Medical and perioperative management of acute and subacute heparin-induced thrombocytopenia in adults requiring urgent cardiopulmonary bypass surgery.

Abstract

Abstract Heparin-induced thrombocytopenia (HIT) is a highly prothrombotic disorder and presents a unique challenge for patients undergoing urgent cardiopulmonary bypass (CPB) surgery, where heparin is the mainstay intraoperative (IO) anticoagulant. Clinically relevant HIT occurs in acute and subacute HIT A, both pathologic phases distinguished by a positive functional platelet assay, such as the serotonin release assay (SRA). Presently, data are lacking on the optimal management of SRA+ HIT patients requiring urgent CPB. We undertook this multi-institutional retrospective study to describe the management and outcomes of SRA+ HIT patients undergoing CPB. Our cohort of 25 patients had 17 males and 8 females with a median age of 58 years (range 18-75). The cohort was critically ill, with 76% admitted to an intensive care unit immediately preceding the day of surgery (Day 0). The distribution of surgeries consisted of 40% left ventricular assist device implantation, 28% heart transplant, 8% coronary artery bypass, 8% open valve repair, 4% lung transplant, 4% evacuation of a right atrial thrombus, and 4% pulmonary artery endarterectomy. Most (68%) required short-term mechanical circulatory support. One patient's surgery was scheduled for Day 0 but postponed due to infection and the patient expired before the procedure could be performed; data from this patient was included for preoperative (pre-op) management and omitted from IO and postoperative (post-op) outcomes. The first heparin exposure occurred 12-91 days before Day 0 (median 19). The median baseline inpatient platelet count before HIT diagnosis was 225 and the median nadir attributed to HIT was 54. All patients had positive platelet factor 4/heparin (PF4/H) immunoassays and SRAs (range 22-101%; median 74%). Initial median PF4/H optical density (OD) was 1.81. The range of SRA positivity dates before Day 0 was 1-69 days (mean 17.4 ± 19.4; median 9), with 3 tests resulting more than 50 days prior. Repeat SRAs performed for 3 cases within 3 days of Day 0 remained positive. After diagnosis of HIT, most were started on the alternative anticoagulants argatroban or bivalirudin (bival) in 0-1 days. Preoperative (pre-op) plasma exchange (PLEX) was done in 6/25 between Day -13 to -1 (median 3 sessions), ending on Day -1 or Day 0. Pre-op PLEX was done in 4/25 only on Day 0. In patients with serial OD levels after PLEX (n=4), a decrease in OD was seen after each session (mean 0.32 ± 0.18; median 0.37). Serial OD levels without PLEX (n=8) increased on average by 0.29 ± 0.55 when performed a median of 10 days from the initial PF4/H test. The 9 patients with OD values within 2 days of surgery had a median of 1.84. Treatment groups were further divided by the following subsets: IO bival ± pre-op PLEX (n=5), IO heparin with pre-op PLEX alone (n=6), and IO heparin with IO PLEX alone (n=10). These cases (each n=1) were not grouped: IO heparin+tirofiban with pre-op PLEX+IVIG, IO heparin alone, IO heparin with pre-op plus IO PLEX, and 1 who died before planned surgery. All patients survived surgery and all returns to the operating room were planned. IO transfusion of ≥4 blood products in 1 hour was seen in 80% of the IO bival group compared to 25% of the combined PLEX alone groups. Between both PLEX alone groups, IO transfusion of ≥4 blood products in 1 hour was higher in pre-op PLEX than IO PLEX (67% vs 20%). The IO bival subset also had greater IO packed red blood cell (PRBC) requirements and incidence of acute kidney injury (AKI) compared to PLEX alone (mean units of PRBC 9.8 vs 1.7; incidence AKI 60% vs 25%). The one post-op death attributed to coagulopathy was in the IO bival group. Thrombotic complications were comparable throughout, with 0 IO VTEs and 1 IO arterial embolus after removing an Impella. At post-op Day 90, 1 multi-limb VTE was seen in the IO bival group vs 3 upper extremity VTEs in the PLEX alone subset, with no post-op arterial events. In conclusion, PLEX alone appears to be an effective strategy to manage SRA+ HIT prior to urgent CPB surgery. PF4/H OD levels decrease with PLEX but may not be a sensitive predictor of thrombotic events in isolation. IO bival appears to increase risk of IO and post-op bleeding but conclusions are limited by sample size. Larger, comparative studies are needed to determine and characterize the optimal management approach for this unique clinical scenario.