Challenges to the design and execution of controlled clinical studies of anticoagulants in patients with heparin-induced thrombocytopenia: lessons learned

Abstract

Heparin-induced thrombocytopenia (HIT) is an immunemediated adverse drug event in which antibodies to the platelet-factor-4 (PF4)/heparin complex are produced, resulting in platelet activation and amplified thrombin production. Patients with HIT are at extremely high risk of thrombosis, and prompt initiation of an alternative anticoagulant is recommended to prevent thrombosis, limb amputation, or death [1]. The US Food and Drug Administration (FDA) has approved three direct thrombin inhibitors (DTIs; argatroban, lepirudin, and bivalirudin) for the prevention and/or treatment of thrombosis in patients with HIT (bivalirudin is only approved in patients undergoing percutaneous coronary intervention), but no randomized clinical trials have been conducted with these agents in patients with HIT. Their approval for use in the HIT setting was based on single-arm, historically controlled studies classified as lowquality evidence (i.e., evidence Level C) by the current American College of Chest Physicians Evidence-Based Clinical Practice Guidelines [1]. Comparisons of safety and efficacy data across studies are also difficult owing to fundamental differences in inclusion criteria, duration of treatment, and observation period. For example, a clinical diagnosis or a history of HIT was sufficient for inclusion in the argatroban trials (e.g., a fall in platelet count of 50% without serologic evidence), whereas, patients in the lepirudin trials were required to have serologically confirmed HIT prior to enrollment. The clinical community clearly would prefer to evaluate new technologies based on head-to-head, comparative data from appropriately powered, multicenter, double-blind, randomized clinical trials (i.e., evidence Level A). However, in the case of complex and relatively rare disorders, such as HIT, this approach may be unrealistic. We attempted to conduct the first randomized clinical trial of DTIs in patients with HIT. The purpose of this article is to discuss our experience in the design and execution of the PREVENT-HIT trial [2] and highlight issues of importance to the thrombosis community when evaluating the development of new agents for this serious adverse drug reaction. The PREVENT-HIT (NCT00787332) trial was an openlabel, multicenter study designed to evaluate the clinical and economic utility of twice-daily fixed-dose desirudin administered subcutaneously (SC) compared with a continuous infusion of activated partial thromboplastin time (aPTT)-adjusted argatroban to prevent or treat thrombosis in patients with suspected HIT. Although, we planned to enroll 120 patients, just 16 were randomized over 14 months, at a cost of more than $3 million. The study was stopped in December 2009 owing to poor enrollment. The results of the PREVENT-HIT study have been reported elsewhere [2] and are briefly summarized here. Sixteen patients were randomized to treatment with fixeddose desirudin (15 or 30 mg, depending on the absence or presence of thrombosis, respectively) SC every 12 h or with aPTT–adjusted argatroban administered by intravenous (IV) infusion. The primary efficacy measure was the composite of new or worsening thrombosis (objectively documented), amputation, or death. Secondary outcomes included major and minor bleeding while on study drug, time to platelet count recovery, and pharmacoeconomics. No amputations or deaths occurred. One patient randomized S. W. Boyce (&) Advanced Heart Failure Program, Department of Cardiac Surgery, Washington Hospital Center, Medstar Heart Institute, 110 Irving Street N.W., Suite 1E3, Washington, DC 20815-6048, USA e-mail: steven.w.boyce@medstar.net