PB0348 - Can 5B9, a monoclonal antibody to PF4/H, be useful in improving the performance of the Asserachrom HPIA test for the diagnosis of HIT?

Sunday, June 25, 2023

18:30 – 19:30 ET

Room: Exhibition

Presenting Author(s)

Claire Pouplard (she/her/hers)

Professor
University Hospital of Tours
Tours, Centre, France

Co-Author(s)

Jérôme Rollin, PhD

University of Tours
Tours, Centre, France
Audrey Carlo, MSc

Scientific Marketing Manager
Stago
Asnières-sur-seine,, Ile-de-France, France
AB

Anne Bauters

PharmD
Department of hemostasis, University Hospital of Lille, France
Lille, Nord-Pas-de-Calais, France
Philippe Savard, MD (he/him/his)

MedD
Hématologie Biologique, CHU, Dijon Bourgogne
Dijon, Bourgogne, France
SJ

Simon Jaouen

PharmD
Department of hemostasis, University Hospital of Brest, France
Brest, Bretagne, France
LM

Laure Maucorps

MD
Department of Hematology/Hemostasis, Pontchaillou University Hospital of Rennes, Rennes, France
Rennes, Bretagne, France
CV

Caroline Vayne

Associate Professor
Department of hemostasis, University Hospital of Tours, France
Tours, Centre, France
EG

Eve Anne Guery

PharmD
Department of hemostasis, University Hospital of Tours, France
Tours, Centre, France
YG

Yves Gruel

Professor
EA 4205, University of Tours, Tours, France
Tours, Centre, France

Background: In a recent study, we demonstrated that 5B9, a chimeric IgG monoclonal antibody to H/PF4 with a human Fc fragment, could help to standardize HIT functional tests (Pouplard et al, J Thromb Haemost, 2022).

Aims: Here, we evaluated if 5B9 could be used to improve the accuracy of the Asserachrom HPIA® IgG, a PF4/H IgG specific ELISA.

Methods: 170 patients with intermediate (n = 144) or high (n = 26) clinical probability of HIT were prospectively included. The Asserachrom HPIA® IgG and PF4-SRA were systematically performed, and HIT was confirmed when both tests were positive. During each ELISA run, a calibration curve was done with 5B9 tested at 6 concentrations (0.16 to 10 microg/ml). Results were then expressed as optical density (OD 450 nm) or equivalent 5B9 (Eq5B9) using a 4 parameter logistic regression.

Results: HIT was confirmed in 43 patients in whom the probability of HIT was high (n = 18; 69.2%) or intermediate (n = 25, 17.4%). ROC curves designed according to the diagnosis of HIT with OD or Eq5B9 values showed similar AUC, of 0.986 and 0.980 respectively (Figure 1). When the OD threshold was equal to 1.0, the NPV of the assay was excellent (98.3%), with a good PPV (83.7 %), which reached 100% for OD ≥ 2.0.

Conclusion(s): The use of 5B9 as a calibrator does not increase the diganostic performances of the ELISA. However, this study emphasized that an OD cut off value close to 0.18 as proposed in Asserachrom IgG HPIA® insert sheet, is too low and may lead to unnecessary functional tests. When the clinical probability of HIT is intermediate or high, an OD value lower than 1.0 excludes HIT with high NPV, and HIT is highly probable when OD is ≥ 2.0.