RESUMO
BACKGROUND: Heparin-induced thrombocytopenia (HIT) is an adverse complication of heparin caused by HIT antibodies that recognize platelet factor 4-heparin (PF4/hep) complexes leading to platelet activation. Several methods are available for the identification of HIT antibodies. OBJECTIVES: To evaluate the clinical usefulness of different antigen-binding assays for detection of antibodies against PF4/hep complexes in a prospective study. PATIENTS/METHODS: A prospective cohort of 500 surgical and medical patients suspected of having HIT was evaluated. The laboratory assessment included particle gel immunoassay (PaGIA), polyspecific ELISA recognizing IgG, IgM and IgA antibodies (Poly-ELISA), IgG-specific ELISA (IgG-ELISA) and the HIPA test. The pretest probability of HIT was determined using the 4T's model. Positive and negative predictive values (PPV, NPV) of each immunoassay were determined depending upon the heparin-induced platelet activation (HIPA) results and the clinical scoring. The operating characteristics of each immunoassay were determined using the receiver-operation characteristic (ROC) curve. RESULTS: Platelet-activating antibodies were identified in 35/500 patients, all of whom had intermediate to high clinical probability. PF4/hep antibodies were detected in 124, 86 and 90 sera using Poly-ELISA (PPV = 28), IgG-ELISA (PPV = 40.6) and PaGIA (PPV = 36.6). NPV was 100% for Poly- and IgG-ELISA, but only 99.5% for PaGIA. ROC analysis revealed that PaGIA is less informative than ELISA. The IgG-ELISA provides better diagnostic information than the other assays. In addition, there is a clear correlation between optical density (OD) value and the probability of having HIT. CONCLUSIONS: Our observation indicates that an IgG-ELISA provides the best diagnostic information of all antigen-binding assays.