Study Title:

Rivaroxaban for Treatment of Patients with Suspected or Confirmed Heparin-Induced Thrombocytopenia

ClinicalTrials.gov Identifier:

Sponsor and Collaborator:

McMaster University

Participating Centers:

7 medical centers in Canada

Accrual Goal:

200

Study Design:

This is a prospective, multicenter, single-arm cohort study (Linkins LA et al. J Thromb Thrombolysis. 2014. Epub ahead of print). The target population is adult patients with an intermediate or high clinical probability (4T score ≥ 4) of heparin-induced thrombocytopenia (HIT). Patients with mechanical heart valves, severe renal insufficiency (CrCl <30 mL/min), or moderate or severe hepatic disease (Child-Pugh B and C) are excluded. All enrolled subjects are initially given 15 mg of rivaroxaban twice daily while awaiting the results of HIT laboratory testing. If laboratory testing is negative, rivaroxaban is discontinued. In patients who test positive for HIT, rivaroxaban is continued for 30 days. The dose is changed to 20 mg daily after 21 days in patients with thrombosis, and, in patients without thrombosis, the dose is changed at the time of platelet count recovery (defined as platelet count ≥ 150 x 109/L or recovery to platelet count at the time heparin was initiated). The primary outcome is new symptomatic venous or arterial thromboembolism at day 30. Secondary endpoints include major bleeding and time to platelet count recovery. In addition to local testing, serum from all subjects will be analyzed for platelet activating antibodies in a central laboratory (McMaster Platelet Immunology Laboratory) using a serotonin release assay (SRA).

Rationale:

Drugs licensed for the treatment of HIT (e.g., argatroban, danaparoid, lepirudin) are associated with important drawbacks, including parenteral administration, need for intensive laboratory monitoring, limited availability (in the case of danaparoid and lepirudin), high cost, and incomplete efficacy. These agents decrease thromboembolism by 50 to 65 percent but do not reduce HIT-associated amputation or mortality. Transitioning from argatroban to warfarin is challenging because of the effect of argatroban on the INR and added length of hospital stay.

Rivaroxaban, a direct inhibitor of free and clot-bound factor Xa, is a promising agent for the management of HIT that may overcome some of these limitations. It is administered orally as a fixed dose, is less costly than parenteral agents, does not require routine laboratory monitoring, does not cross-react with HIT antibodies in vitro (Walenga JM et al. Br J Haematol. 2008;143:92-99), and has an established efficacy and safety profile in the management of other thromboembolic disorders.

Comment:

There is an urgent need for novel therapeutics in HIT that improve upon the efficacy, safety, convenience, and cost-effectiveness of existing options. Whether one or more of the target-specific oral anticoagulants will be able to fill this niche is unproven. This study will provide valuable preliminary information but, due to two important limitations in its design, will not definitively address the role of rivaroxaban in the management of HIT.

First, the study lacks a control group against which rivaroxaban may be compared. Ideally, patients would have been allocated to receive either rivaroxaban or an active comparator (e.g., argatroban). However, randomized controlled trials have proven difficult to conduct in HIT; only two such trials have been published. A trial of danaparoid versus dextran 70 was completed in the 1990s prior to the advent of approved agents (Chong BH et al. Thromb Haemost. 2001;86:1170-1175). A more recent trial of desirudin versus argatroban was terminated early due to poor accrual after only 16 subjects had enrolled (Boyce SW et al. Am J Ther. 2011;18:14-22). Comparisons between the rivaroxaban cohort and the pivotal single-arm studies that led to approval of argatroban and lepirudin are unlikely to be meaningful because of differences in patient population, outcome assessment, and adjunctive care.

Second, it is likely that the majority of patients enrolled in the study will not have serologically confirmed HIT. In a recent meta-analysis, HIT was corroborated using a specific washed platelet assay (e.g., SRA) in only 23 percent of patients with suspected HIT and a 4T score ≥ 4 (Cuker A et al. Blood. 2012;120:4160-4167). Thus, it can be anticipated that 50 or fewer of the 200 subjects enrolled in the study will have “true” serologic HIT. Whether this number will be sufficient for meaningful assessment of rivaroxaban for the treatment of HIT remains to be determined.

The aforementioned limitations notwithstanding, this study represents an important step forward in the investigation of HIT. It places renewed focus on a potentially devastating orphan disease with limited treatment options and no new FDA-approved therapies since argatroban received licensure 14 years ago. Despite lack of a control arm, the quality of evidence obtained from this study will undoubtedly exceed that derived from retrospective analyses cited to support off-label use of fondaparinux and bivalirudin for HIT. If results are promising, this study should pave the way for an international, randomized, controlled trial to definitely address the role of rivaroxaban for treatment of HIT.

Competing Interests

Dr. Cuker indicated no relevant conflicts of interest.