997 ACUTE AND LATE GASTROINTESTINAL BLEEDING EVENTS AFTER LEFT VENTRICULAR MECHANICAL SUPPORT IMPLANTATION: THE ISMETT REGISTRY

Abstract

Abstract Background The number of patients with end stage heart failure (HF) which requires a ventricular assist devices (VADs) implantation is constantly increasing. Gastrointestinal bleeding (GIB) represents one of the most common complications that occur in this patient population. GIB risk assessment could have important implications for candidate selection and postimplant therapeutic strategies. A recent model, called UTAH bleeding risk score (UBRS), was created with the aim of predicting the risk of GIB during LVAD support, but its use does not yet have a worldwide validation. The two most popular bleeding risk (BR) scores available, the ARC-HBR and the HAS-BLED scores, have been validated respectively in the setting of percutaneous coronary intervention and atrial fibrillation but their use in VAD patients is limited. Aims to describe clinical outcome and gastrointestinal bleeding complications of patients underwent LVAD implantation at a single tertiary referral hospital for heart transplant. Furthermore, to explore possible correlation among clinical, echocardiographic, hemodynamic parameters and the development of gastrointestinal bleeding complications has been evaluated. Finally, to evaluate the predictive value of UBRS, ARC-HBR and HAS-BLED scores in this population and to propose possible new parameters capable of predicting the risk of GIB in VAD patients. Methods Between November 2010 and December 2021, 75 consecutive patients with end stage heart failure who underwent LVAD implantation have been enrolled in this monocentric retrospective registry. Univariable and multivariable Cox regression analysis were performed to identify factors independently associated with GI bleeding in the study population. Results Patient's mean age was 57,8 ± 8,8 years. Thirty-six patients (48%) died at mean time of 579 ± 481 days after implant of VAD. The overall patient survival was 66,6% at two years. To the mean follow-up time of 737 ± 593 days, a total of 44 bleeding (incidence of 58,7%) occurred in 30 patients. Nineteen patients (25%) experienced 21 GIB events. Mean time to first GIB was 469 ± 520 days and only one event occurred within 30 days.The patient survival at two years did not differ significantly between patients who experienced GIB and those who did not (58% vs 69,6%; p=0,3). While, considering the total bleedings, overall mortality was significantly higher in patients who developed any bleeding event, than in those who did not bleed (67 vs 35%; p= 0,008).The clinical characteristics of patient with and without gastrointestinal bleeding were similar with regard to age, INTERMACS profile, etiology of heart failure and VAD implant indication.The subgroup analysis demonstrated that low cardiac output (P< 0,001), right ventricular disfunction (P=0,01) and severe mitral regurgitation (P= 0,01) were all associated with increased risk of GIB. In our population, high risk profile UBRS was shown to be related to a higher percentage of GIB, compared to patients with a low risk profile UBRS, in which patients without GIB prevailed (p< 0,001). Instead, ARC-HBR and HAS-BLED scores, didn't prove to be effective in predicting the GIB risk of VAD patient population. Conclusions High gastrointestinal bleeding rates associated with LVAD implantation do not impact negatively with patient survival. Total bleedings are correlated with higher mortality. Reduced cardiac output, right ventricular disfunction and severe mitral regurgitation are independent GIB risk factors. UBRS was able to stratify risk, in order to predict GIB, in the population tested, while ARC-HBR and HAS-BLED did not.