Not long ago Merck & Co threw down the gauntlet to its oncology rivals, saying the checkpoint inhibitor space was all about winning in first-line lung cancer. It will soon find out whether its most audacious bid for domination has paid off.
This involves the US accelerated approval filing of a Keytruda plus chemo first-line combo based on Keynote-021 data, an FDA verdict on which is due by May 10. The filing had caught many unawares, but while Merck looks likely to prevail eventually there are reasons to suppose that this challenge might not immediately be successful.
Perhaps the most obvious caveat is the imminent readout of a competitor, Roche’s Tecentriq, in the Impower-150 trial. Impower-150 is due to yield survival data in the second half of 2017, and it could make sense for the FDA to wait for this and other readouts before deciding on the robustness of Keynote-021.
While outright rejection of Merck’s filing would thus be the worst case at this stage, a more likely outcome is a delay of several months to the PDUFA date. Immuno-oncology approvals have tended to come well before scheduled action dates, so the fact that this has not yet happened could also suggest that approval – a knockout scenario – is not imminent.
This is not to say that a delay or even outright rejection would be disastrous. Merck has a larger trial of a Keytruda plus chemo combo, Keynote-189, due to yield progression-free survival data around September; it is possible that the FDA might want to wait for these results before giving a verdict.
The first-line lung cancer space was thrown wide open last last summer when Bristol-Myers Squibb’s Opdivo failed to beat chemo in the hotly awaited Checkmate-026 study.
The shock failure saw many competitors scrambling to amend and enlarge trials to increase chances of success (Roche powers up for first-line lung duel, February 2, 2017). While both Keytruda and Opdivo are approved in second-line lung cancer, the Merck drug later became the first anti-PD-1 MAb to gain a first-line label – albeit only in patients with 50% or greater PD-L1 expression.
So the race is now on to expand into lower PD-L1 expressers, in which combinations look promising since the combo agent can turn the tumour immunogenic and make it amenable for checkpoint blockade. Roche has made chemo the cornerstone of its combo effort, and Impower-150, which adds Avastin into the mix, will be the first of four such studies to read out.
|Imminent shots at first-line NSCLC|
|Merck & Co|
|Keynote-021||Keytruda + chemo||Improved ORR & PFS, but not OS||PDUFA date on accelerated approval 10 May 2017||NCT02039674|
|Keynote-189||Keytruda + chemo||Possible confirmatory for Keynote-021||PFS data in H2 2017||NCT02578680|
|Mystic||Durvalumab +/- tremelimumab||Many scenarios, varying PD-L1 cutoffs||PFS data mid-2017*||NCT02453282|
|Impower-150||Tecentriq + Avastin + chemo||First of four Tecentriq/chemo trials||PFS data in Q3/Q4 2017||NCT02366143|
|Checkmate-227||Opdivo +/- Yervoy||Bristol declined to file on Checkmate-012/568 data||PFS data in PD-L1 expressers in early 2018||NCT02477826|
|Note: *some analysts have suggested that if PFS is negative Astra might make no announcement until OS from Mystic reads out in 2018.|
A separate challenge was thought to be imminent from Bristol itself, based on pooled data from the uncontrolled Checkmate-012 and 568 trials of various Opdivo combos, including with Yervoy and chemo, but this early-filing effort was abandoned in January.
At the same time Bristol made the astonishing admission that it assumed Merck would get US approval on the basis of Keynote-021 (JP Morgan – Merck muscles into Bristol’s combo space, January 11, 2017). Leerink is one bank whose analysts are also betting that this outcome will accelerate Keytruda sales in the second half.
Bristol yesterday reported first quarter Opdivo sales of $1.1bn, slightly beating the lowered post-Checkmate-026 consensus forecast. Merck reports first quarter earnings next week.
A third threat to Keytruda could come from Astrazeneca’s combination of durvalumab and tremelimumab. The twice-amended Mystic trial has given Astra several shots on goal – assuming sufficient powering remains – and this complex study could yield progression-free survival data in mid-2017.
However, it is not clear what PD-L1 cutoffs the UK company is using; if it were to show a benefit in >25% PD-L1 expressers that would still leave others a chunk of the market to go after. But success is not a foregone conclusion – Evercore ISI’s Umer Raffat recently said he doubted that Mystic would show a PFS benefit, but admitted that overall survival benefit in PD-L1-high patients was possible.
This points to the difficulties of interpreting survival data in immuno-oncology: PFS can be ruined by pseudoprogression, where the tumour grows before shrinking, while an overall survival benefit can be masked by failing placebo patients switching to active treatment.
Indeed, in Keynote-021 Keytruda failed to extend overall survival, though overall remission and progression-free survival were improved. This is another reason why the FDA might view it sceptically, and why regulatory action on the filing is the most important immuno-oncology event of the year so far.