This post originally appeared at Forbes.com.
Drug developer and one-time Wall Street pet Portola Pharmaceuticals (PTLA) on Thursday released top-line results from a phase 3 study of the oral anticoagulant betrixaban, sending its stock down by 30%.
The APEX study tested betrixaban against Sanofi’s (SNY) approved injectable anticoagulant, Lovenox (enoxaparin), in patients at risk for blood clots after being hospitalized for a number of common medical conditions, including heart failure or infection. Portola used a unique study design in an attempt to enrich the trial and improve its odds of success.
That might ultimately be betrixaban’s undoing.
Betrixaban outperformed enoxaparin in the study’s overall patient population (an impressive 7,500 patients globally), and in Cohort 2, those with elevated D-dimer (a protein fragment present after a blood clot has developed) or ≥75 years of age. But betrixaban missed in a predefined subset of patients dubbed Cohort 1: high-risk patients with elevated D-dimer levels. The study’s primary endpoint was the relative risk of deep venous thrombosis, non-fatal pulmonary embolism, or venous thromboembolism-related death.
(Corrected to reflect that both the overall population and Cohort 2 achieved a p-value of <0.05.)
The study called for an efficacy analysis of Cohort 1 first, and required a p-value of 0.05 or less in order to test Cohort 2. The second analysis also required a p-value of 0.05 or less in order to test the overall study population.
Cohort 1 achieved a p-value of 0.054, just shy of the threshold for further analysis of the program. Per protocol, that should have been the end. But Portola went ahead with Cohort 2 and overall analysis anyway. Cohort 2 and the overall study population achieved p-values of 0.029 and 0.006, respectively.
There was no statistical difference in major bleeding between the betrixaban and enoxaparin arms in any of these three patient groups, though the betrixaban appears to be numerically inferior. The number of fatal bleeds was balanced between the two arms, and the number of intracranial hemorrhages was numerically lower in the betrixaban arm, according to the company.
Portola characterizes the net benefit of betrixaban as positive, and says it will discuss these data with the U.S. FDA before filing for approval. A failure to follow the statistical analysis plan never sits well with the FDA, or investors, and betrixaban isn’t addressing a completely unmet need. Factor Xa inhibitors, like betrixaban, have been around for some time, thus the need to demonstrate betrixaban’s superiority to current standards. Lovenox did more than $4 billion in sales at peak. Portola and Wall Street have sung betrixaban’s differentiated profile since its IPO in May of 2013: a longer half-life for once-daily dosing, low renal clearance, and the potential for reduced drug-drug interactions. Portola plans to release more data from the study at a medical meeting in the next few months.
Portola’s lead asset actually performs in the opposing role as betrixaban. Andexanet Alfa is a recombinant modified Factor Xa molecule designed as a Factor Xa inhibitor antidote to reverse the effects of Factor Xa inhibitors in patients on the drugs who suffer a major bleeding episode, or who require emergency surgery. The company submitted a marketing application to the FDA last year and expects an approval decision this August.
