— PFS rises by 2 months, OS by 5 months with osimertinib-savolitinib after osimertinib progression
October 27, 2026 • 4 min read
Patients with MET-altered advanced non-small cell lung cancer (NSCLC) lived significantly longer with a targeted combination after progression on standard-of-care osimertinib (Tagrisso), a large randomized trial showed.
Median progression-free survival (PFS) improved from 5.4 months with chemotherapy to 7.6 months with the MET inhibitor savolitinib and osimertinib, a difference that represented a 45% reduction in the hazard for progression or death. Patients with MET overexpression or amplification had greater improvement. A preliminary analysis of overall survival (OS) showed almost a 5-month improvement with the combination.
Rates of treatment-related adverse events (TRAEs) were similar between the treatment groups, reported Shun Lu, MD, of Shanghai Jiao Tong University in China, at the European Society for Medical Oncologyopens in a new tab or window (ESMO) congress in Madrid.
"The SAFFRON study established the clinical value of an all-oral, biomarker-directed treatment approach, targeting both EGFR and MET, following progression on osimertinib, reinforcing osimertinib as a backbone treatment across lines of therapy," said Lu. "Osimertinib plus savolitinib established a new global standard of care for patients with EGFR-mutant advanced non-small cell lung cancer with MET overexpression and/or amplification after progression on osimertinib."
ESMO-invited discussant Natasha Leighl, MD, of Princess Margaret Cancer Centre in Toronto, said the trial was well designed and produced valid results that were also clinically meaningful. She agreed that the study established a new standard of care for a subset of patients with progressive NSCLC.
Savolitinib is not currently approved in the U.S. but data on the combination have been submitted to the FDAopens in a new tab or window for a potential approval in NSCLC.
"This is the first study in osimertinib-resistant lung cancer to demonstrate improved progression-free and overall survival with an oral combination to be preferred over IV chemotherapy in a molecularly defined population," said Leighl. "Tomorrow, when we go back to clinic, we must ensure that every patient gets a biopsy, gets their testing, and that they are then matched to the best possible therapy or clinical trials."
SAFFRON has also "kicked up the molecular testing a notch," validating the role of MET testing by immunohistochemistry and fluorescence in-situ hybridization. The study also achieved an ESMO magnitude of clinical benefitopens in a new tab or window score of 3, which should be incorporated into any ESMO guideline update, she added.
Despite advances in first-line therapy for NSCLC, disease progression remains inevitable, Lu noted. MET dysregulation is a key resistance mechanism to EGFR tyrosine kinase inhibitors, including osimertinib. As many as a third of patients exhibit MET overexpression or amplification after osimertinib progression, which is associated with poor outcomes. Those patients have limited treatment options.
In a phase II trialopens in a new tab or window, osimertinib plus savolitinib led to clinically meaningful and durable responses in patients with EGFR-mutated, MET-altered advanced NSCLC that had progressed on osimertinib. Subsequently, a phase III trialopens in a new tab or window conducted in China showed an improvement in PFS with the combination in a similar population.
The multinational phase III SAFFRON trial involved 338 patients with locally advanced EGFR-mutated NSCLC that had progressed on osimertinib and had documented MET overexpression or amplification. Patients were randomized to continue osimertinib with the addition of savolitinib or to pemetrexed-platinum chemotherapy. The primary endpoint was PFS by independent review.
The study population had a median age of 64, women accounted for over 60%, nearly two-thirds were from Asia, and about 70% were nonsmokers. Almost all of the patients had metastatic disease, and most had received one prior line of therapy.
The primary analysis showed a statistically significant reduction in the risk for disease progression or death in favor of the targeted doublet (HR 0.55, 95% CI 0.42-0.72, P<0.001). The advantage persisted across all subgroups. Prespecified analyses showed significant PFS improvement in patients with MET overexpression (8.5 vs 5.6 months, P<0.001) and numerical improvement in those with MET amplification (7.6 vs 4.2 months).
Patients randomized to osimertinib-savolitinib had a median OS of 17.8 months compared with 12.9 months in the control arm (HR 0.68, 95% CI 0.50-0.91, P=0.010).
Confirmed objective responses occurred almost twice as often with osimertinib and savolitinib (51% vs 27%), disease control rates were 85% versus 65%, and the median duration of response was 9.7 versus 5.7 months. The median reduction in target lesion size was 44% with the targeted combination and 25% with chemotherapy.
Grade ≥3 TRAEs occurred less often in the osimertinib-savolitinib arm versus the chemotherapy arm (40% vs 52%), but more patients on the targeted combination discontinued because of toxicity (12% vs 6%). The most common any-grade TRAEs with osimertinib-savolitinib were peripheral edema (42%), nausea (21%), increased creatine kinase (20%), and paronychia (20%).