ESMO ORAL SESSION: Applying CSO cheklist for Phase III trials supporing EMA Lung Cancer approvals, 2020-2025

For those of you attending ESMO Madrid, in addition to the CSO symposium, Javier Benítez, from the Rio group is giving an oral presentation on Saturday the 24th. 

It covers work we’ve done applying the CSO checklist to EMA approvals for lung cancer between 2020 and 2025. 

we are currently preparing the full paper. Here is the abstract:

Background: Common Sense Oncology (CSO) prioritises meaningful endpoints, appropriate comparators, patient-centred outcomes, toxicity burden, and transparent reporting. We assessed adherence to a full CSO-derived checklist in trials supporting recent European Medicines Agency (EMA) lung cancer approvals.

Methods: We performed a retrospective review of EMA lung cancer phase III approval-relevant trial-indications from 1 January 2020 to 31 December 2025. After regulatory-date and pivotal-evidence verification, 31 trial-indications were included in a single-review working analysis. Seventy checklist elements were scored separately: 25 protocol/design elements using a 0-2 scale and 45 reporting/publication elements using a 0-1 scale. Not applicable items were excluded from denominators.

Results: Median overall adherence was 57.1% (mean 56.7%); median protocol/design and reporting/publication adherence were 58.3% and 56.1%. Comparator selection was fully met in 25/31 trial-indications and partially met in 6/31. Overall survival or a justified validated surrogate was fully met in 12/31, HRQoL in 9/31, and clinically meaningful sample-size assumptions in 10/31. Conventional reporting was stronger: primary endpoint statement in abstract, grade 3-5 toxicity, CTCAE toxicity reporting, funding, sponsor role, and conflicts of interest were each met in 31/31. Recurrent gaps included dropout/informative-censoring mitigation, chronic low-grade toxicity, and time-toxicity (0/31 each). Among non-OS-primary trials, OS HR/CI with absolute benefit was reported in 1/19 abstracts. HRQoL abstract summaries, patient involvement, ESMO-MCBS reporting, and cost/resource statements were absent in 31/31.

Conclusions: Full-element CSO appraisal shows that recent EMA lung cancer approval trials generally satisfy conventional reporting standards but incompletely address patient-centred design, censoring, chronic toxicity, time burden, and benefit contextualisation. Prospective use of granular CSO criteria may improve the clinical interpretability of approval trials.