Amgen’s latest cut of data on its Kras inhibitor should be enough to maintain respectable expectations for the highly valued project, though the results serve as a reminder that AMG 510 will not work in all Kras patients.
At the World Lung Conference this morning the company said that of the 13 subjects treated with a 960mg dose, which is thought to be the most effective, 54% experienced a partial response, with no complete responses - this represents the overall response rate.
That might disappoint some, given how high expectations have been; at Asco all three patients treated with this dose responded, prompting speculation that a much higher number might emerge from this cohort.
Last week, analysts at Evercore ISI speculated that a 70-80% ORR might be achieved, though they admitted that represented a very high bar. The data on AMG 510 are still early, of course, but this number also gives the group’s smaller rival Mirati a clear line to cross.
The update includes 34 patients with non-small cell lung cancer driven by the Kras G12C mutation, and includes another three months of data from the Asco update. Other tumour types are being investigated, and an update later this month at Esmo will provide a further look at the results from this ongoing phase I study (Esmo 2019 preview – Parps on parade, September 5, 2019).
Of those 34 patients, 23 were evaluable for efficacy, Amgen said. Of the 13 treated with the 960mg dose, six (46%) had stable disease, on top of the seven (54%) partial responses.
Across all doses that 100% “disease control” rate drops to 96%, said Dr Ramaswamy Govindan, who presented the results to a press conference this morning. 11 of the 23 evaluable patients (48%) had a partial response at various dose levels, he said, adding that eight of those 11 patients were still in the study. Five subjects are nearly six months along, and one has had a partial response for more than eight months.
“The activity is quite striking and the early indications are that these responses are durable, but only time will tell,” Dr Govindan said.
The real durability of responses to this project is a crucial question that remains to be properly answered, but with three deaths also reported in patients who initially responded it is clear that AMG 510 is far from the final answer. True, this is a very sick patient population with a median of 3.5 prior lines of therapy, and two of those deaths were due to complications of the underlying disease, Amgen pointed out.
But it seems likely that this data set will prompt a resetting of expectations somewhat – on an investor conference call this afternoon the company was asked several questions about the patients that relapsed, and whether any evidence of deepening of responses had been seen in those that remain on treatment.
The answer to both points was essentially that with patient numbers still very small, it was too soon to be able to make definitive statements.
"It remains elusive whether there's one particular epidemiological factor [that drives resistance or response], and it doesn’t appear that prior lines of therapy preclude response," Greg Friberg, Amgen's head of oncology, told analysts.
Mr Friberg said that data from a phase II expansion study, a single-arm trial being conducted with the 960mg dose of AMG 510, should be available next year. The hope is that this will provide sufficient evidence to support an accelerated filing and executives did little to dampen this prospect of this on the call.
However the promise of an early approval is unlikely to completely compensate for concerns about durability of responses, and what ORR might emerge from the larger trial. Focus will also now turn to patients being treated with AMG 510 plus an anti-PD-1 antibody – this data should also be available next year, Mr Friberg said.
These issues aside, many are still likely to view these results as encouraging for such a poorly served disease. Safety also looks sound, which will help with the company's push to move AMG 510 into earlier lines of therapy. Of the 34 patients no dose-limiting toxicities were reported and only nine (27%) reported treatment-related adverse events of grade 1 or 2, which are considered relatively mild. Three patients reported grade 3 events, and none experienced the most severe grade of treatment-related adverse events.
Amgen will need to confirm this safety record in other tumour types at Esmo in a couple of weeks. So far, though, the data on AMG 510's effectiveness are somewhat inconclusive, and much is yet to be gleaned.
This is an update to an article published earlier.