Excluding HIT Diagnosis by a Particle Gel Immunoassay.

Heparin-induced thrombocytopenia (HIT) is a prothrombotic condition, characterized by a platelet drop of >50% and/or thrombosis with a temporal relationship of 1–2 weeks after initiation of heparin treatment. These surrogate markers are useful for the clinical assessment but are hardly applicable in multi-morbid patients with clinical conditions mimicking HIT. Platelet activation assays (heparin-induced platelet aggregation assay, HIPA; serotonin release assay, SRA) and platelet factor-4/polyanion enzyme immunoassays (PF4-ELISA) confirm HIT. HIPA and SRA are highly specific but laborious requiring selected donor platelets and extended experience. High titer IgG antibodies correlate with clinical HIT, but ELISA is also time-consuming. The alternative H/PF4-antigen particle gel immunoassay (ID-PaGIA H/PF4^®) is easy, provides results within one hour and detects mainly IgG, but also IgA/M antibodies. We evaluated specificity and sensitivity of the PaGIA in comparison to HIPA and ELISA (PF4-ENHANCED^®) in 285 patients (median 71yrs, range 1–97, 45% cardiovascular surgery; f/m 56/73, and 55% medical, f/m 78/78) with undetermined likelihood for HIT and 89 healthy controls (median 42yrs, range 26–64, f/m 49/40) to validate it as a rapid assay to exclude HIT. HIPA was positive in 12% of patients. Based on ROC curves (ELISAs vs HIPA), OD-cutoff values for the IgG/A/M- and IgG-ELISA were 0.761 and 0.564, respectively. In controls the OD of IgG-ELISA was 0.066 (0.028–0.500) and lower than from patients’ samples negative by PaGIA and IgG/A/MELISA (n=158; 0.074; 0.009–0.339; p=0.017). PaGIA was positive in 70 patients (25%). In both patient groups both ELISA ODs were higher if PaGIA and IgG/A/MELISA were positive compared to negative PaGIA but positive IgG/A/MELISA (p<0.001; data are shown as median; range). Analyzed by ROC curves (ELISAs vs HIPA), IgG/A/MELISA (OD-cutoff 0.761) and IgG-ELISA (OD-cutoff 0.564) had sensitivities of 81% and 86%, and specificities of 75% and 81%. Positive and negative predictive values were 31% and 97% (IgG/A/MELISA) and 39% and 98% (IgG-ELISA), respectively. PaGIA, IgG/A/M- and IgG-ELISA were significant predictors for the results of HIPA (p<0.001) and the IgG-ELISA had the highest explained variance (41%). Based on our results the PaGIA may serve as a rapid test to exclude HIT since a negative PaGIA is rarely associated with high titer IgG PF4-antibodies or a positive HIPA. In cases of positive PaGIA results and in patients highly suspicious for HIT but with negative PaGIA, alternative anticoagulation should be considered until results from functional testing are available.