ESMO 2024 preview – Astellas’s KRAS degrader disappoints

KRAS G12D inhibition has so far failed to live up to the hype, and this seems unlikely to change any time soon. First-in-human data with Astellas’s G12D degrader ASP3082, released this week in an ESMO abstract, look lacklustre, raising questions about whether a degrader approach is any better than inhibition.

An 8% overall response rate among 65 evaluable patients is in line with the 6% ORR seen at last year’s ESMO with Jiangsu Hengrui’s G12D inhibitor HRS-4642. While Astellas is still refining its ideal dose, the group will need to show a lot more to keep up excitement around G12D, which is becoming increasingly crowded. Full data will be presented at ESMO on Sunday.

Dose response?

The phase 1 trial tested ASP3082 at 10-600mg once weekly, in 98 patients with various relapsed KRAS G12D-positive solid tumours, the most common being pancreatic cancer.

The efficacy analysis included 65 patients receiving 10-300mg. Within this group, 35 subjects received a dose of 90mg or under; according to the abstract, the predicted lowest dose for efficacy was over 100mg.

However, excluding these 35 patients still only produced a 17% ORR. And among patients receiving 140-300mg there was no clear dose response, with one partial response at 140mg, none at 200mg, and four at 300mg.

Astellas might not have scope to go too much higher: there were two dose-limiting toxicities at 450mg and one at 600mg – all grade 3 liver enzyme elevations. Despite this, the abstract noted that the maximum tolerated dose hasn’t been reached.

Any-grade treatment-related AST and ALT elevations were seen in 7% and 6% of patients respectively. Other common treatment-related adverse events included fatigue (15%) and infusion-related reactions (14%). There were no treatment-related grade 4 adverse events, and no deaths.

The evidence so far doesn't back the theory that, by destroying a target rather than just blocking it, degraders could be more effective than inhibitors in this setting.

And there are broader doubts about hitting KRAS G12D. At last year’s ESMO Jiangsu Hengrui reported just one partial response among 18 solid tumour patients receiving HRS-4642. When only NSCLC patients were included this rose to one partial response among 10 subjects, but this is still not much to write home about.

Cross-trial comparison of projects targeting KRAS G12D

Source: ESMO.

HRS-4642 is still in development, but combinations now look to be the focus.

Another benchmark, of sorts, comes from Revolution Medicines’ pan-KRAS inhibitor RMC-6236, which is being tested in patients with various G12 mutations, including G12D. That group recently reported an ORR of 20% in pancreatic cancer, although it’s unclear how many G12D subjects were included. These mutations were seen in around half of the patients in an earlier readout.

Revolution also has a specific G12D inhibitor, RMC-9805, with initial phase 1 data due in the fourth quarter. A project from Mirati, MRTX1133, had been expected to yield its first clinical results this year, but things have gone quiet since Bristol Myers Squibb acquired that company. Preclinical data on MRTX1133 will be presented at ESMO.

Others developing KRAS G12D degraders include Arvinas, Risen Pharmaceuticals and Seed Therapeutics, but these are all preclinical. It might be a while longer until it becomes apparent whether KRAS G12D degradation – and indeed inhibition – has legs.

ESMO will take place in Barcelona, Spain, on 13-17 September 2024.