Predictors of Gastrointestinal Rebleed Events in Patients With a Ventricular Assist Device (VAD)

Abstract

Introduction: Ventricular assist device (VAD) placement is at the forefront of advanced heart failure treatment as a bridge to cardiac transplant or destination therapy. Gastrointestinal bleeding (GIB) continues to be one of the most common associated adverse events. Nearly 30 to 50% of VAD patients will have recurrent hemorrhage. We provide the first descriptive analysis of potential predictive factors for recurrent GIB in VAD patients. Methods: We conducted a retrospective, single center, observational, cohort study of VAD patients admitted with GIB between 2008-2016. We compared characteristics of those with and without recurrent bleeding to identify predictors of rebleeding. Continuous variables were reported as means, medians, and standard deviations, while categorical variables were reported as frequencies and percentages. Univariable analysis was performed using Fisher's exact test for categorical variables. All analyses were performed using STATA v14 (College Station, TX). Results: 106 VAD patients were identified, of which 26 (24.5%) had an initial bleed. Of those patients, 17 (65.4%) had a second GIB. Patients with repeat GIB had a mean age of 60.1 years, were more likely to be male (76.4%), and on average had 3.5 (2-9) bleed events (Table 1, Figure 1). A majority of rebleed events occurred within the first year of the herald bleed (Figure 2). Those who rebled were significantly more likely to have comorbidities of hypertension (HTN), coronary artery disease (CAD), hyperlipidemia (HLD), post VAD ejection fraction (EF) >30% (Table 1) and likely to be kept on an anticoagulant post initial bleed (Table 1). However, international normalized ratio (INR) levels were not significantly different on days 7 and 14 (P=0.621 & P=0.267, respectively) post initial bleed. On logistic regression, a higher age (odds ratio [OR] 2.58; 95% CI 2.45-2.72; P=0.049) and was associated with a decreased time to a rebleed event.Figure: Bar graph illustrating the total number of gastrointestinal bleed events in patients who have a second bleed.Figure: A Kaplan Meier survival analysis showing a substantial re-bleed rate occurring predominantly within one year of the initial bleed.Table: Table. Demographic and Clinical Parameters of VAD CohortConclusion: We report a substantial rate of recurrent GIB among VAD patients. Though exploratory, we identify potential risk factors associated with recurrent GIB and note no difference in recurrent GIB as related to the INR level. Identifying VAD patients who are at risk for recurrent GIB may help risk stratify VAD candidates. Larger multicenter prospective studies investigating the frequency and predictors of recurrent GIB in VAD patients are necessary.