Durvalumab for Resectable non-small cell lung cancer (NSCLC)

While surgery remains the primary curative-intent treatment for early-stage NSCLC, 30 to 55% of patients have tumour recurrence within 5 years. Neoadjuvant or adjuvant chemotherapy offers only 5% improvement in 5 year survival as compared with surgery alone. The AEGEAN trial (N Eng J Med 2023;389(18):1672–1684) was a phase 3, double-blind, placebo-controlled trail to investigate durvalumab. Patients with resectable stage II to IIIB NSCLC were randomly assigned to receive platinum-based chemotherapy plus durvalumab (400 patients) or placebo (402 patients) before surgery, followed by adjuvant durvalumab or placebo.

At the first interim analysis, significantly longer duration of event-free survival was observed with durvalumab than with placebo; the HR for disease progression, recurrence, or death was 0.68 (95% CI [95%confidence interval], 0.53 to 0.88; p=0.004). At 12 months, event-free survival was 73.4% in durvalumab group, compared with 64.5% with placebo. Furthermore, better rates of complete pathological response were noted with durvalumab than with placebo. Importantly, these benefits were observed regardless of stage and PD-L1 expression. Grade 3–4 severity adverse events occurred in 42.4% of patients with durvalumab and in 43.2% with placebo.

Perioperative durvalumab plus neoadjuvant chemotherapy should be considered as a potential new treatment option for resectable NSCLC given the greater event-free survival and pathological complete response with similar safety profiles.

Selpercatinib for advanced Ret fusion NSCLC

RET gene fusions are oncogenic drivers in multiple cancers and are present in 1–2% of NSCLC. Selpercatinib, which has central nervous system penetration, is a highly selective, potent RET kinase inhibitor with encouraging results in non-randomised studies. Zhou and colleagues (N Engl J Med 2023;389(20):1839–1850) conducted a randomised phase three trial in patients with newly diagnosed advanced RET fusion–positive NSCLC. They aimed to evaluate selpercatinib monotherapy compared with platinum-based chemotherapy with or without pembrolizumab.

A total of 261 patients with RET fusion–positive advanced NSCLC were randomised. The primary endpoint of progression free survival was sequentially tested, first in 212 participants who were planned for pembrolizumab therapy after randomisation if allocated to control (intention-to-treat–pembrolizumab), before testing in the overall population. Participants had stage IIIB to IV unresectable NSCLC and most had an ECOG performance status of 0–1 (97%).

At the time of the pre-planned interim analysis, selpercatinib significantly increased median progression-free survival (24.8 months (95% CI, 16.9 to not estimable)), compared with control (11.2 months (95% CI, 8.8 to 16.8), HR 0.46; 95% CI, 0.31 to 0.70; p<0.001) in the intention-to-treat pembrolizumab population. Results were similar in the overall intention-to-treat analysis. Analyses assessing central nervous system progression and treatment responses also favoured selpercatinib.

Grade 3–4 adverse events were more common with selpercatinib, including deranged liver function, hypertension, and oedema. As expected, chemotherapy related side-effects of neutropaenia, anaemia and thrombocytopaenia, were more common in the control group. In advanced RET fusion positive NSCLC, treatment with selpercatinib has shown significantly longer progression-free survival. However, the adverse effect profile needs to be considered.

The ‘Canpredict’ model and lung cancer risk

Lung cancer screening with low dose CT reduces mortality. Risk prediction models to select people at high risk of lung cancer for screening programmes are potentially important to minimise wastage of resources spent on screening low risk people. Liao and colleagues (Lancet Respir Med 2023;1;1:685–97) conducted a retrospective, population-based, cohort study using electronic healthcare records in the UK to develop, validate and test a new lung cancer prediction model, the ‘CanPredict’ model.

The authors included people aged 25–84 years without a diagnosis of lung cancer before cohort entry, to estimate the risk of being diagnosed with lung cancer in the next 10 years. The authors included primary care electronic healthcare records from approximately 12.99 million in a derivation cohort and approximately 6.95 million in a validation cohort.

During follow-up, 73 380 (0.56%) people in the derivation cohort and 22 838 (0.63%) people in the validation cohort developed lung cancer. The ‘CanPredict’ model included age, sex, ethnicity, socioeconomic status, smoking status and intensity, alcohol status, BMI, comorbidities, asbestos exposure, family history of lung cancer and personal history of cancer.

The CanPredict model had excellent discrimination in both the derivation in the overall population aged 25–84 years and performed better than seven other screening tools in those meeting the targeted lung health check criteria (aged 55–74 years, ever smokers). However, in the validation cohort, while performance was still good overall, the CanPredict model produced similar results to other screening models in the targeted lung health check population. More work is needed to see whether the CanPredict model, may prove a useful tool to target lung cancer screening.

Air pollution and lung cancer in never-Smokers (Lcins)

LCINS are increasing in number and are the eighth most common cause of cancer death in the UK. LCINS are more likely to be adenocarcinomas with EGFR mutations, have a female preponderance, and a predilection to East Asian populations. Air pollution is thought to be important in the development of LCINS. Particulate matter (PM) - everything in the air that is not a gas - is a key component of air pollution and is classified by size. Fine particles≤2.5 µm (PM2.5) are mainly responsible for lung cancer as their small size enables them to travel deep into the lung.

Hill and colleagues (Nature 2023;616: 159–167) investigated whether air pollutants promote inflammatory change in lung tissues leading to an expansion of pre-existing mutated clones, promoting cancer formation. They aimed to identify potential mechanisms of air-pollution-induced lung tumour promotion by incorporating epidemiological evidence with functional preclinical models and clinical cohorts.

They found significant associations between geographical PM2.5 levels and the incidence of EGFR positive lung cancer in three separate cohorts. Functional mouse models showed that air pollutants cause an influx of macrophages and release of interleukin-1β in the lung leading to progenitor-like cell state in EGFR mutant lung alveolar type II epithelial cells, suggesting a tumour promoting role for PM2.5.

Even though this study has limitations, the results support a public health mandate to restrict particulate emissions in urban areas.