ECLS and ECMO in HIT II Patients: Is Direct Thrombin-Antagonism A Safe Strategy?

Objectives: ECLS patients may develop heparin (Hep)-induced thrombocytopenia type II (HIT II). Drug approval for direct thrombin antagonism in ECLS patients is lacking. Here, we analyze if treatment of ECLS patients with Argatroban (Arg) is feasible.

Methods: “Urgent” or “emergency” ECLS patients (venoarterial)/ECMO (veno-venous) were studied retrospectively. ICU morbidity, survival, therapeutic stability of anticoagulation, bleeding, thrombosis and technical integrity was controlled.

Results: Forty-three ECMO and 157 ECLS patients received Hep or were changed to Arg due to suspected (18%) or proven HIT II (11.5%). ECMO patients were younger ([y] Hep: 49.8 ± 2.7; Arg: 50.7 ± 3.9 vs. ECLS: Hep: h61.3 ± 1.8; Arg: 60.6 ± 2.5). ECLS hospitalization was longer (p = 0.0028) but not related to medication (ECMO Hep: 36.9 ± 10.8 vs. Arg 28.2 ± 8.3; p = 0.236/ECLS: Hep: 11.6 ± 1.9 vs. Arg 21.5 ± 3.4; p = 0.087). Arg-ECLS stayed longer on ICU (ECMO Hep: 31.9 ± 6.6 vs. Arg 28.9 ± 7.2; p = 0.225/ECLS: Hep: 11.1 ± 1.6 vs. Arg 22.6 ± 3.9; p = 0.007). ECMO support was longer (13.54 ± 1.92 days vs. 5.65 ± 0.49 days; p < 0.001) but independent from medication (p = 0.742). Dosage adjustment was rare ([n/day] ECMO 0.7 ± 0.1 vs. ECLS 1.5 ± 0.2; p = 0.022) and reduced by Arg initiation (ECLS: 1.1 ± 0.1; p = 0.53). Oxygenator exchange was required during Hep (ECMO: 2 thrombosis, 1 technical dysfunction/ECLS: 2 thrombosis). aPTT[sec] (Hep vs. Arg; 62.1 ± 1.6 vs. 58.4 ± 1.4; p = 0.146) and ACT [sec] (158.1 ± 5.9 vs. 162.9 ± 4.8; p = 0.671) were alike. Subtherapeutic dosage was rare yet frequent in HITII ([per patient]; Hep 5.3 ± 0.6 vs. Arg 5.1 ± 1.7 vs. HITII 9.6 ± 1.7; p = 0.001). Bleeding was higher in thrombocytopenia and lower after Arg ([l] Hep2.5 ± 0.3 vs. Hep-Arg: 5.3 ± 0.8 vs. Arg 4.7 ± 1.0; p = 0.003). No patient-thromboembolism occurred. 56% (ECMO), 54% (ECLS) patients were weaned (p = 0.997). Cause of death (cardial, pulmonal, multiorgan failure, sepsis) differed by system (p < 0.001) not by medication (ECLS, p = 0.470; ECMO p = 0.801). Neither 30-day (Hep vs. Arg: ECMO 35.7 vs 28.6; p = 0.738/ECLS 27.3 vs 33.3; p = 0.615) nor 1-year survival (ECMO 9.5 vs 22.5; p = 0.125/ECLS 27.0 vs 23.1; p = 1.0) was medication dependent.

Conclusions: Argatroban is not inferior during ECLS/ECMO therapy in patients with proven or suspected HITII. In case of (suspected) HIT II, direct thrombin-antagonism on ECLS/ECMO is a practicable therapeutic strategy.