Limited concordance of heparin/PF4 antibody assays for the diagnosis of heparin-induced thrombocytopenia: an analysis of the TORADI-HIT study.

BACKGROUND Anecdotal reports suggest that the correlation between heparin/PF4 antibody assays for the diagnosis of heparin-induced thrombocytopenia (HIT) is limited. To address this issue, we investigated the correlation between widely used assays and examined possible factors contributing to variability. METHODS This is a large, prospective cohort study with 10 participating tertiary care hospitals including 1'393 patients with suspected HIT in clinical practice. HIT was defined by a positive heparin-induced platelet activation assay (HIPA; washed-platelet reference standard test). Three different immunoassays were used to measure heparin/PF4 antibodies: chemiluminescent immunoassay (CLIA), enzyme-linked immunosorbent assay (ELISA), and particle gel immunoassay (PaGIA). Various factors that could influence the assays were examined: sex (male, female), age (< 65 years, ≥ 65 years), unfractionated heparin exposure, presence of thrombosis, cardiovascular surgery, and intensive care unit. Spearman's correlation coefficients were calculated. Z-scores, and diagnostic odds ratios (DOR) were determined in above-mentioned subgroups of patients. RESULTS Among 1,393 patients, 119 were classified as HIT positive (prevalence 8.5%). The median 4Ts score was 5 in patients with HIT (interquartile range, IQR, 4-6), compared to 3 in patients without (IQR 2-4). Correlations (rs) between immunoassays were weak (0.53 to 0.65). Inconsistencies between immunoassays could not be explained by further analyses of z-scored test results and DOR in subgroups of patients. CONCLUSIONS The correlation between widely used heparin/PF4 antibody assays was weak and key factors could not explain this variability. Standardization of immunoassays is requested to improve comparability.