While anti-PD-1 antibodies had piqued interest at last year’s Asco meeting, this year they look set to steal the show. And among the three projects fighting for attention when the Asco abstracts were unveiled yesterday Bristol-Myers Squibb’s nivolumab might be ahead by a nose.
This is thanks to results from a hotly awaited study combining it with another of Bristol’s immune therapies, Yervoy, and a compelling mechanistic thesis backing just such a combo. In a day of frenzied trading yesterday Bristol’s stock reached an 11-year high to close at $44.34. Still, the firm’s competitor, Merck & Co, is not far behind, and it has plenty to play for, as has Roche.
Supporting phase III
While these agents are in development for a number of cancer types, the key data revealed concerned melanoma. So far the numbers to beat in this setting were an objective response rate (ORR) of 50% and complete response (CR) of 7% for monotherapy with Merck’s anti-PD-1, lambrolizumab.
Nivolumab monotherapy itself had generated ORR as high as 41%, while Yervoy monotherapy has shown modest 11% ORR and 1.5% CR in melanoma, for which it was launched in 2011. According to yesterday’s presentation a phase I study has shown the two agents given concurrently to 52 evaluable patients to yield a 40% ORR, 11% CR and 31% rate of ≥80% tumour shrinkage.
But what excited was one dose, which in 17 patients showed 53% ORR and an impressive 17% CR. True, these are small patient numbers, cross-study comparisons are notoriously unreliable, and some optimists had expected an ORR as high as 60%; but the key 50% ORR threshold has been crossed – for one dose, anyway – and 17% tumour clearance is particularly impressive.
ISI Group’s Mark Schoenebaum said that while the data might have been at the low end of some expectations, the “near” CRs – meaning the ≥80% tumour reduction numbers – were impressive. Bernstein analysts said the data supported a move into phase III.
Yervoy sold $706m last year, and EvaluatePharma cites consensus sales forecasts of $1.5bn in 2018 for nivolumab monotherapy, which is in phase III for NSCLC and renal cell carcinoma, as well as melanoma.
PD-1 is a so-called programmed-death protein that is expressed on the surface of T-cells, and can bind to PD-L1, a ligand found on tumours (Therapeutic focus – PD-1s show worth at Asco, June 7, 2012).
This interaction is thought to allow tumour cells to evade a normal T-cell killing mechanism, and blocking PD-1 is thus thought to reverse this. Adding Yervoy into the mix might potentiate the effect, given that Yervoy acts to block CTLA-4, a receptor that also acts to downregulate immune response.
Of course, the PD-1/PD-L1 interaction can also be inhibited by blocking PD-L1, and this is how Roche’s MPDL3280A acts. A trial unveiled yesterday, in 141 patients, gave the first phase I data ever for the Roche antibody, and showed it to yield a 21% ORR across several tumour types.
More impressively, this rose to 36% in PD-L1-positive tumours – logical given the purported mechanism of action. In PD-L1-negative tumours the response rate was 13%; the fact there was some response here might be accounted by laboratory error or an incomplete characterisation of PD-L1, although the data do back the view that this ligand is a predictor of response.
Roche pointedly boasted a lack of grade 3-5 pneumonitis – a known safety issue of PD-1 therapy. With this key differentiator of safety, and the known side effects of Yervoy, not to mention its high price, there is still plenty at stake.
Interestingly, Bristol was developing an anti-PD-L1 of its own, BMS-936559, but has dropped it in cancer in favour of infectious diseases. No consensus forecasts are available for MPDL3280A, while Merck’s lambrolizumab carries modest $106m in consensus 2018 sales.
The bulls will seize on the fact that the Yervoy combo has brought nivolumab’s efficacy up to and perhaps over that of lambrolizumab. However, in terms of the two agents’ timeline there is little to tell them apart, not to mention that bigger studies are needed to show the robustness of the effect.
And if analysts are to be believed the numbers attached to nivolumab, a project originated by Medarex, make Bristol’s $2.4bn acquisition of that biotech company seem more than justified.
|Melanoma studies with selected anti-PD-1/PD-L1 antibodies|
|nivolumab||Bristol-Myers Squibb||Phase III||Data in 2015; also in phase III for NSCLC and RCC||NCT01721746|
|lambrolizumab||Merck & Co||Large phase II||Data in 2014/15; possibly pivotal||NCT01704287|
|nivolumab + Yervoy||Bristol-Myers Squibb||Phase I||Clinicaltrials.gov lists phase III study about to start||NCT01024231|
|MPDL3280A||Roche||Phase I||Various tumour types; phase II likely in NSCLC||NCT01375842|