Type II heparin-induced thrombocytopenia (HIT) is an antibody-mediated complication of heparin therapy that is associated with a consumptive thrombocytopenia and a high risk of thromboembolism. It is now known that antibodies associated with type II HIT recognize sites on PF4 when complexed with heparin. Due to technical challenges associated with the serotonin release assay, most clinicians rely on the PF4 ELISA to establish a laboratory diagnosis of HIT. However, it is unclear whether the ELISA has predictive value in identifying those patients who have clinically apparent disease and whether ELISA optical density (OD) measurements correlate with the development of thrombosis. We retrospectively reviewed the medical records of 103 sequential patients who tested positive for HIT by a commercial PF4 ELISA from 2000–2002. While blinded to ELISA OD values, we recorded patient age and gender, admission diagnosis, hospital course, type of heparin administered, route of administration, baseline and nadir platelet counts, timing of thrombocytopenia, HIT therapy, clinical thrombotic events within 30 days, and other potential causes of thrombocytopenia to determine the clinical likelihood of HIT. The association between OD values and both the clinical likelihood of HIT and objectively confirmed thrombotic events was analyzed using the Pearson Chi-square test. Median patient age was 62 yrs and 52% of patients were female. Those with cardiac disease, cancer, and venous thromboembolism constituted 44%, 19% and 9% of the study population, respectively. Median platelet count nadir was 40K/mL (range, 2K–90K), and average onset of thrombocytopenia was 6.5 days. Median ELISA OD value was 0.77 (0.40–3.81). The clinical likelihood of HIT was noted in 24/103(23%) patients, and 22(21%) suffered a thrombotic event. An ELISA OD >1.0 was significantly associated with the clinical likelihood of HIT (p=0.03), though OD values ranged from 0.41 to 3.81 among these patients. Neither an OD threshold of 1.0 or 1.5 was associated with thromboembolism (p=0.2 and p=0.1, respectively). In contrast, the clinical likelihood of HIT was significantly associated with subsequent thrombosis (p=0.002). ELISA results were not significantly associated with the extent or onset of thrombocytopenia, type of heparin or route administered, patient age or gender, admission diagnosis, or hospital course. In conclusion, PF4 ELISA OD values are significantly associated with the clinical likelihood of HIT but are not predictive of HIT on an individual patient basis. Clinical judgment but not ELISA OD values is associated with subsequent thromboembolism. In summary, the clinical likelihood of HIT, and not ELISA results, remains the best parameter by which to modify anticoagulation therapy in patients being assessed for HIT.

Figure
View largeDownload slide

Figure

Topics:
clinical diagnosis, enzyme-linked immunosorbent assay, platelet factor 4, thrombocytopenia, heparin-induced, thrombocytopenia, heparin, thromboembolism, thrombus, antibodies, drug administration routes

Corresponding author

2005, The American Society of Hematology
2005
Sign in via your Institution