The Real-World Performance of the PF4-Dependent P-Selectin Expression Assay (PEA) As the First-Line Functional Assay for the Diagnosis of Heparin-Induced Thrombocytopenia (HIT)

Introduction:

The serotonin release assay (SRA) serves as the gold standard confirmatory test to diagnose heparin-induced thrombocytopenia (HIT). The platelet factor 4 (PF4)-dependent P-selectin (PS) expression assay (PEA), a less complex test, has emerged as a promising alternative with reported comparable accuracy in the diagnosis of patients with HIT. In our previous work (Hoesley J, et al. Blood, 2022;140(1):1215-6), we analyzed the real-world performance of the SRA and the PEA at a large academic medical center. Based on those results, including the ability of the PEA to provide a conclusive answer in cases when the SRA was inconclusive, we implemented the PEA as the first-line functional assay for all patients with a positive HIT screening test. Herein, we present the performance of the PEA from the first two years since implementation and compare it to the historical performance of the SRA. We also compared the results of both tests in the workup of patients when both assays were requested since the PEA was introduced in our institution over a four-year period.

Methods:

We performed a retrospective investigation of three time periods of HIT testing at the University of Alabama at Birmingham, each encompassing two years: (1) the SRA-only period (1/2018-12/2019), (2) the transitional period, where the SRA was first-line but the PEA was also used in cases of diagnostic uncertainty (1/2020-12/31/2021), and (3) the PEA period, where the PEA was first-line but the SRA was used in cases of diagnostic uncertainty (6/16/2022-6/16/2024).

Laboratory testing: All patients with suspected HIT were screened in-house with the enzyme immunoassay (EIA) Asserachrom® HPIA, Diagnostica Stago (France). The SRA and PEA assays were performed at Versiti (Milwaukee, WI). SRA was positive if there was >20% release of serotonin with low dose heparin (0.1 U/mL) and <20% with high dose heparin (100 U/mL). The PEA was positive when there was ≥35% PS expression in the presence of the patient's serum and 30 µg of PF4 together with inhibition of PS expression by 50% in the presence of 100 U/mL heparin. Inconclusive functional results were defined as a percentage of serotonin release and/or P-selectin expression not meeting criteria for a positive or negative result.

Results:

A total of 3826 EIAs were performed during the six years of the study: 1252 in the SRA period, 1403 in the transitional period, and 1171 in the PEA period, with similar positivity rates of 14%, 15% and 16%, respectively. Functional assays were ordered 558 times: 176 in the SRA-only period, 203 in the transitional period, and 179 in the PEA period. Similarly, the rate of positivity of the functional tests was also consistent across all two-year time-periods examined with 24%, 20% and 21% respectively.

In the PEA period, 4% of patients had an inconclusive result (n=7) compared to 6% when the SRA was used as a first-line assay (n=24). Of the patients with inconclusive results, 47% were diagnosed with HIT when only the SRA was available (n=7), compared to 37% when both the SRA and PEA were available (n=6).

In the overall 6-year study period, 30 patients had both SRA and PEA results. In the 14 cases where the first-line test was inconclusive, the alternative test provided a definitive positive or negative result 86% of the time (n=12). In those patients where the first-line functional test was definitive, but the result was questioned based on the judgement of the clinical team (N=16), 44% (n=7) of results were discordant between the two assays with one negative and one positive result. Of those patients, five were diagnosed with HIT and two were not.

Conclusions:

In our experience, the real-world performance of the PEA as the first-line functional assay for the diagnosis of HIT did not differ significantly from the performance of the SRA. Specifically, inconclusive results occurred just as frequently with the PEA (4%) as with the SRA (6%). Access to a second functional assay in cases where the first-line assay is inconclusive, successfully confirmed or refuted the diagnosis of HIT in the majority of cases. However, providers should be aware that discordant results between these two assays can occur, and clinical judgement with incorporation of other diagnostic tools like the 4Ts Score and the EIA optical density remains foundational in the diagnosis of HIT.

No relevant conflicts of interest to declare.