Abstract

Abstract 2189

Heparin-induced thrombocytopenia (HIT), caused by antibodies against heparin/platelet factor 4 (HPF4) complex, is a rare but potentially serious side effect of heparin therapy where due to high mortality, rapid diagnosis is crucial. For the detection of HPF4 antibodies we compared the new nanoparticle-based lateral-flow immunoassay (LFI-HIT, Milenia Biotec, Germany) and a particle gel immunoassay (PaGIA, BioRad, Germany) with an IgG-specific-PF4/polyanion enzyme-linked immunosorbent assay (IgG-ELISA, GTI Diagnostics, USA).

Sera from 121 patients (54/67 f/m, median 73 years, range 14–94) with suspected HIT were prospectively tested. The LFI-HIT and the PaGIA were evaluated visually, the IgG-ELISA was positive at an optical density (OD) cutoff > 0.4. For most of the positive samples, the functional heparin-induced platelet activation (HIPA) assay was additionally performed to detect false positive serological results and to confirm a clinically relevant HIT by in vitro platelet-activation. Regarding HIT as a clinico-pathological syndrome, characteristics for HIT were evaluated for each patient by the 4Ts scoring system and divided into high, intermediate or low risk.

Results of serological analyses and OD values are summarized in the table. Ten of 121 samples were positive in the LFI-HIT, 10/10 positive in the PaGIA and 8/10 positive in the IgG-ELISA. The HIPA was tested in 9/10 samples and was positive in 8/9 samples. Of the 2 samples positive for LFI-HIT and PaGIA but negative in the ELISA, 1 was HIPA positive, 1 HIPA negative, resulting in a specificity of 88.9% for the LFI-HIT assay correlated to the HIPA.

From 111/121 LFI-HIT-negative samples, 2 were positive in the PaGIA, the IgG-ELISA (OD 1.318 and 2,019) and in the HIPA. Seven of the 111 LFI-HIT negative samples were positive only in the IgG-ELISA. Due to marginal positive reactions of 5/7 samples in the ELISA with OD values between 0.4 to 0.5, only 2 LIF-HIT negative IgG-ELISA positive samples were tested by HIPA and 1/2 was positive.

Based on the ELISA, the sensitivity of the LFI-HIT was 91.9% (102/111 negative samples also negative in the ELISA) in contrast to 93.1% of the PaGIA. The specificity of the LFI-HIT was 80% (LFI-HIT and IgG-ELISA positive), compared to 57.9% of the PaGIA. Notably, the clinical risk estimated by the 4Ts score system (received from 92/121 patients) did not correlate with laboratory diagnosis of HIT, probably due to inadequate evaluation.

Concluding our data, a reliable exclusion of HIT by rapid testing with the LFI-HIT only seems possible with additional analysis of HPF4 antibodies by IgG-ELISA and/or HIPA assay.

LFI-HITPaGIAIgG-ELISAOD IgG-ELISAHIPA assay
Median (range)
Samples n=121 Pos 10 Pos 10 Pos 8 2.366 (0.902-3.000) 7/7 pos 
Neg 2 0.199 and 0.170 1/2 pos, 1/2 neg 
Neg 0  
 
Neg 111 Pos 9 Pos 2 1.318 and 2.019 2/2 pos 
Neg 7 0.110 (0.054-0.139) 6/6 neg 
Neg 102 Pos 7 0.436 (0.404-1.463) 1/2 pos, 1/2 neg 
Neg 95 0.082 (0.013-0.376)  
LFI-HITPaGIAIgG-ELISAOD IgG-ELISAHIPA assay
Median (range)
Samples n=121 Pos 10 Pos 10 Pos 8 2.366 (0.902-3.000) 7/7 pos 
Neg 2 0.199 and 0.170 1/2 pos, 1/2 neg 
Neg 0  
 
Neg 111 Pos 9 Pos 2 1.318 and 2.019 2/2 pos 
Neg 7 0.110 (0.054-0.139) 6/6 neg 
Neg 102 Pos 7 0.436 (0.404-1.463) 1/2 pos, 1/2 neg 
Neg 95 0.082 (0.013-0.376)  

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.