Amid the excitement over this week’s positive overall survival data from the Pacific study of Astrazeneca’s PD-L1 inhibitor, Imfinzi, was one largely ignored red flag: an arguably negative effect in PD-L1-negative patients.
The finding, from a post-hoc analysis, was far from emphatic. But it could make doctors think twice about prescribing Imfinzi, at least without first testing patients’ PD-L1 status. Astra already has a small patient population to shoot for in stage III non-small cell lung cancer, so anything that could hinder the uptake of Imfinzi would be bad news.
The full Pacific data, presented at the World Congress on Lung Cancer, found an impressive overall survival benefit in the total trial population, sending Astra’s stock up 2% on Tuesday (World Lung 2018 – Merck looks to spoil Astrazeneca’s Pacific party, September 25, 2018).
But the benefit was strongly driven by patients with high PD-L1 expression. Zooming in on patients with PD-L1 levels of less than 1% revealed a hazard ratio of 1.36, suggesting that these patients, when receiving Imfinzi, had an increased risk of death than those in the control cohort.
Astra pointed out that the data came from a post-hoc analysis requested by the EMA, telling Vantage: “We believe the finding … was as a result of a small patient sample, few events and abnormally favourable placebo performance, and therefore drawing a conclusion from this analysis is inappropriate.”
Still, the data are already sowing doubts among doctors about Imfinzi’s use in PD-L1-negative subjects. As well as making the pool of treatable patients smaller, PD-L1 testing itself could become a barrier to the drug’s uptake.
And this might not just be a question of time or convenience; other trials presented at World Lung this year, such as the Impower-133 study of Roche’s Tecentriq in small-cell lung cancer, highlighted the difficulty of getting tissue samples of sufficient quality to enable PD-L1 status to be assayed.
More data needed
Some doctors are already planning to implement PD-L1 testing of their patients based on the Pacific findings. Dr Alastair Greystoke of the Freeman Hospital in Newcastle, UK, told Vantage: “The patients I already have on [Imfinzi], I’ll be PD-L1 testing them and having a discussion about the pros and cons.”
He was not involved in the Pacific study but has prescribed Imfinzi through a named-patient programme. His stance is not controversial, at least in Europe, where Astra’s drug is only indicated in patients expressing PD-L1 at 1% or higher.
But in the US, where Imfinzi’s label is PD-L1 agnostic, doctors might become similarly cautious. When asked if US physicians would want to test patients’ PD-L1 status, Dr Greystoke replied: “If I were practising in America I would. If you’re going to consent the patient and talk to them about what they can achieve I think that’s difficult to do without PD-L1 [testing].”
Dr Greystoke would like to see more data on the PD-L1-negative subgroup, a call echoed by Dr Tim Clay, an oncologist at the St John of God Subiaco Hospital in Perth, Australia, on Twitter.
And US physicians also weighed in on the social media site, with Gregory Riely of the Memorial Sloan Kettering Cancer Center asking for “biomarker data to help us avoid lots of extra drug/wasted time for our patients”. He admitted that the Pacific findings in PD-L1-negatives were “concerning”, with the caveat that this came from an unplanned analysis.
While the worst-case scenario could see Astra losing a portion of the currently small stage III NSCLC patient pool, this population could get bigger if a push towards screening leads to more patients being diagnosed at earlier stages of disease.
Another study presented at World Lung, Nelson, found that CT screening of people at high risk of developing lung cancer reduced the risk of death by 26% in men and by up to 61% in women.
At the meeting’s closing press conference yesterday Professor Paul Bunn, of the University of Colorado School of Medicine, and Dr Anna Farago, of Massachusetts General Hospital, both highlighted the Nelson trial as one of a handful of truly practice-changing studies presented at the meeting.
Dr Greystoke agreed, telling Vantage: “A number of people I spoke to thought that was the highlight of the conference.”
He said he would be “very disappointed” if there was not a shift towards screening and earlier diagnosis in the next five years – a shift that could benefit Astra. Imfinzi is currently the only PD-(L)1 targeting agent approved in the stage III disease, with other immuno-oncology players focused on later-stage, metastatic NSCLC, though Merck is studying Keytruda in the earlier setting.
Still, cost considerations look likely to hinder screening efforts. And getting public funding could be doubly difficult as programmes that focused on at-risk populations like smokers would be politically sensitive, Dr Greystoke pointed out.
Even if lung cancer screening becomes the norm, this will not happen overnight. Astra’s more pressing problem is whether the Pacific data will lose it a chunk of its current stage III patient base.