A Historical Comparison of Danaparoid (with or without Coumarin) Versus Ancrod (with or without Coumarin), or Coumarin Alone, for the Treatment of Heparin-Induced Thrombocytopenia (HIT): Evidence That Treatment Experience Can Influence Choice of Anticoagulant Rather Than Approval Status

Abstract number: P1727

Lubenow¹ N, Warkentin² TE, Greinacher¹ A, Wessel¹ A, Sloane² DA, Krahn² EL, Magnani³ HN

¹¹Ernst-Moritz-Arndt University, Greifswald, Germany ²²McMaster University, Hamilton, Canada ²²McMaster University, Hamilton, Canada ³³Organon NV, Oss, the Netherlands

Due to its low frequency and variable clinical picture, prospective randomized trials of HIT are difficult to perform. During the 1980's, the defibrinogenating snake venom, ancrod (± warfarin, Canada), or coumarin (warfarin, Canada; phenprocoumon, Germany) alone, were often used to treat HIT. During the 1990's, danaparoid (mixture of anticoagulant GAGs) ± coumarin began to be used. By 1995, danaparoid had replaced these treatments in some areas, such as Hamilton (Canada), despite lack of comparative data, and despite not being approved for treatment of HIT (although ancrod was approved for HIT in Canada). We combined two databases to perform a retrospective evaluation of treatment outcomes (1986–1999) comparing danaparoid ± coumarin vs ancrod ± coumarin or coumarin alone (controls), reaching a total of 62 danaparoid-treated patients and 56 controls who met the study criteria (e.g., HIT antibodies; definite indication for anticoagulation). The predefined composite endpoint of adjudicated new, progressive, or recurrent thrombosis (including thrombotic death), or limb amputation, at day 7 (maximum, one event per patient) was significantly lower in danaparoid-treated patients, compared with controls: 8/62 (12.9%, 95% CI, 4.3–21.5) vs 22/56 (39.3%, 95% CI, 26.1–52.5); P= 0.0014. We also found a lower frequency in danaparoid-treated patients of the composite endpoint at end of study (day 35): 12/62 (19.4%) vs 24/56 (42.9%); P= 0.0088. Significant differences in these endpoints were also found in the subgroup of patients treated in Hamilton. Major bleeding (by day 7) occurred in 7/62 (11.3%) and 16/56 (28.6%) of danaparoid-treated and control patients, respectively; P= 0.02. No significant differences (day 35) were seen in all-cause mortality (9.7% vs 14.3%; P= 0.57) or limb amputation (4.8% vs 7.1%; P= 0.71). We conclude that the replacement of ancrod ± coumarin, or coumarin alone, by danaparoid (± coumarin) during the mid-1990's was justified by improved efficacy and safety.