EGFR inhibitors: two improve survival, none wins outright

Design

systematic review and frequentist random-effects network meta-analysis of randomised controlled trials, treatments ranked by P-score

Population

nine randomised trials for progression-free survival, eight for overall survival and safety, in treatment-naive EGFR-mutant advanced non-small-cell lung cancer

Primary outcome

overall survival, progression-free survival and grade ≥3 adverse events versus first-generation inhibitors

Effect

overall survival — dacomitinib HR 0.75 (95% CI 0.59–0.95), osimertinib 0.80 (0.67–0.96); progression-free survival — furmonertinib 0.44 (0.34–0.57); no significant safety differences

First-line choice among EGFR tyrosine kinase inhibitors in advanced non-small-cell lung cancer rests on very few head-to-head trials, which is what network meta-analysis exists for. This one pooled nine randomised trials for progression-free survival and eight for overall survival and grade 3 or worse adverse events, comparing first, second and third-generation agents in treatment-naive EGFR-mutant disease.

For overall survival, only dacomitinib (hazard ratio 0.75, 95% CI 0.59–0.95) and osimertinib (0.80, 0.67–0.96) beat first-generation inhibitors significantly. For progression-free survival, every second and third-generation agent did, with furmonertinib ranked highest at 0.44 (0.34–0.57). On safety, no agent differed significantly from first-generation inhibitors in grade 3 or worse adverse events — befotertinib and dacomitinib had numerically higher odds with wide intervals crossing the null.

The authors' conclusion is appropriately flat: several agents improve on first-generation therapy, none is clearly superior across all outcomes, and osimertinib's efficacy was the most consistent. Note the one specific caveat they raise — osimertinib's safety advantage appeared only in a sensitivity analysis excluding FLAURA China. A finding that depends on which trial you remove is a finding to hold loosely.

• Progression-free survival rankings are not survival rankings — furmonertinib topped the first and has no overall survival advantage demonstrated.
• Indirect comparison is the whole method here; these agents were mostly not tested against each other.
• Generic gefitinib and erlotinib remain widely used in India on cost grounds, and this quantifies what that costs in overall survival terms — a hazard ratio of about 0.8 against osimertinib.
• Central nervous system activity and resistance mechanisms at progression differ between generations and are not captured by these three outcomes.
• Access, not evidence, is the limiting factor for third-generation agents in most Indian practice.

Why it matters

The choice between generations is usually made on price, and this puts a survival number on that decision.

Don't overread it

Mostly indirect comparisons across nine trials — the rankings are not a head-to-head result.

The statistics, in plain English

P-scores rank treatments on a continuum from 0 to 1 and are widely misread as a probability that one drug is best — they are not, and a high P-score with overlapping confidence intervals means very little. Furmonertinib's progression-free survival hazard ratio of 0.44 comes from a small number of trials, and network meta-analysis borrows strength across indirect comparisons, which assumes the trial populations were similar enough to compare — an assumption that is questionable across trials run in different countries over a decade.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.