Harmoni spoils Summit's pre-ASCO splash

For the second time in two years the results of a key trial of Summit/Akeso's ivonescimab have been unveiled on the eve of the ASCO conference, but this time the market response has been rather harsher than in 2024. Revelation that an interim read of immature overall survival data from the Harmoni trial failed to hit statistical significance sent Summit shares down 30% on Friday.

That's a severe reaction, but it has some basis given the expectations riding on ivonescimab, and the fact Harmoni is the first global study to attempt to show that Chinese data generated by Akeso aren't a fluke. PFS results appear highly positive, but here too there is reason for concern about the drug's benefit in patients outside China given a large overlap with the population in Akeso's Harmoni-A trial.

Harmoni compared ivonescimab plus chemo against chemo alone in patients with non-squamous EGFR-mutant lung cancer who progressed on a third-generation EGFR inhibitor like Tagrisso, but so far it's impossible to gauge the real-world relevance of the results. Only the hazard ratios have been revealed: 0.52 (p<0.00001) for PFS and 0.79 (p=0.057) for OS.

No US filing, for now

While describing the PFS result as clinically meaningful, Summit disclosed that the FDA has said regulatory submission can't be made without a statistically significant OS result. Accordingly, no filing can take place at this point, and while Summit says it intends to submit one the timing is still under consideration.

Whether Harmoni can ever deliver a significant OS result is now a key uncertainty. Akeso's Harmoni-A trial, in an analogous but Chinese population, doesn't give much confidence: that showed a 28% reduction in risk of death at first interim, but this shrunk to 20% at the second read, suggesting that the OS result there is deteriorating as the data mature.

Trials of ivonescimab + chemo in 2nd-line EGFRm NSCLC

Note: *the FDA allowed Harmoni-A patients who had progressed on a 3rd-generation EGFR inhibitor to be included in the Harmoni analysis; 276 of Harmoni-A’s 322 patients met this criterion. Source: OncologyPipeline & Summit.

A little-appreciated fact is that a large proportion of Harmoni patients were actually from Harmoni-A, and the fact this overlap existed is another reason for investors to be especially cautious.

The FDA allowed Summit's trial to include all Harmoni-A patients who had received a third-generation kinase inhibitor, and this amounted to 276 patients. These patients then made up around two thirds of Harmoni's population.

The glaring fact is that the hazard ratio for PFS in Harmoni-A was 0.46, but the number for Harmoni has come in at 0.52. So the implied hazard ratio number for the non-Harmoni-A patients (meaning most of the ex-China enrollees) is worse, perhaps of the order of 0.60 or higher.

In its statement Summit said the PFS benefit was consistent between Asian and western patients, and that the OS results were consistent between Harmoni-A and Harmoni, demonstrating ivonescimab's potential to benefit patients including those in the US.

In a note to clients Evercore ISI's Cory Kasimov called the OS miss narrow, and said the data were better than the market reaction implied. As for PFS, he wrote that Summit management indicated that the current analysis comprised a new cut of the China data.

However, this marks the second setback for Summit, after April's unscheduled analysis of OS data from the separate Harmoni-2 trial also revealed a non-significant interim result, and Summit stock plunged 36%. Investors now have a nervous wait for the full Harmoni results and for more news on OS.