This week’s surprisingly positive readout of Bristol-Myers Squibb’s Checkmate-9LA study put the spotlight on Astrazeneca’s similar Poseidon trial. Today the UK group confirmed that, despite an apparent timeline slippage, should Poseidon read out positively it would be used for a US approval filing before the end of the year.
Remarkably, the basis for such a filing will be progression-free survival, a possibly tall order in a first-line lung cancer setting already well served by an approved drug, Merck & Co’s Keytruda. Astra was also remarkably bullish about its lupus project anifrolumab, which it wants to file even though only one of its two pivotal trials worked.
Either way, the lung cancer submission is expected to come first. Doubts around Astra’s first-line NSCLC strategy had arisen because Poseidon’s clinicaltrials.gov entry was changed in August to reflect a new planned primary completion date: April 2021 instead of September 2019.
But on Astra’s third-quarter call today José Baselga, vice-president of oncology R&D, stated: “The results of Poseidon are expected in the fourth quarter of this current year. What we will have ... is the final PFS analysis; the expected analysis for final OS is going to come later, in 2021.”
Poseidon tests four cycles of chemo on top of Imfinzi with or without tremelimumab, versus chemo alone, and has OS and PFS as co-primary endpoints. The clear implication from Astra’s timeline, therefore, is that the company thinks PFS alone will be sufficient for a regulatory filing in first-line NSCLC.
Even if, as expected, such a filing is made in all-comers, such patients already have the option of Keytruda’s chemo combo, backed by OS data from the Keynote-189 trial. And Bristol might soon file its Imfinzi/Yervoy/chemo triplet in this setting, backed by an apparently positive OS hit in Checkmate-9LA (Bristol might play a role in front-line lung cancer after all, October 23, 2019).
As such it could be asking a lot to expect the US FDA to wave Astra’s combo through on the strength of PFS alone, the risk being that the agency will first want to see full OS data from Poseidon. Mr Baselga refused to say whether Bristol’s Checkmate-9LA success had influenced Astra’s view on Poseidon’s analysis or chances of success.
Meanwhile, in lupus, where Astra recently scored a win with anifrolumab, the group could be on safer ground, given that apart from Glaxosmithkline’s Benlysta nothing new has come to market recently.
Still, anifrolumab’s pivotal success came in one study, Tulip-2, which had been overhauled based on the failure of a first phase III trial, Tulip-1. And the success was due to the use of a lupus measure called BICLA, an endpoint of which the FDA’s view is not entirely clear (Second time lucky for Astra in lupus, August 29, 2019).
But today Mene Pangalos, Astra’s vice-president of biopharmaceutical R&D, was unambiguous, stating: “BICLA is a regulatory-approved endpoint.” He also said that if Tulip-1 was analysed post-hoc according to BICLA it too would yield a positive result, so the issue is more subtle than simply asking for approval on the basis of a single phase III trial.
“We have two phase III studies – one positive with BICLA, one positive in a post-hoc analysis – and we also have a very large phase II study that also was positive,” said Mr Pangalos. A US filing is planned for the second half of next year.
Given the unmet need this could give anifrolumab a good shot at approval – an argument that is more difficult to make for Imfinzi/tremelimumab in NSCLC.