Abstract
Heparin‐induced thrombocytopenia (HIT) is caused by platelet‐activating antibodies that recognize platelet factor 4//heparin (PF4/H) complexes. According to the “iceberg model,” only a subset of anti‐PF4/heparin antibodies of IgG class evincing strong platelet‐activating properties cause clinical HIT. Since many centers rely predominantly on an anti‐PF4/polyanion enzyme‐immunoassay (EIA) to diagnose HIT, we estimated the potential for overdiagnosis when only this single test is available. We examined a database of 100 patients in whom the probability of HIT had been estimated using a clinical scoring system (4Ts), and where patients underwent systematic testing for HIT antibodies using three assays: the platelet serotonin release assay (SRA), an “in‐house” EIA that detects IgG anti‐PF4/heparin antibodies (EIA‐IgG), and a commercial EIA that detects anti‐PF4/polyanion antibodies of all three immunoglobulin classes (EIA‐GTI). Whereas 16 of 100 patients fulfilled a “classic” definition of HIT (intermediate/high probability plus strong platelet‐activating anti‐PF4/heparin IgG antibodies), an additional 16 patients fulfilled a “liberal” definition in which any investigated patient (irrespective of the pretest probability) who had a positive EIA‐GTI was considered to have HIT. The clinical features of these 16 additional patients—including generally weak antibodies and low risk for thrombosis—suggest underlying non‐HIT explanations for thrombocytopenia. Patients with a positive SRA generally corresponded to those with intermediate or high pretest probability of HIT who also had strong EIA‐GTI reactivity (>1.20 OD units). We conclude there is the potential to overdiagnose HIT by ∼100% if any positive EIA is considered to “confirm” the diagnosis of HIT irrespective of the clinical scenario. Am. J. Hematol., 2007. © 2007 Wiley‐Liss, Inc.