Association of Infections in Patients Supported with Left Ventricular Assist Device with Vasoplegia and Post-Cardiac Transplant Outcomes

BackgroundLVAD-specific infections (LSI) are common in patients on LVAD support. The impact of LSI on post-heart transplant (HT) outcomes is not clear. We hypothesized that LSI are associated with higher rates of acute renal failure (ARF), rejection and mortality in the first year post-HT with perioperative vasoplegia serving as a mediating factor.MethodsA retrospective analysis at our institution was conducted for all LVAD patients from 10/2012 to 9/2017 who received HT. Total artificial heart patients were excluded. LSI was defined as driveline infection, pump pocket, pump or cannula infection. Vasoplegia was considered present if the following criteria were met: epinephrine/norepinephrine ≥ 150 ng/kg/min, dopamine ≥ 10 ug/kg/min or vasopressin ≥ 4 U/hr at any time point or duration of inotropic support > 14 days with a cardiac index > 2.1 L/min/m2 and systemic vascular resistance less than < 800 dyn⋅s⋅cm−5 without hemodynamic right heart failure (CVP > 15 mmHg), or any use of methylene blue or vitamin B12 complex. The primary outcome was a composite of ARF, rejection (antibody-mediated or acute cellular rejection ≥ 2R or hemodynamic compromise), and mortality in the first year post-HT.ResultsA total of 105 patients were bridged to HT with LVAD therapy. Of these, 63 (60%) had evidence of an LSI pre-HT with a majority (92%) being driveline-associated. These patients were more likely to be men (85% vs 69%), have higher BSA (2.05 m2 vs 1.89 m2), and be UNOS 1A (97% vs 64%), but no differences were noted in BMI, etiology of cardiomyopathy, device type, duration on LVAD support, or medical therapy. Among patients with LSI, there was an equal distribution of gram positive and negative organisms identified and 56% were on oral antibiotics vs 41% on IV antibiotics. The incidence of vasoplegia in the overall cohort was 12% (13/105), with 16% (10/63) in the LSI group vs 7% (3/42) in the no LSI group (p=0.183). Over half of the patients (56/105) experienced a composite of ARF, rejection or death post HT, 59% (37/63) in the LSI group vs 45% (19/42) in the no LSI group (p=0.175). Similarly, there was no difference in the individual components of the composite outcomes or in the secondary outcome hospital length of stay.ConclusionsIn our cohort, LSI were common among LVAD patients but not associated with vasoplegia or the clinical outcome of ARF, graft rejection, or mortality in the first year post HT. This may be limited by a small sample size. LVAD-specific infections (LSI) are common in patients on LVAD support. The impact of LSI on post-heart transplant (HT) outcomes is not clear. We hypothesized that LSI are associated with higher rates of acute renal failure (ARF), rejection and mortality in the first year post-HT with perioperative vasoplegia serving as a mediating factor. A retrospective analysis at our institution was conducted for all LVAD patients from 10/2012 to 9/2017 who received HT. Total artificial heart patients were excluded. LSI was defined as driveline infection, pump pocket, pump or cannula infection. Vasoplegia was considered present if the following criteria were met: epinephrine/norepinephrine ≥ 150 ng/kg/min, dopamine ≥ 10 ug/kg/min or vasopressin ≥ 4 U/hr at any time point or duration of inotropic support > 14 days with a cardiac index > 2.1 L/min/m2 and systemic vascular resistance less than < 800 dyn⋅s⋅cm−5 without hemodynamic right heart failure (CVP > 15 mmHg), or any use of methylene blue or vitamin B12 complex. The primary outcome was a composite of ARF, rejection (antibody-mediated or acute cellular rejection ≥ 2R or hemodynamic compromise), and mortality in the first year post-HT. A total of 105 patients were bridged to HT with LVAD therapy. Of these, 63 (60%) had evidence of an LSI pre-HT with a majority (92%) being driveline-associated. These patients were more likely to be men (85% vs 69%), have higher BSA (2.05 m2 vs 1.89 m2), and be UNOS 1A (97% vs 64%), but no differences were noted in BMI, etiology of cardiomyopathy, device type, duration on LVAD support, or medical therapy. Among patients with LSI, there was an equal distribution of gram positive and negative organisms identified and 56% were on oral antibiotics vs 41% on IV antibiotics. The incidence of vasoplegia in the overall cohort was 12% (13/105), with 16% (10/63) in the LSI group vs 7% (3/42) in the no LSI group (p=0.183). Over half of the patients (56/105) experienced a composite of ARF, rejection or death post HT, 59% (37/63) in the LSI group vs 45% (19/42) in the no LSI group (p=0.175). Similarly, there was no difference in the individual components of the composite outcomes or in the secondary outcome hospital length of stay. In our cohort, LSI were common among LVAD patients but not associated with vasoplegia or the clinical outcome of ARF, graft rejection, or mortality in the first year post HT. This may be limited by a small sample size.