Vascular Reactivity Analysis in Patients with Continuous Flow Left Ventricular Assist Devices (CF-LVADs) - the Role of Endothelial Function in Continuous Flow Physiology

Abstract

Endothelial dysfunction associated with heart failure should improve after a CF-LVAD placement due to correction of the heart failure environment. Yet, a loss of pulsatility could counteract such improvement as endothelial cells need a pulse pressure for normal function. Mixed results of endothelial dysfunction in CF-LVAD when compared to pulsatile VAD have been reported utilizing both brachial reactivity and EndoPAT® machines. Both these tests rely on physical pulse volume and might be impacted by the mere loss of pulsatility. Hence, we set out to analyze endothelial function utilizing thermal digital monitoring (VENDYS) in patients with CF-LVADs. We recruited stable patients during their routine visits and performed the VENDYS test in a soothing environment. Aortic valve opening was determined with a portable echo (GE V-scan) and graded as per previously described classification. Rest of the data was collected from the patient charts. Patients were classified into three groups based on parameters of vascular reactivity index (VRI) described in normal population: poor (<1) intermediate (1 - 2) and good VRI (2 - 3) with higher VRI correlated with better endothelial function, hence better outcomes. Pearsons correlation and T-test were performed, P-values < 0.05 were considered significant. 40 patients had Heartmate II and 2 had Heartware devices. Table 1 describes the characteristics of the 3 cohorts of endothelial function. Overall 81% had aortic valve opening consistently at the time of endothelial function assessment. Only 19% had a poor VRI, while majority had an intermediate VRI. Older patients had worse endothelial function while rest of the baseline characteristics including aortic valve opening and flow were not different between the groups. Correlation between VRI and VAD flow (r= -0.2766 P value= 0.08) and high Doppler blood pressure (r= -0.2656 P value=0.09) showed no significance. However, 50% of the patients with a poor VRI had adverse events while only 23% of those with good endothelial function did (Table 1). The 2 patients with poor VRI who had a GI bleed had 5 episodes while the 3 patients with good VRI had 4 episodes. In a cohort of CF-LVAD patients where the majority had their aortic valve open, endothelial function, as reflected by DTM was poor in only a few. It is not clear if endothelial dysfunction contributes to any of adverse events for CF-LVAD patients but there seem to be a higher recurrence of GI bleeding in patients who had a poor endothelial function.Table 1Total<11 to 22 to 3p valueN42 (100%)8 (19%)21 (50%)13 (31%)VENDYS score1.71±0.780.45±0.351.69±0.282.5±0.35Age57±12.860.3±7.559.6±10.150.2±16.70.073Male34 (81%)5 (62.5%)19 (90%)10 (77%)0.2082Caucasian17(40%)2 (25%)9 (43%)6 (46%)0.6009DT31 (74%)6 (75%)17 (81%)8 (61%)0.45AV opened34 (81%)7 (87%)17 (81%)10 (77%)0.835Days under support585±567244±276672±586497±5820.17BUN275.47±282.425.25±16.4724.8±7.2320.6±7.540.41Hgb11.06±2.229.77±1.4711.56±2.4511.04±2.010.15Flow5.49±1.346.4±1.475.21±1.515.37±0.510.098Patients with events17(100%)4 (23%)10 (59%)3 (18%)0.3GIB10 (58%)2530.99Infections4 (23%)1300.36Arrhythmia1 (6%)0100.59ADHF3 (18%)0300.19Death3 (14%)1200.46VENDYS score: Endothelial function based thermal vascular reactivity. DT: Destination Therapy. AV: Aortic valve. GIB: Gastrointestinal bleeding. ADHF: Acute decompensated heart failure. Open table in a new tab