ASAIO Journal 2014
Clinical Cardiovascular
Improvement in Glycemic Control After Left Ventricular Assist Device Implantation in Advanced Heart Failure Patients With Diabetes Mellitus Naila Choudhary,* Leway Chen,* Lisa Kotyra,* Steven D. Wittlin,† and Jeffrey D. Alexis*
Left ventricular assist devices (LVADs) have been shown to improve outcomes in advanced heart failure (HF). We hypothesized that LVADs improve glycemic control in HF patients with diabetes mellitus (DM). During a 6 year time period, 202 patients underwent mechanical circulatory support. Of these, 50 patients with DM were included. Data were collected within 2 months before LVAD implantation and at 5.6 ± 1.1 months post-LVAD implant. There was no significant difference in body mass index, hemoglobin, hematocrit, and renal function pre-LVAD and post-LVAD. Fasting blood glucose improved from 136 ± 35 to 108 ± 29 mg/dl post-LVAD (p < 0.001). In 18 patients taking insulin only, daily insulin dose decreased from 43 ± 37 to 29 ± 24 units (p = 0.02). Of the 17 patients taking oral hypoglycemic agents, four did not require antidiabetic medications, six continued the same dose, two required higher doses, and five patients were switched to insulin post-LVAD. In a subset of 22 patients with available data, hemoglobin A1c improved significantly post-LVAD (p 125 mg/dl, and 6) hemoglobin A1c (HbA1c) >6.5%. Patients were excluded if they died during LVAD implantation hospitalization, were diagnosed with DM at the time of LVAD implantation, or received only right ventricular assist devices or only extracorporeal membrane oxygenation. Baseline characteristics, laboratory data, medication usage, and LVAD data were collected using INTERMACS (Interagency Registry for Mechanically Assisted Circulatory Support) registry and review of electronic medical records. Laboratory data and data on the use of antidiabetic medications were collected within 2 months before LVAD implantation and up to 6 months post-LVAD implantation (5.6 ± 1.1 months post-LVAD implantation). In patients with multiple available FBG laboratory data pre- and post-LVAD, the level that was more consistent with other recordings was selected. Measurements for HbA1c were made using highperformance liquid chromatography. C-reactive protein (CRP) levels were measured using immunoturbidimetric method.
Key Words: glycemic improvement, heart failure, left ventricular assist devices
The relationship between diabetes mellitus (DM) and con-
gestive heart failure (HF) is closely linked and seemingly bidirectional. Diabetes mellitus is a well-known and independent risk factor for congestive HF.1 At the same time, congestive HF has also been identified as a risk factor for development and worsening of insulin resistance (IR).2,3 Insulin resistance is an independent risk factor for mortality in HF patients and is also associated with the severity of HF.4 Several mechanisms have been proposed for IR in the setting of HF, including poor From the *Division of Cardiology, Department of Medicine, University of Rochester School of Medicine and Dentistry, Rochester, New York and †Division of Endocrinology and Metabolism, Department of Medicine, University of Rochester School of Medicine and Dentistry, Rochester, New York. Submitted for consideration March 2014; accepted for publication in revised form July 2014. Disclosure: The authors have no conflict of interest to disclose. Correspondence: Jeffrey D. Alexis, MD, University of Rochester School of Medicine and Dentistry, 601 Elmwood Ave., Box 679, Rochester, NY 14642. E-mail: [email protected]. Copyright © 2014 by the American Society for Artificial Internal Organs
Statistical Analysis Continuous variables were compared using paired sample t-test and were presented as mean ± standard deviation. Categorical variables were expressed as number of patients and percentages and analyzed using Pearson Chi-squared test or Fisher exact test. The statistical analysis was performed using the Statistical Program of Social Sciences (IBM SPSS statistics,
DOI: 10.1097/MAT.0000000000000127
675
676 CHOUDHARY et al. version 21.0, IBM Corp., Armonk, NY). All p values were two tailed and were considered to be significant when 125 mg/dl (n=1) • HbA1c > 6.5% (n= 4)
Excluded (n=11) • Death during LVAD implant hospitalization (n=6) • Diagnosis of diabetes at the time of LVAD implant (n= 2) • Implant of RVAD only or ECMO only (n=3)
50 Patients
Figure 1. Inclusion and exclusion criteria. DM, diabetes mellitus; ECMO, extracorporeal membrane oxygenation; FBG, fasting blood glucose; HbA1c, hemoglobin A1c; MCS, mechanical circulatory support; OHA, oral hypoglycemic agent.
Table 1. Baseline Characteristics of Study Population (n = 50) Age at the time of LVAD (years) Men Race African American Caucasian Heart failure etiology Ischemic Nonischemic Valvular Arrhythmic Congenital Hypertrophic cardiomyopathy LVAD type Continuous-flow HeartMate II Continuous-flow VentrAssist Pulsatile-flow paracorporeal BiVAD
56 ± 10.9 42 (84%) 8 (16%) 42 (84%) 31 (62%) 15 (30%) 1 (2%) 1 (2%) 1 (2%) 1 (2%) 48 (96%) 1 (2%) 1 (2%)
antagonists, diuretics, antiarrhythmic agents, nitrates, and statins pre-LVAD and post-LVAD. There was less use of angiotensinconverting enzyme inhibitors or angiotensin receptor blockers (ACEIs/ARBs) after LVAD implantation (22% vs. 64%; p < 0.001). Calcium channel blockers were used more frequently after LVAD implantation (2% vs. 22%; p = 0.004). Amlodipine was the most commonly used calcium channel blocker. Clinical and Metabolic Characteristics Several clinical and laboratory characteristics were compared pre-LVAD and at 5.6 ± 1.1 months post-LVAD implantation (Table 3). There was no significant change in weight and body mass index (BMI) post-LVAD in comparison with pre-LVAD. Fasting blood glucose improved significantly from 136 ± 35 mg/dl pre-LVAD to 108 ± 29 mg/dl post-LVAD (p < 0.001). There was no significant change in hepatic aminotransferases, total bilirubin, total protein, and albumin pre-LVAD and post-LVAD. Hemoglobin and hematocrit did not change significantly after LVAD implantation. There was no difference in glomerular filtration rate pre- and post-LVAD implantation (67 ± 27 ml/min/1.73 m2 vs. 68 ± 34 ml/min/1.73m2; p = 0.83). Diabetes Management Of the 50 study patients, 17 were taking OHA, 18 were managed with subcutaneous insulin, five were on both OHA and insulin, and 10 patients were not on any medications for management of DM pre-LVAD. Table 2. Use of Cardiac Medications in the Study Population Before and After LVAD Medications
Pre-LVAD (n = 50)
Post-LVAD (n = 50)
p Value
Beta-blockers ACEI/ARB Aldosterone antagonist Diuretics Antiarrhythmic agents Nitrates CCB Statin
43 (86%) 32 (64%) 29 (58%) 42 (84%) 12 (24%) 9 (18%) 1 (2%) 31 (62%)
43 (86%) 11 (22%) 26 (52%) 42 (84%) 18 (36%) 13 (26%) 11 (22%) 23 (46%)
NS
Clinical Cardiovascular
Improvement in Glycemic Control After Left Ventricular Assist Device Implantation in Advanced Heart Failure Patients With Diabetes Mellitus Naila Choudhary,* Leway Chen,* Lisa Kotyra,* Steven D. Wittlin,† and Jeffrey D. Alexis*
Left ventricular assist devices (LVADs) have been shown to improve outcomes in advanced heart failure (HF). We hypothesized that LVADs improve glycemic control in HF patients with diabetes mellitus (DM). During a 6 year time period, 202 patients underwent mechanical circulatory support. Of these, 50 patients with DM were included. Data were collected within 2 months before LVAD implantation and at 5.6 ± 1.1 months post-LVAD implant. There was no significant difference in body mass index, hemoglobin, hematocrit, and renal function pre-LVAD and post-LVAD. Fasting blood glucose improved from 136 ± 35 to 108 ± 29 mg/dl post-LVAD (p < 0.001). In 18 patients taking insulin only, daily insulin dose decreased from 43 ± 37 to 29 ± 24 units (p = 0.02). Of the 17 patients taking oral hypoglycemic agents, four did not require antidiabetic medications, six continued the same dose, two required higher doses, and five patients were switched to insulin post-LVAD. In a subset of 22 patients with available data, hemoglobin A1c improved significantly post-LVAD (p 125 mg/dl, and 6) hemoglobin A1c (HbA1c) >6.5%. Patients were excluded if they died during LVAD implantation hospitalization, were diagnosed with DM at the time of LVAD implantation, or received only right ventricular assist devices or only extracorporeal membrane oxygenation. Baseline characteristics, laboratory data, medication usage, and LVAD data were collected using INTERMACS (Interagency Registry for Mechanically Assisted Circulatory Support) registry and review of electronic medical records. Laboratory data and data on the use of antidiabetic medications were collected within 2 months before LVAD implantation and up to 6 months post-LVAD implantation (5.6 ± 1.1 months post-LVAD implantation). In patients with multiple available FBG laboratory data pre- and post-LVAD, the level that was more consistent with other recordings was selected. Measurements for HbA1c were made using highperformance liquid chromatography. C-reactive protein (CRP) levels were measured using immunoturbidimetric method.
Key Words: glycemic improvement, heart failure, left ventricular assist devices
The relationship between diabetes mellitus (DM) and con-
gestive heart failure (HF) is closely linked and seemingly bidirectional. Diabetes mellitus is a well-known and independent risk factor for congestive HF.1 At the same time, congestive HF has also been identified as a risk factor for development and worsening of insulin resistance (IR).2,3 Insulin resistance is an independent risk factor for mortality in HF patients and is also associated with the severity of HF.4 Several mechanisms have been proposed for IR in the setting of HF, including poor From the *Division of Cardiology, Department of Medicine, University of Rochester School of Medicine and Dentistry, Rochester, New York and †Division of Endocrinology and Metabolism, Department of Medicine, University of Rochester School of Medicine and Dentistry, Rochester, New York. Submitted for consideration March 2014; accepted for publication in revised form July 2014. Disclosure: The authors have no conflict of interest to disclose. Correspondence: Jeffrey D. Alexis, MD, University of Rochester School of Medicine and Dentistry, 601 Elmwood Ave., Box 679, Rochester, NY 14642. E-mail: [email protected]. Copyright © 2014 by the American Society for Artificial Internal Organs
Statistical Analysis Continuous variables were compared using paired sample t-test and were presented as mean ± standard deviation. Categorical variables were expressed as number of patients and percentages and analyzed using Pearson Chi-squared test or Fisher exact test. The statistical analysis was performed using the Statistical Program of Social Sciences (IBM SPSS statistics,
DOI: 10.1097/MAT.0000000000000127
675
676 CHOUDHARY et al. version 21.0, IBM Corp., Armonk, NY). All p values were two tailed and were considered to be significant when 125 mg/dl (n=1) • HbA1c > 6.5% (n= 4)
Excluded (n=11) • Death during LVAD implant hospitalization (n=6) • Diagnosis of diabetes at the time of LVAD implant (n= 2) • Implant of RVAD only or ECMO only (n=3)
50 Patients
Figure 1. Inclusion and exclusion criteria. DM, diabetes mellitus; ECMO, extracorporeal membrane oxygenation; FBG, fasting blood glucose; HbA1c, hemoglobin A1c; MCS, mechanical circulatory support; OHA, oral hypoglycemic agent.
Table 1. Baseline Characteristics of Study Population (n = 50) Age at the time of LVAD (years) Men Race African American Caucasian Heart failure etiology Ischemic Nonischemic Valvular Arrhythmic Congenital Hypertrophic cardiomyopathy LVAD type Continuous-flow HeartMate II Continuous-flow VentrAssist Pulsatile-flow paracorporeal BiVAD
56 ± 10.9 42 (84%) 8 (16%) 42 (84%) 31 (62%) 15 (30%) 1 (2%) 1 (2%) 1 (2%) 1 (2%) 48 (96%) 1 (2%) 1 (2%)
antagonists, diuretics, antiarrhythmic agents, nitrates, and statins pre-LVAD and post-LVAD. There was less use of angiotensinconverting enzyme inhibitors or angiotensin receptor blockers (ACEIs/ARBs) after LVAD implantation (22% vs. 64%; p < 0.001). Calcium channel blockers were used more frequently after LVAD implantation (2% vs. 22%; p = 0.004). Amlodipine was the most commonly used calcium channel blocker. Clinical and Metabolic Characteristics Several clinical and laboratory characteristics were compared pre-LVAD and at 5.6 ± 1.1 months post-LVAD implantation (Table 3). There was no significant change in weight and body mass index (BMI) post-LVAD in comparison with pre-LVAD. Fasting blood glucose improved significantly from 136 ± 35 mg/dl pre-LVAD to 108 ± 29 mg/dl post-LVAD (p < 0.001). There was no significant change in hepatic aminotransferases, total bilirubin, total protein, and albumin pre-LVAD and post-LVAD. Hemoglobin and hematocrit did not change significantly after LVAD implantation. There was no difference in glomerular filtration rate pre- and post-LVAD implantation (67 ± 27 ml/min/1.73 m2 vs. 68 ± 34 ml/min/1.73m2; p = 0.83). Diabetes Management Of the 50 study patients, 17 were taking OHA, 18 were managed with subcutaneous insulin, five were on both OHA and insulin, and 10 patients were not on any medications for management of DM pre-LVAD. Table 2. Use of Cardiac Medications in the Study Population Before and After LVAD Medications
Pre-LVAD (n = 50)
Post-LVAD (n = 50)
p Value
Beta-blockers ACEI/ARB Aldosterone antagonist Diuretics Antiarrhythmic agents Nitrates CCB Statin
43 (86%) 32 (64%) 29 (58%) 42 (84%) 12 (24%) 9 (18%) 1 (2%) 31 (62%)
43 (86%) 11 (22%) 26 (52%) 42 (84%) 18 (36%) 13 (26%) 11 (22%) 23 (46%)
NS
