In EGFR-mutant lung cancer, TP53 co-mutation identifies who needs chemotherapy too

Design

multicentre, randomised, open-label phase 3 trial at 17 sites in China, 1:1

Population

294 treatment-naive patients with stage IV or recurrent non-squamous NSCLC carrying concurrent TP53 and EGFR-sensitising mutations; median age 57, 54.1% female

Primary outcome

investigator-assessed progression-free survival

Effect

34.0 vs 15.6 months, difference 18.4 months (95% CI 9.9-22.3), hazard ratio 0.44 (0.32-0.60), P<0.001; overall survival 30.6% mature

Adding chemotherapy to osimertinib in EGFR-mutant non-small-cell lung cancer improves progression-free survival for the population as a whole, at the cost of real toxicity — which leaves the clinician deciding who is worth treating that way. This trial tested one candidate answer: concurrent TP53 mutation, a well-recognised marker of worse outcome on tyrosine kinase inhibitor monotherapy.

Across 17 Chinese sites, 294 patients with treatment-naive stage IV or recurrent non-squamous disease and both an EGFR-sensitising and a TP53 mutation were randomised to osimertinib plus pemetrexed and carboplatin for four cycles then osimertinib-pemetrexed maintenance, or osimertinib alone. Median progression-free survival was 34.0 against 15.6 months, a difference of 18.4 months (95% CI 9.9-22.3), hazard ratio 0.44 (0.32-0.60). Benefit held in prespecified subgroups including brain metastases and L858R. Grade 3 or higher treatment-related adverse events were commoner with combination therapy, with no new safety signals.

Doubling median progression-free survival in a biomarker-defined group is a larger effect than the all-comer combination trials produced, which is the point: the co-mutation is doing selection work. The practical requirement is that TP53 status be known at diagnosis, which means broad panel sequencing rather than an EGFR hotspot test — a real constraint where sequencing is paid for out of pocket.

• Request broad panel next-generation sequencing at diagnosis, not an EGFR-only test, if combination therapy is an option locally
• Overall survival data are 30.6% mature — the progression-free survival gain is not yet a survival gain
• Warn patients that combination therapy carries more grade 3 or higher toxicity, and that four cycles of platinum is the price
• Benefit held with brain metastases and with L858R, the subgroups that usually do worse
• Where sequencing is unaffordable, say so plainly rather than treating TP53 status as unknown-and-therefore-absent

Why it matters

It turns a whole-population toxicity trade-off into a selectable one, if the sequencing is done up front.

Don't overread it

Open-label, single-country, and progression-free survival assessed by investigators — the survival question remains open.

The statistics, in plain English

A hazard ratio of 0.44 means the rate of progression was roughly halved over the follow-up period — an unusually large effect for a combination question that has been argued about for years. Median progression-free survival of 34.0 months in the combination arm is based on events accumulated to a November 2025 cutoff; with survival only 30.6% mature, the 'trend toward overall survival benefit' the authors describe is not yet a result.

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.