AACR 2024 – Akeso impresses in stomach cancer

China’s bispecific antibody specialist Akeso made a name for itself after licensing ivonescimab to Summit for $500m up front in 2022. Now this cash is helping fund development in China of Akeso’s cadonilimab, which shortly before that deal became the world’s first approved anti-PD-1 x CTLA-4 bispecific.

In January China’s NMPA accepted cadonilimab’s second filing, as part of a chemo combo in front-line stomach cancer, and yesterday at AACR full data were presented from Compassion-15, the study backing this use. With the big caveat that the trial took place in China, its most intriguing feature is that overall survival results have outperformed those seen with Keytruda and Opdivo on a cross-study basis.

Chemo combos of Keytruda and Opdivo are both FDA-approved for the specific setting in question, first-line, HER2-negative gastric/gastroesophageal junction adenocarcinoma, based on the Keynote-859 and Checkmate-649 trials respectively. 

At first sight this cross-trial win might not be surprising given that cadonilimab hits CTLA-4 in addition to PD-1. However, what is surprising is that cadonilimab doesn’t appear to add significant toxicity. In Compassion-15 the severe treatment-related adverse event rate was 66%, versus 60% for Checkmate-649 and 59% for Keynote-859; events leading to discontinuation were lower in the Akeso study versus the Bristol and Merck trials, AACR heard.

And a second relevant point is that cadonilimab plus chemo in Compassion-15 was able to beat the result of Checkmate-649’s Opdivo plus Yervoy cohort, which (perhaps as a result of Yervoy’s known toxicities) drew a blank versus chemo on overall survival.

Low PD-L1

It’s also notable that Keytruda is approved in all-comers even though Keynote-859 showed no benefit in PD-L1-negative disease. Not so cadonilimab, argued Peking University Cancer Hospital’s Dr Jiafu Ji, who told AACR that Compassion-15 positioned the Akeso drug as a “new potential first-line standard ... especially in patients with low PD-L1 expression”.

Whether the data bear this out is up for debate, however, as survival curves weren’t presented for <1% PD-L1 expressers; OS curves for <5% expressers showed separation at nine months, leading to a 30% reduction in risk of death across the trial, and a nominal p value of 0.011.

Ji did reveal the hazard ratio for the <1% (PD-L1-negative) subgroup, which in Compassion-15 came in at 0.77. This, he said, compared favourably against 0.95 in Checkmate-649 and 0.92 in Keynote-859, though no further statistical analysis was provided.

As for the all-comers OS benefit, a median of 15 months for cadonilimab plus chemo compared favourably versus Keytruda plus chemo’s 12.9 months and Opdivo plus chemo’s 13.8 months. Compassion-15’s chemo comparator scored 10.8 months of median OS, slightly underperforming the 11-12 months seen in the Merck and Bristol studies.

Cross-trial OS comparison

Source: AACR, ASCO, ESMO & prescribing information.

All that said, debate will continue about the relevance to the US of data generated in China, where treatment choices and stomach cancer demographics differ versus the west. There is, moreover, no sign that Akeso is planning US development, for which it would likely need a western partner.

The company did attempt one earlier US approval, filing its anti-PD-1 MAb penpulimab with the FDA in May 2021 for third-line nasopharyngeal carcinoma. But the PDUFA date was missed because of Covid-related travel delays, and Akeso tells ApexOnco that penpulimab has now taken a back seat to its primary focus on bispecifics.

Clearly cadonilimab is much more important, and front-line gastric cancer approval is its next goal. Akeso yesterday said a first-line cervical cancer filing in China would be made by the year-end, and four other key pivotal studies are ongoing.

Selected cadonilimab trials

Note: *ph2 trial, others are ph3. Source: OncologyPipeline.