The China-based SANOVO phase III trial found that adding savolitinib to osimertinib cut the risk of progression or death by 44% in previously untreated EGFR-mutated lung cancer with MET overexpression, according to data presented at ESMO 2026. HUTCHMED (LSE:HCM) plans a supplemental marketing application in China.

What was announced

HUTCHMED (LSE:HCM) (China) Limited said on 26 October 2026 that full results from SANOVO were presented that morning at 08:30 CET in a Proffered Paper session at the ESMO Congress 2026 in Madrid, as abstract LBA69. The presentation followed the SAFFRON phase III data, which were shown in Presidential Symposium II on 25 October.

SANOVO is a blinded, randomised, controlled phase III trial in China. It enrolled patients with treatment-naïve locally advanced or metastatic non-small cell lung cancer (NSCLC) carrying activating EGFR mutations (exon 19 deletion or L858R) and MET overexpression. A total of 326 patients were randomised 1:1 to ORPATHYS (savolitinib) plus TAGRISSO (osimertinib) (n=162) or placebo plus TAGRISSO (n=164). TAGRISSO was given at 80 mg once daily and ORPATHYS at 300 mg or 200 mg twice daily depending on body weight. ORPATHYS is jointly developed by AstraZeneca and HUTCHMED and commercialised by AstraZeneca.

The numbers

The primary endpoint was investigator-assessed progression-free survival (PFS). In the intention-to-treat population (MET IHC 2+ and 3+), median PFS was 24.9 months (95% CI 22.1–27.6) for the combination compared with 13.8 months (95% CI 10.9–16.6) for TAGRISSO alone. The hazard ratio was 0.56 (95% CI 0.43–0.74; p<0.0001), which the company describes as a statistically significant 44% reduction in risk. There were 101 events (62.3%) in the combination arm and 124 (75.6%) in the control arm. The data cut-off for PFS was 31 July 2026.

In patients with high MET overexpression (IHC 3+ >90%; n=98 per arm), median PFS was 24.8 months versus 11.1 months, with a hazard ratio of 0.45 (95% CI 0.31–0.66; p<0.0001), described as a statistically significant 55% reduction in risk.

Overall survival (OS), a secondary endpoint, reached a median of 44.5 months (95% CI 41.3–not estimable) for the combination versus 34.9 months (95% CI 26.3–41.1) for the control arm, with a hazard ratio of 0.48 (95% CI 0.33–0.68). There were 49 deaths (30.2%) versus 83 (50.6%), and estimated three-year OS was 67.4% versus 47.4%. HUTCHMED describes the OS benefit as "clinically meaningful"; the release does not report a p-value for OS or claim statistical significance for it. The company said the PFS and OS benefits were consistent across all prespecified subgroups.

Adverse events led to treatment discontinuation in 14.8% of patients in the combination arm. Fatal adverse events, excluding disease progression, occurred in 4.3% versus 4.8%. The company said there were no new safety concerns.

What the company and partners said

Professor Yi-Long Wu of Guangdong Provincial People's Hospital, the lead principal investigator, said the results "mark a pivotal evolution" in first-line treatment of EGFR-mutated lung cancer. Johnny Cheng, Acting Chief Executive Officer and Chief Financial Officer of HUTCHMED, said: "SANOVO reinforces our strategy to improve long-term clinical outcomes for patients." Dr Jing He, Senior Vice President and Head of R&D China at AstraZeneca, said the results "allow us to further advance our strategy to redefine standards of care."

Regulatory position and what to watch next

In China, ORPATHYS plus TAGRISSO is already approved for MET-amplified, EGFR-mutated non-squamous NSCLC after progression on EGFR tyrosine kinase inhibitor therapy, based on the SACHI phase III trial. HUTCHMED plans to submit a supplemental New Drug Application in China for the first-line setting based on SANOVO, but has not given a date. In Switzerland, the combination holds temporary authorisation in the post-TAGRISSO setting, based on the phase II SAVANNAH trial.

Investors will watch for the timing of the Chinese filing and any regulatory acceptance. SANOVO was conducted only in China, and the release does not say whether the data will be used to support filings elsewhere.