Dual guided heparin monitoring with anti-Xa and APTT during micro-axial flow pump support for cardiogenic shock

Abstract

Abstract Background Precise management of unfractionated heparin (UFH) is key to avoid coagulopathic complications in patients on percutaneous mechanical circulatory support (pMCS) for cardiogenic shock (CS). While the activated partial thromboplastin time (APTT) is commonly used for this purpose, it can be affected by various laboratory factors during critical illness. Alternatively, the anti-Xa test directly measures UFH activity. We proposed a parallel anti-Xa/APTT-based protocol for UFH titration during pMCS. (1) Purpose We aimed to investigate the correlation between parallel anti-Xa/APTT measurements in a CS population on pMCS. We also evaluated mortality correlation in samples exhibiting prolonged APTT and in-range anti-Xa levels, based on fibrinogen levels and INR. Methods We assessed all CS patients with isolated left-sided ImpellaTM-support (>24h) and UFH anticoagulation monitoring per protocol at two pMCS-institutions from April-2017 to June-2022. Overall correlation between parallel anti-Xa/APTT samples was assessed by Pearson correlation coefficient (with ≤0.7 considered to confer a weak correlation). Anti-Xa levels were defined as 0.20-0.30IU/mL or 0.31-0.50IU/mL, with corresponding APTT values of 40-55s and 55-80s, respectively. Samples with prolonged APTT as compared to the in-range anti-Xa were stratified based on fibrinogen (<1.5g/dL or ≥ 1.5g/dL) and INR (<1.5 or ≥1.5) on the same day as the index sample. Mortality of patients with normal fibrinogen ànd INR was then compared to those with abnormal fibrinogen and/or INR via a Chi² test. Results Ninety-three ImpellaTM-runs (77 men, 83%) with a median age of 56 years (IQR 46-66) were analyzed. Median support was 6 days (IQR 4-12). In 2447 parallel anti-Xa/APTT samples, the Pearson correlation coefficient was 0.50 (p<0.001). Among the 1914 in-range anti-Xa samples, APPT was shortened in 460 (24%), in-range in 986 (52%) and prolonged in 468 (24%) samples. In the 59 patients with samples with prolonged APTT, two patients (with 80 samples) were excluded due to absence of fibrinogen and/or INR measurements on the same day. The mortality in patients with prolonged APTT but normal fibrinogen levels and normal INR (N=29; APTT prolongation due to FXI/FXII-consumption) was significantly lower compared to the remaining 28 patients who showed samples with reduced fibrinogen levels and/or prolonged INR (suffering from concomitant conditions) (10% vs. 32%; p=0.043). Conclusion We highlight a weak correlation between anti-Xa and APTT in an anti-Xa guided UFH anticoagulated CS population during left-sided ImpellaTM support. Patients with APTT-prolongation due to concomitant conditions with INR prolongation and/or fibrinogen shortage are at highest mortality risk. The APTT/Anti-Xa coefficient can serve as an independent risk factor whilst managing critically CS patients.