Diabetes Increases Risk of Mortality in Heart Failure Patients Who Undergo Left Ventricular Assist Device Implantation

Presentation Number: SAT 523
Date of Presentation: April 1st, 2017

Chinenye Usoh*1, Saadia Sherazi2, Barbara Szepietowska2, Valentina Kutyifa2, Scott Mcnitt2, Anna Papernov2, Meng Wang2, Stephen R Hammes3 and Jeffrey Alexis2
1University of Rochester School of Medicine and Dentistry, Rochester, NY, 2University of Rochester Medical Center, 3University of Rochester Medical Center, Rochester, NY


Prior studies have shown that patients with diabetes have worse surgical outcomes when compared to non-diabetics. Despite previous studies, the mortality risk of diabetic patients post left ventricular assist device (LVAD) implant, remains unclear. In addition, the relationship between the degree of glycemic control and long-term mortality risk in LVAD patients with diabetes has not been established.

We hypothesized that diabetic LVAD patients would have a higher mortality rate than non-diabetic LVAD patients; and amongst LVAD diabetic patients, mortality would increase with worse diabetes control (defined as higher hemoglobin A1c).

Ninety-five non-diabetic and ninety-six diabetic patients from the University of Rochester Medical Center, who received a HeartMate II continuous-flow LVAD between August 26th, 2007 and June 30th, 2014 were included in this study. Diabetics were defined as having a diagnosis of diabetes in medical records, or hemoglobin A1c greater than or equal to 6.5%, or random glucose greater than 200 mg/dL on more than one occasion prior to LVAD implantation. The primary outcome was all-cause mortality. Secondary outcomes included rates of infection, neurological dysfunction, renal dysfunction, and re-hospitalization. Kaplan-Meier cumulative probabilities of long-term all-cause mortality were assessed by diabetes and by the degree of glycemic control.

During follow-up, 32 (33%) diabetics and 15 (16%) non-diabetics died following LVAD implantation (p=0.005). Cumulative probability of death was higher in diabetics when compared to non-diabetics (42% vs. 21% at 3 years, p=0.008). There was no difference in overall rates of infection, neurological dysfunction, re-hospitalization, or renal dysfunction between the two groups. However, after initial outcome event, diabetics had a higher mortality rate when compared to non-diabetics. There was no statistically significant difference in cumulative probability of death between diabetics with pre-LVAD hemoglobin A1c < 7.5% and diabetics with pre-LVAD hemoglobin A1c ≥ 7.5% (p=0.198).

Diabetics who undergo LVAD implantation have a higher probability of death compared to non-diabetic patients. Overall rates of selected outcomes did not differ, but diabetics had increased mortality after initial event which could be contributing to the higher total mortality. Finally, the degree of glycemic control in diabetics prior to LVAD did not influence mortality.


Nothing to Disclose: CU, SS, BS, VK, SM, AP, MW, SRH, JA