"Autoimmune HIT", a Potential Diagnostic Artifact Due to Remnant Heparin in HIT Samples, Is Not Associated with More Severe Outcomes

In recent years, there has been increasing use of the term "autoimmune" heparin-induced thrombocytopenia (HIT) to describe patients with HIT antibodies that activate platelets in the absence of exogenous heparin (i.e., the "zero heparin" serotonin release assay, SRA), with the proposed implication that such patients present with more severe phenotypes. Platelet activation in the SRA in the absence of exogenous heparin is likely a real phenomenon in rare/infrequent patients who did not receive proximate heparin therapy (Spontaneous HIT) and in patients who develop HIT after heparin treatment has ceased (delayed-onset HIT). However, in other patients deemed to have "autoimmune" HIT based on reactivity in the "zero heparin" SRA, we hypothesized that platelet reactivity in this SRA test system might result from remnant heparin in the patient's diagnostic blood sample. To further complicate the interpretation of autoimmune HIT testing, there is a lack of rigorous data demonstrating a correlation between this diagnosis and disease severity. This study was undertaken to evaluate the technical validity and clinical relevance of an autoimmune HIT diagnosis.

First, thirty "real-world" diagnostic HIT patient samples were subjected to anti-Xa testing. Remarkably, fourteen of the 30 samples (47%) had detectable heparin, with a range of 0.11-0.55 IU/mL for thirteen samples, while one sample had heparin <0.1 IU/mL (data not shown due to Abstract 2-figure panel limit). Next, to determine the extent to which low concentrations of unfractionated or low molecular weight heparin spiked into patient samples impact reactivity in the SRA, spiked samples were tested in the "conventional" low dose heparin (0.1 IU/ml) SRA and the "zero heparin" SRA (autoimmune HIT test). Fig 1A demonstrates that even ultra-low concentrations of unfractionated heparin (0.06 IU/mL) spiked into the sample facilitate serotonin release in the "zero heparin" SRA. On the other hand, the conventional SRA was not impacted by heparin spiked into samples unless concentrations were 5 IU/mL or higher, several-fold greater than what would be seen in patient samples. Similarly, LMWH induced platelet reactivity in the zero heparin SRA at concentrations between 0.25-1.0 IU/mL, while there was no effect on the conventional SRA at these concentrations (data not shown). As proof-of-concept that low levels of heparin mediate reactivity in a zero-heparin functional test, a HIT patient sample devoid of heparin was spiked with 0.05 IU/ml of heparin and evaluated with and without digestion with heparinase I. Heparin-spiked samples induced platelet activation in a zero-heparin functional assay which was abrogated upon heparinase I digestion (data not shown). Together, these results are consistent with the hypothesis that at least some cases of "autoimmune HIT" are labeled as such due to residual heparin present in patient samples rather than due to the presence of a qualitatively different ("autoimmune") HIT antibody.

To evaluate the clinical relevance of an "autoimmune" HIT diagnosis, platelet nadir and the presence of thrombosis were interrogated as markers of disease severity in a cohort of 48 HIT patient samples with unknown heparin levels. Thirty-six patients were positive in the zero heparin SRA ("Autoimmune HIT" cohort; >50% serotonin release), while 12 were negative ("Classical HIT" cohort; positive in the conventional SRA, but <50% serotonin release in the zero heparin SRA). Fig 1B demonstrates that there was no difference in platelet nadir between the two groups (p=0.562); further, there were no differences between these groups in rates of confirmed thrombosis/skin necrosis and acute system reactions (p=0.324; data not shown).

In summary, "autoimmune HIT" diagnostic testing (zero-heparin SRA) should not be performed until samples are confirmed to be devoid of heparin, and rigorous large-scale studies are needed to evaluate whether results of validly-conducted autoimmune HIT tests correlate with disease severity.

Figure 1. (A) Low concentrations of spiked heparin in a HIT patient sample facilitates platelet activation in the "autoimmune" HIT test (zero-heparin SRA; red line) but does not affect the conventional SRA (black line). (B) Platelet nadir was similar between "autoimmune" (n=36) and conventional HIT (n=12) cohorts.

Garcia:Incyte: Research Funding; Daiichi Sankyo: Research Funding; Janssen: Other: Personal Fees and nonfinancial support; Seattle Genetics: Other: Personal Fees. Pruthi:Bayer Healthcare AG: Honoraria, Membership on an entity's Board of Directors or advisory committees; HEMA Biologics: Honoraria, Membership on an entity's Board of Directors or advisory committees; Instrumentation Laboratory: Honoraria, Membership on an entity's Board of Directors or advisory committees; Merck: Consultancy; Genentech Inc: Honoraria, Membership on an entity's Board of Directors or advisory committees; CSL Behring: Honoraria, Membership on an entity's Board of Directors or advisory committees. Padmanabhan:Retham Technologies: Other: Equity owner and officer; Veralox Therapeutics: Membership on an entity's Board of Directors or advisory committees; Mayo Clinic, Versiti, Retham Technologies: Patents & Royalties.