World Lung 2025 – Bristol seeks PRMT5 biomarkers

Bristol Myers Squibb's perseverance with the PRMT5 inhibitor BMS-986504 has some more backing, courtesy of updated results from its phase 1 study presented at a World Conference on Lung Cancer late-breaker on Monday.

The data show the company zeroing in on 35 pretreated NSCLC patients, around a quarter of the population of a trial that also includes other cancers, all of which have MTAP deletions. While the 29% response rate in NSCLC is in line with previous reports, a new analysis has shown divergence in activity depending on the presence of other mutations, including EGFR.

This could be relevant for other players who have struggled with PRMT5 inhibition, an approach being tested extensively across MTAP-deleted tumours, but one thought still to require refinement based on additional biomarkers to dial up efficacy.

That said, it's clearly too early to say whether EGFR mutation status could hold one key to unlocking the activity of PRMT5 inhibitors, since the World Lung analysis relates to very small numbers of patients. Not only that, but it concerns 200-800mg doses of BMS-986504; Bristol has selected 400mg and 600mg for further study.

Mutations

The headline number at World Lung was a 57% response rate in EGFR-positive MTAP-deleted NSCLC. Though this concerned four responses in just seven patients, among 22 subjects without a documented EGFR mutation there were only six further responses.

Researchers also cut the data by ALK alterations, and while here the responses diverged too there were only four ALK-positive subjects, two of whom responded. An analysis according to histology was also carried out, but whether patients' tumours were squamous or non-squamous didn't appear to make a difference to ORR.

The total NSCLC dataset from this study now comprises 35 patients, in whom there are 10 responses plus a further two awaiting confirmation. Serious treatment-related adverse events across the whole trial were seen in 4% of patients, while those graded severe amounted to 13%.

Addressable population

MTAP-deleted tumours, on which PRMT5 inhibitors are designed to act, are thought to account for up to 27% of NSCLC cases, and there's no suggestion that Bristol intends to cut BMS-986504's addressable population further; its phase 2/3 Mountaintap-29 trial in first-line NSCLC, due to start next month, enrols those with homozygous MTAP deletion or MTAP loss.

However, the group's main competitor in this mechanism, Amgen, is still deciding on a phase 3 dose for AMG 193, having reported toxicity and less impressive efficacy in its phase 1/2 trial at last year's ESMO. At the time, the discussant said there remained a pressing need to identify biomarkers of response and resistance, as MTAP status alone was insufficient.

BMS-986504 was originated by Mirati, a company Bristol acquired two years ago for $4.8bn. Mirati's KRAS G12C inhibitor Krazati was the key reason for this takeover, but some speculated that the PRMT5 work was also something Bristol was interested in; however, data on AMG 193, and Tango's struggles with TNG908, took some gloss off this approach.

Nevertheless, industry continues to put PRMT5 inhibitors into clinical development, and just this week Ideaya filed a US IND for its contender, IDE892. Meanwhile, the next key dataset is expected to come in the fourth quarter from the phase 1/2 Primrose trial of AstraZeneca's AZD3470.