Optical Density of 1.0 Is a Better Cutoff for Positive HIT Ab ELISA.

Abstract

Abstract 3515Poster Board III-452 BackgroundHeparin-induced thrombocytopenia (HIT) is a clinicopathologic diagnosis based on pretest clinical assessment aided by the 4T score and confirmed by laboratory testing for the presence of anti-heparin-platelet factor 4 antibody (HIT Ab). Prompt and accurate diagnosis of HIT is paramount due to an extraordinary high risk of thrombosis and the inherent risk of bleeding and cost with the use direct thrombin inhibitors (DTI). The polyspecific enzyme linked immunosorbent assay (poly-ELISA) for the HIT Ab is the most commonly available test that detects IgG, IgM and IgA HIT Ab. The IgG-specific ELISA detects only IgG HIT Ab, the antibody that is known to cause HIT. The use of a second step ELISA with high-dose heparin in the reagent improves the specificity by demonstrating heparin-dependence of the antibody detected. Serotonin release assay (SRA) is considered the gold standard for detecting HIT Ab, but it is available only at a few select centers in the world. Past studies have shown that the probability of SRA positivity increases with the optical density (OD) unit of the ELISA. In particular, only 1-5% of patients with OD of 0.4 to 1.0 have a positive SRA, and increases to 89-100% when OD is greater than 2.0 (Warkentin et al, JTH 2008). The 4T score was developed to predict the probability of HIT. This score takes into account the severity of thrombocytopenia, timing of platelet fall with relation to heparin use, presence of new thrombosis, and other causes of thrombocytopenia. The high negative predictive value of the 4T score has been validated in multiple studies. However, the polyspecific ELISA was used in most of these studies, increasing the possibility of false positive tests. Study: We have collected a database of patients being tested for HIT at our institution, where the IgG-specific ELISA along with high-dose heparin inhibition is being used to detect the HIT Ab. We performed a retrospective review of the last 165 ELISAs performed and the clinical circumstances of the testing. Clinical data was reviewed in detail and patients were categorized as having definite HIT, probable HIT and not HIT. We hypothesize that the high negative predictive value of the 4T score combined with the more specific IgG-specific ELISA could allow the use of OD unit of 1.0 as a cutoff for positive ELISA rather than the current standard of 0.4. ResultsWe identified 15 patients with OD>0.4; of these 10 had OD>1.0. 6 of the 10 had 4T score of 5 or higher, with 5 of the 6 considered to have definite HIT and 1 with probable HIT. 2 of the 4 patients with OD>1.0 and 4T score of 4 were not considered to have HIT, while 2 had probable HIT. 5 patients had OD of 0.4 to 1.0; 4 with 4T score of 4 were considered to have no HIT, while the last patient only had probable HIT with a 4T score of 6. ConclusionBased on these results, we conclude that the IgG-specific ELISA could use a cutoff of OD 1.0 for positive HIT Ab test. When the OD is between 0.4 and 1.0, the diagnosis of HIT should be entertained when a high risk 4T score is present. Thus the degree of “positivity” of the ELISA in conjunction with the 4T score can be used to properly identify and manage patients with probable or definite HIT. Our data confirms that testing “positive” for HIT Ab with the ELISA does not necessitate a diagnosis of HIT, but simply increases the probability of having HIT. Inherent risk of bleeding with the use of DTI and the lack of a reversal agent would dictate that the diagnosis of HIT is applied with a high degree of certainty. We believe that using a higher cutoff of 1.0 would improve our confidence in the diagnosis of HIT when used in conjunction with the 4T score. Further expansion of our study with more patient numbers would be necessary to confirm our preliminary findings.Table 1Patients Profiles with OD>0.44T ScoreOD% InhibitionClinical DiagnosisPatients with OD>1.042.34186Probable HIT41.969100Probable HIT41.227105Not HIT41.213100Not HIT52.348103Probable HIT51.87797Definite HIT63.46799Definite HIT62.09892Definite HIT61.86396Definite HIT61.83193Definite HITPatients with OD 0.4-1.040.51197Not HIT40.49284Not HIT40.479105Not HIT40.40244Not HIT60.54996Probable HIT Disclosures:No relevant conflicts of interest to declare.