Impaired Insulin Sensitivity as an Underlying Mechanism Linking Hepatitis C and Posttransplant Diabetes Mellitus in Kidney Recipients

Authors


* Corresponding author: Seema Baid-Agrawal, seema.baid-agrawal@charite.de

Abstract

Aim of this study was to investigate the mechanism/s associating hepatitis C virus (HCV) infection and posttransplant diabetes mellitus in kidney recipients.

Twenty HCV-positive and 22 HCV-negative kidney recipients, 14 HCV-positive nontransplant patients and 24 HCV-negative nontransplant (healthy) subjects were analyzed. A 3-h intravenous glucose tolerance test was performed; peripheral insulin sensitivity was assessed by minimal modeling. Pancreatic insulin secretion, hepatic insulin uptake, pancreatic antibodies and proinflammatory cytokines in serum (tumor necrosis factor-α, intereukin-6, high-sensitive C-reactive protein) were also assessed.

HCV-positive recipients showed a significantly lower insulin sensitivity as compared to HCV-negative recipients (3.0 ± 2.1 vs. 4.9 ± 3.0 min−1.μU.mL− 1.104, p = 0.02), however, insulin secretion and hepatic insulin uptake were not significantly different. Pancreatic antibodies were negative in all. HCV status was an independent predictor of impaired insulin sensitivity (multivariate analysis, p = 0.008). The decrease of insulin sensitivity due to HCV was comparable for transplant and non-transplant subjects. No significant correlation was found between any of the cytokines and insulin sensitivity.

Our results suggest that impaired peripheral insulin sensitivity is associated with HCV infection irrespective of the transplant status, and is the most likely pathogenic mechanism involved in the development of type 2 diabetes mellitus associated with HCV infection.

Introduction

Chronic hepatitis C virus (HCV) infection is a multifaceted disease, which is associated with numerous extrahepatic manifestations, of which essential mixed cryoglobulinemia with or without membranoproliferative glomerulonephritis, and porphyria cutanea tarda are well known (1). Recent epidemiological studies have added another clinical condition, type 2 diabetes mellitus (DM), to the spectrum of HCV-associated diseases. In 1994, Allison et al. first reported an increased prevalence of DM in patients with HCV-associated liver cirrhosis as compared to patients with cirrhosis due to hepatitis B virus, alcohol, cholestatic liver disease or autoimmune hepatitis (2). Subsequently, various reports have confirmed the association between DM and HCV infection, even in the absence of cirrhosis (3–9). In these studies, the prevalence of DM in patients with liver disease secondary to HCV infection has been reported to vary from 20% to 50%, in contrast to the 2.5% to 25% prevalence found in patients with non-HCV-related liver disease.

Similar to the general population, several reports have also suggested an association between HCV infection and new onset posttransplant diabetes mellitus (PTDM) after liver and kidney transplantation (10–15). PTDM is a common complication after organ transplantation, and it has been associated with significant deleterious effects on long-term patient and graft survival. A recent meta-analysis of 10 studies in 2502 kidney recipients confirmed a significant and independent relationship between anti-HCV seropositive status and PTDM (16). Moreover, preliminary evidence shows that the induction of a pretransplant-sustained viral response by interferon-alpha treatment in HCV-positive dialysis patients awaiting kidney transplant may be associated with a lower risk of PTDM (17).

Despite the current epidemiologic evidence linking DM with HCV, the precise mechanism of this association remains unclear. Possible pathophysiologic mechanisms that have been suggested include induction of insulin resistance, diminished hepatic glucose uptake and glycogenesis, as well as a direct cytopathic effect of HCV on islet-cells inducing beta-cell dysfunction (18–22). Some evidence indicates that a predominant effect of the virus is the induction of insulin resistance, caused by inhibitory actions of the virus on insulin regulatory pathways within the liver, possibly mediated by proinflammatory cytokines (18,22). However, after transplantation, the mechanism underlying the association between HCV and PTDM is likely to be multifactorial and complex and may be different from those in immunocompetent HCV-positive individuals, owing to concomitant immunosuppression. We have recently found an evidence for increased insulin resistance in HCV-infected liver transplant recipients (23). In another study of kidney transplant recipients, HCV-infected recipients were found to have increased insulin resistance, however, there were no HCV-negative controls for comparison (24).

The aim of our study was to explore the initial mechanisms involved in PTDM in HCV infection in kidney transplant recipients. For this purpose, a 3-h intravenous glucose tolerance (IVGTT) was performed. Insulin sensitivity (the inverse of insulin resistance), pancreatic insulin secretion, pancreatic antibodies and proinflammatory cytokines were compared between nondiabetic HCV-positive and -negative kidney transplant recipients.

Materials and Methods

Design of the study

A cross-sectional study.

Study population

Group 1: HCV-positive kidney transplant recipients (HCV+ Tx+). Twenty adult, clinically stable nondiabetic kidney transplant recipients infected with HCV, presenting to our transplant outpatient clinic for their regular posttransplant follow-up between May 2004 and May 2006 were prospectively enrolled in the study. All subjects who had antibodies against HCV, as measured by ELISA (Ortho HCV 3.0 EIA) and detectable HCV RNA by reverse transcription polymerase chain reaction (COBAS Amplicor HCV qualitative test, Roche Diagnostics, Indianapolis, IN) in serum were considered HCV positive and eligible for enrollment. Amplicor HCV version 2 (Roche Diagnostics, Branchburg, NJ) was used to quantify the HCV RNA levels in serum. No patient had been treated with antiviral drugs and none had a liver biopsy. No patient had clinical evidence of hepatic decompensation (hepatic encephalopathy, ascites, variceal bleeding or serum bilirubin level greater than 2-fold the upper limit of normal). The following conditions were also excluded: previously diagnosed DM or fasting blood glucose levels >7 mmol/L, severe anemia (Hb <10 g%), concurrent active hepatitis B virus or HIV coinfection, unstable angina, myocardial infarction, cerebrovascular accident or coronary artery bypass grafting within last 3 months, chronic heart failure (NYHA functional class 3 or 4), active cancers, active infections or pregnancy.

Group 2: HCV-negative kidney transplant recipients (HCV− Tx+). A total of 22 contemporaneous clinically stable HCV-negative kidney transplant recipients of similar age and gender as the HCV-positive recipients (frequency matching) were enrolled applying the same exclusion criteria as mentioned above.

Group 3: HCV-positive nontransplant patients (HCV+ Tx−). As a second control group, we included 14 HCV-positive nontransplant patients who were suffering from chronic hepatitis C infection with no evidence of cirrhosis, and who had not yet been initiated on antiviral therapy, applying the same exclusion criteria as mentioned above. They were also matched for age and gender with the HCV-positive transplant recipients (frequency matching).

Group 4: HCV-negative nontransplant (healthy) subjects (HCV− Tx−). As a third control group, we included 24 apparently healthy asymptomatic volunteers taking no medications with normal liver function tests and no known history of DM and/or HCV infection who were of similar age to the HCV-positive transplant recipients (frequency matching).

The study protocol was approved by our institution's Ethics Committee, and fully informed written consent was obtained from all participants before entry into the study.

Clinical and laboratory assessment

The following demographic and clinical data were collected at the time of enrollment: age at (last) transplant and at IVGTT, sex, race, weight, height, body mass index (BMI) at transplant and at IVGTT, waist/hip ratio, blood pressure, history of alcohol use, smoking, family history of DM, underlying kidney disease, duration of dialysis before transplant, total number of transplants, type of donor (deceased vs. living), time from transplant to IVGTT, type and dose of immunosuppressive drugs, serum trough levels of calcineurin inhibitor (cyclosporine or tacrolimus), intake of beta-blockers, thiazide diuretics or statins, total number of acute rejection episodes and total methyl prednisolone boluses used.

For routine biochemical, immunologic and cytokine assessments, fasting venous blood samples were obtained after an overnight fasting period and after 15 min supine rest, prior to the IVGTT. HCV RNA was assessed by reverse transcriptase-polymerase chain reaction (RT-PCR) (if not checked within 3 months of the IVGTT). Following pancreatic antibodies were also measured: antiislet cell antibodies using indirect immunofluorescence assay (Binding Site, Heidelberg, Germany), and antiglutamic acid decarboxylase (anti-GAD) and antityrosine phosphatase antibodies using enzyme-linked immunosorbent assay (Euroimmun, Luebeck, Germany).

Insulin sensitivity assessment

All participants underwent IVGTT, as previously described (25). The IVGTT was performed under standardized conditions in a metabolic day ward (air conditioned, quiet room) starting in the morning between 8:00 and 9:00 am following overnight fasting after at least 20 min supine rest. A glucose bolus (50% solution) was administered intravenously, at a dose of 0.5 g/kg body weight. All blood samples were immediately processed and stored in aliquots at −80°C until analysis of glucose, insulin and C-peptide. Serum insulin was measured by AutoDELFIA automatic immunoassay system (Perkin Elmer Wallac, Turku, Finland) and C-peptide was analyzed by use of the Immulite system (DPC Diagnostic Products Corporation, LA, distributed in Germany by DPC Biermann GmbH, Bad Nauheim, Germany). From the glucose and insulin dynamic profiles, the insulin sensitivity index (SI) was calculated using the minimal model approach according to Bergman (26). The minimal model assesses peripheral insulin sensitivity that mainly reflects muscle tissue as the major glucose utilizing tissue of the body. The relatively high glucose dose (0.5 vs. 0.3 g/kg) enables evaluation of insulin sensitivity by the minimal model without the need for augmentation of plasma insulin concentrations by tolbutamide or insulin injection. We have validated insulin sensitivity estimates derived using this approach in patients with chronic heart failure against the euglycemic clamp reference method (27,28). SI is defined as the fraction of the glucose distribution space cleared per minute by insulin-dependent glucose disposal relative to the concentration of insulin and is expressed in min−1.μU.mL− 1.104. Insulin concentrations during the IVGTT were expressed as incremental area under the concentration profile, calculated using the trapezium rule.

Insulin secretion assessment

Pancreatic beta-cell function was assessed in all transplant recipients by the extended combined model according to Watanabe, which uses the data of both plasma insulin and C-peptide kinetics during an intravenous glucose tolerance test (29). Early phase (phase 1: during the first 10 min of IVGTT) and late phase (phase 2: from 10 to 180 min of IVGTT)) insulin secretions were calculated separately. The extended combined model incorporates the correct two-compartmental structure for C-peptide disappearance and provides an accurate measure of prehepatic insulin secretion rates without separate quantification of C-peptide kinetics or a priori assumption of kinetic parameters of the model.

Proinflammatory cytokines

A batch analysis for tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6) and high sensitive C-reactive protein (hs-CRP) was done on the baseline serum samples stored at −80°C. Enzyme-linked immunosorbent assay was used for TNF-α and IL-6 (Human Quantikine HS; R&D Systems, MN). Analysis for hs-CRP was performed on the BN II nephelometer (Behring) by particle-enhanced immunonephelometry with a detection limit of 0.0175 mg/L and a measuring range of 0.0175–7 mg/L.

Body composition

In all subjects, BMI was calculated as the ratio of weight (kg) and squared height (m2). For body composition assessment, dual energy x-ray absorptiometry (DEXA) was performed employing a Lunar Prodigy scanner (GE Lunar Corp., Madison, WI). Total body scans were analyzed to obtain total and regional (legs, arms and trunk) measurements of fat and lean tissue. Precision of assessments in total body mode are <1%. Fat mass of the trunk, termed as ‘central fat mass’, includes both visceral and subcutaneous fat of this anatomical region. The sum of fat mass of the legs and arms was termed as ‘peripheral fat mass’. The distribution of fat mass was calculated as the ratio of central fat mass/peripheral fat mass.

Statistical analyses

Results of continuous measured data are presented as means ± standard deviation if not stated otherwise. Unpaired t-tests or Mann–Whitney U-tests and analysis of variance (ANOVA) followed by t-tests used as appropriate to compare the continuous variables between the groups, and Fisher's exact test or chi-square test was used for categorical variables. Distribution for biochemical variables was evaluated for normality using the Kolmogorov–Smirnov test, and logarithmic transformation was applied where necessary to allow a parametric statistical approach. SI was square-root transformed in accord with previous evaluation of the distributional characteristics of model-derived variables (30). Pearson's product moment correlation was used for correlation analysis.

Univariate and multivariate linear regression analyses were performed to determine the factors associated with impaired insulin sensitivity. In addition to HCV status, other potential predictor variables for insulin sensitivity identified in Tables 1 and 2 that were different between two transplant groups with p-value ≤0.15 were examined. A stepwise variable selection (backward selection, inclusion p ≤ 0.05, exclusion p ≥ 0.1) was applied. Unstandardized regression coefficients with two-sided 95% limits of confidence are reported. A two-sided p-value ≤0.05 was considered statistically significant.

Table 1.  Baseline characteristics
 HCV+ Tx+ (n = 20)HCV− Tx+ (n = 22)HCV+ Tx− (n = 14)HCV−Tx− (n = 24)p-Value (HCV+ Tx+ vs. HCV− Tx+)
  1. BMI = body mass index; BP = blood pressure; DEXA = dual energy x-ray absorptiometry; DM = diabetes mellitus; HCV = hepatitis C virus; IVGTT = intravenous glucose tolerance test; NA = not available; Tx = transplant.

Age at IVGTT (years)48 ± 8 51 ± 13 46 ± 11 52 ± 120.4
Gender: male (%)10 (50)15 (68)7 (50)22 (92)0.3
Race: Caucasian (%)19 (95) 22 (100)14 (100)22 (92)1.0
Alcohol abuse (%)1 (5) 1 (4.5)2 (14)01.0
Smoking (%) 4 (20)10 (46)10 (71) 00.1
+ Family history of DM (%) 8 (40) 7 (32)2 (14)NA0.7
Hypertensive (%)16 (80)16 (73)0 (0) 00.7
Systolic BP (mmHg)134 ± 20129 ± 15114 ± 12122 ± 100.3
Diastolic BP (mmHg) 82 ± 13 75 ± 1171 ± 9 75 ± 6  0.06
BMI at IVGTT (kg/m2)25 ± 425 ± 526 ± 325 ± 40.8
Waist/hip ratio at IVGTT 0.9 ± 0.1 0.9 ± 0.1 1.0 ± 0.1NA0.3
Central/peripheral fat mass ratio at IVGTT by DEXA 1.51 ± 0.52  1.6 ± 0.61 1.17 ± 0.371.37 ± 0.31 0.95
Table 2.  Transplant-related clinical and laboratory data
 HCV+ Tx+ (n = 20)HCV− Tx+ (n = 22)p-Value
  1. BMI = body mass index; CyA = cyclosporine; HCV = hepatitis C virus; IVGTT = intravenous glucose tolerance test; MMF = mycophenolate mofetil; MP = methyl prednisolone; Tac = tacrolimus.

Time from (last)Tx to IVGTT (years)7.9 ± 5.75.1 ± 3.40.06
Age at Tx (years)39 ± 1146 ± 130.1
Type of donor: deceased (%)19 (95)16 (73)0.1
Patients with >1 Tx (%) 9 (41)0 (0)0.003
Duration of cumulative immunosuppression12.6 ± 7.4 5.1 ± 3.40.0003
Duration of dialysis before Tx (years)5.1 ± 2.63.5 ± 2.60.07
BMI at Tx (kg/m2)22 ± 4 25 ± 3 0.07
Increase in BMI at IVGTT from at Tx2.5 ± 1.21.9 ± 2.90.13
Acute rejection in the past (%)13 (65)9 (41)0.1
Baseline HbA1c at Tx5.6 ± 0.65.8 ± 0.80.55
HbA1c at IVGTT5.9 ± 0.76.1 ± 0.60.3
Fasting glucose (mg/dL)84 ± 1489 ± 120.3
Creatinine (mg/dL)1.5 ± 0.61.6 ± 0.40.6
AST (GOT) (U/L)42 ± 5023 ± 8 0.09
ALT (GPT) (U/L) 76 ± 10822 ± 130.04
Bilirubin total (mg/dL)0.7 ± 0.30.6 ± 0.20.5
Albumin (g/dL)4.1 ± 0.24.3 ± 0.30.052
Iron (μmol/L)20.5 ± 11.518.5 ± 8.4 0.5
Ferritin (μg/L)563 ± 638 481± 6420.7
Cholesterol (mg/dL)187 ± 41 199 ± 58 0.4
HDL (mg/dL)55 ± 1354 ± 170.8
LDL (mg/dL)103 ± 36 112 ± 50 0.5
Triglycerides (mg/dL)139 ± 81 167 ± 73 0.2
Tac level (SE) μg/L6.7 ± 2.18.4 ± 2.70.07
CyA level (SE) μg/L90 ± 35111 ± 34 0.3
Tac (%)13 (65)15 (68)1.0
CyA (%)6 (30) 7 (32)1.0
MMF (%)9 (45) 9 (41)1.0
Myfortic (%)2 (10) 4 (18)0.7
Azathioprine (%)4 (20) 3 (14)0.7
Sirolimus (%)2 (10)1 (5)0.6
Prednisolone mg/d4.0 ± 2.44.2 ± 0.90.8
MP pulses in the past (%)12 (60) 9 (41)0.4
Beta-blocker (%)11 (55) 22 (100)0.004
Diuretic (%) 6 (30) 9 (41)0.5
Statin (%) 7 (35)16 (73)0.03

Results

Patient characteristics

There were 20 HCV-positive (HCV+ Tx+), 22 HCV-negative (HCV−Tx+) kidney transplant recipients, 14 HCV-positive nontransplant patients (HCV+ Tx−) and 24 HCV-negative healthy asymptomatic subjects (HCV− Tx−) in the study. The mean age of all 42 kidney transplant recipients was 49 ± 11 years (range: 27–70 years), 60% were males and the mean BMI was 25 ± 4.5. The vast majority of patients were Caucasians. None of these patients had a liver biopsy or were treated with interferon-alpha and/or ribavirin anytime before or after transplantation. The baseline characteristics of HCV-positive and -negative transplant recipients were similar and are summarized for all four groups in Table 1. Table 2 shows the transplant-related clinical and laboratory data in HCV-positive and -negative recipients. HCV-positive patients were found to have significantly increased frequency of retransplantation, increased duration of cumulative immunosuppression, higher mean alanine aminotransferase (ALT) levels, and decreased frequency of beta-blocker and statin intake as compared to HCV-negative patients.

No significant differences in bilirubin, iron or ferritin levels were found. Detailed body composition from anthropometric assessment and DEXA scan at the time of IVGTT showed no significant difference between groups with regard to global composition and regional fat tissue distribution. Furthermore, antiislet cell, anti-GAD and antityrosine phosphatase antibodies were negative in all recipients.

Immunosuppressive therapy

At the time of enrollment, the majority of transplant recipients were being treated with prednisolone and tacrolimus in both groups. The daily dose of prednisolone ranged from 2 to 5 mg in most of the patients, reflecting our institutional practice of using lower maintenance dosages of steroids in chronic stable patients. As shown in Table 2, the prednisolone doses, cyclosporine/tacrolimus levels and the use of MMF/azathioprine/myfortic/sirolimus were not significantly different between the HCV-positive and -negative recipients. There were no significant differences between the two groups with respect to incidence of biopsy-proven rejection, use of methylprednisolone boluses or antilymphocyte antibody therapy.

Results of IVGTT

Fasting insulin, glucose and C-peptide, as well as the results of the minimal model analysis of peripheral insulin sensitivity, insulin secretion and hepatic uptake for all four groups are presented in Table 3. Mean fasting glucose and insulin levels were not significantly different between HCV-positive and HCV-negative patients but between transplant and nontransplant patients. Peripheral insulin sensitivity was significantly reduced in the HCV-positive transplant recipients as compared to HCV-negative recipients who had similar levels as healthy controls (Figure 1). The insulin secretion (both phase 1 and 2), and hepatic insulin uptake were also not statistically different between the two groups of kidney transplant recipients. In the two-factorial ANOVA (transplantation: yes/no, HCV status: yes/no), we found significantly lower values of insulin sensitivity for HCV-positive versus HCV-negative subjects but no influence of transplantation. The decrease of insulin sensitivity due to HCV was comparable for transplant and non-transplant subjects, thus no interaction between HCV infection and transplantation was present.

Table 3.  Results of IVGTT
 HCV+ Tx+ (n = 20)HCV− Tx+ (n = 22)HCV+ Tx− (n = 14)HCV−Tx− (n = 24)p-Value ANOVAp-Value all Tx+ vs. Tx−p-Value all HCV+ vs. HCV−p-Value HCV+ Tx+ vs. HCV− Tx+1
  1. 1t-test for independent samples.

  2. HCV = hepatitis C virus; IVGTT = intravenous glucose tolerance test; SI,= insulin sensitivity index; Tx = transplant.

Fasting glucose (mg/dL)84.4 ±14.1 88.9 ±12 96.7 ± 12.7 91 ± 7.1 0.033 0.0090.83 0.26
Fasting insulin (mU/L)7.4 ± 4.65.8 ± 2.29.4 ± 4.05.2 ± 3.10.0090.360.0010.17
SI (min−1.μU.mL−1.104)3.0 ± 2.14.9 ± 3.03.5 ± 1.24.7 ± 1.90.0120.410.0030.02
Insulin secretion phase 1 (pmol/mL)0.9 ± 0.50.9 ± 0.7