FDA Panel to Consider Rivaroxaban's Comparative Efficacy

Interpreting the FDA’s "As Effective" Requirement

The FDA reviewers seek to put the data into the context of FDA policy that requires new therapies be as effective as approved treatments when the disease at issue is life-threatening or capable of causing irreversible morbidity.

"The policy is broadly written and lacks a discussion of operational details," the reviewers said.

"For example, it does not state whether comparable effectiveness to an approved therapy used in an unskilled manner would be adequate for approval. However, if it is essential for a therapy to be as effective as an approved therapy to protect public health, it is logical that the new therapy should be as effective as the approved therapy when the approved therapy drug is used skillfully."

The reviewers said rivaroxaban should not be approved "unless the sponsor submits convincing information that it is as safe and effective for its target indication as warfarin when it is used skillfully (e.g. in the subgroup of patients at centers where TTR [is greater than or equal to about] 67%), or that it is as safe and effective as another approved agent, such as dabigatran."

The reviewers left a door open for rivaroxaban’s approval for those patients whose INR cannot be well controlled on warfarin or who are unwilling to take the vitamin K antagonist.

"However, such patients have an alternative, dabigatran, which is approved for rivaroxaban’s proposed indication," the reviewers pointed out. "Dabigatran was shown to be superior to warfarin in preventing stroke and systemic emboli in the overall results of the large global RE-LY trial (with a median TTR of about 67%), and it was robustly noninferior to warfarin at RE-LY centers with TTR above the median."

Given the availability of dabigatran, "it seems advisable to make rivaroxaban a third-line agent, behind both warfarin and dabigatran," the reviewers said.

In the draft questions to the advisory committee, the FDA notes its "as effective" policy explicitly does not apply if the new therapy is studied in a new population. The agency asks whether this exclusion applies to rivaroxaban based upon differences in the populations of the ROCKET AF and RE-LY trials.

Key among these differences was the enrollment in ROCKET of patients at higher risk for stroke, based upon CHADS scores, than those studied in RE-LY. ROCKET also enrolled more patients who had a prior stroke or systemic embolism.

A Firm "No" On Superiority

If the committee votes in favor of approval, the panel is asked to discuss whether a claim for superiority over warfarin is merited.

FDA reviewers said the sponsors’ request for superiority language should be rejected for several reasons. Among these, only the on-treatment analyses of the safety and per-protocol populations support superiority. "All analyses that include follow-up of patients for at least seven days after the last dose of study, and all ITT analyses do not support superiority. We generally prefer an ITT analyses as the basis of a superiority claim."

More Data Needed on Transition Strategy

In explaining their second reason why the NDA should not be approved, the reviewers pointed to a statistically significant increase in the rate of strokes in the rivaroxaban arm compared with warfarin while subjects were being transitioned from blinded study drug to open-label warfarin at the end of study.

Unlike in the RE-LY and ARISTOTLE trials, no provisions were made for a short period of dual therapy for patients in the rivaroxaban arm to continue anticoagulation during the lag period of INR control at the start of open-label warfarin.

"To ameliorate the risk of events after discontinuation of rivaroxaban, the sponsor has submitted proposed labeling with instructions for the transition from rivaroxaban to warfarin therapy. These instructions call for a period of concomitant treatment with both drugs under INR control (with INR measured at the end of the rivaroxaban dosing interval)," the reviewers said.

However, these instructions were developed after the ROCKET trial was completed and are based on pharmacokinetic/pharmacodynamic modeling; they have not been evaluated in a clinical trial. Given the serious risk of stroke associated with this transition period, "it seems prudent to require the sponsor to demonstrate in a clinical study in A Fib patients receiving rivaroxaban therapy that the proposed transition regimen is safe and effective."

This coverage is provided courtesy of "The Pink Sheet." This news organization and "The Pink Sheet" are owned by Elsevier.