Shire’s dry eye product lifitegrast has met both of its phase III primary efficacy endpoints. Unfortunately, it has missed them both too.
Results of the Opus-2 study showed that the LFA-1 inhibitor lifitegrast improved symptoms of eye dryness but failed to show an improvement in inferior corneal staining score. In the earlier phase III Opus-1 study, the findings were the other way round (Event – Shire faces unusual binary outcome, December 2, 2013). Shire must now convince the FDA that the symptom win justifies approval – without an additional confirmatory trial.
The results are surprising on two counts: improvement in symptoms is generally thought of as the more demanding outcome, and Opus-1 met the inferior corneal staining score outcome with strong statistical significance. In Opus-2 this measure missed significance by a long way, with a p value of 0.6186.
What has caused the reversal is unclear, though a failure to exclude patients with dry eye symptoms but no dry eye disease could have had an effect. In any case, Shire’s decision to change the symptom measure from ocular surface disease index in Opus-1 to a patient-reported score appears to have been the right one.
Shire will have some points it can make when it meets with the FDA to persuade it to accept a submission based on pooled data. Most importantly, lifitegrast is the first dry eye drug to meet a symptom endpoint in a phase III clinical trial. The lack of options available and the difficulties of developing a dry eye drug are also in its favour (Therapeutic focus – Plenty of dry eyes in the house, December 5, 2013).
But the inconsistency between the two trials is likely to militate against Shire’s hopes of a quick approval. A third phase III trial, Sonata, is also underway and will read out in mid-2014, but as this is focused on safety it is unlikely to add to Shire’s case.
Allergan’s immunomodulator Restasis, currently the bestseller for this indication, was approved without having demonstrated an improvement in symptoms. But that was in 2003, and ten years later the FDA is a much tougher gatekeeper.
Most analysts, including those form Bernstein, UBS and Leerink Swann believe that the likelihood is that the FDA will order another trial, delaying launch by around two years. At best, a third efficacy study could start in the summer, leading to a possible FDA filing in 2016.
Time is short. Santen today filed an MAA for its candidate, Cyclokat, in the EU under the trade name Ikervis, and the coming year will see pivotal data on Mimetogen’s MIM-D3, which showed improvement in both signs and symptoms in phase II. If Shire cannot inveigle regulators into allowing it to proceed with the data it has it could lose out.