A changing of the guard in stomach cancer?

With an ASCO Plenary session yesterday zeroing in on gastric cancer trials of TIGIT and Claudin18.2 blockers, the Chinese group Akeso gatecrashed the party by announcing an overall survival hit in the pivotal chemo combo study of its anti-PD-1 x CTLA-4 bispecific MAb cadonilimab in first-line disease.

The big caveat is that this trial took place in China, where gastric cancer demographics differ considerably versus the west, and used chemo as the comparator. In the US Opdivo plus chemo is already a front-line standard, but cadonilimab’s mechanism is intriguing because adding CTLA-4 blockade, via Yervoy, to Opdivo failed to beat chemo in the Checkmate-649 study.

Checkmate-649 ultimately led to Opdivo plus chemo being approved as, in contrast to the Yervoy combo, this chemo doublet did beat chemo alone. Cadonilimab is notable for having become the first PD-1 x CTLA-4 bispecific to be approved, though this was in China and for second-line cervical cancer.

Akeso’s latest data, from the AK104-302 trial, might be of interest to a western licensee, given that the Chinese company has no US partner for cadonilimab, and apart from one MD Anderson-sponsored phase 2 trial no current cadonilimab studies involve a US hospital. A year ago Akeso licensed ivonescimab, an anti-PD-1 x VEGF bispecific, to Summit for $500m up front.

Clinically meaningful survival

For now all that is known about the AK104-302 data is that cadonilimab plus chemo beat chemo alone at interim analysis, with a benefit said to be statistically significant as well as clinically meaningful.

This effect was seen in all-comers, and PD-L1 ≥5% and <5% expressers, though in PD-L1-negatives the benefit was only numerically superior. Cadonilimab is to be filed with China’s NMPA on the basis of the data.

Once full results are presented at a medical meeting, it might not be the survival benefit cadonilimab adds to chemo alone that will be of interest as much as how the absolute OS number compares against the benchmark set by Opdivo plus chemo in Checkmate-649.

The same was true for Merck & Co’s Keytruda, which now awaits a 16 December FDA verdict in this cancer based on Keynote-859, and is also the case for Gilead/Arcus’s domvanalimab/zimberelimab/chemo triplet in the Edge-Gastric trial, which at the ASCO Plenary posted arguably positive results in terms of overall response rates, at least in PD-L1 ≥5% expressing patients.

Ultimately, the ASCO discussant, Dr Elizabeth Smyth of Oxford University, dismissed the Edge-Gastric data as involving numbers of patients that were too small for an accurate evaluation. Akeso’s trial did at least enrol over 500 patients so the same criticism will not apply here, though the relevance of a China-only population to a US setting will no doubt be debated.

Some OS benchmarks for cadonilimab’s AK104-302 trial

Source: ASCO, ESMO & prescribing information.