Diagnosing heparin-induced thrombocytopenia (HIT) in 30 minutes : A prospective evaluation of a rapid diagnostic work-up.
M. Marchetti, S. Barelli, F. Monnin, N. Nicolas, E. Matthey, C. Gerschheimer, L. Alberio (Lausanne, Switzerland)
Time: 08:00 - 09:15
Objective: We assessed 3 immunoassays (IAs) detecting anti-PF4/heparin antibodies for their ability to predict the result of the functional gold standard Heparin Induced Platelet Aggregation Test (HIPA). Our aim was to prospectively evaluate a rapid algorithm able to confirm or exclude HIT within 30 minutes.
Methods: We conducted a retrospective (05.2014 – 08.2015; n=220) and a prospective (09.2015 – 08.2016; n=234) evaluation study in patients with clinical HIT suspicion. Plasma samples were analyzed by 3 IAs (ELISA Zymutest HIA monostrip IgG, AcuStar HIT IgG and ID-H/PF4-PaGIA). Employing ROC analysis we compared areas under the curve (AUC) and determined cut-offs with 100% negative (NPV) and positive (PPV) predictive values for a positive HIPA.
Results: HIT diagnosis could be proven by positive HIPA in 10% (22 patients) in the retrospective and in 9.4% (22 patients) in the prospective study. AcuStar showed an AUC of 0.98 both in the retrospective (p=0.02 compared to Zymutest; p=0.59 to PaGIA) and in the prospective study (p=0.06 to Zymutest; p=0.52 to PaGIa). A 100% PPV was observed with a result >1.37 U/ml (identifying 18/22 of HIT cases) and >0.77 U/ml (20/22 of HIT cases), respectively. A 100% NPV was observed with a result ≤0.12 U/ml (excluding 18/24 of HIPA-neg cases) and ≤0.18U/ml (excluding 200/212 of non-HIT cases). PaGIA showed AUC of 0.99 and 0.97. In both studies a titer ≥16 had a 100% PPV (identifying 28/44 of HIT cases) and a titer ≤1 had a 100% NPV (excluding 210/244 of non-HIT cases). In the prospective study, applying the cut-off values <0.12 U/ml = HIT-neg and >1.00 U/ml = HIT-pos, AcuStar excluded HIT in 192/234 (82%) and diagnosed it in 18/234 (8%) patients, recognizing 18/22 of HIT cases. 24 out of 234 results (10%) were in the AcuStar “grey zone” (0.12 – 1.00). Among them, 12 were correctly solved by PaGIA (HIT excluded by titer ≤1) while 12 (5% of the total) remained unclear until HIPA was performed. There were no false positive or negative results.
Conclusion: The sequential application of two rapid immunoassays (AcuStar HIT IgG and ID-H/PF4-PaGIA) with in-house cut-off values with 100% NPV and PPV enables a reliable and conclusive diagnostic work-up for 95% of patients with clinical suspicion of HIT, with a laboratory turn-around-time of 30 minutes. We are now undertaking a prospective validation of this rapid diagnostic algorithm.