The association between novel clinical factors and gastrointestinal bleeding among patients supported with continuous-flow left ventricular assist device therapy


Aurora Cardiovascular Services, Center for Integrative Research on Cardiovascular Aging (CIRCA), Aurora Research Institute


BACKGROUND: This study explores novel preimplantation risk factors associated with gastrointestinal bleeding (GIB) after continuous-flow left ventricular assist device (CF-LVAD) implantation. CF-LVAD therapy implantation for patients with advanced heart failure is associated with a 20% to 40% incidence of GIB.

METHODS: This study includes patients receiving CF-LVAD at a quaternary medical center from 2006 to 2014 (n = 254). The primary endpoint was GIB within 12 months after implantation; the secondary outcome was 3-year all-cause mortality. The Student t test or the χ2 test compared continuous or categorical variables. Competing risks analysis calculated the cumulative incidence of GIB postimplantation. Cox proportional hazards model was used for univariate/multivariate models predicting GIB.

RESULTS: Sixty-four patients had GIB, with incidence rates at 1, 3, and 12 months of 11.8%, 19.3%, and 25.2%, respectively. Endoscopy revealed no identified source of bleeding in 41%; 33% of lesions were localized in the upper gastrointestinal tract, with the bulk (39%) categorized as vascular. Patients with prior gastrointestinal abnormalities (n = 98) had a greater risk of GIB post-CF-LVAD (HR 1.85 [1.11-3.09]; P = 0.02) than those with normal gastrointestinal evaluation results (n = 45) and those without preimplantation gastrointestinal evaluation (n = 111). Baseline blood urea nitrogen, chronic obstructive pulmonary disease, and prior percutaneous coronary intervention were statistically associated with post-CF-LVAD GIB. The presence of GIB within 12 months of CF-LVAD implantation was associated with an increased risk of 3-year all-cause mortality (HR 2.57 [1.57-4.15]; P < 0.01).

CONCLUSIONS: First-year GIB is associated with increased mortality post-CF-LVAD. We advocate a closer examination of several GIB risk factors when evaluating CF-LVAD candidates.

Document Type


PubMed ID


Link to Full Text