutation (c.37G > T, p.E13X)	MLH1	Yes
25	62	MSH2 pathogenic mutation (c.229_230del, p.S77Cfs*4)	MSH2	Yes
27	47	MSH2 pathogenic mutation (exons 1–6 deletion)	MSH2	Yes
29	47	MSH6 pathogenic mutation (c.1109 T > C, p.L370S)	MSH6	Yes
30	51	MSH6 pathogenic mutation (c.3840_3846del, p.E1281Lfs*44)	MSH6	Yes
31	38	MSH2 pathogenic mutation (c.425C > G, p.Ser142X)	MSH2	Yes
32	71	MSH2 pathogenic mutation (c.942 + 3A > T, Skip exon 5)	MSH2	Yes
33	47	MLH1 pathogenic mutation (c.1778_1779delCA, p.Pro593ArgfsX16)	MLH1	Yes
35	46	MLH1 pathogenic mutation (c.1489delC, p.Arg497GlufsX11)	MLH1	Yes
36	32	MSH2 pathogenic mutation (c.942 + 3A > T, Skip exon 5)	MSH2	Yes
37	46	PMS2 pathogenic mutation (736_741del6ins11)	PMS2	Yes
38	46	MLH1 pathogenic mutation (c.790 + 1G > A, Skip exon 9)	MLH1	Yes
39	67	MSH6 pathogenic mutation (c.3261delC, p.Phe1088SerfsX2)	MSH6	Yes
40	37	MSH2 pathogenic mutation (c. 830 T > G, L277X)	MSH2	Yes
42	52	MLH1 pathogenic mutation (exon 6–12 duplication)	MLH1	Yes

dMMR, Mismatch Repair Protein Deficient; IHC , Immunohistochemistry; LOH , Loss of Heterozygosity; NNG , Nonneoplastic Colonic Crypts.

MMR-deficient Nonneoplastic Glands in Patients with Endometrial Cancer

Case	Age	BMI	LS vs. DS	Genetic Testing Results	MMR IHC Performed	dMMR NNGs
1	61	32	DS	MSH6: Exon 7 deletion; c.3261dup; p.F1088fs*5	MSH6	No
2	67	36	DS	MSH6: c.721del; p.S241Vfs*5; c.3261dup; p.F1088Lfs*5	MSH6	No
3	58	43	DS	MLH1: c.1833_1842del; p.V612X; c.676C > T; p.R226X	MLH1	Yes (LUS)
4	47	62	DS	MSH2: c.1203dupA; p.Q402Tfs*15; c.1294_1298del; p.L432Gfs*9	MSH2	No
5	63	27	DS	MSH2: c.1046C > A; p.P349H; c.2089 T > C; p.C697R	MSH2	No
6	76	25	DS	MSH2: c.790C > T; p.Q264X; c.2038C > T; p.R680X	MSH2	No
7	66	29	DS	MLH1: c.442A > C; p.T148P; LOH	MLH1	No
8	57	47	DS	MLH1: c.707_711del; p.K236Tfs*69; c.1990-3C > G	MLH1	No
9	63	28	DS	MLH1: c.199G > A; p.G67R; LOH	MLH1	No
10	56	33	DS	MSH2: c.1705_1706del; p.E569Ifs*2; LOH	MSH2	No
11	65	25	DS	MSH2: c.1250 T > C; p.V4717A; c.2276G > T; p.G759V	MSH2	No
12	39	36	LS	MSH2 likely pathogenic mutation (c.2006-1G > T)	MSH2	Yes (LUS)
13	41	24	LS	MSH2 likely pathogenic mutation (c.2211-1G > C)	MSH2	Yes
14	53	21	LS	MSH2 pathogenic mutation (c.2132G > C, p.R711P)	MSH2	No
15	44	34	LS	MSH6 pathogenic mutation (c.1109 T > C, p.L370S)	MSH6	No
16	44	21	LS	MSH6 pathogenic mutation (c.1109 T > C, p.L370S)	MSH6	Yes
17	52	29	LS	MSH6 pathogenic mutation (c.3768 T > G, p.Y1256X)	MSH6	Yes
18	57	27	LS	MSH6 pathogenic mutation (c.3939_3957dup, p.A1320Sfs*5)	MSH6	No
19	53	40	LS	MSH6 pathogenic mutation (exon 3–6 deletion)	MSH6	Yes
20	50	28	LS	PMS2 pathogenic mutation (c.1831dup, p.I611Nfs*2)	PMS2	Yes

BMI, Body Mass Index; dMMR, Mismatch Repair Protein Deficient; DS, Double Somatic Mutation; IHC, Immunohistochemistry; LOH, Loss of Heterozygosity; LS, Lynch Syndrome; LUS, Lower Uterine Segment; NNG, Nonneoplastic Endometrial Glands.

PP-007: Chemoprevention

The chemopreventive effect of lithium on adenoma development in patients with Familial Adenomatous Polyposis (FAP); a pilot study

Jasmijn D. G. Linssen ¹, Sanne M. Van Neerven², Arthur S. Aelvoet³, Christine C. Cohen³, Louis Vermeulen², Evelien Dekker³

¹Department of Gastroenterology and Hepatology, Laboratory for Experimental Oncology and Radiobiology, Center for Experimental and Molecular Medicine, Cancer Center Amsterdam and Amsterdam Gastroenterology and Metabolism, Oncode Institute, Amsterdam University Medical Centers, Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands. ²Laboratory for Experimental Oncology and Radiobiology, Center for Experimental and Molecular Medicine, Cancer Center Amsterdam and Amsterdam Gastroenterology and Metabolism, Oncode Institute, Amsterdam University Medical Centers, Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands. ³Department of Gastroenterology and Hepatology, Cancer Center Amsterdam and Amsterdam Gastroenterology and Metabolism, Amsterdam University Medical Centers, Meibergdreef 9, 1105 AZ, Amsterdam, The Netherlands

Background and aim FAP is a rare autosomal dominant disease characterized by germline mutations in the APC gene, resulting in the development of numerous premalignant polyps in the colorectum. As these patients have a high risk of developing CRC, guidelines suggest prophylactic colectomy during early adulthood, however, adenoma development is still observed in the remaining intestinal tract. Therefore, FAP patients would greatly benefit from chemoprevention strategies that reduce the development of adenomas. Our recent work in mice reveals a chemopreventive effect of the Wnt pathway agonist lithium on the development of adenomas by inhibiting the expansion of Apc mutated cells within the crypts of normal intestinal mucosa. Here, we aim to investigate the effect of lithium on the spread of APC mutant cells within the human intestinal epithelium.

Method This prospective phase II single arm pilot trial will include FAP patients (18–35 years) who have a genetically confirmed APC mutation and did not undergo colectomy. The trial has a duration of 18 months in which a colonoscopy with biopsies of normal intestinal mucosa will be performed at baseline and every six months. Patients will be treated with lithium carbonate orally between month 6 and 12 achieving a serum level of 0,2–0,4 mmol/l. Primary endpoint is the effect of lithium on the spread of APC mutant cells within intestinal crypts over time by using APC specific marker NOTUM in situ hybridization. Secondary endpoints include difference in number and size of polyps between baseline and end of study, patient reported side effects of lithium and safety outcomes. Total sample size is 12 patients and recruitment will take place in the Amsterdam UMC.

Results We have currently demonstrated the specific expression of NOTUM in a panel of FAP adenomas and detected the presence of NOTUM-positive cells in normal-appearing intestinal crypts, indicating that we can successfully trace the spread of APC mutant cells over time.

Conclusions So far, there is no clinically applied chemoprevention therapy available for FAP patients. Lithium is used in the clinic as treatment for mood disorders for several decades, therefore many studies have been conducted on pharmacodynamics and pharmacokinetics. If lithium treatment indeed reduces on the spread of APC mutated ISCs within crypts and inhibits adenoma development in FAP patients, invasive surgery for the prevention of CRC in these patients could potentially be postponed or even avoided.

Keywords Lithium, Chemoprevention, FAP, adenomatous polyps, CRC.

PP-008: Clinical: Counselling, Psychosocial

Quality of Life in Familial Adenomatous Polyposis patients after colorectal preventive surgery

Stefano Signoroni¹, Chiara Maura Ciniselli², Chiara Ghidoli ¹, Laura Sophie Boer³, Loïs Oldhoff³, Laura Dijk³, Franka Van Luijk³, Maria Teresa Ricci¹, Guglielmo Niccolò Piozzi⁴, Clorinda Brignola¹, Emanuele Rausa¹, Claudia Borreani⁵, Sara Alfieri⁵, Giovanni Apolone⁶, Paolo Verderio², Marco Vitellaro¹

¹Unit of Hereditary Digestive Tract Tumors, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ²Unit of Bioinformatics and Biostatistics, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ³Department of Medical Sciences, University of Groningen, Netherlands – Erasmus project. ⁴Colorectal Surgery Division, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁵Clinical Psychology Unit, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁶Scientific Directorate, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy

Background and aim To evaluate post-operative quality of life (QoL) in patients with Familial Adenomatous Polyposis (FAP), submitted to preventive colorectal surgery. Provide future FAP patients a detailed view on post-surgical QoL.

Method This was a monocentric observational study evaluating post surgical QoL of 232 FAP patients submitted to prophylactic surgery between January 2000 and October 2018. The SF-36 and an ad-hoc designed questionnaire (IT-quest) were used. The SF‐36 investigates general health perceptions, physical function, social function, mental health, bodily pain and vitality, role limitations due to physical health and emotional problems. The IT-quest analyses 4-main aspects immediately after surgery and at the time of the survey: daily life (diet changes; work/school resuming problems; emotional bonding problems; sexual activity problems; libido problems; mood disorders), faecal and urinary incontinence, bowel movements and general perception of the surgery (satisfaction of the surgery and general satisfaction). A total of 152 patients contacted on the phone, agreed to receive the survey and to participate to the study.

Results A total of 102 patients completed the survey. 80% underwent colectomy with ileo-rectal anastomosis and 20% proctocolectomy with ileal pouch anal anastomosis. Main results showed mood disorders reported in 12% of patients, work/school resuming problems in 29% of patients, change of diet habits in 74% of patients. Sexual activity problems has been reported in 14% of patients, daily bowel movements less than 5 times a day in 71% of patients and more than 5 times a day in 27% of patients. Faecal incontinence immediately after the surgical procedure were reported seldom in 32% and often in 19%, while at the time of survey has been reported seldom in 21% and often 7% of patients. Symptoms relief was time dependent. At the time of the survey 90% reported at least good general health perception.

Conclusions Few studies reported emotional and social aspects on the QoL of FAP patients. The data of this study allowed to highlight the outcomes related to these surgical procedures to better understand the needs and where to intervene in order to improve the quality of life in patients with familial polyposis.

Keywords Familial Polyposis; Quality of Life; Preventive Surgery; SF 36; FAP patients needs.

PP-009: Clinical: Counselling, Psychosocial

Psychological well-being of people living with a colorectal cancer predisposition syndrome: evidence from a systematic review

Laura Monje Garcia¹, Jai Vairale², Hilary Watt², Sondus Hassounah², Heidi Lai², Kevin Monahan ¹

¹St Mark’s Hospital Centre for Familial Intestinal Cancer, Imperial College London, London, UK. ²Imperial College London, London, UK

Background and aim About 5–10% of newly diagnosed colorectal cancers have a genetic cause. Early detection is key to facilitate effective treatments and prevent future cancers. When an inherited colorectal cancer is diagnosed, predictive genetic testing can be cascaded to first degree relatives. Genetic and genomic health information increasingly informs routine clinical care. Some health care professionals have concerns about the potential for genetic information to inflict psychological harm on patients. Therefore, it was decided to explore this area in depth within the scope of inherited colorectal cancer.

Method A systematic review was carried out using PRISMA guidelines. Six relevant databases were searched systematically, using MeSH and other search terms. Inclusion criteria followed those in our Prospero-published protocol: peer-reviewed quantitative and qualitative studies on the psychological well-being of adult asymptomatic individuals living with colorectal cancer.

Results The search strategy yielded 1590 studies of which 33 were eligible. Eight studies were qualitative interviews of probands and families, and 25 were quantitative. Psychological distress, anxiety, depression, cancer worry, risk perception, quality of life, resilience, coping, and hopelessness were assessed using psychometric scales. Quantitative studies showed that there is a transient increase in anxiety, worry, and depression following genetic results which drop back to baseline level one year after the genetic results are given. Baseline is defined as immediately prior to genetic testing, when related worries may already be present. Overall, important factors that influence individuals’ psychological well-being are decision making, risk perception, cancer worry, family relationships and coping mechanisms. Related positive outcomes are hopefulness, optimism, good communication within the family, and positive attitude.

Conclusions Risk factors, such as experience of nursing a relative through colorectal cancer, significant family history, lack of close family connections, poor family dynamics, and ‘feeling’ alone in the genetic diagnosis, can help clinicians to identify patients that are more likely to need additional support when they receive a genetic diagnosis. We recommend further research using uniform, standardized tools to assess essential domains and positive outcomes across different cultures, incorporating wider determinants for meaningful interpretations and clinical applicability.

Keywords Inherited colorectal cancer, psychological Impact, Wellbeing, colorectal cancer, psychological distress, uncertainty.

PP-010: Clinical: Counselling, Psychosocial

Patient Uptake of Updated Genetic Testing Following Nondiagnostic Hereditary Cancer Results

Sarah Mantia ¹, Kristin Clift², Douglas Riegert Johnson¹, Stephanie Hines³

¹Department of Clinical Genomics, Mayo Clinic Florida, Jacksonville, United States. ²Department of Clinical Studies, Mayo Clinic Florida, Jacksonville, United States. ³Department of Internal Medicine, Mayo Clinic Arizona, Scottsdale, United States

Background and aim Updates to clinical practice and technology have made the genetic testing initially offered to evaluate for hereditary breast and ovarian cancer risks obsolete. Previous research has shown low natural completion of a second genetic test, even when patients were notified of potential benefit. This study reviewed incidence of additional testing after a nondiagnostic BRCA1/2 result and offered updated testing to those that had not completed. The aim was to better examine patient interest and which factors may influence completion.

Method A review of the Mayo Clinic Florida Hereditary Cancer Clinical Registry was done to identify patients that had completed BRCA1/2 only analysis between 2001 and 2020. Patients with nondiagnostic results, email address, and U.S. address were invited to complete additional testing through the study. After review of an educational video and consenting, they were mailed a saliva collection kit to complete an 84 gene hereditary cancer panel at no personal charge.

Results Fifteen percent (N = 102) of the 671 patients with a nondiagnostic result had already completed expanded testing. Those that had completed additional testing were more likely to have a history of cancer (p < 0.01), have initial testing prior to the BRCA1/2 patent removal in 2013 (p < 0.01), have a BRCA1/2 variant of uncertain significance (p = 0.01), and live closer to Mayo Clinic FL (p = 0.03). Study information was sent to 372 people, and 116 (31%) consented to participate. Confirmation that study information was received was available for 142 of those that did not proceed (38%), and 13 patients elected to test outside of the study. Between those that chose to participate in the research study or not, the differences were not as clear. Whether BRCA1/2 analysis had been comprehensive did not impact whether additional testing was completed before or during the study.

Conclusions The need for updated analysis applies similarly to most areas of genetic testing, including hereditary gastrointestinal tumor evaluation, and it will continue as an issue. Here, many of the potential associated barriers of retesting, like costs, need for travel, and unawareness of clinical need, were removed and still only about 30% of patients agreed to participate. Further research should focus on understanding what additional factors influence this decision.

Keywords clinical genetic testing, multi-gene panel testing.

Patient Recruitment Flow Chart

PP-011: Clinical: Diagnosis, Endoscopy, Management, Implementation

Patient-reported outcome measures—a complement to routine hospital visits for patients with Lynch syndrome?

Kaisa Fritzell¹, Ann Sofie Backman ², Jan Björk²

¹Hereditary Cancer Clinic, Theme Cancer, Karolinska University Hospital; NVS, Division of Nursing, Karolinska Institutet. ²Hereditary Cancer Clinic, Theme Cancer, Karolinska University Hospital; Department of Medicine, Solna, Karolinska Institutet

Background and aim This study is part of a larger project that strives towards a more personalized care at the Hereditary outpatient clinic, Karolinska University Hospital, that provides care for approximately 30% of all individuals with hereditary colorectal cancer in Sweden. To do that the project include the development and implementation of patient-reported outcome measures (PROM), as a complement to yearly routine visits to the physician. The aim of the current study was to investigate if patients with Lynch syndrome respond to PROM as an alternative to hospital visits and to investigate what kind of care this procedure generates.

Method Pre-defined patients with uncomplicated illness received a letter, instead of a yearly routine visit to the physician, with a brief information letter about Lynch syndrome and the questionnaire. The responded questionnaires were evaluated by a specialist nurse and actions were taken according to a manual with pre-defined cut of levels. All actions taken were documented in an excel sheet. PROM included measures of health literacy, illness perception, health related quality of life, sexual health, symptoms, and endoscopy experience.

Results Preliminary results show that during the period April 2021 – January 2022, 150 of 241 patients (63%) responded to the questionnaire. Of those, actions were taken in 95%, 67% were contacted by the specialist nurse and the content of the calls were mainly due to worries associated with endoscopic examinations and the laxative procedure, general well-being and impact on life due to Lynch syndrome; 16% were contacted by the physician and the content of calls were mainly due to gastrointestinal symptoms and newly discovered cancer in family members with Lynch syndrome; 11% were contacted by the genetic counsellor, the content of the calls were mainly due to lack of Lynch specific knowledge, worry of own health and the health of family members with Lynch syndrome and genetic testing of off-springs.

Conclusions This was a first attempt to include PROM as an alternative to yearly routine visits to the hospital. The response rate was, however, quite low and there are probably several explanations for that. The patients might not be used to respond to questions about their health and there were quite many questions included. Moreover, the questionnaires are not validated for patients with Lynch syndrome, so the relevancy is unknown. Next step is to shorten and validate PROM with qualitative methods such as individual interviews.

Keywords Lynch syndrome, PROM, person center care.

PP-012: Clinical: Diagnosis, Endoscopy, Management, Implementation

Functional Outcome Differences between Males and Females Who have Undergone Reconstructive Surgery with Ileorectal Anastomoses (IRA) or Ileal Pouch-Anal Anastomoses (IPAA) due to Familial Adenomatosis Polyposis—A Prospective Cohort Study

Daniel Mira, Jan Björk, Johan Reutfors, Rolf Hultcrantz, Ann Sofie Backman.

Department of Medicine, Karolinska Institute, Stockholm, Sweden

Background and aim Familial adenomatous polyposis (FAP), an inherited polyposis syndrome causes ~ 1% of colorectal cancers (CRC). Patients develop 100–100,000 polyps which may develop into CRC. Colectomy is inevitable. Pan-proctocolectomy with incontinent ileostomy was the recommended prophylactic surgery for decades. Ileorectal Anastomoses (IRA) was introduced in 1950s; However, IRA patients remained at risk of CRC. Ileal Pouch-Anal Anastomoses (IPAA) was introduced 1984 in Sweden as an alternative. Functional outcome was better after IRA vs IPAA in follow-up studies. Outcome differences have not been studied between males and females. Long-term studies of functional outcome, and CRC risk after IRA or IPAA are needed. The aim of the study is to compare postoperative functional outcome up to 45 years after IRA or IPAA and identify sex differences.

Method Structured protocols between 2001–2021 were used to evaluate anal function at Hereditary Gastrointestinal unit at Karolinska university hospital.

Results 71 patients had at least one protocol. Females (n = 44) were on average 2 years older than males (21.5 vs 23.5) at first surgery. Higher fraction of females with IPAA had daily bowel movements (61.9% female-IPAA, 40% male-IPAA, 13.6% female-IRA), nighttime bowel movements (76.2%, 46.7%, 40.9%), nighttime incontinence (40%, 20%, 0%), and daytime incontinence (21.1%, 0%, 0%) compared to IPAA males and IRA females. Six females had secondary surgery (27%) and one male (8%) post-IRA. Median duration to conversion, 26y (4-41y) for females and 34y for the male.

Conclusions IPAA females had worse functional outcome compared to IPAA males and IRA females. Females were overrepresented in incidence of secondary surgery after IRA.

Keywords Familial adenomatous polyposis, Ileal Pouch-Anal Anastomosis, Ileorectal Anastomosis, Functional outcome, Gender, Restorative proctocolectomy.

Figure

Table

PP-013: Clinical: Diagnosis, Endoscopy, Management, Implementation

Unexplained mismatch repair deficiency: case closed

Ellis Eikenboom ¹, Sarah Moen², Lotte Van Leeuwen³, Willemina Geurts Giele³, Carli Tops⁴, Tjakko Van Ham³, Winand Dinjens⁵, Hendrikus Jan Dubbink⁵, Manon Spaander², Anja Wagner³

¹Erasmus MC Cancer Institute, University Medical Center Rotterdam, The Netherlands, Departments of Clinical Genetics and Gastroenterology and Hepatology. ²Erasmus MC Cancer Institute, University Medical Center Rotterdam, The Netherlands, Department of Gastroenterology and Hepatology. ³Erasmus MC Cancer Institute, University Medical Center Rotterdam, The Netherlands, Department of Clinical Genetics. ⁴Leiden University Medical Center, The Netherlands, Department of Clinical Genetics. ⁵Erasmus MC Cancer Institute, University Medical Center Rotterdam, The Netherlands, Department of Pathology

Background and aim To identify Lynch syndrome (LS) carriers, DNA mismatch repair (MMR) immunohistochemistry (IHC) is performed on colorectal cancers (CRC). Upon subsequent LS diagnostics, MMR deficiency sometimes remains unexplained (UMMRd). Recently, the importance of performing complete LS diagnostics to explain UMMRd, involving MMR methylation, germline, and somatic analysis, was stressed. To explore why some MMRd CRCs still remain unsolved, we performed a systematic review of the literature and mapped UMMRd patients diagnosed in our center.

Method A systematic literature search was performed in Ovid Medline, Embase, Web of Science, Cochrane CENTRAL, Google Scholar, for articles on UMMRd CRCs after complete LS diagnostics, published until December 15, 2021. Additionally, UMMRd CRCs diagnosed in our center since 1993 were mapped.

Results Of 754 identified articles, 17 were included, covering 74 UMMRd patients. Five CRCs were microsatellite stable. Upon complete diagnostics, 39 patients had single somatic MMR hits and six an MMR germline variant of unknown significance (VUS). Ten had somatic pathogenic variants (PVs) in POLD1, MLH3, MSH3, APC. The remaining 14 patients were the only identifiable cases in the literature in whom no plausible cause of the UMMRd could be identified. Of those, nine were suspect LS. In our center, complete LS diagnostics in approximately 5,000 CRCs left seven CRCs UMMRd. All had a somatic MMR hit or MMR germline VUS, indicative of a missed second MMR hit.

Conclusions In nearly all UMMRd patients, complete LS diagnostics suggest MMR gene involvement. Optimizing detection of currently undetectable PVs and VUS interpretation might explain virtually all UMMRd CRCs, considering UMMRd a case closed.

Keywords Lynch syndrome, Lynch-like syndrome, colorectal cancer, unexplained mismatch repair deficiency, somatic tumor testing.

PP-014: Clinical: Diagnosis, Endoscopy, Management, Implementation

Worldwide Prevalence of Lynch Syndrome in Colorectal Cancer Patients: Systematic Review and Meta-analysis

Nadine Abu Ghazaleh¹, Varun Kaushik², Alexandra Gorelik³, Mark Jenkins⁴, Finlay Macrae ²

¹Department of Medicine, University of Melbourne. ²Department of Colorectal Cancer and Genetics, Royal Melbourne Hospital. ³Monash Department of Clinical Epidemiology, Cabrini Institute. ⁴School of Population and Global Health, University of Melbourne

Background and aim Lynch Syndrome (LS) is the most common hereditary colorectal cancer (CRC) syndrome with an estimated prevalence of 2–3% of CRC. Universal screening of individuals with newly diagnosed CRC for LS has been considered. Although clinically prudent, this will place extra demands on the limited assets available to global healthcare systems. A prevalence study is needed to provide healthcare leaders with an accurate estimate of the true prevalence of LS. Establishing robust prevalence estimates in different populations and ethnicities provides a benchmark for programs aiming to identify LS carriers in their own populations. Furthermore, a review of the current landscape of screening and diagnostic strategies for LS identification will inform workforce and colonoscopy provision planning and develop optimal strategies for effective detection and preventative actions.

Method In this study, we perform a systematic review to identify the global prevalence of LS in CRC patients as well as assess multiple parameters that influence the prevalence of LS. These include age, ethnicity, intervention and diagnostic strategy, pathogenicity assignment criteria, number of MMR genes tested, and subject setting selection. MEDLINE (Ovid), Embase, Web of Science, were searched for all studies that involved screening for LS using germline testing, published between 2000 and 2020. After initial abstract and title screening, two reviewers reviewed full text articles and extracted data independently with disagreements being settled through consensus. Prevalence was calculated by random effects meta-analysis models. I2 score was used to assess heterogeneity across studies. Meta regression was performed for between study variance. To our knowledge, this is the first systematic review and meta-analysis of the worldwide prevalence of LS in CRC.

Results 51 studies were included in this review. The overall pooled yield of LS screening was 2.2% based on all methods of detection. Studies performing germline tests on all CRC participants reported higher prevalence (5.1%) as opposed to studies only performing germline tests on participants with tumours with MMR deficiency (1.6%) or MSI (1.1%). Selected cohorts of CRC had a higher prevalence of germline LS diagnoses.

Conclusions LS prevalence across multiple ethnic, geographic, and clinical populations is remarkably similar. Universal germline testing of patients presenting with cancer identifies the most CRC attributed to LS. Young patients presenting with CRC and those that fulfil a criteria referring a familial risk provide the highest returns for LS identification. Our study supports the universal germline CRC screening for LS.

Keywords Lynch Syndrome, Colorectal Cancer, Germline mutations, Prevalence.

Prevalence Map

This prevalence map is based on the 38 unselected studies. USA country had the highest number of eligible studies, followed by Australia and Japan. The colour corresponds to the Lynch Syndrome prevalence.

PRISMA Chart

PP-015: Clinical: Diagnosis, Endoscopy, Management, Implementation

A systematic review of clinical effectiveness, detection rates, test accuracy and harms related to gynaecological surveillance in Lynch Syndrome

Helen Coelho¹, Simon Briscoe¹, Nia Morrish¹, Kate Boddy¹, Tracy Smith², Fiona Lalloo ³, Emma Crosbie⁴, Neil Ryan⁵, Claire Hulme¹, Tristan Snowsill¹

¹University of Exeter Medical School, University of Exeter, Exeter, United Kingdom. ²Lynch Syndrome UK (UK registered charity 1,161,840), United Kingdom. ³Manchester University NHS Foundation Trust, Manchester, United Kingdom. ⁴Division of Cancer Sciences, University of Manchester, Manchester, United Kingdom. ⁵University Hospitals Bristol and Weston NHS Foundation Trust, Bristol, United Kingdom

Background and aim Lynch syndrome is an inherited autosomal dominant condition, primarily associated with an increased risk of colorectal, endometrial (EC), and ovarian (OC) cancers. Colorectal surveillance is common, but surveillance for EC/OC is not always offered and risk-reducing surgery (RRS) is usually recommended. For those who have not yet completed their families, there is a need to know how effective surveillance is. This research aimed to systematically review the research evidence related to clinical effectiveness, detection rates, diagnostic test accuracy (DTA) and harms of gynaecological surveillance in Lynch syndrome. Funding: NIHR HTA programme (NIHR129713).

Method Studies were identified through systematic searches in CENTRAL (via the Cochrane Library), CINAHL (via EBSCO), MEDLINE (via Ovid), Embase (via Ovid), Web of Science SCI and CPCI—S (via Clarivate Analytics), and through citation chasing, ongoing trial registries and conference proceedings. Both title and abstract screening and screening of full text articles were performed independently by two reviewers using pre-specified inclusion criteria. Data extraction and risk of bias assessments (using the CASP cohort study checklist and additionally ROBINS-I for comparative studies and QUADAS-2 for studies providing DTA data) were performed in the same way. Data were tabulated and summarised narratively. Test accuracy data were validated in STATA using the diagti command.

Results 30 studies were eligible for inclusion (ten compared different surveillance time periods or surveillance with RRS or no surveillance). These studies were poorly reported and generally at serious risk of bias. EC and OC detection rates were low across studies. Although one study reported lower rates of EC and OC and lower all-cause mortality with surveillance than with no surveillance, this may be due to sampling biases. Limited data also suggested that all-cause mortality might be lower with RRS than with surveillance. There was some indication of reduced HADS depression and anxiety scores and an improvement in SF-36 QoL scores over time in people undergoing surveillance, but this was based on single-arm data. There were insufficient DTA data reported to draw any firm conclusions about the accuracy of any of the individual tests. There was evidence to suggest that patients found biopsies moderately painful, and this might impact upon the desire to undergo the procedure. Concerns about the impact of hysteroscopy on infections and fertility were also found.

Conclusions There is no good evidence to suggest that gynaecological surveillance can effectively detect EC or OC that would not otherwise be detected, or impact upon mortality and survival. However, there is some limited and preliminary indication that this might be the case. This may be an issue with biases in the research evidence, so large-scale, prospective, comparative studies are needed.

Disclaimer: This report presents independent research commissioned by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, MRC, CCF, NETSCC, the Health Technology Assessment programme or the Department of Health.

Keywords Lynch syndrome, surveillance, gynaecological cancer, endometrial cancer, ovarian cancer.

PP-016: Clinical: Diagnosis, Endoscopy, Management, Implementation

Adenomatous Polyposis Phenotype in BMPR1A and SMAD4 Mutation Carriers

Guy Rosner ¹, Yael Petel Galil¹, Ido Laish², Zohar Levi³, Revital Kariv¹, Hana Strul¹, Ophir Gilad¹, Nathan Gluck¹

¹Department of Gastroenterology and Liver Diseases, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel and Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel. ²Gastroenterology Institute, Sheba Medical Center, Tel Hashomer, Ramat Gan, Israel and Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel. ³Department of Gastroenterology, Beilinson Hospital, Rabin Medical Center, Petach Tikva, Israel and Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel

Background and aim Juvenile polyposis syndrome (JPS) is a rare autosomal dominant condition caused by SMAD4 or BMPR1A mutations and characterized by multiple gastrointestinal (GI) hamartomatous polyps. A clinical phenotype of attenuated adenomatous polyposis without hamartomatous polyps is rarely described in JPS.

Method Retrospective cohort study of patients with SMAD4 or BMPR1A genetic variants having multiple colorectal adenomas without hamartomatous polyps. All patients had multigene panel and MLPA analysis to exclude other genetic syndromes.

Results Study cohort included eight unrelated patients with ≥ 10 cumulative colonic adenomas found to be heterozygotes for a germline variant in SMAD4 or BMPR1A genes. Variants detected included four missense mutations, two large scale deletions, one splice site mutation, and one nonsense mutation. Co-segregation of the variant with polyposis or colorectal cancer (CRC) was obtained in six of the eight families. Three of the eight patients had CRC (age < 50) in addition to polyposis phenotype. Two patients had neoplastic extraintestinal involvement (pancreas and papilla of Vater).

Conclusions The clinical phenotype of SMAD4 and BMPR1A mutations extends, in few cases, beyond the classical JPS phenotype. Applying multigene panel analysis of hereditary cancer-related genes provides syndrome-based clinical surveillance for the carriers and their family members.

Keywords Polyposis, Juvenile, Genetics, Cancer, Phenotype.

PP-017: Clinical: Diagnosis, Endoscopy, Management, Implementation

Intestinal and extraintestinal neoplasms in patients with NTHL1 tumor syndrome: a systematic review

Søren Hammershøj Beck¹, Anne Marie Jelsig², Halimo Mohamed Yassin¹, Lars Joachim Lindberg³, Karin Wadt⁴, John Gásdal Karstensen ⁵

¹University of Copenhagen, Faculty of Health and Medical Sciences, Denmark. ²Department of Clinical Genetics, Copenhagen University Hospital, Rigshospitalet, Denmark. ³Danish HNPCC register, Gastrounit, Copenhagen University Hospital – Amager and Hvidovre, Hvidovre, Denmark. ⁴Department of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark. ⁵Danish Polyposis Registry, Gastrounit, Copenhagen University Hospital – Amager and Hvidovre, Hvidovre, Denmark

Background and aim Germline biallelic pathogenic variants (PVs) in NTHL1 have since 2015 been associated with the autosomal recessive tumor predisposition syndrome: NTHL1 tumor syndrome or NTHL1-associated polyposis. In this systematic review, we aim to systematically investigate the phenotypic and genotypic spectrum of the condition including occurrence of both benign and malignant tumors.

Method The databases PubMed, EMBASE, and Scopus were searched. The search was conducted the 25th of august 2021. We included patients with germline PVs, both heterozygous and homo-/compound heterozygous carriers.

Results Twenty-one papers were selected including 47 patients with biallelic PVs in NTHL1 in 32 families. Twenty-three out of 47 patients (49%) were diagnosed with colorectal cancer (CRC) (mean age: 55, range: 31–73) and 12 out of 22 female patients (55%) were diagnosed with breast cancer (mean age: 49, range: 36–63). Apart from three, all patients who underwent a colonoscopy, had colonic adenomas (93%), and three patients (6%) had duodenal adenomatosis. We also identified 158 heterozygous carriers of germline PVs in NTHL1. Twenty-six out of 68 (38%) heterozygous carriers, who underwent colonoscopy, had colonic polyps or adenomas. Twenty-nine heterozygous carriers (18%) were diagnosed with CRC and 59 (49%) with breast cancer.

Conclusions We observed a high frequency of early onset CRC and breast cancer in patients with NTHL1 tumor syndrome. Subsequently, colorectal, breast, and endometrial cancer screening programs are recommended for NTHL1 biallelic carriers.

Keywords NTHL1-associated polyposis, NTHL1 tumor syndrome, Intestinal neoplasms, extraintestinal neoplasms, colorectal cancer, breast cancer.

PP-018: Clinical: Diagnosis, Endoscopy, Management, Implementation

Young-Onset Colorectal Neoplasia Risk in Individuals Undergoing Diagnostic Colonoscopy

Tara Anastasia Russell¹, Avia Wilkerson¹, Emre Gorgun¹, Carol A Burke³, Carole Macaron³, Suneel Kamath⁴, Kanika Nair⁴, Joshua Sommovilla ², Michael Valente¹, Alok Khorana⁴, David Liska²

¹Department of Colorectal Surgery, Digestive Disease & Surgery Institute, Cleveland Clinic, Cleveland, OH, USA. ²DeBartolo Cancer for Young-Onset Colorectal Cancer, Cleveland Clinic, Cleveland, OH, USA. ³Department of Gastroenterology Hepatology and Nutrition, Digestive Disease & Surgery Institute, Cleveland Clinic, Cleveland, OH, USA. ⁴Department of Hematology and Medical Oncology, Taussig Cancer Institute and Case Comprehensive Cancer Center, Cleveland Clinic, Cleveland, OH

Background and aim Given the increasing incidence of young-onset colorectal cancer (YOCRC), individuals with symptoms concerning for colorectal cancer, regardless of age, are encouraged to seek medical attention. There is currently a paucity of data on the frequency of neoplastic colorectal lesions in symptomatic patients < 50 years of age. This study aimed to evaluate the prevalence of neoplastic findings among young adults undergoing diagnostic colonoscopy.

Method Retrospective review of all adults age 18–49 undergoing diagnostic colonoscopy from 2011–2021 at a tertiary academic medical center and affiliate community sites. Patients undergoing high risk screening due to a family history of polyposis or cancer, surveillance due to a personal history of polyps or cancer, or evaluation for known inflammatory bowel disease were excluded. Patient demographics, indications for colonoscopy and pathologic findings were analyzed using descriptive statistics. A positive finding was defined as any adenomatous or serrated lesion, excluding hyperplastic polyps < 10 mm. Advanced neoplasia included advanced adenomas (high-grade dysplasia, ≥ 25% villous component, or ≥ 10 mm diameter), advanced serrated lesions (dysplasia or ≥ 10 mm diameter), and invasive adenocarcinoma.

Results 50,270 young adults underwent colonoscopy, of which 30,933 (61.5%) were diagnostic examinations. The median age was 39 and 60.7% were female. The most common indications for diagnostic colonoscopy were change in bowel habits (37.9%), bleeding (31.3%), and pain (17.2%). 2,965 (9.6%) had neoplastic lesions, 677 (2.2%) had advanced neoplasia and 12 (0.04%) had invasive adenocarcinoma. The prevalence of neoplastic lesions and detection of advanced neoplasia increased with age at the time of colonoscopy (p < 0.001). The prevalence of neoplastic lesions or advanced neoplasia was greatest amongst patients with bleeding, anemia, weight loss, and abdominal mass. Over the study period the prevalence of neoplastic lesions and advanced neoplasia increased (p < 0.05).

Conclusions Although the incidence of invasive adenocarcinoma is low amongst young adults undergoing diagnostic colonoscopy, the prevalence of advanced neoplasia amongst those with more classic alarm symptoms (bleeding or weight loss) exceeds 3%. Regardless of age, in the setting of these symptoms, colonoscopy should be strongly considered.

Keywords Neoplasia, Colonoscopy, Prevalence.

PP-019: Clinical: Diagnosis, Endoscopy, Management, Implementation

From Diagnosis of Colorectal cancer to Diagnosis of Lynch Syndrome: The RM Partners Quality Improvement Project

Laura Monje Garcia¹, Kevin Monahan ¹, Timothy Bill², Lindsay Farthing², Nate Hill², Ricki Ostrov², Emma Kipps², Angela Brady³, Zoe Kemp⁴, Katie Snape⁵

¹St Mark’s Hospital Centre for Familial Intestinal Cancer, Imperial College London, London, UK. ²RM Partners Cancer Alliance, London, UK. ³North West Thames Regional Genetics Service, London, UK. ⁴The Royal Marsden Hospital Cancer Genetics Unit, London, UK. ⁵South West Thames Centre for Genomics, London, UK

Background and aim The UK NICE (National Institute for Health & Care Excellence) DG27 guidelines recommends universal testing for Lynch syndrome (LS) in all newly diagnosed Colorectal Cancer (CRC) cases. However, implementation of the DG27 guideline varies significantly by geography. This Quality Improvement Project (QIP) was developed to measure variation and deliver an effective diagnostic pathway from diagnosis of colorectal cancer to diagnosis of LS within the RM Partners (RMP) cancer alliance geography.

Method RMP covers a population of 4 million people and incorporates 9 CRC multidisciplinary teams (MDT) overseen by a Pathway Group, and 3 regional genetic services, managing approximately 1500 new CRC cases annually. A responsible LS champion was nominated within each MDT. A regional project manager and nurse practitioner were appointed to support the LS champions, to develop online training packages and patient consultation workshops. MDTs were supported to develop an ‘in-house’ mainstreaming service to offer genetic testing in their routine oncology clinics. Baseline data was collected through completion of the LS pathway audit of the testing pathway in 30 consecutive CRC patients from each CRC MDT. This information identified areas for improvement in each MDT.

Results In baseline audit the first step of the testing pathway, tumour mismatch repair testing, was performed in 93% of 270 CRC patients. However only 23% (7/30) of eligible patients underwent methylation testing, and only 9% (2/22) eligible for constitutional testing were referred to a genetics service, with high levels of variation between each MDT. During the QIP we have developed new mainstreaming services and demonstrated implementation of systematic and robust testing pathways across the cancer alliance. We have implemented systematic approaches to facilitate identification and diagnosis of eligible patients for LS testing after CRC diagnosis across each cancer team.

Conclusions The LS project will be completed by April 2022. This work has led to the development of a new NICE standard QS30 in England which recommends local leadership within cancer teams to ensure delivery of diagnosis of LS. We have implemented a systematic approach with workforce transformation to facilitate identification and ‘mainstreamed’ genetic diagnosis of LS. This programme has now evolved to be one of the UK national genomics transformational projects which will integrate genomics into clinical practice.

Keywords Lynch syndrome, Quality Improvement Project, diagnosis, MMR IHC, mainstreaming.

PP-020: Clinical: Diagnosis, Endoscopy, Management, Implementation

Performance of InheRET™ Genetic Risk Assessment Tool for Identifying Pathogenic Germline Variants among Individuals Referred for Cancer Genetics Evaluation

Marie Louise Henry, Kyra Pialtos, Stacy Fry, Erika Koeppe, Jenae Osborne, Amanda Cook, Bailey Hulswit, Kara J. Milliron, Sofia Merajver, Elena Stoffel

Cancer Genetics Clinic, University of Michigan Rogel Cancer Center

Background and aim Identification of individuals with inherited susceptibility to cancer offers opportunities for early detection and prevention. The National Comprehensive Cancer Network (NCCN) has guidelines for genetic referral, however these complex criteria can be challenging to apply. InheRET™ employs a patient-facing health survey and a risk assessment algorithm to identify individuals who meet NCCN criteria for genetic evaluation. Our goal was to examine the performance of InheRET™ for identifying individuals with a pathogenic germline variant (PGV) conferring inherited susceptibility to cancer.

Method Individuals referred for genetics evaluation at an academic cancer genetics clinic between 8/2021 – 2/2022 were invited to complete InheRET™ prior to their appointment. InheRET™ reports (met vs. did not meet genetic referral criteria) were compared with outcomes of genetic testing.

Results Seventy individuals completed InheRET™ and underwent germline genetic testing; mean age was 51.5 years [23–76]; 54.3% were female and 35.7% had a personal history of cancer. Forty-one of 70 (58.6%) were flagged by InheRET™ as meeting NCCN criteria for genetics referral; 20 (28.6%) did not meet criteria, and 9 (12.8%) were classified as “clinician discretion.” PGVs in cancer susceptibility genes were identified in 12 (17%) (ATM n = 2, BRCA2 n = 1, CDKN2A n = 1, CHEK2 n = 4, PMS2 n = 1, SDHC n = 1, SMAD4 n = 1, TP53 n = 1). Eight PGV carriers were flagged as meeting genetics referral criteria (n = 6) or clinician discretion (N = 2, both with a CHEK2 low-penetrance variant). Three PGV carriers were not flagged due to user data entry error (failed to indicate that PGV had been diagnosed in a family member). The one PGV carrier truly missed by InheRET had a history of 12 colorectal adenomas, no personal or family history of cancer, with an unexpected PGV in BRCA2 identified on multigene panel testing. All 11 individuals with personal history of gastrointestinal cancer (colorectal n = 6, pancreatic n = 4, appendiceal n = 1) were appropriately flagged as meeting genetics referral criteria.

Conclusions InheRET™’s NCCN-based algorithm identified cancer-affected and unaffected individuals meeting criteria for genetic evaluation for various hereditary cancer syndromes.

Keywords Genetic risk assessment.

PP-021: Clinical: Diagnosis, Endoscopy, Management, Implementation

Colonoscopy findings in patients with a germline pathogenic variant in CDH1

Arjun Chaterjee¹, Lady Katherine Mejia Perez ², Margaret O'malley³, Lisa Laguardia³, David Liska³, Susan Milicia³, Joshua Sommovilla³, Nicholas Smith³, Carole Macaron², Carol Burke²

¹Department of Internal Medicine, Community Care Institute, Cleveland Clinic, Cleveland, Ohio, USA. ²Department of Gastroenterology, Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgery InstituteCleveland Clinic, Cleveland, Ohio, USA. ³Department of Colorectal Surgery, Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, Ohio, USA

Background and aim Germline pathogenic variants in CDH1 predispose to hereditary cancer. The relationship between pathogenic variants in CDH1 and hereditary diffuse gastric cancer and lobular breast cancer is well established. There are no data on the risk of colorectal neoplasia in carriers of CDH1 pathogenic variants. Our aim was to investigate the colonoscopy findings in patients with a pathogenic variant in CDH1 undergoing colorectal cancer screening.

Method This IRB-approved study identified patients with germline pathogenic variants in CDH1 throughout the Cologene™ database in the David G. Jagelman Inherited Colorectal Cancer Registry. The electronic medical record was used to obtain information, including demographics, personal and family history of cancer, and colonoscopy findings, including adenomas, advanced adenoma (≥ 3 adenomas, or an adenoma ≥ 10 mm, or with villous features or high-grade dysplasia), and invasive adenocarcinoma.

Results 34 patients (70.6% female) with a mean age of 53 years and ≥ 1 colonoscopy were included (table). 29/34 (85.2%) patients had a personal history of gastric or breast cancer and 11 patients (32.4%) had a family history of CRC or advanced adenoma. 18 patients (52.9%) were diagnosed with any type of polyp including adenomas, advanced adenomas, and hyperplastic polyps. Adenomas were found in 12 patients (35.3%, at a mean age 54.6 years), advanced adenomas in 4 patients (11.8%, at a mean age 54.1 years), and CRC in 1 patient (3%) at age 49.2 years.

Conclusions In our small cohort of carriers of pathogenic variants in CDH1 undergoing screening colonoscopy, we found one patient (3%) with early onset CRC and a relatively high incidence of early onset adenomas, and advanced adenomas. Larger studies from other centers are recommended to expand the knowledge of the prevalence of colorectal neoplasia in carriers of pathogenic CDH1 carriers.

Keywords CDH1 pathogenic variants; colorectal cancer; advanced adenoma; screening colonoscopy.

Demographics and colonoscopy findings of patients with CDH1 pathogenic variants

Factors	Number of Carriers = 34
Age (years)	52.6 [48.7;60.7]
Gender: Female/Male	24 (70.6%)/ 10 (29.4%)
Personal History of Gastric Cancer/Breast Cancer	16 (47.1%)/13 (38.2%)
Family History of Colorectal Cancer or Advanced Adenoma	11 (32.4%)
Age at 1st colonoscopy (years)	50.1 [46.7;57.8]
Number of colonoscopies per patient
1	24 (71%)
2	7 (21%)
3	3 (9%)
Number of Patients with Polyps	18 (53%)
Age at First Polyp (years)	51.6 [49.4;61.8]
Number of Patients with Hyperplastic Polyps	4 (12%)
Number of Patients with Tubular Adenomas < 10 mm	12 (35%)
Number of Patients with Advanced Adenomas	4 (12%)
Tubulovillous/Villous adenoma	3 (9%)
High-Grade Dysplasia	1 (3%)
Number of Patients with CRC	1 (3%)
Age at first Tubular Adenoma (years)	54.6 [47.4;63.3]
Age at first Advanced Adenoma (years)	54.1 [50.4;58.0]
Age at CRC (years)	49.2

Data is presented as median and quartiles [25th, 75th percentiles] or frequency (percent). *Advanced adenomas: ≥ 3 tubular adenomas, or adenoma ≥ 10 mm or with villous features, or high grade dysplasia

PP-022: Clinical: Diagnosis, Endoscopy, Management, Implementation

Prevalence of Adenomatous Polyposis in Lynch Syndrome

Ayushi Jain¹, Maryam Alimirah², Heather Hampel³, Rachel Pearlman⁴, Jianing Ma⁵, Jing Peng⁵, Matthew F. Kalady⁶, Peter P. Stanich ²

¹Department of Internal Medicine, The Ohio State University Wexner Medical Center, Columbus, OH, USA. ²Division of Gastroenterology, Hepatology & Nutrition, The Ohio State University Wexner Medical Center, Columbus, OH, USA. ³Division of Cancer Genomics, City of Hope National Medical Center, Duarte, CA, USA. ⁴Division of Human Genetics, The Ohio State University Wexner Medical Center, Columbus, OH, USA. ⁵Center of Biostatistics, Department of Biomedical Informatics, The Ohio State University Wexner Medical Center, Columbus, OH, USA. ⁶Division of Colorectal Surgery, The Ohio State University Wexner Medical Center, Columbus, OH, USA

Background and aim Adenomatous oligopolyposis is defined as 10 or more cumulative colorectal adenomas and is associated with many hereditary polyposis syndromes. Lynch syndrome (LS), previously hereditary nonpolyposis colorectal cancer syndrome, has not traditionally been considered in this category. However, as adenoma detection rates rise in the general population, cumulative adenoma counts in LS may also be increasing. Our aim was to assess the prevalence of adenomatous oligopolyposis in LS.

Method A retrospective review of patients followed in the Hereditary and High-Risk GI Clinic for LS from August 2014 to December 2020 was approved by the IRB. Inclusion criteria included age ≥ 18 years, pathogenic/likely pathogenic variant in a mismatch repair gene and at least one colonoscopy. Exclusion criteria included multiple hereditary cancer syndromes and total colectomy prior to identification of first adenoma. Advanced neoplasia was defined as colorectal cancer, advanced adenoma, or advanced sessile serrated lesion.

Results 222 patients met study criteria. The mean age was 48 years (standard deviation 13 years) and the majority were women (68.0%) and Caucasian (95%). The breakdown of mismatch repair gene mutations was: MSH6 (34%), MSH2 (26%), PMS2 (22%) and MLH1 (18%). A personal history of colorectal cancer was noted in 18% of patients and a family history of colorectal cancer in a first degree relative was noted in 46% of patients.

Of the 222 patients included, 142 (64.0%) had documented adenomas. 14 patients (6.3%) met criteria for adenomatous oligopolyposis with a median of 12 cumulative adenomas (range 10–28). They had a mean age of 62 and the most common mutation types were MSH6 (8/14) and MSH2 (4/14). Most notably, patients with adenomatous oligopolyposis were at a significantly increased risk of advanced neoplasia [OR 10, 95% CI: 2.7–66.7] (Table 1).

Conclusions Adenomatous oligopolyposis is not unusual in LS and is associated with a tenfold increase in risk for advanced colon neoplasia. Consideration should be given to differentiating colonoscopy intervals based on the presence of oligopolyposis in LS, especially given the recent trend towards longer intervals in guidelines. In addition, while oligopolyposis generally raises concern for traditional polyposis syndromes, the mismatch repair genes should be included in the multigene panel testing recommended for these patients.

Keywords Lynch syndrome, polyposis, adenomas.

Table 1. Characteristics of Patients with Adenomatous Oligopolyposis

PP-023: Clinical: Diagnosis, Endoscopy, Management, Implementation

Increased Extracolonic Gastrointestinal Cancers in older Lynch Syndrome patients justify reappraisal of surveillance recommendations

Samara Rifkin, Tannaz Guivatchian, Anthony Scott, Erika Koeppe, Kara Jencks, Danielle Kim Turgeon, Elena M. Stoffel

University of Michigan Cancer Genetics Clinic

Background and aim Lynch syndrome (LS) is a highly penetrant CRC syndrome caused by inherited defects in DNA mismatch repair genes. LS is also associated with increased risk for extra colonic gastrointestinal cancers. Our aim was to describe the prevalence of these cancers and associated clinical characteristics in a longitudinal LS cohort.

Method Patients with genetically confirmed LS seen through our institution’s cancer genetics clinic between 1998–2021 were included in the study. Medical records were audited for cancer diagnoses, genetic variant, sex, age at diagnoses, medical comorbidities, and outcomes of surveillance tests. Chi-square and t-tests were utilized to identify clinical features associated with extracolonic gastrointestinal cancers.

Results Our cohort included 501 individuals diagnosed with LS (MLH1 112, MSH2 196, MSH6 116, PMS2 70, EPCAM 8). The average age at last follow up was 46.6 and 60% were female. 265/494 (54%) individuals had a personal history of cancer, including 155 (31%) with CRC. 37/494 (7.5%) had been diagnosed with extracolonic GI cancers, including 7 with gastric (1.4%), 14 with small bowel (2.8%), 11 with pancreatic (1.8%), 3 with esophageal (0.08%) and 2 (0.6%) with bile duct cancer. Ages at diagnosis for extracolonic GI cancers ranged from 24 (duodenal cancer) to 81 (gastric cancer). Pathogenic variants in MLH1 (26% of vs. 20%) and MSH2 (66% of vs. 36%, p-value =  < 0.001) were more frequent among LS patients with Extracolonic GI cancers. LS patient with extracolonic GI cancer were more often male (58% vs. 42%, %, p-value = 0.02), tobacco users (55% vs. 27%, p-value = 0.01) and older age (56 vs. 47, %, p-value = 0.001). Of the 7 patients with gastric cancer, all were age > 60, and 5 had undergone EGD within 18 months of diagnosis. Four of the 11 small bowel cancers were located in the duodenum and were found during diagnostic work up for symptoms. Four of the 11 patients with pancreatic cancer had family history. None of the patients diagnosed with pancreatic or bile duct cancers had undergone prior imaging.

Conclusions Nearly 1 in 14 LS patients developed extracolonic GI cancers. All incident gastric cancers were diagnosed age > 60 and 70% had undergone an EGD within 17 months prior to diagnosis. While larger prospective studies are needed to confirm the most appropriate intervals for endoscopic surveillance, annual EGD should be considered for MLH1/MSH2 carriers age > 60.

Keywords Lynch Syndrome (LS), extracolonic gastrointestinal cancers.

Table 1

PP-024: Clinical: Diagnosis, Endoscopy, Management, Implementation

Reviewing the clinical management of adrenal incidentaloma in familial adenomatous polyposis

Vicky Cuthill, Jeshu Chauhan, Susan Clark, Andrew Latchford

St Mark's Centre for Familial Intestinal Cancer, St Mark's Hospital, London North West University Healthcare NHS Trust, London, UK

Background and aim Adrenal incidentaloma is a known extraintestinal manifestation of familial adenomatous polyposis (FAP) but there is little information about their natural history. We instituted a management algorithm in 2008 (Fig. 1). We aimed to review the outcomes and audit the management algorithm.

Method Patients with a confirmed APC pathogenic variant and adrenal incidentaloma were identified from a prospectively maintained database. Medical records, pathology and radiology reports were reviewed.

Results Ninety-one (49 male) patients were identified. The median age at diagnosis of adrenal mass was 45 years (range 19–72). All were discovered incidentally, in patients undergoing CT scan for abdominal pain, desmoid surveillance or colorectal cancer follow-up. Patients were followed up for a median of six years radiologically (range 4 – 16) and six years clinically (range 1–18). Thirty-three patients had reached the end of the planned radiology surveillance period (Table 1).

The median adrenal mass size at identification was 18 mm (range 3–60). Sixty-six patients had more than one scan and median maximum diameter during follow up was 21 mm (range 5 – 60). Interval growth was observed in 18/66; median size increase was 9 mm (range 4 – 20).

Nine patients underwent adrenalectomy, all confirmed to have benign histology. In two, synchronous adrenalectomy was performed at the time of nephrectomy for renal cell cancer. Six patients were referred to endocrinology and underwent surgery based on incidentaloma size. One patient had surgery for a hyperfunctioning 19 mm adenoma, having presented with hypertension and hypokalaemia and diagnosed with primary hyperaldosteronism. This was the only patient with a known hyperfunctioning adenoma.

Seventy-five patients were diagnosed after our current management algorithm was initiated in 2008. Complete functional screening was available in only 8/75 with 67 patients having absent or incomplete functional screening (Table Tab2).

Conclusions Our data shows that there are often barriers to functional screening. No growth on surveillance was seen in 73%. Our data supports modifying our current algorithm to ensure correlation with the current European Society of Endocrinology Current Practice Guideline which would overcome these barriers and reduce radiological follow-up.

Keywords Adrenal incidentaloma, adrenalectomy, FAP, clinical management.

Fig. 1

Current management algorithm for adrenal incidentaloma

Table 1

Reason for not completing radiological surveillance	Number of patients
Not reached end of scanning period	33
Died/palliative diagnosis prior to surveillance end	7
Adrenalectomy	9
Discharged to local follow-up	4
Lost to follow-up	4
Small bowel transplant	1
Total	58

Reasons for not completing radiology surveillance

Table 2

Reason for unavailability of function screening results	Number of patients
Functional screening not performed	48
ACTH requested but not performed by laboratory	7
Metastatic disease	3
Managed locally	2
Urinalysis performed but not ACTH	2
ACTH performed but not urinalysis	2
Referred directly to endocrinology	1
Results requested via GP, results not available	2
Total	67

Reasons for unavailability of function screening results

PP-025: Clinical: Diagnosis, Endoscopy, Management, Implementation

Development and External Validation of a Novel Nomogram for Screening Chinese Lynch Syndrome: Based on A Multicenter, Population Study

Ying Yuan ¹, Mengyuan Yang¹, Dan Li¹, Wu Jiang², Lizhen Zhu¹, Haixing Ju³, Yan Sun⁴, Yuqiang Shan⁵, Chunkang Yang⁶, Jian Dong⁷, Lin Wang⁸, Baoping Wu⁹, Meng Qiu¹⁰, Xianli Yin¹¹, Xicheng Wang¹², Bin Xiong¹³, Wei Yan¹⁴, Tao Liu¹⁵, Chenglin Liu¹⁶, Xinru Mao¹⁷, Kefeng Ding¹⁸, Suzhan Zhang¹⁸, Shu Zheng¹⁸, Dong Xu¹⁸, Peirong Ding²

¹Department of Medical Oncology, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China; Cancer Institute, Key Laboratory of Cancer Prevention and Intervention, Ministry of Education, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China. ²State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China; Department of Colorectal Surgery, Sun Yat-sen University Cancer Center, Guangzhou, China. ³Department of Colorectal Surgery,Cancer Hospital of the University of Chinese Academy of Sciences (Zhejiang Cancer Hospital), Hangzhou, China. ⁴Department of Pathology, Tianjin Medical University Cancer Institute and Hospital, National Clinical Research Center for Cancer, Key Laboratory of Cancer Prevention and Therapy, Tianjin, China. ⁵Department of General Surgery, Affiliated Hangzhou First People's Hospital, Zhejiang University School of Medicine, Hangzhou, China. ⁶Department of Gastrointestinal Surgical Oncology, Fujian Cancer Hospital and Fujian Medical University Cancer Hospital, Fuzhou, China. ⁷Department of Colorectal Surgery, The Third Affiliated Hospital of Kunming Medical University, Yunnan Cancer Hospital, Kunming, China. ⁸Department of Medical Oncology, 81st Hospital Affiliated to Nanjing University of Chinese Medicine, Nanjing, China. ⁹Department of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou, China. ¹⁰Department of Medical Oncology, Cancer Center, West China Hospital, West China Medical School, Sichuan University, Chengdu, China. ¹¹Department of Gastroenterology and Urology, Hunan Cancer Hospital, The Affiliated Cancer Hospital of Xiangya School of Medicine, Central South University, Changsha, China. ¹²Department of Gastrointestinal Oncology, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education), Peking University Cancer Hospital and Institute, Beijing, China. ¹³Department of Gastrointestinal Surgery, Zhongnan Hospital of Wuhan University, Wuhan, China. ¹⁴Department of Pathology, The First Affiliated Hospital of Air Force Medical University, Xian, China. ¹⁵Department of Gastrointestinal Surgery, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China. ¹⁶Department of Bioinformatics, Burning Rock Biotech, Guangzhou, China. ¹⁷Medical Department, Burning Rock Biotech, Guangzhou, China. ¹⁸Department of Colorectal Surgery and Oncology, Key Laboratory of Cancer Prevention and Intervention, Ministry of Education, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China

Background and aim This multicenter study aimed to reveal the genetic spectrum of colorectal cancer (CRC) with deficient mismatch repair (dMMR) and build a screening model for Lynch syndrome (LS).

Method Through the immunohistochemical (IHC) screening of mismatch repair protein results in postoperative CRC patients, 311 dMMR cases, whose germline and somatic variants were detected using the ColonCore panel, were collected. Univariate and multivariate logistic regression analysis was performed on the clinical characteristics of these dMMR individuals, and a clinical nomogram, incorporating statistically significant factors identified using multivariate logistic regression analysis, was constructed to predict the probability of LS. The model was externally validated by an independent cohort containing 259 dMMR cases.

Results In total, 311 CRC patients with IHC dMMR included 95 identified MMR germline variant (LS) cases and 216 cases without pathogenic or likely pathogenic variants in MMR genes (non-Lynch-associated dMMR). Of the 95 individuals, approximately 51.6%, 28.4%, 14.7%, and 5.3% cases carried germline MLH1, MSH2, MSH6, and PMS2 pathogenic or likely pathogenic variants, respectively.

A novel nomogram was then built to predict the probability of LS for CRC patients with dMMR intuitively (Fig. 1A). In order to make this screening model more directly available to the public, a small program has been developed based on this model (Fig. 1B). Whether it is a patient, a family member or a doctor, this small program can be easily and conveniently used to estimate the probability of LS.

The receiver operating characteristic (ROC) curve informed that this nomogram-based screening model could identify LS with a higher specificity and sensitivity with an area under curve (AUC) of 0.87 than current screening criteria based on family history (Fig. 2A). In the external validation cohort, the AUC of the ROC curve reached 0.804, inferring the screening model’s universal applicability (Fig. 2B). We recommend that dMMR-CRC patients with a probability of LS greater than 0.435 should receive a further germline sequencing.

Conclusions This novel screening model based on the clinical characteristic differences between LS and non-Lynch-associated dMMR may assist clinicians to preliminarily screen LS and refer suspectable patients to experienced specialists.

Keywords deficient mismatch repair, Lynch syndrome, non-Lynch-associated dMMR, colorectal cancer, nomogram.

PP-026: Early Onset Colorectal Cance

Norwegian PMS2 mutation spectra and prevalence of cancer in carriers of pathogenic variants

Hanne Kjensli Hyldebrandt ¹, Wenche Sjursen², Liss Anne Solberg Lavik³, Bjørn Ivar Haukanes⁴, Sarah Louise Ariansen¹, Siri Briskemyr⁵, Anna Elisabeth Sylvander³, Marianne Tveit Haavind⁴, Maren Fridtjofsen Olsen³, Elin Synnøve Røyset⁶, Hildegunn Høberg Vetti⁴, Astrid Tenden Stormorken¹, Eli Marie Grindedal¹

¹Department of Medical Genetics, Oslo University Hospital, Oslo, Norway. ²Department of Medical Genetics, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway; Department of Clinical and Molecular Medicine, Norwegian University of Science and Technology, Trondheim, Norway. ³Department of Medical Genetics, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway. ⁴Department of Medical Genetics, Haukeland University Hospital, Bergen, Norway. ⁵Department of Medical Genetics, University Hospital of North Norway, Tromsø, Norway. ⁶Department of Pathology, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway; Clinic of Laboratory Medicine, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway

Background and aim The PMS2-gene is the least described of the four MMR-genes. In Norway we have offered testing of PMS2 since 2006, and inclusion criteria for testing have not been limited to only those fulfilling strict clinical criteria like the Amsterdam criteria. After more than 15 years of testing we have a large cohort of PMS2 carriers in Norway. The aim of this study was to describe all identified PMS2 variants in Norway, and to describe prevalence of cancer in carriers of pathogenic PMS2 variants in our cohort. The study is a collaboration between all Departments of Medical Genetics offering genetic testing of PMS2 in Norway.

Method All detected PMS2 variants were collected from the diagnostic laboratories and reclassified according to the ACMG criteria. Clinical information about all carriers and obligate carriers of likely pathogenic or pathogenic variants in PMS2 was collected: gender, age at last follow-up, cancer diagnosis and age at diagnosis, and for women whether they were hysterectomized or not.

Results In total 214 different PMS2 variants were detected, 29 likely pathogenic/pathogenic (class 4 and 5), 44 variants of uncertain clinical significance (class 3) and 141 benign/likely benign (class 1 and 2). The most common variant was c.989-1G > T, which was detected in 194 patients from 58 families, most of them originating from Mid-Norway. In total, 444 heterozygous carriers and 38 obligate carriers of class 4 or 5 PMS2 variants were identified, 262 women and 220 men. Colorectal cancer (CRC) was the most common cancer in both men and women. Eighty seven (87/482 = 18.0%) had been diagnosed with CRC, mean age of diagnosis was 54.0 (range 24–86). Two of these (2/87 = 2.3%) had CRC before the age of 30, five (5/87 = 5.7%) had CRC before the age of 35, and 34 (34/87 = 39%) were diagnosed before the age of 50. Thirty two women (32/262 = 12.2%) had been diagnosed with endometrial cancer. Mean age of diagnosis was 55.4 years (range 21–86).

Conclusions Almost half of the PMS2 carriers in Norway have the founder variant 989-1G > T. Of the carriers, 18% had been diagnosed with CRC. The mean age for CRC in the carriers was 54 years old, but almost 6% of the affected had been diagnosed with CRC before the age of 35. Our findings confirm the variable penetrance of PMS2.

Keywords PMS2, colorectal cancer, endometrial cancer.

PP-027: Early Onset Colorectal Cance

Early-onset colorectal cancer: Geographical disparities according to phenotype. Data from GEOCODE- Europe

Jose Perea ¹, Luis A Corchete¹, Kevin Monahan², Antonino Spinelli³, Noel Fcc De Miranda⁴, Claire Palles⁵, Francesc Balaguer⁶, Marc Martí⁷, Alfredo Vivas⁸, Andrew Latchford⁹, Wieslaw Tarnowski¹⁰, Marek Szczepkowski¹¹, Silviu Tiberiu Makkai Popa¹², Rogelio González Sarmiento¹, Andreana N Holowatyj¹³

¹Department of Molecular Medicine, Institute of Biomedical Research of Salamanca, Salamanca, Spain. ²Chelsea and Westminster Hospital NHS Foundation Trust, London, London, United Kingdom. ³Colorectal Deparment. Humanitas Mirasole SpA, Rozzano, Lombardia, Italy. ⁴Pathology Department. Leids Universitair Medisch Centrum, Leiden, Zuid-Holland, Netherlands. ⁵Birmingham City University, Birmingham, West Midlands, United Kingdom. ⁶Deparment of Gastroenterology. Clinic Hospital of Barcelona. Barcelona, Spain. ⁷Surgery Department. Vall d´Hebrón University Hospital. Barcelona, Spain. ⁸Surgery Department. 12 de Octubre University Hospital. Madrid, Spain. ⁹London North West Healthcare NHS Trust, Harrow, London, United Kingdom. ¹⁰Samodzielny Publiczny Szpital Kliniczny im prof Witolda Orrowskiego, Warszawa, Poland. ¹¹Szpital Bielanski im ks Jerzego Popieluszki, Warszawa, Warszawa, Poland. ¹²Centre Hospitalier de Luxembourg Eich, Luxembourg, Luxembourg. ¹³Vanderbilt University Medical Center, Nashville, TN, United States

Background and aim The incidence associated with early-onset colorectal cancer (EOCRC) (age: 18–49 years) have been increasing, but this increase in incidence is heterogenous in different countries worldwide. Thus, we sought to explore disease patterns of non-hereditary EOCRC in Europe.

Method Retrospective multicenter study, including cases of EOCRC between 2010–2017, excluding hereditary syndromes, with centers from Spain (3), United Kingdom (3), Poland (2), Italy (1), the Netherlands (1) and Luxembourg (1). Data were captured and differences in clinical and demographic features were compared using chi-square and ANOVA tests between countries. To appreciate broad geographical phenotypic associations, we defined 3 population comparisons: Mediterranean countries vs others; by means of EuroVoc subregions; and by United Nations geoscheme.

Results A total of 851 cases were collected. The youngest cases were from UK (Median, 40), while the eldest were from the Netherlands (44) (P = 0.004). Earlier stages (I-II) were more frequent in cases from Poland (49.2%) and UK (42.5%), while more advanced (III-IV) in Luxembourg (81.2%) and the Netherlands (68.3%) (P = 0.00001). Non-familial cases comprised over 40% of cases in Italy, Spain and the UK (43%, 44% and 56%, respectively) (P = 0.000001). In the UK, Colorectal cancer (CRC) in 1st degree relatives was more frequent (31%) (P = 0.000001). Finally, familial early-onset cancer cases were more prevalent in UK (18.4%) and the Netherlands (17.3%) (P = 0.0000001). Concerning the aggrupation approach, the most prominent geographical classification was the one applying the United Nations geoscheme, with 16 statistically differential variables (age at diagnosis by intervals, median age at diagnosis, median of Body Mass Index, stage at diagnosis, grade of differentiation, metastasis sites at diagnosis, history, median and type of polyps, CRC in 1st degree relatives, other tumors in 1st degree relatives, digestive cancers different from CRC only in 2nd degree relatives, gynecologic and breast cancers only in 2nd degree relatives, other tumors only in 2nd degree relatives, early-onset cancers in the family, and non-familial cases).

Conclusions Within the EOCRC population, clinical and demographic features differ by geographic region in Europe. Further study of genetic, environmental and lifestyle patterns of EOCRC that may partly explain geographic differences in EOCRC presentation and outcomes is needed.

Keywords Early-onset colorectal cancer, disparities, geography, phenotypes.

PP-028: Epidemiology and Registries

Hereditary colorectal cancer syndromes and the COVID-19 pandemic: results from a survey conducted in patients enrolled in a dedicated registry

Stefano Signoroni¹, Chiara Maura Ciniselli², Guglielmo Niccolò Piozzi³, Maria Teresa Ricci¹, Marta Pastori², Paolo Merola⁴, Marco Bellazzi⁵, Tatiana Smaldone¹, Clorinda Brignola ¹, Giuseppe Rosito⁵, Claudia Borreani⁶, Paolo Verderio², Giovanni Apolone⁴, Marco Vitellaro¹

¹Unit of Hereditary Digestive Tract Tumors, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ²Unit of Bioinformatics and Biostatistics, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ³Colorectal Surgery Division, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁴Scientific Directorate, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁵Information and communication technology Unit, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy. ⁶Clinical Psycology Unit, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy

Background and aim The coronavirus 2019 (COVID-19) pandemic has had profound consequences also for non-infected patients. This study aimed to evaluate the impact of the pandemic on the quality of life of a population with hereditary gastrointestinal cancer predisposition syndromes and on the surveillance/oncological care program of patients enrolled in a dedicated registry.

Method The study was conducted by means of an online self-report survey during the first Italian national lockdown. The survey comprised four sections: demographics; perception/knowledge of COVID-19; impact of the COVID-19 pandemic on surveillance and cancer care; health status (SF-12 questionnaire).

Results 211 complete questionnaires were considered. 25.12% of respondents reported being not at all frightened by COVID-19, 63.98% felt "not at all" or "a little" more fragile than the healthy general population, and 66.82% felt the coronavirus to be no more dangerous to them than the healthy general population. 88.15% of respondents felt protected knowing they were monitored by a team of dedicated professionals.

Conclusions Patients with hereditary gastrointestinal cancer predisposition syndromes reported experiencing less fear related to COVID-19 than the healthy general population. The study results suggest that being enrolled in a dedicated registry can reassure patients, especially during health crises.

Keywords Coronavirus pandemic; Familial polyposis; Hereditary gastrointestinal cancer syndromes; Lynch syndrome; Registry; SF-12 questionnaire.

PP-029: Epidemiology and Registries

The PRECEDE Consortium: A Longitudinal International Cohort Study of Individuals with Genetic Risk or Familial Pancreatic Cancer

George Zogopoulos ¹, Yan Bi³, Randall E. Brand⁴, Teresa A. Brentall⁵, Daniel C. Chung⁶, Julie Earl⁷, James Farrell⁸, Srinivas Gaddam⁹, John J. Graff¹⁰, Talia Golan¹¹, Joanne M. Jeter¹², Vivek Kaul¹³, Fay Kastrinos¹⁴, Bryson W. Katona¹⁵, Kelsey A. Klute¹⁶, Sonia S. Kupfer¹⁷, Richard S. Kwon¹⁸, James M. Lindberg¹⁹, Andrew M. Lowy²⁰, Aimee Lucas²¹, Salvatore Paiella²², Jennifer B. Permuth²³, Intan Schrader²⁴, Rosalie C. Sears²⁵, Daniel A. Sussman²⁶, Raymond C. Wadlow²⁷, Diane M. Simeone²

¹Research Institute of the McGill University Health Centre Montréal, QC, Canada. ²Perlmutter Cancer Center, New York University Langone Health, New York, NY, USA. ³Mayo Clinic, Jacksonville, FL, USA. ⁴University of Pittsburgh, Pittsburgh, PA, USA. ⁵University of Washington Fred Hutchinson Cancer Research Center, Seattle, WA, USA. ⁶Massachusetts General Hospital Boston, MA, USA. ⁷Medical Oncology Department, Ramón y Cajal University Hospital Madrid, Spain. ⁸Yale School of Medicine, Yale University New Haven, CT, USA. ⁹Cedars-Sinai Medical Center, Los Angeles, CA, USA. ¹⁰Arbor Research Ann Arbor, MI, USA. ¹¹Sheba Medical Center, Tel Aviv University, Israel. ¹²Huntsman Cancer Institute, University of Utah, Salt Lake City, UT, USA. ¹³University of Rochester, Rochester, NY, USA. ¹⁴Columbia University Medical Center, NY, NY, USA. ¹⁵University of Pennsylvania Perelman School of Medicine, Philadelphia, PA, USA. ¹⁶University of Nebraska Medical Center, Omaha, NE, USA. ¹⁷University of Chicago, Chicago, IL, USA. ¹⁸University of Michigan Health Ann Arbor, MI, USA. ¹⁹University of Massachusetts Memorial Health Worcester, MA, USA. ²⁰University of California San Diego Moores Cancer Center, La Jolla CA, USA. ²¹Icahn School of Medicine at Mount Sinai, NY, NY, USA. ²²Unit of General and Pancreatic Surgery, Pancreas Institute, University of Verona, Verona, Italy. ²³Moffitt Cancer Center Tampa, FL, USA. ²⁴British Columbia Cancer Agency, British Columbia, Canada. ²⁵Oregon Health & Science University Portland, OR, USA. ²⁶University of Miami Health System, Miami, FL, USA. ²⁷Inova Schar Cancer Institute, Fairfax, VA, USA

Background and aim Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal disease with lack of effective early detection strategies. There is an incomplete understanding of who is at risk for PDAC development and the contribution of heritability to that risk. Further, efforts at biomarker development for detection of early stage disease have been hampered by small sample sizes, lack of coordination, and inadequate access to high quality clinical data and biospecimens in relevant clinical populations. The PRECEDE Consortium was established to serve as a collaborative international network of PDAC clinical and research centers to accelerate early detection advances by standardizing collection of clinical data and biospecimens from patients at increased risk for PDAC. The consortium goal is to increase the overall survival rate for PDAC to 50% in 10 years by enabling transformative biomarker-driven discoveries in early detection of high-risk premalignant lesions and early stage cancers.

Method The PRECEDE Consortium (NCT04970056; precedestudy.org) launched in 2019 and began enrollment in May, 2020. Data and biospecimen sharing are required for centers to join the consortium, which is facilitated through use of standardized data and biospecimen collection, and a centralized database (PRECEDELink) managed by a data coordinating center (Arbor Research). Imaging and clinical sequencing data will be stored and analyzed via a PRECEDE solution in the Amazon Web Services cloud. Participants age 18–90 are enrolled into one of seven cohorts based on personal and/or family history of PDAC and carrier status of pathogenic germline variants (PGV) in cancer predisposition genes (CPG). Three-generation pedigrees are collected at enrolment from participants, and standardized clinical germline testing is offered. Blood sample collection for DNA, plasma, and serum is completed at enrollment, and repeated annually for individuals meeting guidelines for annual surveillance.

Results To date, 26 clinical sites have enrolled 2370 participants, with a target of 10,000 participants enrolled from 100 sites over the next 5 years. Among enrolled patients, 55% meet criteria for annual surveillance by MRI or endoscopic ultrasound. Demographics of the cohort to date: 56% female; 73% white; 35% CPG PGV carriers; 32% meet criteria for familial pancreatic cancer.

Conclusions The PRECEDE Consortium study is a large international, longitudinal, prospective cohort study designed to accelerate the pace and scale of early diagnosis. Planned projects will address modifiers of risk, penetrance of disease, creating comprehensive risk models for clinical decision-making, and development and validation of biomarker assays. The PRECEDE Consortium provides a unique, innovative platform to bring together key stakeholders (academia, patients, public and private sector) to effect progress.

Keywords pancreatic cancer, early detection, genetic testing, genetic counseling, family history, germline variants.

PRECEDE Cohorts and Enrollment Table

PP-030: Epidemiology and Registries

Rates of Familial Adenomatous Polyposis amongst individuals referred to the New Zealand Familial Gastrointestinal Cancer Service with Congenital Hypertrophy of the Retinal Pigment Epithelium

Harriet Claire Marshall ¹, Christopher John Wakeman¹, Teresa Alice Chalmers Watson², Alison Louise Moore³

¹Department of General Surgery, Canterbury District Health Board, Christchurch, New Zealand. ²Department of Gastroenterology, Canterbury District Health Board, Christchurch, New Zealand. ³New Zealand Familial Gastrointestinal Cancer Service, Christchurch, New Zealand

Background and aim Familial adenomatous polyposis (FAP) is a heritable condition which predisposes individuals to the development of colorectal cancer. It is important to identify early, as the median age of development of colorectal cancer is 39 years of age and there is a lifetime risk of colorectal cancer of almost 100%.

There are some extracolonic manifestations of FAP, including congenital hypertrophy of retinal pigmented epithelium (CHRPE) of the eye. Atypical variants and bilateral CHRPEs are more commonly associated with FAP. The New Zealand Familial Gastrointestinal Cancer Service (NZFGICS) accepts nationwide referrals for possible familial gastrointestinal cancers. This includes new CHRPEs and suspected FAP.

We propose that bilateral and atypical CHRPEs referrals to the NZFGICS, particularly those identified by an ophthalmologist, could be useful in diagnosing new cases of FAP.

Method We flagged nationwide referrals to the NZFGICS with CHRPE from 2014 onwards. We then performed a retrospective audit of these referrals to identify individuals with confirmed FAP.

Results Between the years of 2014 and 2021 there were a total of 31 referrals to the NZFCICS with CHRPE. Of the 31 referrals with CHRPE there were zero confirmed cases of FAP.

At this junction, 10 of 23 CHRPEs were bilateral and 12 had been described as atypical. Eight were referred by an ophthalmologist and 16 of 20 individuals had fundoscopy performed by an ophthalmologist. At least two individuals were thought not to have CHRPE or to have low risk CHRPEs after fundoscopy by a specialist ophthalmologist.

We have confirmed endoscopy of 21 individuals. Three are awaited as the individual is too young for endoscopy. Only two individuals were found to have polyps on endoscopy; these were hyperplastic or low-grade in nature.

Conclusions Based on the NZFGICS registry, referrals with CHRPE has not been a fruitful way to identify new cases of FAP. None of bilateralism, atypical nature of CHRPE, nor referral from ophthalmologist were good predictors of diagnosis of FAP.

There is an argument for specialist ophthalmology review but this is frequently not feasible as the referrals are nationwide and specialist ophthalmologists are not.

The NZFGICS will continue to accept CHRPE referrals for work-up regarding FAP. We will try and streamline the referral process with the help of referrers, to reduce the burden of further investigation.

Keywords FAP, CHRPE, New Zealand.

PP-031: Epidemiology and Registries

The English National Lynch syndrome transformation project: An NHS Genomic Medicine Service Programme

Kevin Monahan ¹, Adam Shaw², Laura Monje-Garcia¹, Fracesca Faravelli³, Frances Elmslie², Anna Kim², Simone Gelinas²

¹St Mark’s Hospital Centre for Familial Intestinal Cancer, Imperial College London, North Thames GMSA, London, UK. ²South East Genomic Medicine Cancer Alliance, London, UK. ³North Thames Genomic Medicine Cancer Alliance, London, UK

Background and aim In the UK, NICE (The National Institute for Health & Care Excellence) guidelines recommend universal testing of newly diagnosed colorectal and endometrial cancer for Lynch syndrome (LS), however there is strong evidence of variation in delivery of these guidelines by clinical services. In England, through 7 regional Genomics Medicine Service Alliances (GMSA), a transformation project aims to establish robust pathways to improve guideline compliance.

Method A national oversight group was formed in May 2021, with membership drawn from 21 sub-regional Cancer Alliances (CA) and 7 GMSA, charities and other stakeholders. Each CA was tasked with identifying and supporting a responsible ‘Lynch syndrome champion’ within each local cancer team, and we performed a baseline survey to identify barriers to the testing pathway. Workforce training focused on overcoming barriers to testing, identification of eligible patients and mainstreamed constitutional gene testing, and this training is delivered via online modules, workshops, and face-to-face peer-support and co-consultation. Data analysis is performed in conjunction with the National Disease Registration Service (NDRS), and includes clinicopathological data for all cancer patients diagnosed across England, including somatic and constitutional testing outcomes.

Results Baseline data from NDRS and from the survey demonstrates that although cancer teams self report that 71% offer universal testing for LS, in 2019 only 41% of colorectal or endometrial cancer patients received any form of MMR testing. The main barriers to testing identified relate to funding streams and systematic approaches to testing. Lynch champions have been identified in 195 local cancer teams, and LS nurses are being appointed in each GMSA to support workforce development. Subgroups have been established in primary care, nursing, pathology, training, and to pilot testing in other Lynch-related tumour types. Each GMSA has identified LS patients diagnosed via their service, which will be used to ascertain people for a Nationally coordinated screening programme from 2023, and the development of a National LS Registry.

Conclusions This ongoing transformational project is supported by high levels of engagement across stakeholders in England. Despite barriers, significant quality improvement has been implemented, facilitating systematic delivery of universal testing for LS nationally, with reduction in variation in care.

Keywords Lynch syndrome, Transformation project, diagnosis, MMR IHC, mainstreaming.

PP-032: Epidemiology and Registries

The Danish HNPCC Register – High Completeness on CRC

Lars Joachim Lindberg ¹, Inge Bernstein², Henrik Møller³, Lone Sunde⁴, Christina Therkildsen¹

¹The Danish HNPCC Register, Gastro unit, Copenhagen University Hospital, Hvidovre, Denmark. ²Department of Gastroenterology, Aalborg University Hospital, and Faculty of Medicine, Aalborg University, Aalborg, Denmark. ³Research Unit for General Practice, Research Centre for Cancer Diagnosis in Primary Care (CaP), Department of Public Health, Aarhus University, and The Danish Clinical Registries (RKKP), Aarhus, Denmark. ⁴Department of Clinical Genetics, Aalborg University Hospital, Aalborg, Denmark

Background and aim The Danish Hereditary Non-polyposis Colorectal Cancer (HNPCC) Register has through the last three decades collected documentation from clinicians on diagnosed cancers in families with a suspected or proven hereditary predisposition to colorectal cancer (CRC). The reporting is voluntary, and the completeness of cancer data remains unknown. We aimed to evaluate the agreement between the Danish HNPCC Register and the Danish Cancer Register, which is receiving mandatory reports on cancer diagnoses from hospitals treating CRC in Denmark, and we aimed to estimate the completeness of each register.

Method Individuals registered in the HNPCC Register and with valid social security numbers were considered eligible for the study. Data on CRC diagnosed from 1943 to 2014 (including date of diagnosis, location, and histology) were extracted from the HNPCC Register and from the Cancer Register. Cancer diagnoses were matched between the two registers by date of diagnosis and location. Agreement between databases and completeness of each database were calculated and analyzed in relation to familial predisposition, time period of diagnosis, first or subsequent CRC, and diagnosis before or after registration of the individuals in the HNPCC Register.

Results In total, 9160 documented CRCs in 49,799 individuals were identified using both registers. Overall agreement between the registers was 85.1%. Subgroup analyses showed increasing agreement over time (p < 0.0001), better agreement on first CRC (p < 0.0001), and on CRCs diagnosed before registration of the individual in the HNPCC Register (p < 0.0001). Familial predisposition did not influence agreement (p = 0.66).

The HNPCC Register was generally more complete than the Cancer Register (95.4% compared to 89.7%, p < 0.001), but especially regarding synchronous/metachronous CRCs (92.8% compared to 49.8%, p < 0.001) affecting completeness in Lynch syndrome (93.3% compared to 81.4%, p < 0.001) due to more multiple CRCs in Lynch syndrome than in familial colorectal cancer and moderate familial risk.

Conclusions Overall, there was a good agreement between the registers, however the HNPCC Register was more complete than the Cancer Register, and thus a better data source for studies on CRC – especially in Lynch Syndrome, though a combination of both registers would secure the most optimal dataset.

Keywords Registries, Epidemiology, Agreement, Completeness, Colorectal Cancer, Lynch Syndrome.

Table 1. Comparisons of CRCs recorded in the Danish HNPCC Register (HNPCC) and in the Danish Cancer Register (CAR)

FCC: Familial Colorectal Cancer; MFR : Mo d erate Familial Risk

PP-033: Epidemiology and Registries

APC I1307K carriers at higher risk of cancer: not just colorectal cancer or Ashkenazi Jewish population

Esther Forkosh¹, Michael Bergel¹, Kathryn E. Hatchell², Sarah M. Nielsen², Brandie Heald², Ariel Benson¹, Eitan Friedman³, Edward D. Epslin², Lior H. Katz ¹

¹Department of Gastroenterology, Hadassah Medical Center, Jerusalem, Israel. ²Medical Affairs, Invitae, San Francisco, CA, USA. ³The Meirav Center High Risk Clinic, Sheba Medical Center, Ramat-Gan, Israel

Background and aim APC I1307K has a higher prevalence among Ashkenazi Jews (AJ), with a two-fold increased risk for colorectal cancer (CRC) than the population. We assessed CRC and extracolonic malignancies among I1307K carriers from AJ and non-AJ Caucasians (NAC).

Method We compared the rate of I1307K in cancer patients who underwent genetic testing via a multi-gene panel at a single commercial laboratory with genetic data of healthy subjects retrieved from the gnomAD database.

Results Overall, 586/7,624 (7.6%) AJ patients with cancer carried I1307K compared to 342/4918 (6.9%) in the AJ control group (p = NS). In the NAC population, 318/141,673 (0.2%) cancer patients and 73/58,918 (0.1%) controls carried the variant (p < 0.001). Within the NAC group, carrying I1307K was associated with increased overall risk of any cancer [OR = 1.8], CRC [OR 1.95], melanoma [OR 2.54], breast [females, OR 1.73] and prostate cancer [males, OR 2.42]. Among AJ, the variant increased the risk for CRC [OR 1.67] and renal cancer [OR 1.64]. AJ men also had a higher risk for any cancer.

[OR 1.3] and melanoma [OR 2.04].

Conclusions This is the most extensive study conducted on I1307K. The study showed that the NAC population carrying I1307K had a higher risk of any cancer and several specific cancer types. Based on these findings we recommend that cancer screening should be considered in I1307K carriers’ families.

Keywords APC, I1307K, cancer, Ashkenazi Jewish, Non Askenazi Caucasian.

PP-034: Epidemiology and Registries

If You Build It They Will Come: Downstream Benefit to the Healthcare System from a Hereditary Colorectal Cancer Registry

Lisa Ann Laguardia ¹, Margaret O'malley¹, Susan Milicia¹, Sarah Mcgee³, Carole Macaron², Amit Bhatt², Carol Burke², Scott Steele¹, Nicholas Smith¹, Joshua Sommovilla¹, David Liska¹

¹Department of Colorectal Surgery, Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, Ohio, USA. ²Department of Gastroenterology, Hepatology, and Nutrition, Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, Ohio, USA. ³Genomic Medicine Institute, Hematology and Medical Oncology, Center for Personalized Genetic Healthcare, Cleveland Clinic, Cleveland, Ohio, USA

Background and aim The Jagelman Cancer Registry (JCR) is a hereditary colorectal cancer registry (HCCR) that was established at the Cleveland Clinic in 1979 with the goal of preventing death from colorectal cancer. Patient’s clinical encounters for treatment and surveillance are managed through the registry. This study examines the downstream benefits of a hereditary cancer registry to the healthcare system.

Method A prospective hereditary colorectal cancer database (Cologene™) was used to identify patients enrolled in the JCR during a single 12-month period. Data regarding hereditary syndrome related office visits, procedures, surgeries, imaging, and pathology requisitions during the subsequent 5 years (2015–2020) were abstracted from the database and supplemented with the electronic medical record (EMR).

Results The JCR enrolled 107 patients in 2015. A total of 87 patients who had a confirmed pathogenic/likely pathogenic variant in a hereditary colorectal susceptibility gene or who met clinical criteria for a hereditary colorectal cancer syndrome were included. Patients with serrated polyposis syndrome, and hereditary non-polyposis colorectal cancer without a pathogenic variant were excluded. There were 55 patients with a polyposis syndrome (FAP 47, MAP 1, JPS 4, PJS 1, and Cowden 2) and 32 with Lynch syndrome. A personal history of cancer was present in 14 patients at the time of enrollment in the registry and no patients developed a new cancer during an average of 3.4 years of follow up. There were 20 patients that did not return for follow up after one year, including 3 patient deaths. A total of 2431 billable appointments (average ~ 28 per patient) were completed between 2015 and 2020. There were 1091 office visits, 425 endoscopic procedures, 127 surgeries, and 412 imaging studies. (Table 1).

Conclusions Patients enrolled in a HCCR require intense treatment and surveillance, including many diagnostic and therapeutic interventions. Establishing a registry results in significant downstream benefit to the health care system that can outweigh the initial costs for registry creation and maintenance.

Keywords Registry, Hereditary.

Table 1 .

	Encounter Type	n (average per patient)
Office Visits	Gastroenterology	63 (0.7)
	Colorectal Surgery	230 (2.6)
	Other	798 (9.1)
Procedures	Upper Endoscopy	158 (1.8)
	Lower Endoscopy	267 (3.1)
Surgery	Inpatient	73 (0.8)
	Outpatient	54 (0.6)
Pathology	Surgical Pathology	372 (4.3)
Radiology	MRI	38 (0.4)
	CT Scan	235 (2.7)
	X-Ray	139 (1.6)
	Ultrasound	131 (1.5)

PP-035: Epidemiology and Registries

Evaluation of upper gastrointestinal tract surveillance in individuals with Lynch syndrome. An international, multicenter registry – EARLY registry

Robert Hüneburg ¹, Tim Marwitz¹, Matthias Kloor², Aysel Ahadova², Christian P Strassburg¹, Christoph Engel³, Jacob Nattermann¹

¹National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Bonn, Germany; Department of Internal Medicine I, University Hospital Bonn, Bonn, Germany. ²Department of Applied Tumour Biology, Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany. ³Institute for Medical Informatics, Statistics and Epidemiology, University of Leipzig, Leipzig, Germany

Background and aim Lynch syndrome (LS) is the most common hereditary colorectal cancer (CRC) syndrome and accounts for ~ 3% of all CRCs. This autosomal dominant disorder is caused by germline mutations in DNA mismatch repair (MMR) genes (MLH1, MSH2, MSH6, PMS2 and EPCAM). LS includes a variety of extracolonic malignancies such as gastric carcinoma (life-time risk 13%) or small bowel cancer (life-time risk 8%). Whereas several national or European guidelines propose no screening, others propose a baseline esophagogastruoduodenoscopy (EGD) and testing for Helicobacter pylori. Also risk factors for upper gastrointestinal cancer remain unclear. All studies so far were either retrospective or monocentric, which leads to a selection bias.

Method Patients with a proven (likely-) pathogenic germline variant in the DNA mismatch repair genes (MLH1, MSH2, MSH6, PMS2 or EPCAM) are included in a prospective registry. All patients will receive a baseline EGD. This is followed by standard of clinical care according to the local approach. Patients will either be in a surveillance program or without any upper GI surveillance. Patients can receive symptom triggered EGD at all times. Patients will be documented in a remote eletronic case report form. The EARLY registry is supported by the European Hereditary Tumor Group (EHTG) and the Collaborative Group of the Americas on Inherited Gastrointestinal Cancer (CGA-IGC).

Results After signing an informed consent form, a baseline EGD is carried out. Subsequently, surveillance is carried out in accordance with local requirements. All patients are followed up.

Conclusions This prospective endoscopic registry will evaluate surveillance of the upper GI tract in a large international registry of LS patients. The primary endpoint is to compare the 10-year cumulative incidence of upper GI tract neoplasms (gastric cancer and duodenal cancer) between surveillance strategy and symptom-triggered endoscopic examinations. Further, we want to identify factors associated with a higher risk to develop upper GI cancer. We intend to provide the basis for systematically assessing the relation of immune surveillance with upper GI cancer risk.

Keywords Lynch syndrome, EGD, upper GI surveillance.

MMR deficient crypt adjacent to duodenal cancer

A mismatch repair deficient crypt right next to a duodenal cancer in a MSH2 patient

PP-036: Genetics: Research, Germline, Genomics, Variomics

Stakeholder Views of the Future Needs to Improve Implementation of Screening for Lynch Syndrome

Alanna Kulchak Rahm ¹, Catherine M. Vickerman¹, Heather Hampel², Brandie Leach³, Debra Duqette⁴, Erin Peterson⁵, Robin Dubin⁶, David Dubin⁶

¹Geisinger. ²City of Hope. ³Invitae. ⁴Northwestern Univeristy. ⁵Colon Cancer Coalition. ⁶AliveAndKickn

Background and aim Lynch syndrome is a hereditary cancer condition that increases risk for colorectal and endometrial cancer. While it is estimated 1 in every 280 people have Lynch syndrome, 95 percent are not diagnosed. The PCORI SPRINTS (Strategic Planning Retreats in Tumor Screening) Project is a collaboration between the Lynch Syndrome Screening Network and AliveAndKickn designed to identify the current clinical and patient-centered issues, gaps, barriers, and priorities for Lynch syndrome screening programs to improve the identification of individuals with Lynch syndrome.

Method Two activities were conducted using a deliberative engagement format. One activity was conducted in conjunction with the Living with Lynch Patient Workshop September 10–12, 2021 to gain perspectives of individuals with Lynch syndrome and caregiver stakeholders, the second was conducted during the 2021 Collaborative Group of the Americas on Inherited Gastrointestinal Cancer (CGA-IGC) Annual Meeting November 13–15 with clinician and researcher stakeholders. Benefits, challenges, and solutions were discussed by patient stakeholders, and clinician/researcher stakeholders ranked top priorities as important and feasible and generated action plans for each priority. Main themes were extracted from both groups and action plans were mapped to priority areas with the LSSN board in a virtual whiteboard session.

Results Sixteen individual stakeholders (12 patients and four caregivers/spouses) and 25 clinician/researcher stakeholders participated. Three main areas for action were identified across all stakeholder groups to which the action plans were mapped (Table). Cascade testing was the biggest priority for patient stakeholders, while facilitating true universal tumor screening/identification was the biggest priority for clinician/researcher stakeholders, followed by cascade testing. The third priority area identified was around creation and use of registries. Action plans included educational interventions, EHR tools, and policy-level changes.

Conclusions Engagement of stakeholders is important to ensuring the programs and research priorities promoted by organizations like the Lynch Syndrome Screening Network meet the needs and address barriers experienced by key stakeholders experiencing Lynch syndrome. Engagement has identified important and feasible priorities with potential action plans.

Keywords Lynch syndrome, stakeholder engagement.

Priority areas and actions identified

Priority Area	Identified Actions
Cascade Screening	Increase awareness Increase accessibility to family members Fix insurance coverage issues Change/utilize genetic counselor roles Use EHR platforms for family sharing
True Universal Tumor Screening	Make reportable outcome for cancer center certifications Improve education Streamline processes Require ongoing data review
Create Registries	Create deidentified patient registries Harmonize data points between systems Use EHR tools to facilitate registry data collection

PP-037: Genetics: Research, Germline, Genomics, Variomics

Identification of multiple pathogenic variants in multi-gene panels: considerations for clinical care

Ambreen Khan, Journey Bly, Joanne Jeter, Wendy Kohlmann

Family Cancer Assessment Clinic, Huntsman Cancer Institute, Salt Lake City, UT

Background and aim Advances in technology and awareness of the role of precision medicine have made simultaneous testing of multiple genes more accessible and affordable. Compared to single-gene testing, multi-gene panels (MGP) increase the likelihood of identifying a pathogenic variant (PV) by two-fold, and may identify more than one PV, not all of which are associated with the patient’s phenotype or directly relevant to the patient’s health (i.e. carrier status of recessive conditions). However, there is sparse data about the frequency of multiple PVs in individuals undergoing MGP testing.

Method The clinical genetic testing database at Huntsman Cancer Institute (Salt Lake City, UT) was queried for individuals with two or more PVs in cancer predisposition genes who were tested between January 2012 and March 2022. The query was limited to genes commonly included in clinically available MGP tests (Fig. 1).

Results Over 2800 individuals were tested between January 2012 and March 2022. A total of 63 individuals were identified to have two or more PVs. Of these individuals, 89% (56/63) were probands and 11% (7/63) were incidentally identified with another PV when undergoing cascade testing for a single familial variant using MGP. 21% (13/63) had PVs in two or more high penetrance genes, 28% (18/63) in a combination of high and moderate penetrance genes, and 11% (7/63) in a combination of moderate penetrance genes (Fig. 1). 54% (34/63) were found to have at least one PV in a gene related to increased risk for colorectal cancer (CRC) (APC, CHEK2, EPCAM, MLH1, MSH2, MSH6, PMS2, PTEN, TP53, BMPR1A, SMAD4). Of these individuals, 24% (8/34) had a diagnosis of CRC at the time of genetic testing.

Conclusions A significant proportion of MGPs result in identification of multiple PVs in the same individual. The quantification of risk in this situation has not been well described and current conventions are to provide risk estimates and screening recommendations based on the gene that confers the higher cancer risk, as well as any additional cancer risks not covered by recommendations for the higher-risk PV. The impact of patient and provider perceptions of multiple PVs on risk has not been studied. Incidental findings may complicate the process of cascade testing for clinically relevant PVs, which is already suboptimal. Cascade testing resources that address multiple PVs need to be streamlined to facilitate appropriate family communication and risk management. Future studies can focus on whether multiple PVs impact risk, risk perception, and clinical management, as well as possible gene–gene interactions in these individuals.

Keywords multi-gene panels, genetic testing, cascade testing, gene–gene interaction, family communication, colorectal cancer.

PP-038: Genetics: Research, Germline, Genomics, Variomics

A large family with MSH3-related polyposis

Arthur S. Aelvoet ¹, Daniël R. Hoekman², Bert J. Redeker², Jitske Weegenaar³, Evelien Dekker¹, Carel J. Noesel³, Floor A. Duijkers²

¹Amsterdam UMC location University of Amsterdam, Department of Gastroenterology and Hepatology, Amsterdam Gastroenterology Endocrinology Gastroenterology, Cancer Center Amsterdam, Amsterdam, the Netherlands. ²Amsterdam UMC location University of Amsterdam, Department of Human Genetics, Amsterdam, the Netherlands. ³Amsterdam UMC location University of Amsterdam, Department of Pathology, Amsterdam, the Netherlands

Background and aim Biallelic MSH3 mutations are a rare cause of adenomatous polyposis. Only two small families have been reported to date. We describe the phenotype of the largest family reported thus far with MSH3-related adenomatous polyposis. The index patient was a 55-year old male diagnosed with rectal cancer and adenomatous polyposis (cumulatively 52 polyps), with a family history of colorectal polyposis of unknown cause. Next-generation sequencing and copy number variation analysis of a panel of genes associated with colorectal cancer and polyposis revealed compound heterozygous germline pathogenic variants in the MSH3 gene.

Method We performed co-segregation analysis and documented the phenotype of the affected family members. All 11 siblings were invited for targeted analysis of the MSH3-gene using Sanger sequencing and multiplex ligation-dependent probe amplification (MLPA). Two siblings declined participation, resulting in 9 siblings being genotyped.

Results Three siblings carried the same compound heterozygous MSH3 mutation as the index patient. A mean of four colonoscopies in all four compound heterozygous family members showed adenomatous polyposis with a cumulative number of colorectal adenomas ranging from 18 to 54 and age at first polyp detection ranging from 46 to 59 years. In all four patients, no adenomas were detected in the duodenum and stomach with age at last esophagogastroduodenoscopy ranging from 46 to 59 years. Microsatellite analysis demonstrated Elevated Microsatellite Alterations at Selected Tetranucleotide (EMAST) repeats in DNA retrieved from the rectal adenocarcinoma, colorectal adenomas as well as of normal colonic mucosa (Fig. 1). Extra-intestinal findings included a ductal adenocarcinoma in ectopic breast tissue in one female sibling at the age of 46 and liver cysts in three siblings. None of the three heterozygous or wild type siblings who previously underwent colonoscopy had adenomatous polyposis.

Conclusions Compound heterozygous mutations in MSH3 are a cause of attenuated adenomatous polyposis with middle age onset. Frequent endoscopic surveillance seems warranted. Although extra-intestinal findings might be more prevalent than in the general population (Table 1), available data is too scarce for a clear recommendation on screening for extra-intestinal manifestations. Our findings underline the importance of genetic testing for rare polyposis predisposing genes in yet unexplained polyposis cases.

Keywords Adenomatous polyposis, Colorectal cancer, MSH3, MSI, EMAST.

Table 1. All three reported families with MSH3-related polyposis coli

PP-039: Genetics: Research, Germline, Genomics, Variomics

Underutilization of a Medical Policy Covering Universal Microsatellite Instability, Immunohistochemistry, and Genetic Testing in All Patients Diagnosed with Colorectal Cancer—A Real World Data Analysis

Brandie Leach ¹, Chad Moretz¹, Sarah M. Nielsen¹, Joline Dalton², Kathryn E. Hatchell¹, Scott T. Michalski¹, Stacey Dacosta Byfield², Peter Onglao², Lillian Hang², Pamela Hansen², Edward D. Esplin¹

¹Invitae, San Francisco, CA, USA. ²OptumLabs, Minnetonka, MN, USA

Background and aim Since 2009, recommendations for universal tumor screening via microsatellite instability (MSI) and/or immunohistochemistry (IHC) of all newly diagnosed colorectal cancers (CRC) have been in place. Professional societies recommend stratifying CRC patients for Lynch syndrome (LS) germline testing (GGT) based on MSI/IHC results, age at diagnosis and family cancer history. Identification of LS is critical to inform cancer screening, cascade testing, and guide therapy selection, including use of immune checkpoint inhibitors (ICI). In 2020, a large national US health plan instituted a medical policy covering GGT for all CRC patients. The objective of this study was to assess LS testing and treatments for patients diagnosed with CRC under this medical policy.

Method Retrospective analyses were conducted using a longitudinal, real-world data set of de-identified administrative claims. The study cohort included adults (≥ 18 years) with newly diagnosed CRC (≥ 2 claims for ICD10 C18.x, C19.x, C20.x, C21.8) from 1/2017–12/2020; the earliest CRC claim was the index diagnosis date. Continuous enrollment in a commercial or Medicare Advantage plan ≥ 12 months pre-diagnosis and ≥ 6 months post-diagnosis was required with ≥ 1 claim for initial antineoplastic systemic therapy ≤ 6 months post-diagnosis. Patients were stratified into four cohorts based on receipt of GGT/MSI/IHC before initial systemic therapy. Treatments and rates of ICI use were examined by cohort during the variable follow up period (6–48 months continuous enrollment post index date).

Results Among 9,066 CRC patients, 903 (10%) had both GGT and MSI/IHC, 5,745 (63.4%) had MSI/IHC only, 130 (1.4%) had GGT only, and 2,288 (25.2%) had no testing (Figure and Table). Patients with GGT were younger than those without (p < 0.05). ICI utilization was higher among patients who had GGT and MSI/IHC (6.4%) compared to those who had MSI/IHC only (2.3%).

Conclusions Universal MSI/IHC was not followed in > 25% of insurance covered CRC patients. Despite a universal GGT policy for CRC patients, testing was completed proportionally more frequently in younger patients, but was only performed in 11.4% of CRC patients. Underutilization of universally authorized MSI/IHC/GGT highlights the need for systematic implementation of medically necessary GGT to expand access to precision therapies and improve patient care.

Keywords Lynch syndrome, MSI, IHC, germline genetic testing.

Figure. Germline Genetic Testing and/or Microsatellite Instability/Immunohistochemistry Tumor Screening in Colorectal Cancer Patients Prior to Initial Systemic Therapy.

Table. Demographics, Healthcare Utilization, and Follow-Up Procedures for Enrollees Diagnosed with Colorectal Cancer in 2017–2020

PP-040: Genetics: Research, Germline, Genomics, Variomics

MSH3 as a new predisposing gene for adenomatous polyposis, and beyond

Marie Charlotte Villy¹, Julien Masliah-Planchon¹, Sophie Vacher¹, Hélène Delhomelle¹, Lisa Golmard¹, Samia Melaabi¹, Anne Schnitzler¹, Maud Blanluet¹, Voreak Suybeng¹, Marion Dhooge², Nadim Hamzaoui³, Solenne Farelly², Amal Ait Omar⁴, Robert Benamouzig⁴, Vincent Caumette⁵, Michel Bahuau⁵, Joël Cucherousset⁶, Yves Allory⁷, Dominique Stoppa Lyonnet⁸, Ivan Bièche⁸, Chrystelle Colas ¹

¹Department of Genetics, Institut Curie, PSL Research University, Paris, France. ²Oncogenetic Unit, Department of Gastroenterology, AP-HP Centre-Université de Paris, Hôpital Cochin, Paris, France. ³Department of Genetics, AP-HP Centre-Université de Paris, Hôpital Cochin, Paris, France. ⁴Department of Gastroenterology, AP-HP, Hôpital Avicenne, Bobigny, France. ⁵Department of Genetics, AP-HP, Hôpitaux universitaires Henri-Mondor, Créteil, France. ⁶Department of Pathology, GHI le Raincy-Montfermeil, Montfermeil, France. ⁷Department of Pathology, Institut Curie, Université Paris Saclay, Saint-Cloud, France. ⁸Department of Genetics, Institut Curie, Faculty of Medicine, Université de Paris, Paris, France

Background and aim MSH3 is part of the DNA mismatch repair system, deficiency thereof leads to microsatellite instability as in Lynch syndrome. Lynch syndrome has never been related to MSH3 mutation. However, a report of 4 patients from 2 families, bearing biallelic MSH3 germline variants, and a phenotype of attenuated colorectal adenomatous polyposis with some extra-intestinal tumours raises the question of its implication in hereditary predisposition. The patients’ tumours exhibited the so-called EMAST (instability of microsatellites at loci with tetranucleotide repeats), a hallmark of MSH3 deficiency. Since this singular report, MSH3 has been added to diagnostic gene panels.

Method We report 4 unrelated patients with biallelic MSH3 germline variants and no mutation at other known predisposing genes. Personal and family history was collected and molecular features including EMAST were determined in various normal and tumour samples.

Results The patients, 2 males and 2 females, had attenuated colorectal adenomatous polyposis with colorectal cancer in 2, and duodenal polyposis in 2. A man also had lung carcinoma and both women had breast carcinoma. None had a family history of digestive cancer. Three patients were homozygous for novel pathogenic variants (PVs) in MSH3. The further patient was a compound heterozygote with a PV and a variant of unknown significance (VUS). EMAST was present at various levels in all samples except leucocytes, with a gradient in polyps depending on their degree of dysplasia.

Conclusions This report lends further credence to biallelic MSH3 germline PVs being causative in colorectal adenomatous polyposis. Larger-scale studies may help clarify the tumour spectrum and cancer risks in biallelic and monoallelic carriers, although hampered by the scarcity of MSH3 variants. Moreover, ascertainment of EMAST in extra-digestive tumours can help delineate the tumour spectrum, just as it can help in the interpretation of VUS. We recommend adding MSH3 to dedicated diagnostic gene panels.

Keywords Oncogenetics, MSH3, polyposis.

PP-041: Genetics: Research, Germline, Genomics, Variomics

High Prevalence of MUTYH Associated Polyposis Among Minority Populations in Israel, due to Rare Founder Pathogenic Variants

Gili Reznick Levi, Yael Goldberg, Hanna Segev, Itai Maza, Yuri Gorelik, Ido Laish, Zohar Levi, Inbal Kedar, Sonia Naftali Nathan, Nitzan Sharon Swartzman, Naim Abu Freha, Maya Paritzki, Gad Rennert, Hagit Baris Feldman, Tamar Paperna, Karin Weiss, Elizabeth E. Half.

Gastroenterology Department, Rambam Health Care Campus, Haifa, Israel

Background and aim The genetic diagnosis of MUTYH polyposis (MAP), an autosomal recessive polyposis syndrome, in Israel has focused on two common Caucasian founder mutations the p.Gly396Asp and p.Tyr179Cys variants, also frequent in the North African Jewish population, but not in the Arab population. Next Generation Sequencing utilization revealed rare and novel variants. As Non-Jewish population in Israel has high consanguinity rate, it is expected to have relatively high frequency of autosomal recessive disorders. AIM: to evaluate the rate and spectrum of MUTYH pathogenic variants (PVs) in this unique population and to increase identification of MAP in this population.

Method We reviewed health records of all Arab and Druze polyposis patients referred for counseling between 2013–2020, who fulfilled the Israeli Genetic Society criteria for MUTYH/APC testing (Northern Cohort). Data on MUTYH PVs in Arab patients from 4 additional gastro\genetic clinics was also collected.

Results The Northern cohort included 37 patients from 30 unrelated families. Eight families (26.6%) carried bi-allelic PVs; The major variant p.Glu452del was detected in 6/8 Druze and Muslim families, these families shared the same haplotype. Other PVs detected in Israel included p.Tyr56Ter in two families, p.His57Arg, c.849 + 3A > C, p.Ala357fs and p.Tyr151Cys. Among bi-allelic carriers, 87.5% reported consanguinity, 100% had positive family history for polyposis or CRC, and age of CRC was significantly younger (mean 38 years, range 28–49). Two patients were diagnosed with extra colonic cancers: one with liposarcoma and papillary thyroid carcinoma at age 39 and 43 respectively, and one with metastatic melanoma at 38 years. None had duodenal or periampullary cancers or polyps.

Conclusions MAP is frequent among the Arab polyposis individuals in Northern Israel. Mutation spectrum is unique, with high frequency of the founder p.Glu452del PV. These individuals develop CRC at a young age and lack duodenal and periampullary polyps. Since the end of 2020, all Israeli patients suspected of hereditary cancer syndromes, are tested first for a 51 Israeli (Jewish and Arab) frequent mutation panel funded by the HMO. The p.Glu452del variant has been included in this panel.

Keywords MUTYH, Polyposis, Colorectal cancer, rare mutations.

PP-042: Genetics: Research, Germline, Genomics, Variomics

Functional analysis of variants of uncertain significance in the MSH6 mismatch repair gene

Elizabeth Szabo, Christopher Heinen

Department of Molecular Biology and Biochemistry, University of Connecticut Health Center, Farmington, CT, United States

Background and aim Lynch syndrome (LS) is a hereditary condition that increases patients’ lifetime risk of cancer, primarily colorectal cancer. LS is caused by germline mutations in the DNA mismatch repair (MMR) genes. Identification of these mutations is important for definitive diagnosis of LS in order to guide treatment plans and preventative care for patients’ and their families. However, the consequences of some germline variants in these genes are not immediately obvious, granting them classification as variants of uncertain significance (VUS). Half of known variants in the MMR gene MSH6 are VUS and, thus, the goal of my project is to test the ability of a human cell based assay to provide functional evidence for a subset of these variants. Laboratory functional analysis to determine whether these variants disrupt MMR function in human cells can provide important evidence toward eventual classification of the MSH6 VUS.

Method With information provided on InSiGHT, ClinVar, and current literature, we created a list of 35 MSH6 VUS to be tested. Using CRISPR gene editing we will knock-in the 35 MSH6 VUS into the endogenous MSH6 loci in human embryonic stem cells (hESCs). We will also create known benign and pathogenic variant cell lines to serve as calibration controls. We will test the impact of the variant on protein stability and RNA splicing, the ability to maintain microsatellite stability, and the ability to induce a DNA damage response to alkylating agents. The resulting data will be used to calculate a quantitative functional odds of pathogenicity (OddsPath) score.

Results The results from the functional assays will provide insight on the impact that each of the 35 MSH6 VUS has on MMR. Variants that have a greater impact on MMR may show decreased protein stability, aberrant RNA splicing, high microsatellite instability, and/or little to no cell death when treated with an alkylating agent. These assays will provide data that can be used to aid in determining which VUS are likely to be pathogenic, benign, or may remain VUS.

Conclusions The OddsPath scores may contribute to evidence for or against pathogenicity that can then be utilized by the InSiGHT variant expert curation panel to potentially reclassify these MSH6 VUS. The reclassifying of variants will allow for proper diagnosis and treatment management for suspected LS patients.

Keywords MSH6, Variants of uncertain significance, Lynch syndrome, DNA mismatch repair, CRISPR, human embryonic stem cells.

PP-043: Genetics: Research, Germline, Genomics, Variomics

SMAD4 and its Genotype–Phenotype Correlations to Juvenile Polyposis Syndrome

Kimberley Kai Cao¹, John Paul Plazzer², Finlay Macrae ¹

¹Royal Melbourne Hospital Department of Colorectal Medicine and Genetics, Parkville, Victoria, Australia & University of Melbourne Medicine, Parkville, Victoria, Australia. ²Royal Melbourne Hospital Department of Colorectal Medicine and Genetics, Parkville, Victoria, Australia

Background and aim Juvenile polyposis syndrome (JPS) is a rare autosomal dominant hamartomatous syndrome affecting one per 100 000 births, increasing lifetime cancer risk by 20–40%. It is estimated that 40–60% of JPS cases are caused by disease-causing variants (DCV) in SMAD4 or BMPR1A genes, where SMAD4 accounts for 20–30%. Current gaps in JPS research involve further characterising genotype–phenotype correlations between DCV of SMAD4 with JPS and other related conditions.

The objective of this review is thus to better characterise genotype–phenotype correlations relating to sites and types of variants within the SMAD4 gene, focusing on phenotypes of JPS, and with this knowledge, better optimise diagnosis, screening, and management.

Method Online search databases included Ovid MEDLINE, Embase Classic + Embase and PubMed, using search terms classified by MeSH on Demand. Adjacency operators, word truncation and Boolean operators including AND and OR were utilised. 100 articles were included in the review, collating 291 variants from the literature.

Results In SMAD4 + JPS patients, most variants are in the MH2 domain (3’ end) of the gene, where most prevalent DCVs are deletions, small deletions and missense mutations. Extracolonic involvement, massive gastric polyposis and a more aggressive phenotype have been associated with this population, predisposing them to gastric cancer and gastrectomy. This has contributed to an overall higher incidence of GI cancers compared to other genes, with DCVs mostly all within MH2 domain. Genetically related allelic disorders of SMAD4 also have variants in this region, including hereditary haemorrhagic telangiectasia (HHT), occurring in up to 80% SMAD4 + JPS patients, together with Myhre syndrome, existing independently of JPS. Similarly, with DCVs in the MH2 domain, Ménétrier’s disease (MD), hypertrophic osteoarthropathy (HOA) and juvenile idiopathic arthritis (JIA) have been seen in this population, whereas cardiac pathologies have occurred alongside JPS-HHT, but also independently of it with DCVs in the MH1 domain.

Conclusions In conclusion, truncating and missense variants around MH2 region of SMAD4 are most prevalent and pathogenic, thus should undergo careful surveillance. Given association with extracolonic polyposis and higher GI cancer risk, gastroscopic and endoscopic screening should occur more frequently and at an earlier age in SMAD4 + JPS patients, with consideration of Ménétrier’s disease on upper GI endoscopy. In addition, HHT should be evaluated within 6 months of diagnosis, alongside targeted clinical examination monitoring for symptoms of HHT, cardiac pathologies, HOA and JIA. Where HHT is suspected or diagnosed, further investigations should include bubble echocardiogram and an MRI brain.

This review may help modify clinical diagnosis and management of SMAD4 + JPS patients, and aid pathogenicity classification for SMAD4 DCVs.

Keywords juvenile polyposis syndrome, SMAD4 protein, genotype–phenotype correlation, gene analysis.

Informal schematic summarising genotype–phenotype correlations between SMAD4 and JPS

Sites and Types of SMAD4 DCVs, protein change and phenotypes, self-made via Microsoft Visio.

Sites and Types of SMAD4 DCVs, protein change and phenotypes, self-made via Microsoft Visio. All 291 variants were collated from the literature. In the top section, DCVs causing JPS phenotypes have been picturised with lines extending below the diagram in brown (JPS), green (massive gastric polyposis in JPS) and red (GI cancer in JPS). Above the illustrations represent extra-JPS phenotypes, including HHT (dark red), aortopathy (pink), IBD (purple), Myhre syndrome (dark blue), JIA (light blue), hypertrophic osteoarthropathy (yellow), Ménétrier’s disease (grey) and concurrent GI cancer boxed in bright red. In the bottom section, large deletions and chromosomal translocations are picturised via solid lines, surrounding the SMAD4 gene. c.1245_1248delCAGA (p.Asp415Glufs*20), a highly prevalent DCV, is represented by asterisks (*). Variants of uncertain significance (VUS) are marked with adjacent stars, sourced from JPS registries and case reports

PP-044: Genetics: Research, Germline, Genomics, Variomics

Genotype–Phenotype Correlation of BMPR1a Disease Causing Variants in Juvenile Polyposis Syndrome

Marie Papadopulos¹, John Paul Plazzer², Finlay Macrae ¹

¹University of Melbourne Department of Medicine The Royal Melbourne Hospital; Department of Colorectal Medicine and Genetics The Royal Melbourne Hospital. ²Department of Colorectal Medicine and Genetics The Royal Melbourne Hospital

Background and aim Juvenile Polyposis Syndrome (JPS) is an autosomal dominant condition with hamartomatous polyps in the gastrointestinal tract, associated with an increased risk of gastrointestinal malignancy. Most cases of JPS are caused by disease causing variants (DCVs) in BMPR1a and SMAD4 genes. Within those with either a BMPR1a or SMAD4 DCV, there is phenotypic variability in location of polyps, risk of malignancy and extra-intestinal manifestations, with limited identified genotype–phenotype correlation.

The NCBI’s ClinVar database archives DCVs derived from sequencing, which is evaluated by NHGRI’s ClinGen. LOVD archives DCVs for the international Global Variome. The American College of Medical Genetics (ACMG) have developed criteria for classifying pathogenicity across the genome.

We aimed to identify any genotype–phenotype correlation in BMPR1a to inform surveillance recommendations and further developments or modification of the classification of pathogenicity of DCVs.

Method A literature search was performed through EMBASE, MEDLINE and PubMed. Studies that were included explored BMPR1a DCV-related JPS or contiguous deletion of BMPR1a and PTEN. Data was also drawn from the BMPR1a specific database on LOVD and ClinVar.

Results There were 211 DCVs in BMPR1a identified, 82 from the literature in case reports or cohorts of patients with JPS, and 17 from LOVD and 112 from ClinVar classified as pathogenic or likely pathogenic (see the Figure). These included missense, nonsense and frameshift variants and large deletions, occurring across all functional domains of the gene, as well as non-coding regions. There was no upper GI phenotype associated with BMPR1a DCVs. From this data, no specific BMPR1a variant to phenotype correlation could be ascertained including by variant type or functional domain.

DCVs in BMPR1a or SMAD4 account for 45–60% of JPS cases, with BMPR1a DCVs accounting for 17–38% of JPS cases. Unlike in SMAD4 carriers, gastric polyposis and malignancy were not identified in our review in BMPR1A carriers, but colonic polyposis and malignancy occurred in carriers of either BMPR1a or SMAD4 DCVs. Those with contiguous deletion of PTEN and BMPR1a can present with JPS of infancy, with a severe phenotype of GI bleeding, diarrhoea, exudative enteropathy and rectal prolapse. Hereditary Haemorrhagic Telangiectasia only occurs in carriers of SMAD4 DCVs and is not seen in JPS patients with BMPR1a DCVs.

Conclusions Phenotypic characteristics cannot be used to inform variant location in BMPR1A. However, the phenotypic characteristics, being almost exclusively related to the colon and rectum, can assist in pathogenicity assessment of BMPR1A.

Given these findings, we propose that carriers of BMPR1a DCVs should only require surveillance for colorectal polyps and malignancy, and that surveillance for gastric polyps and malignancy is unnecessary. However variant location within BMPR1a does not support differential surveillance recommendations.

This review also supports the InSiGHT ClinGen Variant Curation Expert Panel task of pathogenicity classification using the ACMG Criteria.

Keywords BMPR1a, Juvenile Polyposis Syndrome, Juvenile polyps, GI malignancy.

Location and type of DCVs in BMPR1a identified in the literature and pathogenic or likely pathogenic variants from LOVD and ClinVar

PP-045: Genetics: Research, Germline, Genomics, Variomics

Best Poster Award:

APC Promoter 1B variants……mind the GAPPS?

Isabel Martin ¹, Cherryl Cabalit¹, James Cockburn¹, Victoria Cuthill¹, Susan K Clark², Andrew Latchford²

¹sSt Mark's Hospital Polyposis Registry, London, UK. ²St Mark's Hospital Polyposis Registry & Imperial College London, UK

Background and aim Genotype–phenotype correlation is well described in FAP, but there is significant. Gastric adenocarcinoma and proximal polyposis of the stomach (GAPPS) is associated with pathogenic variants (PV) in promoter 1B of the APC gene. However, there are sparse data about the phenotype associated with PV at this site: it is not clear if only a GAPPS phenotype occurs with 1B promoter PV, or whether a more traditional FAP phenotype is seen. We reviewed our patients with promoter 1B PV and describe their phenotype.

Method Patients were identified from a prospectively maintained registry database. Details of genotype, upper and lower GI tract phenotypes, desmoid disease and surgery were collated.

Results Sixteen patients (nine female) with a median age 50 years (37–85) were identified. Fifteen had a deletion of promoter 1B detected on MLPA.

One patient had a family history of GAPPS and underwent a risk reducing gastrectomy in his twentys; his duodenum was not accessible for upper gastrointestinal (UGI) surveillance due to the nature of the Roux limb reconstruction. Eight patients died before the implementation of UGI surveillance, none from gastric or duodenal cancer. UGI endoscopy reports were available for 7 patients. None of the patients had a GAPPS phenotype (i.e. > 100 fundic gland polyps (FGP) and only two patients had around 20 FGPs. None of the patients had gastric adenomas or a diagnosis of desmoid disease.

Ampullary disease was present in two patients (one major, one minor) and five patients had duodenal polyps with a Spigelman stage of II-III at a median age of 42 years (range 39–65). One patient underwent a pancreaticoduodenectomy for unknown reasons at another institution.

Eleven patients underwent colectomy (8 colectomy and ileorectal anastomosis, three restorative proctocolectomy) where the median colonic polyp count was 275 polyps (range 0–5410).

Conclusions Although GAPPS is described as associated with promoter 1b variants, variants at this site can in fact lead to a wide spectrum of phenotype.

Promoter 1B variants should not be used to diagnose GAPPS. Patients with variants at this site need a full phenotypic assessment, which will dictate their personalised management. A better understanding of why some patients develop a GAPPS phenotype and why some develop a more typical FAP phenotype requires further investigation.

Keywords Familial adenomatous polyposis, APC gene, gastric adenocarcinoma, proximal polyposis of the stomach, GAPPS.

PP-046: Genetics: Research, Germline, Genomics, Variomics

NOMO1 loss promotes cell migration in early-onset colorectal cancer

Abel Martel-Martel¹, Paula García-Vallés², Jésica Pérez-García², Luis Corchete³, Juan L. García⁴, Ana B. Herrero², José Perea ³, Rogelio González-Sarmiento²

¹Institute of Biomedical Research of Salamanca (IBSAL), Salamanca, Spain; Molecular Medicine Unit, Department of Medicine, University of Salamanca, Salamanca, Spain; Medical Oncology Service, University Hospital of Salamanca, Salamanca, Spain. ²Institute of Biomedical Research of Salamanca (IBSAL), Salamanca, Spain; Molecular Medicine Unit, Department of Medicine, University of Salamanca, Salamanca, Spain; Institute of Molecular and Cellular Biology of Cancer (IBMCC), University of Salamanca-CSIC, Salamanca, Spain. ³Institute of Biomedical Research of Salamanca (IBSAL), Salamanca, Spain. ⁴Institute of Biomedical Research of Salamanca (IBSAL), Salamanca, Spain; Institute of Molecular and Cellular Biology of Cancer (IBMCC), University of Salamanca-CSIC, Salamanca, Spain

Background and aim Early-onset colorectal cancer (EOCRC; age younger than 50 years) incidence has been steadily increasing over the last decades worldwide. Recently, we have detected a homozygous deletion of the NOMO1 gene (16p13.11-p13.12) in more than 80% of EOCRC patients, non-hereditary and microsatellite stable (MSS). This suggest that NOMO1 could be a molecular marker associated with EOCRC, particularly with MSS. Yet the mechanisms of NOMO1 in early-onset colorectal carcinogenesis are currently unknown. In this work, we analyzed the migration capability changes between NOMO1 knockout and NOMO1 wild type cell lines by wound healing and transwell migration assays.

Method CRISPR/Cas9 technology was used to generate the NOMO1-KO HCT-116 (EOCRC) and HS-5 (bone marrow) cell lines. Then, the migratory capability of two NOMO1-KO and two NOMO1-WT clones of each cell line was analyzed at 24 and 48 h. During the wound healing and transwell migration assays, the cells were maintained in serum-free medium to quantify only the cells with migratory capability. In the wound healing assay, the area filled by the migrating cells was quantified as percentage of migration. In the transwell migration assay, 1% crystal violet was used to stain the cells that migrated across a porous 8 µm membrane. These cells were then quantified as positive particles. The ImageJ® and SPSS/IBM software were used to calculate the migratory capability in both assays.

Results Consistent across all cell lines, our results showed that the percentage of migration and the positive particles were significantly higher in the NOMO1-KO clones compared to controls (t-test, p < 0.01), in both assays.

Conclusions NOMO1 loss promotes cell migration. That could explain its implication in the development of invasion and metastasis of EOCRC. Further studies are being performed to explore the specific signaling pathways deregulated by NOMO1 loss that might play a relevant role in the pathogenesis of the disease.

Keywords NOMO1, early-onset colorectal cancer, cell migration.

PP-047: Genetics: Research, Germline, Genomics, Variomics

Likely De Novo MLH1 mutation in a young patient with MSI colon cancer

Karin Adrienne Segers, Ben Flies, Philippe Theis, Daniel Stieber, Barbara Klink

National Center of Genetics, Laboratoire national de santé, Dudelange, Luxembourg

Background and aim Heterozygous germline mutations in the mismatch repair (MMR) genes cause Lynch syndrome (LS) which predisposes to colorectal and endometrial cancer. The MMR defect results in tumors with microsatellite instability (MSI) and/or loss of expression of one of the MMR proteins. LS is mostly inherited. De novo MMR gene mutations have been described but appear to be rare. Therefore somatic mosaicism in LS is thought to be infrequent. Here, we report a case of probable somatic mosaicism by a likely de novo MLH1 pathogenic variant.

Method A patient presented with an adenocarcinoma in the right colon at age 41. Molecular testing of the tumor showed the presence of the KRAS mutation c.35G > A, p.(Gly12Asp), absence of V600E BRAF mutation and microsatellite instability. Immunochemistry showed loss of expression of MLH1-PMS2 in the absence of MLH1 promoter methylation. Regarding the young age of the patient and the characteristics of the tumour a Lynch syndrome was suspected.

Sequencing of genes associated with hereditary cancer has been initiated on peripheral blood of the patient.

Results The pathogenic variant MLH1 (NM_000249.3):c.2092_2093del p.(Ser698Argfs*5) was identified with an allele frequency of 14%. We note that the variant was not retained as relevant variant by the analytical software due to his low allele frequency. In order to exclude allelic drop out explaining the low allele frequency, the variant was sequenced by another method. The allele frequency of 14% was confirmed. Sequencing of the tumor detected the variant with allele frequency of 56 and 42% in the biopsy and the surgical specimen respectively. The low allele frequency in blood is suggestive of mosaicism. Paternal family history of the patient is unknown. There is no history of cancer in the maternal family. Unfortunately, to date we could not tested the parents of the patient in order to confirm the potential de novo status of the variant.

Conclusions Here we describe the identification of a potential de novo pathogenic variant in MLH1 gene in a young patient who developed MSI colon cancer with loss of MLH1-PMS2 at age 41. De novo status could not be confirmed by familial study. If this is not a mosaicism, this case highlights the importance of carefully checking the sequencing data especially when using commercial software as this variant was classified in the low confidence data because of his low allele frequency. Missing this kind of variant can have dramatic consequences for relatives.

Keywords Mosaicism, De Novo, Lynch syndrome.

PP-048: Genetics: Research, Germline, Genomics, Variomics

Germline genetic variants that affect Wnt signaling as cause of serrated and adenomatous polyposis

Isabel Quintana¹, Mariona Terradas¹, Pilar Mur², Iris Te Paske³, David Torrents⁴, Montserrat Puiggròs⁴, Romina Royo⁴, Raul Tonda⁵, Genís Parra⁵, Davide Piscia⁵, Sergi Beltrán⁵, Matilde Navarro², Virginia Piñol⁶, Joan Brunet⁷, Noemi Gonzalez Abuin¹, Sophia Peters⁸, Verena Verena Steinke Lange⁹, Anna Sommer⁸, Isabel Spier⁸, Yasmijn Van Herwaarden¹⁰, Galuh Astuti¹¹, Elke Hollinski⁹, Nicoline Hoogerbrugge³, Richarda De Voer³, Stephan Aretz⁸, Gabriel Capellá², Laura Valle ²

¹Hereditary Cancer Program, Catalan Institute of Oncology; Oncobell Program, IDIBELL, Hospitalet de Llobregat, Barcelona, Spain. ²Hereditary Cancer Program, Catalan Institute of Oncology; Oncobell Program, IDIBELL, Hospitalet de Llobregat, Barcelona; Centro de Investigación Biomédica en Red de Cáncer (CIBERONC), Madrid, Spain. ³Department Human Genetics, Radboud Institute for Molecular Life Sciences, Radboud University Medical Center, Nijmegen, The Netherlands. ⁴Life Sciences Department, Barcelona Supercomputing Center (BSC), Barcelona, Spain. ⁵CNAG-CRG, Centre for Genomic Regulation (CRG), The Barcelona Institute of Science and Technology, Barcelona, Spain. ⁶Gastroenterology Unit, Hospital Universitario de Girona Dr Josep Trueta, Girona, Spain. ⁷Hereditary Cancer Program, Catalan Institute of Oncology, IDIBGi, Girona, Spain. ⁸Institute of Human Genetics, School of Medicine, University of Bonn; National Center for Hereditary Tumor Syndromes, University Hospital Bonn, Bonn, Germany. ⁹Medizinische Klinik Und Poliklinik IV, Campus Innenstadt, Klinikum Der Universität München; MGZ—Medizinisch Genetisches Zentrum, Munich, Germany. ¹⁰Department of Gastroenterology and Hepatology, Radboud University Medical Center, Nijmegen, the Netherlands. ¹¹Department of Human Genetics, Radboud University Medical Center, Nijmegen, The Netherlands

Background and aim Much of the genetic predisposition to colonic polyposis, and particularly to serrated polyposis (SP), remains largely unknown. Only RNF43, a negative regulator of Wnt signaling, has been causally linked to SP, a heterogeneous disease associated with significant personal and familial colorectal cancer risk. Almost all currently known hereditary colorectal cancer and polyposis genes affect DNA repair, BMP/TGF-β pathway, and Wnt signaling, being the latter particularly associated with adenomatous and serrated polyposis phenotypes. Here we aimed to identify new SP predisposition genes following a pathway-centered analysis of exome sequencing data.

Method Exome sequencing data from 44 SP patients was analyzed considering genes involved in DNA repair, BMP/TGF-β and Wnt signaling pathways. Pathway- and gene-centered burden tests were performed to evaluate the association with SP. Positively associated genes were evaluated in 98 additional SP and 105 adenomatous polyposis (AP) patients, and TOP/FOP reporter assays were performed to functionally study the effect on Wnt activation of the identified genes and variants.

Results Mutation burden analyses identified higher frequency of damaging and predicted damaging variants in Wnt-related genes in SP patients compared to controls, particularly in 11 Wnt genes, including RNF43. Validation in SP and AP cohorts led to the selection of DKK1, DKK4, HECW1, ITPR3, PSMC3, PSME4, TLE4 and WNT9B as new candidate polyposis genes. Except for ITPR3, all genes belong to the canonical Wnt signaling and a negative regulatory effect was observed for DKK1, DKK4, HECW1, TLE4 and WNT9B, but not for the proteasome components PSMC3 and PSME4. Deleterious functional effects were observed for several variants in Wnt antagonists identified in patients.

Conclusions Wnt signaling-activating variants in DKK1, DKK4, HECW1, TLE4 and RNF43 predispose to SP, and, at least for HECW1, to adenomatous polyposis. An association with polyposis predisposition was also identified for ITPR3, PSMC3, PSME4 and WNT9B.

Keywords Serrated polyposis, Adenomatous polyposis, Wnt, TGF-beta, DNA repair.

PP-049: Genetics: Research, Germline, Genomics, Variomics

Examining differential gene expression in Juvenile Polyposis Syndrome

Maiah H. Dent ¹, Bridgid E. Greed¹, Michael Xie², Natali S. Sobel Naveh³, Garrett M. Brodeur¹, Zvi Cramer⁴, Melani Duvall¹, Kathryn E Hamilton⁵, James Howe⁶, Tatiana Karakasheva⁵, Petar Mamula⁵, Christopher Lengner⁴, Suzanne P. MacFarland¹

¹Division of Oncology, Department of Pediatrics, Children’s Hospital of Philadelphia, Philadelphia, PA, 19,104. ²Division of Bioinformatics, Children’s Hospital of Philadelphia, Philadelphia, PA, 19,104. ³Department of Pathology and Lab Medicine, Children’s Hospital of Philadelphia, Philadelphia, PA, 19,104. ⁴Department of Biomedical Sciences, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA, 19,104. ⁵Division of Gastroenterology, Hepatology, and Nutrition, Department of Pediatrics, Children’s Hospital of Philadelphia, Philadelphia, PA, 19,104. ⁶Department of Surgery, University of Iowa Health Care, Iowa City, IA, 52,242

Background and aim Juvenile Polyposis Syndrome (JPS) is a hereditary cancer predisposition disorder in which 40–60% of patients have a known driver mutation in BMPR1A or SMAD4, both part of the TGFβ/BMP pathways. These “mutation-positive” patients exhibit a clinically distinct phenotype compared to patients who meet clinical criteria for JPS but lack a germline mutation, or “mutation-negative JPS.” We compared these patients to “mutation-negative” patients to better understand the mechanistic difference between the groups.

Method Our “mutation positive” cohort consisted of patients with a germline mutation in either BMPR1A (n = 3) or SMAD4 (n = 3). One “mutation-negative” individual (n = 1) was identified as having a germline mutation in BMPR2, a proposed candidate driver of JPS, and was also included in this group. RNA was extracted from blood and patient-derived organoids in these patients. Following cDNA conversion, expression of BMPR1A, SMAD4, and BMPR2 was measured by RT-qPCR. Data were normalized to TBP and RPLP0 using the ∆∆Ct method. Protein was extracted from the organoids for BMPR1A and SMAD4 Western Blotting.

Results Western blot analysis of our BMPR2 mutation positive patient showed decreased SMAD4 and BMPR1A expression in organoids derived from polyp tissue. Patients positive for a mutation in BMPR1A or SMAD4 all showed decreased expression in SMAD4. Increased gene expression was seen in both BMPR1A and SMAD4 in the majority of "mutation-positive" patients, regardless of their mutation, through RT-qPCR with plasma-derived RNA. All organoid samples exhibited downregulation in BMPR1A. Organoids from patients with BMPR1A mutations showed increased expression in SMAD4, while those with mutations in BMPR2 and SMAD4 showed downregulation in SMAD4.

Conclusions These data demonstrate that “mutation-positive” patients experience differential gene product regulation on both protein and RNA levels. Western blot analysis shows that all mutations result in downregulation of SMAD4 protein expression in polyp and colon tissue in these patients. RT-qPCR analysis mimics the SMAD4 Western blot data in organoids derived from patients with SMAD4 and BMPR2 mutations, although patients with BMPR1A mutations display upregulation in SMAD4. Plasma-derived RNA exhibits the opposite trend, with upregulation in both BMPR1A and SMAD4 in almost all patients. Additional studies examining gene regulation and gene expression in “mutation negative” patients will be important to further elucidate the molecular mechanism of JPS.

Keywords Juvenile Polyposis Syndrome, TGFβ/BMP pathways.

PP-050: Genetics: Research, Germline, Genomics, Variomics

Profuse Polyposis due to concomitant pathogenic variants in BMPR1A and MSH6 – A case study

Margaret Omalley ¹, Lisa Laguardia¹, Susan Milicia¹, Brandie Leach³, Nicholas Smith¹, Joshua Sommovilla¹, David Liska¹, Carol A Burke²

¹Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Department of Colorectal Surgery, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, USA. ²Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Departments of Gastroenterology, Hepatology and Nutrition, Digestive Disease and Surgery Institute, Cleveland Clinic, Cleveland, USA. ³Sanford R. Weiss, MD Center for Hereditary Colorectal Neoplasia, Department of Genetics, Genomic Medicine Institute, Cleveland Clinic, Cleveland, USA

Background and aim Juvenile polyposis syndrome (JPS) is a rare hamartomatous polyposis syndrome that causes juvenile polyps in the stomach, small bowel, and colorectum and an increased risk of gastric and colorectal cancer. JPS is caused by pathogenic variants in BMPR1A and SMAD4. Lynch syndrome is the most common hereditary colorectal cancer syndrome, caused by pathogenic variants in MLH1, MSH2, MSH6, PMS2, and EPCAM. Individuals with Lynch syndrome are at an increased risk of developing colorectal, endometrial, gastric, small bowel, pancreatic, urothelial, skin and other cancers. While Lynch syndrome is not considered a polyposis syndrome, individuals with Lynch syndrome can develop adenomatous polyps. While colonoscopy and EGD screening recommendations are similar for JPS and Lynch syndrome, patients with Lynch syndrome also require screening outside of the digestive track. Herein, we report a case of an individual with JPS and Lynch syndrome.

Method Patient was enrolled in the Jagelman Registry. The registry and electronic medical record were used to obtain personal and family history, genetic testing, colonoscopy, EGD and operative findings.

Results The patient is a 64 year old male with a personal history diabetes and a thoracic aortic aneurysm. He has a father and brother deceased from cancer (unknown primary or age) as well as two sisters, one brother, three daughters and two sons living and unknown to be affected with polyposis or cancer. The patient was diagnosed with polyposis at age 14 due to symptoms. He underwent multiple colon surgeries, his first at age 14 (report is not available), and ultimately underwent an abdominoperineal resection with end ileostomy due to multiple polyps including inflammatory juvenile type polyps, tubular adenomas and a large tubulovillous adenoma at age 55. The patient was originally diagnosed with and managed as though he had familial adenomatous polyposis. At age 60 he presented to our center and underwent multigene panel testing that revealed a pathogenic variant, c.730C > T (p.Arg244*), in BMPR1A and also had a variant of uncertain significance (VUS) in MSH6 c.2972C > T (p.Pro991Leu). The patient underwent routine screening and was found to have duodenal adenoma. At age 63, the MSH6 VUS was reclassified to likely pathogenic.

Conclusions We present a case report of a JPS patient who was also found to have a pathogenic BMPR1A pathogenic variant as well as a likely pathogenic MSH6 variant. Utilization of a multigene panel in this case helped to establish the correct polyposis diagnosis as well as a second pathogenic variant that required screening for additional cancers. This information is also critical to inform cascade testing for the patient’s at-risk family members.

Keywords Lynch Syndrome, Juvenile polyposis, multigene panel, reclassified.

PP-051: Genetics: Research, Germline, Genomics, Variomics

Highly sensitive characterization of microsatellite instability patterns in non-neoplastic tissue and tumors from biallelic germline mismatch repair mutation carriers

Fátima Marín¹, Júlia Canet¹, Vanesa Pérez Alonso², Nerea Domínguez², Constantino Sábado³, Victoria Fioravantti⁴, Héctor Salvador⁵, Antonio Molinés⁶, Núria Dueñas¹, Estela Carrasco⁷, Marta López⁸, Ángela Pastor⁸, Luis Allende⁸, Gabriel Capellá¹, Daniel Rueda⁹, Luis Ignacio González Granado¹⁰, Marta Pineda ¹

¹Hereditary Cancer Program, Bellvitge Biomedical Research Institute (IDIBELL), ONCOBELL program. Catalan Institute of Oncology. CIBERONC. L’Hospitalet de Llobregat, Barcelona, Spain. ²Pediatrics and Haemato-Oncology Service, Hospital Universitario Doce de Octubre, Research institute i + 12, Madrid, Spain. ³Pediatric Onco-Haematology Service, Hospital Universitari Vall d'Hebron, Barcelona, Spain. ⁴Oncogenomics Unit, Pediatric Onco-Haematology Service, Hospital Infantil Universitario Niño Jesús, Madrid, Spain. ⁵Pediatric Onco-Haematology Service, Hospital Sant Joan de Déu, Esplugues de Llobregat, Barcelona, Spain. ⁶Pediatric Onco-Haematology Service, Hospital Universitario Hospital Materno-Infantil de Canarias, Spain. ⁷High Risk Unit and Cancer Prevention (UARPC), Hospital Materno-Infantil Vall d’Hebron, Barcelona, Spain. ⁸Immunology Service, Hospital Universitario Doce de Octubre, Instituto de Investigación i + 12, Madrid, Spain. ⁹Hereditary Cancer Laboratory, Hospital Universitario Doce de Octubre, Research institute i + 12, Madrid, Spain. ¹⁰Primary Immunodeficiency Unit, Paediatrics, Hospital Universitario Doce de Octubre, Research institute i + 12, Madrid, Spain

Background and aim Constitutional MisMatch Repair Deficiency (CMMRD) is a rare and devastating childhood-onset cancer predisposition syndrome caused by germline biallelic inactivating mutations in mismatch repair genes (MLH1, MSH6, MSH2, PMS2). Its accurate and prompt diagnosis is essential for therapeutic decisions and surveillance recommendations. However, CMMRD diagnosis is particularly challenging mainly due to its extremely low prevalence, the broad spectrum of associated tumors and the difficulties in sequencing PMS2 gene. Detection of microsatellite instability (MSI), a hallmark of MMRD, in blood is pathognomonic of CMMRD and improves CMMRD diagnosis. In this regard, novel highly sensitive NGS approaches have allowed the detection of low MSI levels present in blood from these patients.

Aiming to improve the identification and characterization of CMMRD in Spain, we have stablished a multicentric and multidisciplinary consortium of experts in this syndrome, which have allowed the detection of 3 new patients from the 12 reported in Spain so far. Here, we present the characterization of the MSI patterns in non-neoplastic tissues and tumors from CMMRD patients.

Method MSI at high sensitivity (hs-MSI) was evaluated in DNA samples (15 peripheral blood lymphocytes, 10 other non-neoplastic tissues and 10 tumors) from seven CMMRD patients using a custom NGS targeted panel of microsatellite and a specific bioinformatic pipeline (PMID: 31,494,577). Hs-MSI metrics were determined, including the hs-MSI score (percentage of unstable markers) and hs-MSI mean frequency of unstable markers.

Results All non-neoplastic tissues, including cerebrospinal fluid, bone marrow and buccal mucosa, showed high levels of hs-MSI score above a preestablished threshold in control samples. Moreover, high correlation was observed between unstable markers in serial blood samples from the same patient.

All CMMRD tumors (1 high grade glioma, 7 lymphomas, 1 hepatoblastoma and 1 Wilms tumor) showed high levels of hs-MSI and higher clonality than non-neoplastic tissues (p < 0.05). hs-MSI patterns were enriched mainly in 1-base deletions, while the high glioma tumor with somatic mutations in POLE/POLD1 also showed high frequency of 1-base insertions.

Conclusions The hs-MSI tool allowed the detection of low-level MSI in non-neoplastic tissue, independently of the tissue origin, confirming its utility as ancillary diagnostic test (PMID: 33,622,763). It is has also proven useful for the identification and characterization of CMMRD associated tumors.

Keywords Constitutional Mismatch Repair Deficiency, microsatellite instability, high sensitivity.

PP-052: Genetics: Research, Germline, Genomics, Variomics

An Assessment of Racial and Gender Disparities in Genetic Counseling or Testing for Patients with Mismatch Repair Deficient Tumors

Jewel Tomlinson¹, Jennifer Vazzano, DO, MS², Peter P Stanich, MD ¹, Rachel Pearlman, LGC¹, Wei Chen, MD, PhD², Wendy L Frankel, MD², Heather Hampel, LGC¹

¹Department of Internal Medicine, The Ohio State University Wexner Medical Center, Columbus, OH. ²Department of Pathology, The Ohio State University Wexner Medical Center, Columbus, OH

Background and aim Patients with absence of a mismatch repair protein(s) on immunohistochemistry (IHC) found by universal tumor screening (UTS) are considered to have mismatch repair deficiency (dMMR). Those with absence of MLH1 and PMS2 need follow-up testing for acquired MLH1 hypermethylation (MHM), which is done directly or indirectly by assessing for BRAF mutation. Patients with dMMR tumors with MHM ruled out when indicated should undergo genetic counseling and testing to determine if they have Lynch syndrome (LS). Gender and racial disparities have been seen in multiple health care settings including follow-up genetic counseling and testing in UTS programs. Our goal was to see if there were gender, racial or ethnic disparities in follow-up genetic counseling and testing at a tertiary care institution.

Method Electronic medical records were reviewed for all cases (colorectal, endometrial, others) screened with UTS between 1/9/2009 and 5/27/2021 to note patient gender, race, ethnicity; results for MMR IHC, BRAF mutation or MHM; and appointments with genetics and uptake of genetic testing.

Results 3145 patients were diagnosed and got UTS with MMR IHC including 1538 males and 1607 females. 2718 were white, 295 black, 51 Asian, and 33 Hispanic. Race was not available for 48. Among 3145 cases, 205 (6.5%) had a non-methylated dMMR tumor, without a BRAF mutation or MHM (when indicated), and had a clear indication for genetic counseling. There were 113 males (55.1%) and 92 females (44.9%). Of the 202 with known race, 168 were white (83.2%), 23 black (11.4%), 10 Asian (5.0%), and 1 was Hispanic (0.5%). Of 205 dMMR patients, 96 (46.8%) were seen by genetics; rates were similar for men (44.2%) and women (50%). By a chi-squared test of independence, we did not find any significant differences in patients seen by genetics according to race or ethnicity with 47.0% (79/168) white, 39.1% (9/23) black, 70% (7/10) Asian and 100% (1/1) Hispanic (p = 0.28). Notably, 100% of patients seen by genetics elected to pursue genetic testing.

Conclusions While the overall rate of genetic counseling completion for patients with non-methylated dMMR tumors (46.8%) needs improvement, no significant disparities were seen by gender or race. Rates in black and white patients and males and females were similar. Higher rates were seen in Asian and Hispanic patients but were not statistically significant. Importantly, when seen by genetics patients were equally likely to pursue germline genetic testing, regardless of race or gender.

Keywords Diversity, Equity, Inclusion, Genetic Counseling.

Table 1

PP-053: Genetics: Research, Germline, Genomics, Variomics

Analysis of genetic mosaicism in attenuated adenomatous polyposis

Victor Lorca¹, Carmen Poves², María Luisa González³, Isabel Díaz⁴, Concepción Alonso⁵, Alejandra Rosell⁶, Ada Esteban Sánchez¹, Daniel Rueda⁷, Beatriz Hidalgo⁷, Jose Manuel Sánchez Zapardiel⁷, Luis Robles⁸, Pedro Pérez Segura⁴, Vanesa García Barberán¹, Miguel De La Hoya¹, Pilar Garre ⁹

¹Laboratorio de Oncología Molecular, Hospital Clínico San Carlos, IdISSC, Madrid, Spain. ²Servicio de Aparato Digestivo, Hospital Clínico San Carlos, IdISSC, Madrid, Spain. ³Servicio de Anatomía Patológica, Hospital Clínico San Carlos, IdISSC, Madrid, Spain. ⁴Servicio de Oncología Médica, Hospital Clínico San Carlos, IdISSC, Madrid, Spain. ⁵Genética Clínica, Servicio de Oncología Médica, Hospital de La Princesa, Madrid, Spain. ⁶Servicio de Anatomía Patológica, Hospital de La Princesa, Madrid, Spain. ⁷Laboratorio de Cáncer Hereditario, Servicio de Bioquímica, i + 12, Hospital 12 de Octubre, Madrid, Spain. ⁸Servicio de Oncología Médica, Hospital 12 de Octubre, Madrid, Spain. ⁹Unidad de Diagnóstico Molecular, Servicio de Análisis Clínicos, IML, Hospital Clínico San Carlos IdISSC, Madrid, Spain. Laboratorio de Oncología Molecular, Hospital Clínico San Carlos IdISSC, Madrid, Spain

Background and aim Somatic mosaicism in the APC gene has been described in cases of attenuated adenomatous polyposis (AAP) without a family history. However, mosaicism rate is still underestimated due to the limited sensitivity of conventional techniques and the availability of somatic samples for reliable genetic diagnosis. Recent studies based on NGS manage to increase the detection rate and they place it around 15–50% of unexplained AAPs (20–100 accumulated adenomas among large intestine). The number of samples included in each study and the criteria used for the definition of mosaics differs greatly between studies, which may explain the large variation in the detection rates.

Our main objective is to estimate the contribution of mosaicism phenomena in APC, as well as other polyposis/CRC predisposition genes, in a Spanish AAP study cohort, thorough NGS in multiple adenomatous samples.

Method We recruited 33 patients with more than 40 adenomas detected before the age of 60, or more than 30 before the age of 40, and with non-informative study for APC and MUTYH and no family history of polyposis/CRC. At least, two adenoma samples, normal colonic and peripheral blood DNA per patient, were sequenced using a custom NGS panel for the analysis of 15 polyposis/CRC predisposition genes. Validation was performed on additional adenomatous tissues using NGS and/or digital PCR (dPCR).

We have considered true mosaics those patients with at least 3 independent adenomas harboring the same mutation.

Results The analysis of 15 patients allowed the detection of 5 cases (33%) with a strong suspicion of mosaicism in APC, in no case detectable in germline DNA. Analysis of additional somatic samples confirmed the mosaicism in three cases (20%) and ruled it out in two others.

Conclusions NGS analysis in somatic DNA provides greater sensitivity in the identification of mosaicism phenomena in cases of unexplained AAP and without a family history. Furthermore, the results of this study show that both the number of samples analyzed and their location, are critical for the correct interpretation of the mosaics.

These preliminary data support the convenience of the molecular study in multiple somatic samples in those patients with more than 40 adenomas before the age of 60 without polyposis/CRC family history and with a previous non-informative germline study.

Keywords Mosaicism, Adenomatous Polyposis, CRC, NGS.

PP-054: Genetics: Research, Germline, Genomics, Variomics

Adenoma burden in unselected patients with pathogenic APC variants: A MyCode case series

Rachel Schwiter ¹, Heather Rocha¹, Alicia Johns², Melissa A Kelly¹, Adam H Buchanan¹

¹Genomic Medicine Institute, Geisinger, Danville, USA. ²Department of Population Health Sciences, Geisinger, Danville, USA

Background and aim Genomic screening can lead to improved ascertainment of molecular diagnoses and clinical outcomes for some hereditary cancer syndromes, but the clinical presentation and penetrance of colorectal adenomatous polyposis in the context of genomic screening in an unselected population is unknown. Here we describe the clinical presentation of individuals found to have a germline APC pathogenic or likely pathogenic (P/LP) variant associated with risk for familial adenomatous polyposis (FAP) through the MyCode Genomic Screening and Counseling (GSC) program in an unselected healthcare system cohort.

Method Electronic health record (EHR) review was completed for individuals with an APC P/LP variant identified via the MyCode GSC program between 2015–2021. Outcomes of adenoma burden, endoscopy procedure history, colectomy, and any cancer diagnoses from the MyCode APC P/LP cohort were compared to outcomes in individuals with a clinical presentation of FAP and to matched controls of MyCode participants without EHR evidence of an APC P/LP variant.

Results MyCode patient-participants with no prior clinical or molecular diagnosis of FAP comprised 24 individuals (100% white, 71% female) with a median age of 53 years (range 27–83 years) at the time of result disclosure. No patient-participants with an APC P/LP variant in an expected attenuated region who completed colonoscopy (n = 14/18) had polyposis diagnosed after result disclosure. Similarly, of those with an APC P/LP variant in an expected classic region (n = 6), none who completed colonoscopy showed evidence of polyposis at the time of result disclosure, though 50% (n = 3) had at least one sign of FAP in their personal or family history, one of whom developed oligo-polyposis several years after result disclosure. The rate of polyposis diagnoses among patient-participants with APC P/LP variants detected via MyCode was significantly lower than those with a prior clinical diagnosis and consistent with that of matched controls.

Conclusions The clinical presentation of unselected individuals with APC P/LP variants identified via genomic screening revealed a limited adenoma burden more consistent with that of matched controls than a clinically identified FAP cohort. These findings highlight challenges in applying risk management recommendations in an unselected setting as well as the need to further study the utility of genomic screening for APC P/LP variants in larger, more diverse cohorts.

Keywords familial adenomatous polyposis, attenuated FAP, APC gene, genomic screening, genetic counseling.

Fig. 1

MyCode patient-participants with an APC P/LP variant disclosed through GSC program. CRC = colorectal cancer.

Table 1

Summary of clinical history with genomic data for MyCode patient-participants with newly identified APC P/LP variants (n = 24).

PP-055: Genetics: Research, Germline, Genomics, Variomics

Whole Genome Sequencing Applied in Familial Hamartomatous Polyposis Identifies Novel Structural Variations

Revital Kariv ¹, Dvir Dahary², Yuval Yaron¹, Yael Petel Galil¹, Mira Malcov¹, Guy Rosner¹

¹Tel Aviv Sourasky Medical Center and Tel Aviv University. ²Geneyx LTD, Israel

Background and aim Hamartomatous polyposis syndromes (HPS) are rare cancer-predisposing disorders including Juvenile polyposis (JPS), Peutz-Jeghers (PJS) and PTEN-Hamartomatous-Syndromes (PHS). Penetrant-mutations in corresponding genes (SMAD4, BMPR1A, STK11, PTEN and AKT1), are usually diagnosed by next-generation-sequencing gene panel (NGS-GP) for tailored surveillance and pre-implantation-testing for monogenic-disorders (PGT-M).

We aim to study HPS cases with negative genetic workup by whole genome sequencing and analysis.

Method Five non-related probands with HPS phenotype (four with JPS and one with PJS), that underwent genetic workup without identified pathogenic mutations were offered WGS analysis. Genome sequencing was analyzed by Geneyx platform. First-degree family members were used for validation and segregation analyses. Variants validation was performed by PCR.

Results Whole-genome-sequencing (WGS) identified structural genetic alterations at all five probands (Table): Two novel inversions in BMPRA1 and STK11, two BMPR1A-deletions, known as founder among Bukharan-Jews and BMPR1A micro-deletion. BMPR1A inversion was validated by "junction fragment" amplification and direct testing (Fig. 1). PGT-M was performed by multiplex-PCR and enabled successful birth of non-carrier baby. All variants were validated and segregation revealed phenotype-genotype correlations (Fig. 2).

Conclusions WGS may be considered for HPS patients with no NGS-GP findings to exclude structural alterations.

Keywords Genome sequencing, Hamartomatous polyposis, Juvenile polyposis, Peutz-Jegher's syndrome, Inversion, deletion.

Analysis and validation of BMPR1A inversion:

Figure 1 A. Schematic analysis of BMPR1A inversion: At the wild type state, chr.10 cannot be amplified by the combination of F up and R up or F down and R down primers, due to the distance between them and their orientation. The dashed lines indicate the break points of the inversion. In affected individuals, both primers pairs are in proximity and alignment that allow the amplification of the inversion junction fragment and as a result the detection of a ~ 200 bp PCR products. B. Sample1 represent the affected proband while the other 3 samples are of healthy controls. The normal region of the gene- wild type amplicon, is demonstrated in all DNA samples, while it is amplified in the proband. However, the junction fragment that represents the mutant allele was amplified only in the proband that carries the BMPR1A inversion

Pedigrees and segregation analysis results of the study families

WGS genetic alterations findings of the study probands

Family	Variant
A	INV BMPR1A Chr10:87,852,798–88,575,769 (723 Kb)
B	INV STK11 Chr19:1,206,071–1,274,737 (68 Kb)
C	BMPR1A del exons 3–13 (Chr10:88,611,032- 88,964,753)
D	BMPR1A del exons 3–13 (Chr10:88,611,032- 88,964,753)
E	BMPR1A c.1419delT, p.Val474Cysfs*24

PP-056: Genetics: Research, Germline, Genomics, Variomics

Prevalence of Sebaceous and Non-Sebaceous Cutaneous Manifestations in Patients with Lynch Syndrome – A Review of The Literature

Shahram Hama Ali Aziz¹, Hazel O'Sullivan², Terri McVeigh ²

¹Department of Physiology. University of Sulaimani, Sulaymaniyah, Iraq & Department of Dentistry, Komar University of Science and Technology, Sulaymaniyah, Iraq. ²The Royal Marsden NHS Foundation Trust, London, UK

Background and aim Lynch Syndrome is an autosomal disorder characterized by an increased lifetime risk of malignancy, mainly colorectal and endometrial cancers. A subset of patients also demonstrates cutaneous manifestations of disease. Characterization of Lynch-related skin lesions could help in early recognition of patients with Lynch Syndrome.

Method A broad search of the literature on OVID Medline and Embase was carried out using relevant keywords to capture papers reporting cutaneous manifestations in patients with Lynch Syndrome. The results were uploaded into Mendeley reference management software. The PRISMA workflow was used in the literature selection process.

Results Data from 965 cases from 413 studies were included in this systematic literature review including 383 molecularly confirmed Lynch Syndrome cases. The main skin lesions were: Sebaceous adenomas (33%), sebaceous carcinomas (20%), keratoacanthomas (12%), sebaceomas (11%), squamous cell carcinomas (17%), and basal cell carcinomas (7%). ""MSH2"" genetic variants were the most common underlying genotype in the cases (72%). Assessment of mismatch repair by immunohistochemistry, microsatellite instability analysis, or both were performed on 332 skin lesions from 223 (58%) molecularly confirmed cases. In those skin lesions, 95% of Immunohistochemistry and 91% of the microsatellite instability test results were concordant with the underlying genotype.

Conclusions Sebaceous skin lesions are a well-recognized phenotypic feature of Lynch Syndrome. Our results show that squamous and basal cell carcinomas are relatively common in Lynch syndrome patients; however, available evidence cannot attribute their etiologies to Lynch syndrome. Immunohistochemistry and/or microsatellite instability testing of skin tumors in patients with cutaneous malignancy and a family history of Lynch Syndrome-associated cancers may be a useful approach in identifying patients requiring germline genetic testing for Lynch Syndrome.

Keywords Lynch syndrome, Muir-Torre syndrome, Mismatch repair, Sebaceous tumor; squamous cell cancer; basal cell cancer; keratoacanthoma.

PP-057: Clinical: Diagnosis, Endoscopy, Management, Implementation

A Survey on Universal Tumor Screening for Lynch Syndrome Practices Among US Cancer Hospitals

George Kyle ¹, Wendy Frankel³, Rachel Pearlman⁴, Heather Hampel²

¹College of Medicine, The Ohio State University, Columbus, United States. ²Division of Cancer Genomics, City of Hope National Medical Center, Duarte, United States. ³Department of Pathology, The Ohio State University, Columbus, United States. ⁴Division of Internal Medicine, The Ohio State University, Columbus, United States

Background and aim Universal tumor screening (UTS) for Lynch syndrome (LS) on all colorectal (CRC) and endometrial cancers (EC) is recommended by multiple professional societies. In 2012, it was reported that UTS is not uniformly performed, only occurring in 15% of Community Hospital Cancer Programs (CCPs), 36% of College of Surgeons-Accredited Community Hospital Comprehensive Cancer Programs (COMPs), and 71% of National Cancer Institute–Comprehensive Cancer Centers (NCI-CCCs). Our aims were to ascertain what percentage of hospitals were performing UTS for LS in 2021, and to assess if there was a difference in UTS rates between CRC and EC.

Method All 51 NCI-CCCs and 100 randomly selected CCPs and COMPs were asked to complete a 23-question survey on REDCAP assessing their UTS protocol for CRC, EC, and other LS-associated cancers. Hospitals surveyed in this study were selected from three categories: CCPs, COMPs, and NCI-CCCs. Genetic counselors, tumor registrars, or pathologists completed the survey.

Results 74 survey responses were received, for a response rate of 29.5% [46/51 (90%) of NCI-CCCs, 16/100 (16%) of COMPs, and 12/100 (12%) of CCPs)]. 85.1% (63) performed UTS in all patients with CRC as follows: 89% of NCI-CCCs, 81% of COMPs, and 75% of CCPs. Rates of UTS for CRC patients did not significantly differ based on the type of cancer program (P = 0.41789). Next, 68% (n = 57) of respondents performed UTS in all patients with EC, including 80% of NCI-CCCs, 50% of COMPs, and 50% of CCPs While it was not statistically significant (P = 0.059997), our data suggests that UTS rates for EC were higher among NCI-CCCs than COMPs or CCPs. UTS for Lynch syndrome is significantly (P = 0.00937) more often performed for CRC than for EC patients.

Comparing UTS for LS in CRC patients to past studies, there was a statistically significant increase from Beamer et al. (2012) (P < 0.00001), but no statistically significant difference from Hissong et al. (2018) (P = 0.805952).

Conclusions Our survey of cancer programs provides information about UTS in CRC and EC. There was no significant difference between type of cancer program and UTS rates for CRC. NCI-CCCs are more likely to be performing UTS for EC than COMPs or CCPs, although this difference was approaching significance. CRCs were more likely to undergo UTS than ECs, and there was a statistically significant increase in UTS rates for CRC between our study and Beamer et al. (2012), but not between Hissong et al. (2018).

Keywords Lynch, Screening, Colorectal, Endometrial.

Fig. 2

Kaplan–Meier curve for overall survival of CDKN2A mutation carriers after diagnosis of primary PDAC (n = 31)

Fig. 1

Construction of nomogram and small program to predict the probability of Lynch syndrome. (A) The details of nomogram including the factors of age, gender, personal history, family history and deficient pattern of mismatch repair. Other Lynch syndrome related cancers refer to gastric, endometrial, small bowel, ovarian and so on. (B) The QR code and questionnaire of the small program

Fig. 2

Validation of nomogram to predict the probability of Lynch syndrome. (A) ROC curves comparing the specificity and sensitivity of the nomogram-based model and four current screening criteria to identify Lynch syndrome. The black dot represents the best cut-off value (0.435). (B) ROC curve of the nomogram in the external validation cohort.

Fig. 1

Frequency of multiple pathogenic variants identified through multi-gene panel testing

Fig. 2

EMAST analysis. EMAST analyses using the tetranucleotide the marker vWA performed on normal colonic mucosa, tubular adenomas and (metastatic) adenocarcinoma derived from indexpatient II.8 and sibling II.11 and on normal tissue of three healthy donors. Arrows highlight tetranucleotide repeat shifts in present in normal tissue DNA of II.8 and II.11 and not present in normal DNA of three healthy control individuals (HC1-HC3). Abbreviations: N = normal muscosa, TA = tubular adenoma, LG = low grade, HG = high grade, AC = adenocarcinoma, MET = metastatsis