Savolitinib or placebo combined with osimertinib in treatment-naïve advanced NSCLC with EGFR mutation and MET overexpression: Results from the phase 3 SANOVO study

SANOVO

Patient / Population Intervention / Exposition Vergleich Endpunkt

In treatment-naive patients with EGFR-mutant, MET-overexpressing advanced NSCLC, savolitinib plus osimertinib improved investigator-assessed progression-free survival compared with placebo plus osimertinib.

N

326Patienten

Design

Phase 3, randomised 1:1, double-blind, placebo-controlled trial in China of savolitinib plus osimertinib vs placebo plus osimertinib in treatment-naive locally advanced or metastatic NSCLC with an EGFR mutation (exon 19 deletion or L858R) and MET overexpression (IHC 2+ or 3+), stratified by EGFR mutation type, brain metastases and MET IHC level (NCT05009836).

Endpunkt

Investigator-assessed progression-free survival, tested hierarchically: first in the MET IHC 3+ population, then in the ITT population if the first test was positive.

ErgebnisInvestigator-assessed PFS in the MET IHC 3+ population was 24.8 vs 11.1 months with savolitinib plus osimertinib vs placebo plus osimertinib (HR 0.45; 95% CI 0.31-0.66; P<0.0001). In the ITT population it was 24.9 vs 13.8 months (HR 0.56; 95% CI 0.43-0.74; P<0.0001). Median OS in the ITT population was 44.5 vs 34.9 months (HR 0.48; 95% CI 0.33-0.68) at a median follow-up of 38.7 months. Median treatment exposure was 22.1 vs 12.9 months. AEs leading to discontinuation occurred in 14.8% vs 4.3%.

Wu Y, Zhou Q, Sun L, et al. Presented at ESMO Congress 2026, abstract LBA69, 26 October. Not yet published.

Diskussion & Kritik

This double-blind, placebo-controlled phase 3 trial met its primary endpoint of PFS in both the MET IHC 3+ and ITT populations, so the add-on design isolates the contribution of savolitinib to osimertinib. PFS was assessed by investigators. The abstract reports no blinded independent review, and the OS benefit is given without a P value or event counts, so it should be read as supportive rather than formally tested. Discontinuation for adverse events was more than three times as frequent with the combination, but the abstract gives no grade 3 or higher event rates, response rates or intracranial outcomes. Enrolment was limited to China and MET overexpression was defined by a specific IHC scoring scheme, which may limit how far the results apply to other populations and testing settings.