Neurohormonal regulation and improvement in blood glucose control: reduction of insulin requirement in patients with a nonpulsatile ventricular assist device.

BACKGROUND

Heart failure is associated with prolonged stress and inflammation characterized by elevated levels of cortisol and circulating catecholamines. Persistent sympathetic stimulation secondary to the stress of heart failure causes an induced insulin resistance, which creates a need for higher doses of insulin to adequately manage hyperglycemia in this patient population. We hypothesized that cortisol and catecholamine levels would be elevated in end-stage heart failure patients, however, would be reduced after the implantation of a left ventricular assist device (LVAD). Insulin requirements would therefore be reduced post LVAD implant and control of diabetes improved as compared with pre-implant.

METHODS

Pre- and postoperative cortisol, catecholamine, glycated hemoglobin, and blood glucose levels were evaluated retrospectively in 99 LVAD patients at a single center from January 2007 through November 2011. Serum was collected before LVAD implantation and monthly after implantation for 12 months consecutively. Results were evaluated and compared to insulin requirements, if any, before and after implant. Plasma levels were measured by ELISA.

RESULTS

There were a total of 99 patients (81 men and 18 women). Two patients were implanted twice due to pump dysfunction. Mean age was 59 years, ± 10, with a median of 63 years. Of those patients, 64 had ischemic cardiomyopathy and 35 had dilated cardiomyopathy. The total patient years of LVAD support were 92.5 years. All patients received a continuous flow left ventricular assist device. Type II diabetes mellitus was diagnosed in 28 patients. Of those patients, 24 required daily insulin with an average dose of 45 units/day. Average preoperative glycated hemoglobin (HbA1c) levels were 6.8% with fasting blood glucose measurements of 136 mg/dL. Mean cortisol levels were measured at 24.3 μg/dL before LVAD implantation, with mean plasma catecholamine levels of 1824 μg/mL. Post operatively, average HbA1c levels were 5.38% with fasting blood glucose measurements of 122 mg/dL. Mean cortisol levels were measured at 10.9 μg/dL with average plasma catecholamine levels were 815 μg/mL. There was a significant decrease in both cortisol levels post LVAD implant (P = 0.012) as well as catecholamine levels (P = 0.044). The average insulin requirements post LVAD implant were significantly reduced to 13 units/day (P = 0.001). Six patients no longer required any insulin after implant.

CONCLUSION

Implantation of nonpulsatile LVADs has become a viable option for the treatment of end-stage heart failure, helping to improve patient quality of life by decreasing clinical symptoms associated with poor end-organ perfusion. Frequently, diabetes is a comorbid condition that exists among heart failure patients and with the reduction of the systemic inflammatory and stress response produced by the support of a nonpulsatile LVAD, many patients may benefit from a reduction in their blood glucose levels, as well as insulin requirements.