Osimertinib (osi) + savolitinib (savo) vs platinum–pemetrexed (plat–pem) in EGFRm MET-overexpressed (OE) and/or -amplified (AMP) advanced NSCLC post-osi: SAFFRON phase (Ph) 3 primary results

doi:10.1056/nejmoa2611335

SAFFRON

Patient / Population Intervention / Exposure Comparison Outcome

In patients with EGFR-mutated, MET-overexpressed and/or -amplified advanced NSCLC after osimertinib, osimertinib plus savolitinib improved progression-free and overall survival compared with platinum-pemetrexed.

N

338patients

Design

Global phase 3 trial randomised 1:1, with blinding not stated, comparing osimertinib plus savolitinib vs platinum-pemetrexed in adults with EGFR-mutated, MET-overexpressed and/or -amplified advanced NSCLC after first- or second-line osimertinib (NCT05261399).

Endpoint

Progression-free survival by blinded independent central review (RECIST 1.1).

ResultMedian PFS by BICR 7.6 mo (95% CI 5.6-9.7) with osimertinib plus savolitinib vs 5.4 mo (95% CI 4.2-5.8) with platinum-pemetrexed (HR 0.55; 95% CI 0.42-0.72; P<0.001). Median OS 17.8 mo (95% CI 15.6-23.7) vs 12.9 mo (95% CI 10.6-16.8) (HR 0.68; 95% CI 0.50-0.91; P=0.010). Median investigator-assessed PFS 7.4 vs 4.2 mo (HR 0.47; 95% CI 0.36-0.61). Median duration of response by BICR 9.7 vs 5.7 mo. Grade ≥3 treatment-emergent AEs 64% vs 63%, serious treatment-emergent AEs 41% vs 28%, and discontinuation due to treatment-emergent AEs 14% vs 8%.

Lu S, Girard N, Yang JC, et al. N Engl J Med. 2026; published online 25 October. 10.1056/nejmoa2611335

Discussion & critique

This AstraZeneca-funded phase 3 trial shows that the oral combination prolonged PFS by blinded central review and also improved OS against an active chemotherapy comparator. The abstract does not say whether the trial was blinded, and it gives no median follow-up, OS maturity or information on crossover, all of which affect how the OS result should be read. Serious adverse events and discontinuations were more frequent with the combination, although grade ≥3 treatment-related AEs were less frequent. Response rates, intracranial outcomes, subgroup figures for MET-overexpressed and MET-amplified disease, and quality of life are not reported, so the authors' claim of a new standard of care rests on limited published detail.