| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
Hamilton, ON, Canada
Dr. Warkentin is Professor, Department of Pathology and Molecular Medicine, and Department of Medicine, McMaster University; is Associate Head of Transfusion Medicine, Hamilton Regional Laboratory Medicine Program; and is a Hematologist at Hamilton Health Sciences (Hamilton General Site).
Correspondence to: Theodore W. Warkentin, MD, Room 1-180A, Hamilton Regional Laboratory Medicine Program, Hamilton Health Sciences (Hamilton General Site), 237 Barton St E, Hamilton, ON, L8L 2X2, Canada; e-mail: twarken{at}mcmaster.ca
Heparin-induced thrombocytopenia (HIT) is a prothrombotic disorder for which relatively few alternative anticoagulants have undergone prospective evaluation as therapeutic agents. Accordingly, one seeks to learn as much as possible from those studies that have been performed. In this issue of CHEST (see page 1407), Lewis and colleagues1 report a secondary analysis of the prospective cohort studies23 of argatroban that led to regulatory approval of this direct thrombin inhibitor (DTI) for the treatment of HIT with or without thrombosis. Should we pay attention to secondary analyses of prospective studies? Yes.
HIT has a track record for revealing important information from further analyses of prospective trials. For example, for the DTI lepirudin a secondary analysis4 of three prospective studies found a major bleeding rate of 17.4%, which varied from 10.8 to 33.0%, depending on whether the patient’s serum creatinine level was < 90 µmol/L or > 90 µmol/L, respectively. Based on these (and other) data, the authors suggested that the standard lepirudin dosing regimen (a bolus of 0.4 mg/kg, then 0.15 mg/kg/h) should be revised downward (to 0.10 mg/kg/h without a bolus infusion). Other investigators5 have independently concluded that the approved lepirudin dosing regimen is too high.
Another example: a randomized controlled trial compared unfractionated heparin with low-molecular-weight heparin for thromboprophylaxis after hip replacement surgery.6 Data on the daily platelet counts were collected prospectively, but HIT ascertainment was initiated only after completion of the clinical trial. Nevertheless, the conclusions of this secondary analysis (that the risk of HIT is greatly reduced by therapy with low-molecular-weight heparin in orthopedic surgery patients6) has been confirmed.7
In their new study, Lewis and colleagues1 combined data from three prospective cohort studies, Arg-911, Arg-915, plus its extension phase, Arg-915X, for a total of 697 patients who received argatroban for the treatment of clinically suspected HIT. Patients were classified at baseline into groups depending on whether they had HIT without thrombosis (n = 321) or HIT-associated thrombosis (n = 376). A novel approach was to analyze as their primary end point a thrombotic composite comprising death due to thrombosis, amputation secondary to HIT-associated thrombosis, or new thrombosis within 37 days of baseline (maximum, one event per patient). This end point addresses the limitations of the previous composite end point (as required by the US Food and Drug Administration) of all-cause mortality, limb amputation, or new thrombosis, in which enrollment of patients with life-threatening comorbidities and advanced limb ischemia could have compromised the detection of antithrombotic efficacy.8
What can we learn from this new analysis? First, using the novel end point, the antithrombotic effect of argatroban did appear to be greater (ie, lower hazard ratios) than that estimated from the previous studies. The hazard ratios using the thrombotic composite ranged from 0.33 (HIT) to 0.39 (HIT-associated thrombosis), whereas the corresponding values for the composite end point in the Arg-911 and Arg-915 trials ranged from 0.60 to 0.64 (HIT) and from 0.56 to 0.57 (HIT-associated thrombosis).
Second, limb amputation rates did not differ between argatroban-treated patients and historical control subjects. Although disappointing, this finding should be viewed in its historical context. Guidelines for managing DTI-warfarin overlap did not exist when these trials were performed (1995 to 1998). It is now recognized that DTI-warfarin cotherapy must be carefully managed to minimize the potential for warfarin-associated microvascular thrombosis, including venous limb gangrene.89 For example, warfarin therapy should be postponed until the platelet count has substantially resolved, should be administered only in low initial doses, and should be overlapped for at least 5 days with the DTI.8 The clinician also needs to be aware that therapy with argatroban prolongs the international normalized ratio10 and that the international normalized ratio increases further during DTI-warfarin cotherapy,11 so that an incorrect conclusion is not reached that a patient has prematurely attained therapeutic anticoagulation with warfarin therapy. Further, initiating treatment with a novel anticoagulant within a clinical trial context faces inherent delays, including the time to transfer patients to study sites and obtaining informed consent, not to mention the "learning curve" required for the optimal use of an experimental agent. These factors can jeopardize the salvage of a severely ischemic limb. Thus, a more accurate assessment of the extent to which limb loss might be averted by therapy with argatroban (or any other therapy) awaits data from postapproval studies.
Third, the study provides insights into the potential risk factors for HIT-associated thrombosis. One interesting finding was that women had a higher frequency of HIT-associated thrombosis. Does this reflect a truly greater risk of thrombosis in women with HIT, or rather that female subjects were more likely to have actually had HIT in this study? Given that women may be at greater risk of HIT than men,12 and that patients were enrolled in the study based on a clinical suspicion of HIT rather than a positive test result for HIT antibodies, either interpretation is plausible.
There is probably still more to be learned from the argatroban trials. On my "wish list" are the following questions: how did HIT antibody-positive patients fare compared with antibody-negative patients? Is there evidence for rebound thrombosis on stopping argatroban therapy? The authors deserve congratulations for their insightful studies, and should be encouraged to pursue even more.
Footnotes
Dr. Warkentin has done consultant work for and/or has given lectures on behalf of GlaxoSmithKline, Organon Inc, the Medicines Company, and Berlex Laboratories, all of which manufacture antithrombotic drugs that have been used in the treatment of patients with heparin-induced thrombocytopenia.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
