How China applicants could avoid the sintilimab scenario

Farewell, sintilimab. This seems a realistic assessment of the Lilly/Innovent drug’s chances of gaining US approval in front-line lung cancer after yesterday’s 14-1 drubbing at an unusual advisory committee meeting.

Those hoping to see low-cost competition in the US anti-PD-(L)1 space will be disappointed, but the good news is that sintilimab might not necessarily set a broad precedent. Indeed, much of the panellists’ ire was directed at the way Lilly and Innovent had gone about the filing process, suggesting that other companies might avoid a similar fate as long as they engage the FDA early on and do as the agency says.

This became clear yesterday when Harpreet Singh, an FDA director of oncology, accused Lilly of being “incredibly misleading” in its characterisation of a 2020 meeting with the agency and the resulting guidance. This exchange went to the heart of how Orient-11, a first-line NSCLC trial of sintilimab conducted entirely in China, ended up backing a US filing.

Lilly and Innovent said Orient-11 had been designed to support Chinese approval, and they had no discussion over its design with the FDA until after data were reported. It was only then – and after Richard Pazdur, the director of the FDA's Oncology Center of Excellence, welcomed anti-PD-(L)1 drugs developed by Chinese companies as a means of lowering drug prices – that a US path forward emerged, and so a US filing was made.

Didn’t meet flexibility criteria

It is now abundantly clear that such a strategy will not do. Throughout the proceedings the agency stressed its flexibility regarding data generated outside the US, but said Lilly/Innovent failed to meet any of the criteria to warrant such flexibility, and the lack of early FDA interaction was a major black mark.

The FDA’s criticism centred on four key points: that Orient-11 might not be generalisable to the US population, that there were issues over lack of data integrity, that the trial had the wrong primary endpoint (progression-free survival), and that it used an outdated comparator (chemotherapy).

Lack of generalisability was relevant because Orient-11 was a single-country trial – not necessarily because it was conducted ex-US per se, and the FDA stressed the need for patient diversity. Data integrity came up at site inspections, where the FDA had found adverse event underreporting but “no evidence of fraud”.

Lilly/Innovent argued that PFS was a better primary endpoint than OS because it avoided the confounding effect of subsequent therapies, but this cut little ice. Other anti-PD-(L)1 projects with China trials in major indications like first-line NSCLC that also rely on PFS therefore now look like non-starters for US approval.

This could include Cstone/EQRX’s sugemalimab, Novartis/Beigene’s tislelizumab and Coherus’s toripalimab. However, the last two are pursuing a different approach: Coherus’s US filing is in the niche use of nasopharyngeal carcinoma, and could thus pass FDA muster; tislelizumab has been filed for oesophageal cancer on the basis of a global trial, and Novartis’s near-term NSCLC strategy targets second-line use.

But a huge question for companies seeking approval in a major indication is what to give patients in the comparator cohort when an entrenched incumbent like Keytruda is available, as in the case of front-line NSCLC.

The adcom said that when Orient-11 enrolled its first patient (not under a US IND) in August 2018 Keytruda plus chemo was already the standard of care, based on the Keynote-189 study. Had the FDA been consulted it would likely have recommended a head-to-head comparison showing sintilimab to be non-inferior to an approved anti-PD-(L)1/chemo regimen, the panellists said.

Companies with first-line NSCLC trials that do test OS but which use chemo comparators include Arcus with zimberelimab, and the Fortress subsidiary Checkpoint with cosibelimab. These might today be sitting nervously, but will point in their defence to the Empower-Lung-1 trial that backed US approval of Sanofi/Regeneron’s Libtayo.

However, while Empower-Lung-1 did use a chemo comparator and was conducted ex-US it began in 2017, before Keytruda could be said to have become a standard of care in NSCLC. Lilly/Innovent argued that a head-to-head trial for sintilimab would have required 2,000 patients to be enrolled and taken over seven years to run.

For sintilimab a complete response letter now seems certain, leading some analysts to believe that Lilly will walk away from its deal with Innovent. But if the idea of low-cost anti-PD-(L)1 drug competition has been kicked down the road it has not been killed off entirely.

While pricing and competition were said to be outside the scope of the adcom, the panel meeting clearly had relevance beyond sintilimab. Other companies will have their work cut out persuading the FDA to be flexible in designing development paths but, with the agency apparently wanting to be flexible, early interaction holds the key.

Innovent today said the adcom had helped it gain tremendous experience, and the same surely applies to its peer companies too.