ACE Inhibitor Use is Associated with Reduction in Gastrointestinal Bleeding Events in CF-LVAD Patients

Abstract

Gastrointestinal bleeding (GIB) remains a vexing issue in patients supported with continuous-flow LVADs (CF-LVAD). GIB is associated with hospital readmissions, need for procedures, and blood transfusions. Recent literature has suggested association between angiotensin II antagonism and formation of AVMs in CF-LVAD patients. We hypothesized that ACE inhibitor (ACEi) use is associated with a reduction in GIB. We performed a retrospective analysis of CF-LVAD patients from our institution from 2017 to 2019. Descriptive statistics, hypothesis testing (t-tests and chi square) and logistic regression modeling including propensity-score matching was performed with STATA/SE 15.1. 110 patients were included in the study population (76% men, average age 56 years). 27 patients (25%) had GIB events. Age, creatinine, BP, central venous and pulmonary artery pressure, prior ACEi use, and tricuspid and mitral valve regurgitation were all similar between GIB and non-GIB patients. GIB compared with non-GIB patients were more likely to be ischemic (62% vs 39% p=0.049) and implanted as DT (89% vs 65% p=0.018). Only 45% of the population was on an ACEi at 1 month. Pre-operative Cr and Cr at 1 week were higher in patients not on ACEi (1.4 vs 1.1 p=0.0026 and 1.57 vs 1.1 p=0.0019); however, Cr at 1 month was similar between the groups not on ACEi and on ACEi (1.15 vs 0.97 p=0.07). Unadjusted odds ratio for GIB with ACE inhibitor use at 1 month post implant was 0.35 (p=0.035). Odds ratio when adjusted for covariates was not significant in logistic regression modeling. Using a smaller population with complete data (n= 40), propensity-score matching (age, sex, initial indication, etiology, BMI, CVP and Cr pre-implant, Cr at 1 week and Cr at 1 month) for ACEi at 1 month demonstrated a significant association between ACEi use at 1 month and fewer GIB events (coefficient -.175, p=0.014) in a matched group. There is an association between ACEi use at 1 month and reduction in GIB in an unadjusted model. The association remains significant in propensity-score matching analysis. Although Cr was higher in the group not on ACEi very early post-operatively, Cr was the same in patients on or off ACEi by 1 month suggesting that ongoing assessment about candidacy for ACEi use post-implantation should be performed. Further work to clarify timing of ACEi initiation and dosing is needed.