Association of delayed platelet recovery and lower nadir counts with adverse outcomes in heparin-induced thrombocytopenia

Abstract Introduction: Heparin-induced thrombocytopenia (HIT) is a rare, life-threatening prothrombotic disorder triggered by exposure to heparin. Despite timely diagnosis and initiation of appropriate non-heparin anticoagulation, HIT continues to provoke significant morbidity and mortality. Current treatment strategies are supportive, centered on heparin cessation and transition to non-heparin anticoagulants, which enable gradual platelet count recovery and reduced thrombotic risk. However, the prognostic implications of platelet count recovery kinetics, specifically whether rapid versus delayed recovery impacts clinical outcomes, remains poorly defined. Methods: We conducted a retrospective cohort study of patients with serotonin release assay (SRA)–confirmed HIT at the Oregon Health & Science University between 2014 and 2024. We excluded patients with concurrent participation in clinical trials aimed at improving outcomes in HIT. Medical records were manually reviewed to confirm a diagnosis of HIT, extract platelet count data, and evaluate record outcomes. Platelet count recovery was defined as the number of days from the date of SRA draw to the first documented platelet count exceeding 150,000/μL. Primary outcomes included thrombotic and bleeding complications and survival to hospital discharge. Results: Results: A total of 89 patients with SRA-confirmed HIT were identified. Four patients enrolled in clinical trials were excluded. The median age was 61 years. Sixty seven patients (79%) survived to hospital discharge, while 18 (21%) died. The median time to platelet count recovery was 6 days. At the time of HIT diagnosis, 55 patients (61%) presented with thrombosis and 20 (19%) with bleeding. Following diagnosis, 37 patients (40%) developed new or progressive thrombosis, and 13 (15%) experienced bleeding complications. Sixty-three patients (71%) required intensive care during hospitalization, and the median length of hospital stay was 21 days. Of the 85 patients, all had sufficient longitudinal platelet data for group-based trajectory modeling. Two distinct platelet recovery trajectories were identified: 20 patients (24%) were categorized as “fast responders” and 65 (76%) as “slow responders.” While not statistically significant, slow responders experienced numerically higher rates of both thrombotic and bleeding complications. Thrombotic event rates averaged 0.85 vs. 0.65 events per patient (p = 0.41), and bleeding event rates were 0.25 vs. 0.14 events per patient (p = 0.41) for slow versus fast responders, respectively. Platelet transfusions were also more common among slow responders (32% vs. 26%, p = 0.78), and time to first thrombotic event occurred earlier (mean day 1.9 vs. 5.7, p = 0.058). ICU admission rates were proportionally higher among fast responders (90% vs. 69%, p = 0.08), though more slow responders required ICU care in absolute numbers. Additionally, patients with more severe thrombocytopenia (defined as nadir platelet count ≤77 × 10⁹/L prior to SRA) were more likely to experience subsequent thrombotic events, with 53% developing thrombosis compared to 36% of patients with higher nadir counts. The thrombotic event rate per patient-day was also higher in the low platelet group (0.033 vs. 0.025), supporting a potential link between depth of thrombocytopenia and thrombotic risk in HIT. Conclusion: HIT remains a morbid and life-threatening condition, associated with prolonged hospitalization, high rates of thrombotic and bleeding complications, and substantial intensive care utilization. In this cohort, nearly one in five patients died during hospitalization despite appropriate treatment. While not statistically significant, slower platelet recovery was associated with numerically higher rates of thrombosis, bleeding, and transfusion, as well as earlier thrombotic events, suggesting a trend toward greater morbidity. While a trend may be emerging, it is also possible that platelet recovery may not be the most reliable marker for monitoring treatment response. Additionally, patients with more profound thrombocytopenia prior to diagnosis were more likely to develop subsequent thrombosis. These findings support the hypothesis that both delayed platelet recovery and lower nadir platelet counts may serve as indicators of more aggressive disease and could help inform clinical risk stratification in HIT. Further prospective studies are warranted to validate these associations and clarify their clinical utility.