Serelaxin data judged incomplete as FDA panel sends it back


Serelaxin’s long, strange trip will go on at least through 2016. A unanimous FDA advisory committee vote coupled with a straightforward staff recommendation against approval virtually assures that the Novartis acute heart failure candidate will need to wait until two big confirmatory trials have completed before there is a chance of regulatory approval.

Positive but imperfect signals of symptomatic relief from the Relax-AHF trial need to be backed up with data that the project can delay or prevent cardiovascular death, the panel decided in its 11-0 negative vote. Given that the data raised more questions than they answered, it is a mystery why the FDA even accepted the Swiss group’s application – or, indeed, awarded serelaxin breakthrough therapy designation.

No means no

Not every set of adcom briefing documents comes with a blunt staff recommendation, but serelaxin’s did. The die appeared to be cast at that point – the main contention being the co-primary endpoints from the trial, which had been designed by Corthera, the private California-based group Novartis acquired in 2009 for $120m for access to serelaxin.

A negative vote by the European Union’s human drugs advisory panel should also have been a pretty clear sign that the experts in the field remained unpersuaded by the findings so far (Europe brings serelaxin crashing back to earth, January 24, 2013).

Relief of breathlessness for serelaxin was the symptom targeted by the original Relax-AHF trial, which Corthera began back in 2007. This symptom was assessed by patients via the visual analogue scale (VAS) to describe relief from breathlessness over five days, and a seven-point scale to capture changes in symptoms.

Relax-AHF was a success on one of these two measures – the area under the curve for mean improvement of breathlessness was significantly better over the five days after treatment. Tantalisingly, a retrospective, but not prespecified, review of data showed that serelaxin reduced cardiovascular death by a statistically significant 37% at 180 days after treatment (AHA – Two hits and several misses with Novartis’s serelaxin, November 7, 2012).

Usually a molecule claiming to give symptomatic relief needs two trials to support it, which raises the question why the regulator agreed to review serelaxin in the first place. Furthermore, Novartis’s request to have serelaxin labelled as improving the “symptoms” of acute heart failure fell short because only one was measured – not others like cough, oedema or choking.

The FDA contested the application on some rather basic grounds, questioning the breathlessness endpoint because of the way the Relax-AHF investigators measured patients who worsened after treatment. Reviewers suggested that assigning those patients a zero on the VAS, even in the instance of mild worsening that was easily managed by diuretics, might have exaggerated the effect of cases that did improve.

As for the mortality data, the FDA said this indicated a good safety record, but was not sufficient to support efficacy.

The back burner

The fact that Novartis has already begun recruiting into two major outcomes trials – a global study measuring cardiovascular death at 180 days, and a European one measuring worsening or death at five days – should be taken as a sign that the Swiss giant knows how thin the data are, particularly for a cardiovascular project.

Three trials are under way or planned, with enrolment targeted at 10,000 patients, and all on a background of diuretics, ACE inhibitors or other treatments that are now standard of care. Had serelaxin squeaked through, these studies would no doubt have been useful in improving the label claims, but as it is Novartis will need them to succeed to ever get serelaxin to market – and to justify the hunderds of millions it has already spent.

The trials should begin reading out next year, although the biggest one – the 6,400-patient global trial that has 180-day mortality as its primary endpoint – will probably not read out until mid-2016. With a win, the breakthrough therapy designation could mean a speedy passage through the regulator, but for now this is a project that will have to sit on Novartis's back burner for a couple more years.

Trial ID
Relax-AHF-2 NCT01870778
Relax-AHF-EU NCT02064868

To contact the writer of this story email Jonathan Gardner in London at or follow @JonEPVantage on Twitter

Share This Article