Europe has proved to be a tough customer for Puma Biotechnology and its breast cancer drug Nerlynx, but how consistent will it be in challenging modest survival data? This is the question that emerges given that Roche could be on the verge of receiving the EU's blessing for Perjeta to be used in a similar adjuvant treatment setting.
The two agents look surprisingly alike in terms of their ability to prevent disease recurrence if one is to look past the weaknesses in Nerlynx’s data package, as the US FDA did last year. The EMA's apparent posture is that Puma’s Extenet data are not enough, but only when the regulator rules on Perjeta will we know whether the agency is concerned with Nerlynx’s marginal benefit or with gaps in the data.
A problem of numbers
Puma announced a “negative trend vote” from the EMA’s human medicines committee after market close yesterday, signaling that the agency’s expert advisers would likely vote against Nerlynx's approval in the adjuvant setting at their February meeting. Shares in the California based group fell 28% today.
The question is what data the EMA wants to see in a drug to be used in this setting – in Nerlynx’s case, to be used after a year of Herceptin to prevent disease recurrence. Puma said the CHMP's benefit risk assessment of the drug was negative, because the study results were “based on evidence from a single pivotal trial, and the two and five-year invasive disease free survival (iDFS) benefits observed to date may lack sufficient clinical relevance.”
Nerlynx had three problems: a diarrhoea side effect, a survival benefit that is arguably marginal, and significant omissions in the data package from the sole pivotal trial in this setting, called Extenet. This was thanks to the originator, Wyeth, discontinuing follow-up when it concluded that Nerlynx would fail.
Diarrhoea is managed with loperamide, and the survival benefit – 94.2% of Nerlynx patients were shown to be free of invasive disease at two years, compared with 91.9% of placebo patients – was good enough for the FDA. That was based on follow-up available on just half of the patients enrolled into Extenet.
The FDA’s solution was to conduct a “tipping-point analysis”, which concluded that an implausibly high number of missing patients would need to have experienced a recurrence to alter the statistical significance of the Extenet primary endpoint (Puma pulls within sight of FDA OK, May 24, 2017).
Puma’s account of the CHMP decision, which cited lack of “sufficient clinical relevance”, leaves unclear whether it was the patients lost to follow-up or the modest 2.3 percentage point difference that influenced the EMA’s negative view. This boiled down to 106 placebo patients seeing their disease recur versus 67 Nerlynx subjects, out of a population of 2,840 patients.
If it is the modest benefit, then this might not bode well for Roche’s application for Perjeta’s adjuvant application; this sees Perjeta used in combination with Herceptin and chemotherapy following surgery. This found an even smaller benefit – 94.1% of the Perjeta/Herceptin patients were disease free at the end of three years, against 93.2% of Herceptin-only patients (Asco – Aphinity and Roche’s 1% solution, June 5, 2017).
Again, this did not trouble the US FDA, which extended Perjeta's label to include this setting late last year.
Roche’s EMA application was listed on the agenda for the human medicines advisory committee’s December meeting, but the agency has not announced a decision. The January meeting is under way and wraps up January 25, so more news could be forthcoming then.
Roche would only tell EP Vantage that discussions were ongoing with the EMA, and that it would not speculate on potential approval timelines.
One assumption that has been drawn from the European stance on Nerlynx is that the agency is being tough on therapies that show marginal benefit. That assumption could be shattered should the EMA advisors say yes to Perjeta, in which case one could assume that the issues with Nerlynx run deeper.