CASE ANECDOTES, COMMENTS AND OPINIONS A novel link between G6PD deciency and hemolysis in patients with continuous-ow left ventricular assist devices Hassan Alhosaini, MD, a,b Brian C. Jensen, MD, b,c Patricia P. Chang, MD, MHS, b Brett C. Sheridan, MD, d and Jason N. Katz, MD, MHS b From the a Department of Cardiovascular Sciences, East Carolina Heart Institute, Greenville, North Carolina; b Division of Cardiology, University of North Carolina School of Medicine, Chapel Hill, North Carolina; c UNC McAllister Heart Institute, Chapel Hill, North Carolina; and the d Department of Surgery, University of North Carolina School of Medicine, Chapel Hill, North Carolina Hemolysis is a known complication of durable continuous- ow left ventricular assist devices (CF-LVADs), 1 and is thought to result from shear force exerted on circulating red blood cells (RBCs). This mechanical injury may be inuenced by a number of factors, including device positioning and pump thrombosis, although the pathophysiology of LVAD- related hemolysis remains poorly understood. We reviewed the CF-LVAD experience at our institution and report on a novel association between glucose-6-phosphate dehydrogen- ase (G6PD) deciency and clinically signicant hemolysis. G6PD is a ubiquitous cytosolic enzyme that catalyzes the reduction of nicotinamide adenine dinucleotide phosphate (NADP) to NADPH in the hexose monophosphate shunt. The hexose monophosphate shunt is the only source of NADPH in RBCs, and NADPH is critical in protecting RBCs against injury by reactive oxygen species. 2 Individuals with G6PD deciency are prone to hemolysis in the setting of increased oxidative stress due to inadequate levels of NADPH. We reviewed the records of 85 patients consecutively implanted with a CF-LVAD at the University of North Carolina (UNC) Hospitals from 2005 to 2011 and included all 34 patients with available G6PD levels in our analysis. We began systematically drawing pre-operative G6PD levels on all CF- LVAD patients after an individual with known G6PD deciency was found to have a hemolytic event while on mechanical support. All patients were followed for at least 12 months after LVAD, or until heart transplantation or death. A hemolysis event was dened by the presence of characteristic clinical and laboratory parameters (elevated lactate dehydro- genase [LDH] Z2.5 times upper limit of normal, decreased haptoglobin, elevated indirect bilirubin, anemia, hemoglobinur- ia) and documented as such in the medical record by the treating physician and/or consulting hematologist at the time of the encounter. Seven of the 34 patients were G6PD decient. Those with G6PD deciency were more likely to to be black and have non-ischemic cardiomyopathy, otherwise demographic and clinical characteristics were similar between G6PD-decient and normal patients. The median G6PD level in those with G6PD deciency was 3.0 (interquartile range [IQR] 1.5 to 5.3) IU/g hemoglobin (Hb), whereas the median level in patients without G6PD deciency was 7.3 (IQR 6.6 to 7.8) (p o 0.001). Use of aspirin and warfarin was similar in the two groups (Table 1). Twenty-nine percent of all patients (n ¼ 10) were diagnosed with a hemolysis event during the follow-up period. There were no statistically signicant differences in age, gender, race, anti-coagulant use or indication for LVAD implantation between patients who developed hemolysis and those who did not. LVAD pump thrombosis was conrmed in 2 of the 10 patients with hemolysis. Patients with G6PD deciency had a greater risk of experiencing a hemolysis event than those with normal G6PD levels (71% vs 19%) (odds ratio [OR] 11.0, 95% CI 1.6 to 74.0; p ¼ 0.01). Median peak serum LDH level during follow-up was higher in G6PD-decient patients (5,922 U/liter, IQR 5,369 to 20,735 U/liter) than in patients with normal G6PD levels (1,239 U/liter, IQR 999 to 2,018 U/liter) (p ¼ 0.001) (Figure 1). Median peak serum bilirubin level also was signicantly higher in G6PD-decient patients (3.4 mg/dl, IQR 1.5 to 4.9 mg/dl) than in those with normal G6PD levels (1.1 mg/dl, IQR 0.9 to 1.6 mg/dl) (p ¼ 0.02). Seventy-one percent (n ¼ 5) of patients with G6PD deciency were subsequently hospitalized for hemolysis, compared with 11% (n ¼ 3) of normal patients (OR 20.0, 95% CI 2.6 to 153.0) (p ¼ 0.004) (Table 1). The link between G6PD deciency and hemolysis in the setting of mechanical circulatory support (MCS) is not entirely novel, 3 and we hypothesize that hemolysis arose, at least in part, from an unfavorable oxidative environment after LVAD implantation rather than due to shear stress alone. Indeed, peripheral blood smear ndings (Heinz bodies and bite cells) in many of these patients conrmed oxidative hemolysis rather than mechanical destruction; such ndings were absent in patients with hemolysis and normal G6PD levels. Oxidative stress and endothelial dysfunction are central to the pathobiology of heart failure, and redox state is not improved by CF-LVAD support. 4 Continuous ow aug- ments oxidative stress in endothelial cells and is associated with endothelial activation, dysfunction and decreased nitric oxide production. 5 Patients with G6PD deciency may be particularly prone to these adverse endothelial effects, as NADPH deciency can cause uncoupling of endothelial 1053-2498/$ - see front matter r 2014 International Society for Heart and Lung Transplantation. All rights reserved. http://dx.doi.org/10.1016/j.healun.2013.10.028