Roche was probably hoping to make a bigger splash at this year’s Asco, being the first to release fresh data in the fiercely contested front-line lung cancer space. Results from the Impower-131 study of Tecentriq in squamous disease were numerically underwhelming, however, despite the trial meeting a co-primary endpoint in progression-free survival; an overall survival benefit has yet to be seen.
Given that the squamous histology is tough to treat this probably would not matter, but tomorrow Merck & Co will release data on Keytruda in the same setting – and it has already said that the Keynote-407 study hit both PFS and OS endpoints. Roche executives defended the data to EP Vantage, but it is hard to imagine a scenario where Tecentriq comes out on top in this subset of the disease.
Dan Chen, head of cancer immunotherapy development for Roche, admitted that people would compare the two studies, but stressed the caveats of cross-trial comparisons and said that “statistical fluctuations” could always occur when conducting large clinical programmes, as Roche has done with Tecentriq.
He pointed to the Impower-130 study, which last week Roche said had met PFS and OS readouts; this study was conducted in non-squamous patients, but the chemo combination was the same: carboplatin and paclitaxel or Abraxane.
“So I think the key for everyone looking at [the Impower-131] data is you can look at individual trials, but the real data is in all of the information together…from multiple trials. It’s the totality of the data as well,” he told EP Vantage.
This might be true, but then Merck could also use the same argument – and Keytruda has arguably generated a much more consistently strong dataset in first-line lung cancer. The details of Keynote-407 remain crucial to properly judging the relative situation in squamous disease – it too used PFS and OS as co-primary endpoints.
|Squamous NSCLC – the results so far|
|Impower-131||Keynote-407 (interim only)*|
|Tecentriq+chemo (n=343)||Chemo (n=340)||Keytruda+chemo (n=101)||Chemo (n=103)|
|Median PFS||6.3mth||5.6mth||Not known||Not known|
|HR and p value for PFS||0.715, 0.0001||Not known|
|Median OS||14.0mth||13.9mth||Not known||Not known|
|HR and p value for OS||Not significant||Not known|
|Source: Asco press release and abstracts. *Data from all patients to be presented on June 3.|
Of course the results need to be viewed within the context of squamous disease, which represents around 25% of non-small cell lung cancer and kills around 85% of patients within a year of diagnosis. The chemo control arm above represents physicians’ best treatment option at the moment – and the question for them and their patients is whether Tecentriq’s modest benefit is worth added toxicities.
A survival benefit would help make that decision, though given the response rates seen a miss on this measure cannot be ruled out.
Mr Chen said that “a tail is beginning to form” in the data, but that this would need to be followed further.
It is also notable that data from a third arm in this study, which combined Tecentriq with paclitaxel instead of Abraxane in the chemo backbone, has yet to be reported. Given that Abraxane is viewed as a marginally more effective version of paclitaxel, there is a chance that this arm could perform more poorly.
Merck’s trial also stratified patients by Abraxane or paclitaxel, but it will report data from all control patients combined. Keynote-407 also recruited fewer patients – 560 in total.
The results from the interim analysis above point to similar response rates in the active arms of both trials. So perhaps the type of chemo – paclitaxel or Abraxane – is influencing the control arm. Notably, the abstract for Keynote-407 states that only 32% of patients in the interim analysis got Abraxane.
More data are required to help answer these questions. But whether through trial design or by owning the better drug, Merck looks to be in a strong position in yet another lung cancer niche.