Evolution of Gastrointestinal Bleeding in Patients Supported with Continuous Flow Left Ventricular Assist Devices

A. Nguyen¹, A. Civitello¹, A. Shafii¹, G. Loor², S. Chatterjee¹, K. Liao^1
1Baylor College of Medicine, Houston, Texas ²Baylor St. Luke's Medical Center, Division of Cardiothoracic Surgery, Houston, Texas

Purpose: Left ventricular assist devices (LVADs) have evolved from pulsatile to continuous flow (CF) pumps. CF-LVADs offer better survival and fewer overall complications, despite a higher incidence of gastrointestinal (GI) bleeding. We aimed to assess the characteristics and risk factors of this adverse event at our institution.

Methods: We used our institutional LVAD database submitted to the Society of Thoracic Surgeons Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) from January 1, 2008 to December 31,2023. All LVAD implantations, both primary and pump exchanges, were included in the cohort. GI bleeding was defined by INTERMACS as an episode of suspected internal or external bleeding that results in one or more of the following: death, reoperation, hospitalization or transfusion of red blood cells. Patient characteristics were reported as frequencies and proportions for categorical variables and as median and interquartile range (IQR) for continuous variables. Multivariable Cox proportional hazards regression was used to determine the risk factors for GI bleeding. Kaplan Meier estimates and Log-Rank Tests were done for GI bleeding by the most significant factors. Analyses were performed on Stata version 18.0 (StataCorp LLC, College Station, TX, USA). A p-value of < 0.05 was considered statistically significant.

Results: A total of 1030 CF-LVADs were implanted. The median age was 58.5 years (IQR 50.4-65.4), and 818 patients (79.4%) were male. The majority (80.6%) presented with pre-implant INTERMACS profile ≤3 (high acuity). Axial-flow LVADs were used in 569 (55.2%) patients, partial magnetically levitated centrifugal-flow LVADs in 316 (30.7%) patients, and full magnetically levitated centrifugal-flow LVADs in 145 (14.1%) patients. GI bleeding was reported in 386 cases (37.5% or 13.9 per 100 person-years), with 65.3% being rebleeding. Bleeding sources were as follows: 210 (54.4%) upper GI bleeding, 84 (21.8%) lower GI bleeding and 92 (23.8%) unknown source positive stool. The median time from LVAD implant to bleeding was 11 months (IQR 2-31). Blood transfusion, rehospitalization, surgery, and death resulting from bleeding occurred in 89.9%, 52.3%, 4.9%, and 0.82% of cases, respectively. Patients aged >50 were three times more likely to develop GI bleeding compared to those ≤50. Coronary Artery Disease and Chronic Kidney Disease were significant risk factors for GI bleeding, while magnetically levitated centrifugal-flow LVADs reduced GI bleeding risk by four times compared to axial-flow LVADs. GI bleeding risk was not significantly different by sex, race, INTERMACS profiles, RV dysfunction, diabetes, pulmonary hypertension, or liver dysfunction.

Conclusion: GI bleeding was a common adverse event that required special care for patients with CF-LVADs. Significant risk factors for GI bleeding include older age, Coronary Artery Disease, Chronic Kidney Disease and LVAD types. These GI bleeding risk facts should be weighed when making LVAD decisions especially for destination therapy.

Identify the source of the funding for this research project: None