P1519 - Diagnostic and Therapeutic Yield of Early Push Enteroscopy in LVAD-Associated GI Bleeding

Thomas J. Konturek, MD¹, Andrew Choi, MD¹, Joshua Newman, MD¹, Mary E. Moore, MPH, MS², Max Liebo, MD¹, Abdul Haseeb, MD, MPH^1
1Loyola University Medical Center, Maywood, IL; ²Loyola University Chicago, Maywood, IL

Introduction: Left ventricular assist devices (LVADs) are the standard of care in advanced heart failure patients either as a bridge to cardiac transplantation or as destination therapy. Angiodysplasias predominantly located in the upper and mid GI tract account for 40% of GI bleeding (GIB) events in LVAD patients. Few small single center studies suggest that early enteroscopy in LVAD associated GIB resulted in increased diagnostic and therapeutic yields. In this study, our aim was to evaluate the diagnostic and therapeutic yield of push enteroscopy (PE) when compared to an EGD as the first endoscopic modality in LVAD associated GIB.

Methods: Subjects were identified from a prospectively compiled registry of 275 patients who underwent LVAD implantation at Loyola University Medical Center between 5/2008 and 2/2020. The study received institutional IRB approval. Patients older than 18 years of age on LVAD support admitted with GIB were included. Chi-square tests for independence were used to evaluate associations between categorical variables and timing of recurrence. Wilcoxon-Mann-Whitney tests were used to assess differences between patients who underwent PE versus no PE as initial modality.





Results: 67 patients with 204 GIB admissions while on LVAD support were included. Mean age was 66.7±10.48 years, 76% were male and 78% Caucasian. GIB sources were most frequently encountered in the stomach and proximal small bowel (Figure 1A). Angiodysplasias represented the most common etiology of GIB (Figure 1B). Out of the 204 admissions, 112 patients underwent either EGD or PE as the initial endoscopic modality. 94 (83.9%) of these patients presented with melena. Compared to EGD, PE did not result in increased diagnostic (60.42% vs. 60.94%, p=0.96) or therapeutic yield (41.67% vs. 54.69%, p=0.15). When comparing diagnostic and therapeutic yields of any initial combination of endoscopic procedures that included PE (n=98) with procedures that did not include PE (n=106), still no statistically significant difference could be found (63.27% vs. 61.32%, p=0.8; 44.9% vs. 39.6%, p=0.68).

Discussion: Despite previous studies suggesting a role of PE as initial modality in identifying and treating culprit lesions in LVAD associated GIB, our data supports that an initial approach with EGD results in similar diagnostic and therapeutic yields as most lesions are within reach with a standard endoscope. Choosing this approach did not significantly impact the 30- and 90-day readmission rates due to GIB.


Figure: Location (A) and etiology (B) of 204 LVAD related GI bleeding admissions

Table: First endoscopic procedure EGD vs. PE in LVAD related GI bleeding


Disclosures:
Thomas Konturek indicated no relevant financial relationships.
Andrew Choi indicated no relevant financial relationships.
Joshua Newman indicated no relevant financial relationships.
Mary Moore indicated no relevant financial relationships.
Max Liebo indicated no relevant financial relationships.
Abdul Haseeb indicated no relevant financial relationships.


Thomas J. Konturek, MD¹, Andrew Choi, MD¹, Joshua Newman, MD¹, Mary E. Moore, MPH, MS², Max Liebo, MD¹, Abdul Haseeb, MD, MPH¹. P1519 - Diagnostic and Therapeutic Yield of Early Push Enteroscopy in LVAD-Associated GI Bleeding, ACG 2021 Annual Scientific Meeting Abstracts. Las Vegas, Nevada: American College of Gastroenterology.