- Email: [email protected]
Prevalence of diabetes mellitus in patients with end-stage liver cirrhosis due to hepatitis C, alcohol, or cholestatic disease
Journal offfepatology 2000; 32: 209-217 Printed in Denmark AN rights reserved Mmksgaard . Copenhagm
Jaurd of Hepatology ISSN 01684278
Prevalence of diabetes mellitus in patients with end-stage liver cirrhosis due to hepatitis C, alcohol, or cholestatic disease Nizar N. Zein’, Ahmad S. Abdulkarim2,
Russell H. Wiesner’,
Kathleen
S. Egan3 and David H. Persing
‘Division of Gastroenterology and Hepatology and Internal Medicine, =Department of Laboratory Medicne and Pathology, and 3Section of Biosfatistics, Mayo Clinic and Muyo Foun~tion, Rochester, ~~~nesofu, USA
Background/Aims: The aims were to study: 1) the prevalence of diabetes mellitus in patients with endstage liver cirrhosis due to hepatitis C, alcohol, or cholestatic liver disease, 2) viral and host immunogenetic factors that may predispose to diabetes, and 3) liver ~ansplantation outcome in patients with or without diabetes. MetIrods: Fasting blood glucose values of patients who underwent liver transplantation because of hepatitis C-related cirrhosis (73 patients) were compared with those of patients with cirrhosis due to cholestatic (78 patients) or alcoholic liver disease (53 patients) and to a general population. Data on diabetes prevalence in a population without liver cirrhosis was based on the prevalence of diabetes in Olmsted County, Minnesota, residents. HLA was determined using serologic assays. Hepatitis C virus genotypes were determined with polymerase chain reaction amplification and direct sequencing. Hepatitis G RNA was detected with polymerase chain reaction. Liver transplantation outcome in patients with or without diabetes was determined with rejection, retransplantation, or death at 1 year after transplantation as end points, Results: Of 64 patients with hepatitis C alone, 16 (25%) had diabetes before transplantation compared with 1 of 78 (1.3%) with cholestatic liver disease (p= 0.~1) and 10 of 53 (19%) with alcoholic liver disease
@=0.3(i). Nine patients had hepatitis C plus cholestatic liver disease; one of these (11%) had diabetes. The prevalence of diabetes in patients with cholestatic liver cirrhosis was not different from that of the general population. The frequency of hepatitis G virus coi~e~~on, HLA-DR3, or HLA-DR4 in hepatitis C and diabetes was not different from that of hepatitis C alone. The distribution of hepatitis C virus genotype was similar in those with and those without diabetes. Diabetes was not associated with increased risk of rejection, retransplantation, or death at 1 year after transplantation, and had no impact on overall survival after ~ansplantation. Co~rclusions: 1) The risk of diabetes is not increased in patients with liver cirrhosis due to cholestatic liver disease but is in patients with liver cirrhosis due to hepatitis C or alcoholic liver disease; 2) cofactors (age, sex, body mass index, hepatitis G virus coinfection, hepatitis C virus genotype, or HLA-DR3IDR4) did not explain the increased risk of diabetes in patients with hepatitis C; 3) diabetes before liver transplantation did not change the outcome at 1 year after transplantation or survival.
I
1.5% of the population in the U.S. and Western Europe is infected with this virus (1,2). The possibility of extrahepatic manifestations in the setting of hepatitis C virus infection was suggested first in 1990 (3). Since then, various extrahepatic immunologically mediated syndromes have been recognized to complicate the infection, including essential mixed cryoglobulinemia (4), porphyria cutanea tarda (5), membranous glomerulonephritis (6), Mooren cornea1 ulcer (7), autoimmune
NFECTION
with hepatitis C virus is common throughout the world. It has been estimated that 1.2% to
Received 16 February; revised 26 July; accepted 2 August I999 Corresponderzce: Nizar N. Zein, Division of Gastroenterology and Hepatology and Internal Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA. Tel: 507 284 3335. Fax: 507 284 2107. e-mail: [email protected]
Key wouds: Alcoholic liver disease; Cirrhosis; Diabetes mellitus; Disease; Endpoints, Hepatitis; Liver; Outcome; Tr~splantation.
209
N. N. Zein et al.
thyroiditis (8), Sjiigren syndrome (9), lichen planus (lo), and idiopathic pulmonary fibrosis (11). A possible link between hepatitis C infection and development of diabetes mellitus has been suggested but not completely characterized. The increased aminotransferase levels common in patients with type 2 diabetes have traditionally been attributed to fatty infiltration of the liver (12). However, most of these studies were published before the identification of hepatitis C virus. Recently, 2 studies from the United Kingdom and Spain (13,14) showed a high prevalence of hepatitis C virus antibodies in patients with type 2 diabetes, suggesting a possible link between the two disorders. Another study from the United Kingdom by Allison and colleagues (15) provided evidence for increased prevalence of type 2 diabetes in patients with hepatitis Cassociated liver cirrhosis compared with those with liver cirrhosis due to other causes, supporting the likelihood of a causal relationship. A more recent Italian study provided evidence against a potential association between the two disorders (16). Hepatitis G virus is a positive-stranded RNA virus of the Flaviviridae family (17). The genomic organization of this virus is similar to that of the hepatitis C virus, but phylogenetic studies have demonstrated that the viruses are distinct (18-20). The blood-borne nature of hepatitis G virus has been well demonstrated by analyses showing the appearance of the virus in patients after transfusion (17-20). This mechanism of transmission has also been implied by the high frequency of infection among intravenous drug users (21), hemodialysis patients (19,22), and persons with chronic hepatitis C (23). Although hepatitis G virus has not been clearly associated with any human disease, its high prevalence in patients with HCV raises the possibility of it being a cofactor indevelopment disorders that have been attributed to HCV, including diabetes mellitus. In this study, we report the prevalence of type 2 diabetes mellitus in patients in the U.S. with end-stage liver disease and explore the factors that potentially may be associated with the development of diabetes in patients with hepatitis C virus infection, including coinfection with hepatitis G virus. We also examine the impact of pretransplantation diabetes on liver transplantation outcome in patients with end-stage liver disease.
Materials and Methods Patients and controls All patients with a pretransplantation diagnosis of hepatitis C who underwent liver transplantation at our institution between 1986 and 1996 were identified (73 patients). Diagnosis of hepatitis C was based on the detection of hepatitis C virus antibodies by commercially avail-
210
able enzyme immune antibody assay (EIA 2.0, Ortho Diagnostic Systems, Raritan, NJ, USA) and confirmed by a strip immune assay (SIA 2.0, Ortho Diagnostic Systems, Raritan, NJ, USA). In patients who were evaluated for liver transplantation before the introduction of routine hepatitis C virus testing, the diagnosis was made using stored sera. Of the 73 patients, 64 had hepatitis C as the only diagnosis and nine had hepatitis C plus primary biliary cirrhosis or primary sclerosing cholangitis. The demographic data collected at the time of pretransplantation evaluation included age, sex, alcohol use, and the estimated time of acquisition of hepatitis C virus. Data on alcohol consumption in patients with HCV infection were obtained from a standard questionnaire given to all patients undergoing a pretransplantation evaluation. Patients were considered to have a significant alcohol consumption if their mean daily consumption was equal to or greater than 30 g. If accurate data were not recorded on the questionnaire, medical records were reviewed. Patients with hepatitis C were compared with two other groups of patients who had similarly advanced liver disease. The first group included all patients who underwent transplantation at our institution for end-stage liver disease due to cholestatic liver disease between 1986 and 1990 (79 patients, of whom 37 had primary biliary cirrhosis and 42 had primary sclerosing cholangitis) who had no serologic or virologic markers for hepatitis B or C. The second group included those who underwent transplantation at our institution between 1986 and 1996 for alcoholic liver disease in the absence of serologic and virologic markers of hepatitis B or C or evidence for any other liver disease (53 patients). Blood glucose values of all patients and controls were reviewed retrospectively to identify those with diabetes. The prevalence of this disease among patients with liver cirrhosis of different causes was compared with that expected in an age- and sex-matched population. The data on diabetes prevalence in a population without liver cirrhosis were based on the prevalence of diabetes among Olmsted County, Minnesota, residents, which was published recently as part of an epidemiologic study performed at our institution and included all the residents of Olmsted County (24). The diagnosis of diabetes was based on the status at the time of assessment for liver transplantation candidacy. Cases were identified using criteria recommended by the National Diabetes Data Group (25) for the diagnosis of diabetes. These included fasting blood glucose of ~140 mg/dl on 2 separate occasions in those not receiving hypoglycemic medications or corticosteroids. Medications used during the pretransplantation follow-up period were obtained from medical records. Medications were recorded during each weekly visit before transplantation on a special form to ensure accurate records of medications used by each patient. Patients who were receiving an oral hypoglycemic agent or insulin at the time of evaluation were assumed to have diabetes. Patients were excluded if they were receiving treatment with corticosteroids or interferon at the time of blood glucose determination or if 2 blood glucose determinations within a year before transplantation were not available. Although C-peptide or insulin levels were not measured to differentiate type 1 diabetes from type 2 diabetes, these patients were prospectively followed at our institution over a long period and were not prone to ketosis. Patients who had a history of diabetic ketoacidosis or those younger than 30 years presenting with a clinical requirement for insulin were excluded. Based on the clinical information, diabetic patients in this study were assumed to have type 2 diabetes mellitus. Patients who had transplantation for hepatitis B-associated liver disease were excluded because of their small number at our institution. We did not use the new criteria for diagnosis of diabetes mellitus (fasting blood glucose value of >126 mg/dl) that was proposed by the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus (26). This allowed us to compare our findings with those published to date, including the prevalence of diabetes mellitus in the general population and in patients with liver cirrhosis.
Hepatitis C virus RNA detection Viremia was confirmed in all patients with hepatitis C virus-positive antibodies by the detection of hepatitis C virus RNA using a poly-
Risk of diabetes in liver cirrhosis merase chain reaction assay developed in our laboratory (27). In summary, hepatitis C virus RNA was isolated from 100 pl of serum using a commercially available reagent (RNAzol B, Biotecx Laboratories, Houston, TX, USA) according to the manufacturer’s instructions, with 1 modification, the addition of 20 pg of glycogen to each sample as a carrier during isopropanol precipitation. Hepatitis C virus RNA was reverse transcribed and partially amplified with the polymerase chain reaction using primers specific for the 5’ untranslated region of the viral genome. Amplication products were detected by Southern blotting. Hepatitis C virus genotyping For the purpose of genotyping hepatitis C virus, polymerase chain reaction amplification with primers specific for the NS 5 region of the hepatitis C virus genome was performed according to our previously published protocols (28). To avoid possible contamination with exogenous sequences during extraction or amplification, nucleic acid extraction, amplification, and detection steps were performed in separate laboratories. Negative and positive controls were extracted, reverse transcribed, and amplified in each batch of samples tested with the polymerase chain reaction. Each fragment of approximately 401-bp (370 to 410 bp) polymerase chain reaction product was desalted before sequencing using a direct column purification method (Wizard PCR preps DNA purification system, Promega, Madison, WI, USA). Automated sequencing was performed with a standard Sanger procedure involving the incorporation of fluorescein-labeled dideoxynucleotides and detection on an acrylamide gel (ABI model No. 373 A, Applied Biosystems, Hercules, CA, USA). Nucleotide sequences were aligned and compiled with the previously reported sequences by using the Pileup program (courtesy of Wisconsin Genetic Computer Group, Madison, WI, USA [29]). Cluster analysis was performed using the unweighted pair mean group average (UPMGA) included in the program. These methods allowed comparison of a 222-bp fragment of DNA homologous to nucleotide posititions 7975 to 8196 in the prototype virus. A standardized system recently proposed for the classification of hepatitis C virus genotypes and subtypes was adopted for our study (30). According to this system, hepatitis C virus can be classified into 6 major genotypes comprising a total of 11 subtypes. Hepatitis G virus RNA extraction and detection Hepatitis G virus RNA was extracted and then reverse transcribed following the same procedure for hepatitis C virus RNA and using previously published primers (17). Polymerase chain reaction amplification was accomplished by addition of a second mixture containing 0.8XPCR buffer II (500 mM KCl, 100 mM Tris-HCl, pH 8.3, Perkin Elmer, Norwalk, CT, USA), 2.5 pl of Amplitaq DNA polymerase (Perkin Elmer, Norwalk, CT, USA), 10% (vol/vol) glycerol, 1.75 mM magnesium chloride, 190 mM dUTP, 10 mM dig dUTP (Boehringer Mannheim, Mannheim, Germany), UNG (1 pl/bl, Epicentre Technologies, Madison, WI, USA), and 0.5 PM sense primer 77F (5’-CTC TTT GTG GTA GTA GCC GAG AGA T-3’) (21). Amplifications were carried out in a DNA thermal cycler 480 as follows: 50°C for 10 min, 94°C for 4 min, then 50 cycles of 94°C for 1 min and 58°C for 1 min, followed by a final extension step of 72°C for 5 min. A 40-~1 aliquot of each reverse transcriptase polymerase chain reaction was then tested using the polymerase chain reaction ELISA DIG Detection Kit (Boehringer Mannheim, Mannheim, Germany) and capture probe 152F (S’TCG GTT ACT GAG AGC AGC TCA GAT GAG-3’) for detection of the 156-bp amplification product (21). HLA dermination A standard microlymphocytotoxicity technique was used for class I and class II typing in patients with hepatitis C virus. Several locally derived and commercially available sera were used to define each specificity, as described elsewhere (31,32). Post-transplantation outcome To compare liver transplantation outcome in patients with hepatitis C and diabetes with those without diabetes, we used rejection, re-
transplantation, or death at 1 year after transplantation as the end point. We also evaluated the effect of different variables, including diabetes, on overall survival after liver transplantation. Statistical analysis Two-sample t-tests were used to compare the continuous variables for 2 groups, and the Fisher exact test was used to compare binary outcomes. Because of the small number of patients infected with each hepatitis C virus genotype, all tests of associations between genotype and other factors were based on data that were collapsed into 4 groups: genotypes la, lb, 2a-b, and 3a-4a. The SAS statistical analysis package (SAS Institute, Cary, NC, USA) and S-PLUS (MathSoft, Seattle, WA, USA) were used for all calculations. The Cox proportional hazards model was used to assess the effect of the independent variables on survival. The expected number of patients with diabetes was calculated by applying the age- and sexspecific rates obtained from the Olmsted County, Minnesota, general population. Confidence intervals and p-values for comparing observed and expected number of diabetes cases were based on the Poisson distribution,
Results Patient characteristics and prevalence of diabetes Of 204 patients with end-stage liver disease, 28 (13.7%)
had diabetes. Diabetes was present in 16 of 64 patients (25%) with hepatitis C (not including the nine patients with hepatitis C plus either primary biliary cirrhosis or primary sclerosing cholangitis), one of 78 (1.3%) with cholestatic liver disease (primary biliary cirrhosis or primary sclerosing cholangitis), and ten of 53 (19%) with alcoholic cirrhosis. Of the nine patients with hepatitis C plus cholestatic liver disease, one (11%) had diabetes mellitus. The prevalence of diabetes among patients with hepatitis C was significantly higher than among those with cholestatic liver disease (p=O.OOOl). Similarly, diabetes was more common among those with alcoholic liver cirrhosis than cholestatic liver cirrhosis (p=O.O04). Although diabetes was more prevalent among patients with hepatitis C than among those with alcohol-induced cirrhosis, the difference was not significant (p=O.36). The mean age of patients with hepatitis C (48.72 1.2 years) was marginally higher (p=O.O5) than for those with cholestatic liver cirrhosis (45.7+ 1.1 years) but not different from that of those with alcoholic cirrhosis (52.32 1.l years). Sex distribution was also different: 25/78 (32.1%) of those with cholestatic liver disease were males compared with 44/73 (60.3%) of those with hepatitis C (~~0.001) and 40/53 (75.5%) of those with alcoholic liver cirrhosis (~~0.001). Sex distribution was not different between the hepatitis C and alcoholic cirrhosis groups. No differences were noted among the 3 groups with regard to family history of diabetes. The body mass index of patients with hepatitis C and those with alcoholic liver disease (2820.6 and 27.3kO.7, respectively) was slightly higher than that of patients with cholestatic liver disease (23.450.5). However, the former was also more likely to have detectable ascites 211
N. N. Zein et al,
(67% and 88% of patients with hepatitis C and alcoholic liver disease, respectively, had ascites, compared with 54% of patients with cholestatic liver disease), precluding the ability to compare the body mass index among the 3 groups. Risk factors for diabetes Age, sex, family history, and body mass index were not different between those with diabetes and those without diabetes for either the cholestatic or alcoholic groups {Table 1). For the hepatitis C group, the mean age of those with diabetes (53.0~2.3 years) was marginally higher (p=O.O4) than that of those without diabetes (47.321.3). The difference between the 2 groups for the number of years from exposure to hepatitis C virus to liver transplantation, based on the available data, was not significant (Table 1). Also, there were no significant differences between the 2 groups with respect to sex distribution, excessive alcohol intake, hepatitis G virus coinfection, or HLA-DR3 and HLADR4 (Table 1). Five of 17 (29%) patients with hepatitis C and diabetes had a family history of diabetes, as compared with one of 56 (1.8%) patients with hepatitis C alone (p=O.O002). These data may suggest an important role for family history as a risk factor for the development of diabetes in patients with hepatitis C cirrhosis. The hepatitis C virus genotype was determined for 12 of the 17 (70%) patients with hepatitis C and diabetes and for 47 of the 56 (83%) patients without dia-
betes (Table 1). Genotyping was not possible on the remaining samples because of poor sequencing results. The genotype distribution in patients with diabetes was la in three (25O/), lb in five (42%), and 2b in two (17%), and in patients without diabetes, la in 22 (47%), lb in 16 (34%), 2a + b in seven (15%), and 3a and 4a in two (4%). Genotype distribution was not different among those with and those without diabetes (p= 0.74). Observed versus expected number with diabetes mellitus Table 2 shows the observed number of cases of diabetes by sex for the hepatitis C, cholestatic liver disease, and alcoholic cirrhosis groups. The expected age- and sexmatched prevalence rates of diabetes were obtained from prevalence of adult-onset diabetes mellitus among Olmsted County, Minnesota, residents older than 45 years on January 1, 1990. Our samples included people younger than 45 years, Their expected number was calculated using rates for those 45 to 54 years old. This results in a higher number of expected, because the prevalence of diabetes increases with age until age 65 to 74. Thus, the results presented here are conservative. The observed number of cases of diabetes among patients with hepatitis C was significantly higher than the expected number for both men and women (6.2 times higher for men and 10.1 times higher for women, p
TABLE 1 Characteristics of patients with liver cirrhosis* Characteristic
Mean age, years Males Body mass index Family history of DM Years of hepatitis C virus? Excessive alcohol intake Hepatitis G virus coinfection Hepatitis C virus genotype la lb 2a+b 3a+4a HLA-DR3 HLA-DR4 Rejection at 1 year ~transplan~tion at 1 year Death at any time after liver transplantation
PBCX’SC
Hepatitis C
Alcohol
Without DM (n=56)
With DM (n= 17)
Without DM (n=77)
With DM (n=l)
Without DM (n=43)
With DM (n= 10)
47.32 1.3 35 (62.5) 28.120.7 1 (1.8) 24.521.3 7 (12.5) 6 (11)
53.022.3 9 (53) 27.52 1.5 5 (29) 25.423.0 5 (29) 4 (23.5)
45.421.1 25 (32) 23.5kO.S I (1.3)
57.6 0 18 0
51.721.3 33 (77) 27.5z0.8, 2 (4.7)
55121.7 7 (70) 26.5?2.0 0
26 (60.5) 5 (11.6) 13 (30)
7 (70) 1 (10) 2 (20)
22147(47) 16147(34) 7i47 (15) 2147 (4) 9 (16) 8 (14.3) 35 (62.5) 8 (14.2) 11 (19.6)
3112 (25) S/l2 (42) 202 (16.7) 2l12 (16.7) 1 (6) 4 (23.5) 8 (47) 1 (5.9) 1 (5.9)
37 (48) 6 (7.8) 22 (28.6)
0 0 1 (LOO)
DM, diabetes mellitus; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis. * Values are number of patients (%) or value. t Years from the estimated time of exposure to hepatitis C virus to transplantation (available for 52173patients).
212
Risk of diabetes in liver cirrhosis TABLE 2 Observed and expected cases by sex of diabetes mellitus in cholestatic liver disease and hepatitis C* Characteristic
Number
Hepatitis C group Males Females Totalt
Diabetes mellitus
Observed/expected
Observed
Expected
Ratio
95% CI
p-value
1.46 0.79 2.24
6.2 10.1 7.6
(2.8, 11.7) (4.4, 20.0) (4.4, 12.i)
44 29
9 8
73
17
PBCYPSC group Males Females Totalf
25 53 78
0
1 1
0.72 1.16 1.88
0.00 0.86
0.53
(0.0, 4.2) (0.02, 4.8) (0.01, 2.3)
Alcoholic cirrhosis group Males Females Total
40 13 53
7 3 10
1.51 0.39 1.90
4.6 7.7 5.3
(1.9, 9.5) (1.6, 22.5) (2.5, 9.7)
CI, confidence interval; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis. * Expected number of cases was derived from data from residents of Olmsted County, Minnesota (24). + No significant differences by sex (p=O.31). 1 No significant differences by age (p=O.32).
liver disease showed no difference from the expected number of cases in the age- and sex-matched general population without liver disease (Table 2). The observed number of diabetes cases in patients with endstage alcoholic liver disease was also si~ifi~antly higher than that expected for both males and females (4.6 times higher for men and 7.7 times higher for women, p
outcome
and overall survival
Of 28 patients with diabetes (all groups), 15 (53.6%) had at least 1 episode of rejection at 1 year after liver transplantation compared with 98 of 176 patients (55.7%) without diabetes @=O.l). No death occurred in either group at 1 year, but 19 of 176 patients (10.8%) without diabetes required retransplantation compared with two of 28 patients (7.1%) with diabetes (p=O.3) (Table 1). Indications for retransplantation included hepatic artery thrombosis (6 patients), severe cholestasislhepatitis C recurrence (4 patients), ischemic biliary complications (4 patients), primary graft dysfunction (3 patients), and others (4 patients). There were no significant differences in outcome or overall survival between patients with diabetes mellitus and those without it when similar analysis was applied on the basis of the underlying liver disease (Table 1). The overall survival of the group was evaluated. Because follow-up time was different for the groups, Cox proportional hazards models were run to determine if there were any differences in survival time. Sex was the only significant factor for survival of patients with cho-
lestatic liver disease (males at higher risk for death) (p=O.O15) but not for the hepatitis C group or alcoholic cirrhosis group (p=O.16). Neither age at the time of transplantation nor diabetes was significant for overall survival in any of the 3 groups during followup* In the hepatitis C group, a univariate model was performed for the variables specific for this group (alcohol intake, hepatitis G virus, coinfection, hepatitis C virus genotype, HLA-DR3 or HLA-DR4). None of these variables was significant for survival or for death of patients with hepatitis C.
1,000 1 800
;:
-
HCV
- - Aicot~olic -*.-a Cholestatic ---- General population
/
;
0-
0.) . .. . . . . 20-44
. ., . . yyy.~:“:-
,.e”=E:‘-______““‘“’
55-64
45.54
65-74
Age group
Fig. 1. Age-dependent prevalence of diabetes meliitus among Ulmsted County, Minnesota, residents (general population) and in those with liver cirrhosis. Alcoholic, alcoholic liver disease; HCV, hepatitis C; PBCIPSC, cholestatic liver disease (primary biliary cir~hosislpri~ary sclerosing cirrhosis).
213
N. N. Zein et al.
Discussion Impaired glucose tolerance and liver disease frequently occur concurrently (33-35). As many as 80% of patients with liver cirrhosis are glucose intolerant, and 10% to 20% of these patients have diabetes (36,37). It has been proposed that peripheral resistance to insulin as well as impairment in pancreatic insulin secretion may be the underlying mechanisms for diabetes in patients with liver cirrhosis (33-37). In our study, we found that 13.7% of patients with liver cirrhosis had diabetes, consistent with previously published results. However, the prevalence of diabetes was significantly greater among patients with hepatitis C- or alcoholrelated cirrhosis compared with those with cholestatic liver disease. In fact, patients with cholestatic liver cirrhosis had a prevalence of diabetes similar to that of an age- and sex-matched general population of Olmsted County, Minnesota, suggesting that the mechanism of diabetes in patients with liver cirrhosis is related more closely to the underlying cause of liver cirrhosis. The increased risk of diabetes has also been reported in a retrospective study of patients with chronic hepatitis C without liver cirrhosis (38), supporting the role of hepatitis C virus in the pathogenesis of diabetes. These findings may provide an explanation for the discrepancies in the literature based on the selection of the control group. de1 Olmo and colleagues (39) from Spain found no increase in the risk of diabetes in hepatitis Crelated cirrhosis compared with those with alcohol- or hepatitis B-related cirrhosis. However, those with cholestatic liver disease were not included as controls, as in our study. Although both patients with hepatitis C and patients with alcoholic liver disease had a high prevalence of diabetes, patients with both hepatitis C and heavy alcohol consumption did not have increased prevalence of diabetes in comparison with those with hepatitis C and low or no alcohol consumption. This may suggest that hepatitis C and excessive alcohol consumption are not likely to have an additive effect in relation to the risk of having diabetes in any given patient. Family history of diabetes was an important predictor for the development of diabetes in HCV patients but not in the alcoholic or cholestatic disease groups and did not appear to explain the majority of diabetes cases. After adjusting for the family history, patients with HCV continued to have a significantly higher prevalence of diabetes than patients with cholestatic liver disease or than Olmsted County residents. A less likely explanation of the high prevalence of diabetes in patients with end-stage liver disease due to HCV or alcohol is that diabetes accelerates the progression of liver disease in such patients. Our study was
214
not designed to address this issue, but future prospective studies may shed more light on the relationship between diabetes and the underlying liver disease. However, diabetes did not appear to be associated with accelerated disease in our patients after liver transplantation. One of the shortcomings of this study is the selection of patients with decompensating liver cirrhosis. This may not be an accurate representation of all patients with chronic liver disease or compensated liver cirrhosis. Future studies may need to include patients with different stages of liver disease. The small number of patients in each of the three groups of patients with liver diseases is another major shortcoming of our study and may have potentially underestimated or overestimated the prevalence of diabetes in these patients. However, a recent retrospective study that included 1117 patients with chronic viral hepatitis showed a 21% prevalence rate of diabetes in patients with chronic HCV, which is similar to the findings of our study (40). The pathogenesis of diabetes in patients with hepatitis C- or alcohol-associated cirrhosis is not well understood. Features that are present in these 2 disorders and absent from cholestatic liver disease are likely to provide clues to the pathogenesis of “hepatic diabetes.” Increased fat and iron deposition in the liver are common features among patients with hepatitis C virus infection and those with alcoholic liver disease (4143). It has been suggested that liver fat may contribute to insulin resistance (44), which may lead to loss of the restraining effect of insulin on hepatocyte production of glucose, leading to diabetes. This hypothesis is supported by the increased rate of hepatic glucose production in patients with diabetes (45). It also may provide insight into the link between diabetes and nonalcoholic steatohepatitis (46,47). The relationship between iron and the development of diabetes is most evident in hemochromatosis. The mechanism is thought to be related to damage of the exocrine pancreas by excess iron. Excessive hepatic iron deposition in patients with alcohol-related or hepatitis C liver disease is possibly associated with excessive iron deposition in the exocrine pancreas as well, but this hypothesis needs to be evaluated in future studies. Conversely, the pathogenesis of diabetes in patients with hepatitis C may be different from that of patients with alcoholic cirrhosis. We were not able to account for the increased prevalence of diabetes on the basis of family history alone. Body mass index was not different between hepatitis C patients with diabetes and
Risk of diabetes in liver cirrhosis
those without diabetes, providing evidence against it as a major factor in the development of diabetes in these patients. The risk of diabetes increases with age (24), but age alone does not seem to explain the differences found in the present study. Although the mean age of patients with hepatitis C or alcoholic cirrhosis was higher than that of those with cholestatic liver cirrhosis, the difference was marginal (approximately 3 years). Moreover, patients with hepatitis C or alcoholic cirrhosis had a much higher prevalence of diabetes than an age-matched group of patients with cholestatic liver disease or an age-matched general population in Olmsted County, Minnesota (Fig. 1). Several potentially confounding factors in patients with hepatitis C were evaluated, including the infecting hepatitis C virus genotype and coinfection with hepatitis G virus. Previous reports have suggested a relationship between specific hepatitis C virus genotypes and some of the extrahepatic manifestations of hepatitis C virus infection (48). In the present study, hepatitis C virus genotype distribution was similar among the patients with diabetes and those without, suggesting that diabetes is not a genotype-dependent manifestation of the hepatitis C virus. However, a recent study (40) suggested a potential role for HCV genotype 2a in the development of diabetes in patients with chronic HCV infection. Because only three patients were infected with HCV genotype 2a in our study, we were unable to analyze the role of this specific genotype in our patients. Similarly, coinfection with hepatitis G virus was not associated with an increased risk of diabetes in hepatitis C virus-infected patients. Currently, the role of hepatitis G virus in the pathogenesis of any human disease is unclear. The finding that excessive alcohol intake by patients with hepatitis C was not an important factor in the pathogenesis of diabetes may suggest that the effects of alcohol and hepatitis C are not additive with respect to the development of diabetes. Uncommon forms of immune-mediated diabetes have been described. These include stiff-man syndrome, an autoi~une disorder of the central nervous system associated with high titers of antibodies to glutamic acid decarboxylase and diabetes (49). Anti-insulin antibodies can cause diabetes by binding to the insulin receptor, thereby blocking the binding of insulin to its receptors in target tissues (50). We previously reported significant increase in the production of autoantibodies in autoimmune disorders in patients with hepatitis C (51). The production of autoantibodies was strongly associated with HLA-DR3, whereas autoimmune disorders were associated with HLA-DR4 (51). In the present study, patients with hepatitis C and dia-
betes were not more likely to have HLA-DR3 or HLADR4, suggesting that autoimmunity is not a likely cause of diabetes in these patients. The results of orthotopic liver transplantation in patients with diabetes are controversial. A study from Spain suggested that the prognosis of patients with posttransplantation diabetes is worse than that of those without diabetes (52). A study from the U.S. showed that transplantation could be accomplished successfully in patients with diabetes, with results similar to those of patients without diabetes, although infections and renal dysfunction were more common in the former group (53). Because liver transplantation outcome in patients with hepatitis C or alcoholic cirrhosis is less favorable than that of those with cholestatic liver disease, it is important to rule out diabetes as a negative prognostic sign for transplantation outcome, We compared outcomes, including rejection, retransplantation, and death at 1 year, as well as overall survival among those with and those without diabetes. Pretransplantation diabetes had no impact on the outcome of transplantation and did not increase the likelihood of allograft rejection. It is important to understand the limitations of these findings, because patients with complicated diabetes (presence of end-organ damage, including retinopathy, nephropathy, or vasculitis) were excluded from transplantation. Therefore, our data are applicable only to those with pretransplantation diabetes without significant diabetes-related complications. In summary, our results suggest that the underlying cause of liver cirrhosis is an important determinant of the risk of diabetes. The cause of diabetes in patients with hepatitis C or alcoholic cirrhosis is not clear. Direct viral-induced injury, insulin resistance related to excess liver fat, or excess iron deposition may have a role and need to be investigated in future studies. If the findings of this study are confirmed in future investigations, diabetes should not be included in the clinical spectrum of liver cirrhosis; however, it should be added to the list of extrahepatic manifestations of hepatitis C and alcoholic liver cirrhosis.
Acknowledgements This study was supported in part by a grant from the American Liver Foundation (NNZ). This study was presented in part at the annual meeting of the American Association for the Study of Liver Diseases, Chicago, Illinois, November 1996.
References 1. Alter MJ. Epidemiology of hepatitis C in the West. Semin Liver Dis 1995; 15: 5-14. 2. Shakil AO, Conry-Cantilena C, Alter HJ, Hayashi P, Kleiner DE, Tedeschi V, et al. Volunteer blood donors with antibody to hepa-
21.5
N. N. Zein et al.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12. 13.
14.
15.
16.
17.
18. 19. 20. 21.
22.
23.
24.
216
titis C virus: clinical, biochemical, virologic, and histologic features. The Hepatitis C Study Group. Ann Intern Med 1995; 123: 330-7. Pascual M, Perrin L, Giostra E, Schifferli JA. Hepatitis C virus in patients with cryoglobulinemia type II [letter]. J Infect Dis 1990; 162: 569. Gumber SC, Chopra S. Hepatitis C: a multifaceted disease. Review of extrahepatic manifestations. Ann Intern Med 1995; 123: 615-20. Ferri C, Greco F, Longombardo G, Palla P Moretti A, Marzo E, et al. Association between hepatitis C virus and mixed cryoglobulinemia. Clin Exp Rheumatol 1991; 9: 6214. DeCastro M, Sanchez J, Herrera JF, Chaves A, Duran R, GarciaBuey L, et al. Hepatitis C virus antibodies and liver disease in patients with porphyria cutanea tarda. Hepatology 1993; 17: 551-7. Johnson RJ, Gretch DR, Yamabe H, Hart J, Bacchi CE, Hartwell P, et al. Membranoproliferative glomerulonephritis associated with hepatitis C virus infection. N Engl J Med 1993; 328: 46570. Wilson SE, Lee WM, Murakami C, Weng J, Moninger GA. Mooren-type hepatitis C virus-associated cornea1 ulceration. Ophthalmology 1994; 101: 73645. Haddad J, Deny P Munz-Gotheil C, Ambrosini JC, Trinchet JC, Pateron D, et al. Lymphocytic sialadenitis of Sjiigren’s syndrome associated with chronic hepatitis C virus liver disease. Lancet 1992; 339: 321-3. Gruppo Italian0 Studi Epidemiologici in Dermatologia (GISED). Lichen planus and liver diseases: a multicentre case-control study. BMJ 1990; 300: 227730. Ueda T, Ohta K, Suzuki N, Yamaguchi M, Hirai K, Horiuchi T, et al. Idiopathic pulmonary fibrosis and high prevalence of serum antibodies to hepatitis C virus. Am Rev Respir Dis 1992; 146: 2668. Stone BG, Van Thiel DH. Diabetes mellitus and the liver. Semin Liver Dis 1985; 5: 8-28. Gray H, Wreghitt T, Stratton IM, Alexander GJ, Turner RC, O’Rahilly S. High prevalence of hepatitis C infection in AfroCaribbean patients with type 2 diabetes and abnormal liver function tests. Diabet Med 1995; 12: 2449. Simo R, Hernandez C, Genesca J, Jardi R, Mesa J. High prevalence of hepatitis C virus infection in diabetic patients. Diabetes Care 1996; 19: 998-1000. Allison ME, Wreghitt T, Palmer CR, Alexander GJ. Evidence for a link between hepatitis C virus infection and diabetes mellitus in a cirrhotic population. J Hepatol 1994; 21: 1135-9. Mangia A, Schiavone G, Lezzi G, Marmo R, Bruno F, Villani MR, et al. HCV and diabetes mellitus: evidence for a negative association. Am J Gastroenterol 1998; 93: 2363-7. Linnen J, Wages J Jr, ZhangKeck ZY, Fry KE, Krawczynski KZ, Alter H, et al. Molecular cloning and disease association of hepatitis G virus: a transfusion-transmissible agent. Science 1996; 271: 505-8. Pessoa MG, Wright TL. Hepatitis G: a virus in search of a disease. Hepatology 1996; 24: 461-3. Alter HJ. The cloning and clinical implications of HGV and HGBV-C (editorial). N Engl J Med 1996; 334: 1536-7. Di Bisceglie AM. Hepatitis G virus infection: a work in progress (editorial). Ann Intern Med 1996; 125: 772-3. Aikawa T, Sugai Y, Okamoto H. Hepatitis G infection in drug abusers with chronic hepatitis C [letter]. N Engl J Med 1996; 334: 195-6. Masuko K, Mitsui T, Iwano K, Yamazaki C, Okuda K, Meguro T, et al. Infection with hepatitis GB virus C in patients on maintenance hemodialysis. N Engl J Med 1996; 334: 1485-90. Tanaka E, Alter HJ, Nakatsuji Y, Shih JW, Kim JP, Matsumoto A, et al. Effect of hepatitis G virus infection on chronic hepatitis C. Ann Intern Med 1996; 125: 74&3. Leibson CL, O’Brien PC, Atkinson E, Palumbo PJ, Melton LJ III. Relative contributions of incidence and survival to increasing
25.
26.
27.
28.
29. 30.
31.
32.
33.
34. 35.
36.
37. 38.
39. 40.
41.
42.
43.
44.
45.
46.
prevalence of adult-onset diabetes mellitus: a population-based study. Am J Epidemiol 1997; 146: 12-22. National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979; 28: 1039-57. Gavin JR, Alberti KGMM, Davidson MB, DeFronzo RA, Drash A, Gabbe SG, et al. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 1998; 21 Suppl 1: S5-19. Zein NN, McGreger CG, Wendt NK, Schwab K, Mitchell PS, Persing DH, et al. Prevalence and outcome of hepatitis C infection among heart transplant recipients. J Heart Lung Transplant 1995; 14: 865-9. Zein NN, Rakela J, Krawitt EL, Reddy KR, Tominaga T, Persing DH. Hepatitis C virus genotypes in the United States: epidemiology, pathogenicity, and response to interferon therapy. Collaborative Study Group. Ann Intern Med 1996; 125: 6349. Genetics computer group. Program manual for GCG package, version 7. Madison,Wisconsin 53711. April 1991. Simmonds P, Alberti A, Alter HJ, Bonino F, Bradley DW, Brechat C, et al. A proposed system for the nomenclature of hepatitis C viral genotypes [letter]. Hepatology 1994; 19: 13214. Czaja AJ, Rakela J, Hay JE, Moore SB. Clinical and prognostic implications of HLA B8 in corticosteroid-treated severe autoimmune chronic active hepatitis. Gastroenterology 1990; 98: 158793. Czaja AJ, Strettell MD, Thomson LJ, Santrach PJ, Moore SB, Donaldson PT, et al. Associations between alleles of the major histocompatibility complex and type 1 autoimmune hepatitis. Hepatology 1997; 25: 317-23. Petrides AS, DePronzo RA. Glucose metabolism in cirrhosis: a review with some perspectives for the future. Diabetes Metab Rev 1989; 5: 691-709. Megyesi C, Samols E, Marks V Glucose tolerance and diabetes in chronic liver disease. Lancet 1967; 2: 1051-6. Proietto J, Alford FP, Dudley FJ. The mechanism of the carbohydrate intolerance of cirrhosis. J Clin Endocrinol Metab 1980; 51: 103&6. Kruszynska YT, Home PD, McIntyre N. Relationship between insulin sensitivity, insulin secretion and glucose tolerance in cirrhosis. Hepatology 1991; 14: 103-l 1. Nolte W, Hartmann H, Ramadori G. Glucose metabolism and liver cirrhosis. Exp Clin Endocrinol Diabetes 1995; 103: 63-74. Fraser GM, Harman I, Meller N, Niv Y, Porath A. Diabetes mellitus is associated with chronic hepatitis C but not chronic hepatitis B infection. Isr J Med Sci 1996; 32: 52630. de1 Olmo JA, Serra MA, Rodrigo JM. Liver cirrhosis and diabetes mellitus [letter]. J Hepatol 1996; 24: 645. Mason AL, Lau JY, Hoang N, Qian K, Alexander GJ, Xu L, et al. Association of diabetes mellitus and chronic hepatitis C virus infection. Hepatology 1999; 29: 328-33. Fiore G, Fera G, Napoli N, Vella F, Schiraldi 0. Liver steatosis and chronic hepatitis C: a spurious association? Eur J Gastroenter01 Hepatol 1996; 8: 125-9. Riggio 0, Montagnese F, Fiore P, Folino S, Giambartolomei S, Gandin C, et al. Iron overload in patients with chronic viral hepatitis: how common is it? Am J Gastroenterol 1997; 92: 1298301. Eaton S, Record CO, Bartlett K. Multiple biochemical effects in the pathogenesis of alcoholic fatty liver. Eur J Clin Invest 1997; 27: 719-22. Banerji MA, Buckley MC, Chaiken RL, Gordon D, Lebovitz HE, Kral JG. Liver fat, serum triglycerides and visceral adipose tissue in insulin-sensitive and insulin-resistant black men with NIDDM. Int J Obes Relat Metab Disord 1995; 19: 84650. Petrides AS, Vogt C, Schulze-Berge D, Matthews 4 Strohmeyer G. Pathogenesis of glucose intolerance and diabetes mellitus in cirrhosis. Hepatology 1994; 19: 61627. Ludwig J, McGill DB, Lindor KD. Review: nonalcoholic steatohepatitis. J Gastroenterol Hepatol 1997; 12: 398403.
Risk of diabetes in liver cirrhosis 47. Sheth SG, Gordon FD, Chopra ‘S.Nonalcoholic steatohepatitis. Ann Intern Med 1997; 126: 13745. 48. Chin CF, Zein NN. Extrahepatic syndromes associated with hepatitis C infection. Gastroenterol Int 1998; 11: 26-35. 49. Bingley PJ, Bonifacio E, Williams AJ, Genovese S, Bottazzo GF, Gale EA. Prediction of IDDM in the general population: strategies based on combinations of autoantibody markers. Diabetes 1997; 46: 1701-10. 50. Taylor SI. Lilly Lecture: molecular mechanisms of insulin resistance. Lessons from patients with mutations in the insulin-receptor gene. Diabetes 1992; 41: 1473-90.
51. Zein NN, Persing DH, Czaja AJ. Viral and genetic factors as determinants of autoimmune expression in chronic hepatitis C [abstract]. Gastroenterology 1996; 1lO(Supp1): A1367. 52. Navasa M, Bustamante J, Marroni C, Gonzalez E, Andreu H, Esmatjes E, et al. Diabetes mellitus after liver transplantation: prevalence and predictive factors. J Hepatol 1996; 25: 64-71. 53. Trail KC, Stratta RJ, Larsen JL, Ruby EI, Patil KD, Langnas AN, et al. Results of liver transplantation in diabetic recipients. Surgery 1993; 1 14: 650-6.
Jaurd of Hepatology ISSN 01684278
Prevalence of diabetes mellitus in patients with end-stage liver cirrhosis due to hepatitis C, alcohol, or cholestatic disease Nizar N. Zein’, Ahmad S. Abdulkarim2,
Russell H. Wiesner’,
Kathleen
S. Egan3 and David H. Persing
‘Division of Gastroenterology and Hepatology and Internal Medicine, =Department of Laboratory Medicne and Pathology, and 3Section of Biosfatistics, Mayo Clinic and Muyo Foun~tion, Rochester, ~~~nesofu, USA
Background/Aims: The aims were to study: 1) the prevalence of diabetes mellitus in patients with endstage liver cirrhosis due to hepatitis C, alcohol, or cholestatic liver disease, 2) viral and host immunogenetic factors that may predispose to diabetes, and 3) liver ~ansplantation outcome in patients with or without diabetes. MetIrods: Fasting blood glucose values of patients who underwent liver transplantation because of hepatitis C-related cirrhosis (73 patients) were compared with those of patients with cirrhosis due to cholestatic (78 patients) or alcoholic liver disease (53 patients) and to a general population. Data on diabetes prevalence in a population without liver cirrhosis was based on the prevalence of diabetes in Olmsted County, Minnesota, residents. HLA was determined using serologic assays. Hepatitis C virus genotypes were determined with polymerase chain reaction amplification and direct sequencing. Hepatitis G RNA was detected with polymerase chain reaction. Liver transplantation outcome in patients with or without diabetes was determined with rejection, retransplantation, or death at 1 year after transplantation as end points, Results: Of 64 patients with hepatitis C alone, 16 (25%) had diabetes before transplantation compared with 1 of 78 (1.3%) with cholestatic liver disease (p= 0.~1) and 10 of 53 (19%) with alcoholic liver disease
@=0.3(i). Nine patients had hepatitis C plus cholestatic liver disease; one of these (11%) had diabetes. The prevalence of diabetes in patients with cholestatic liver cirrhosis was not different from that of the general population. The frequency of hepatitis G virus coi~e~~on, HLA-DR3, or HLA-DR4 in hepatitis C and diabetes was not different from that of hepatitis C alone. The distribution of hepatitis C virus genotype was similar in those with and those without diabetes. Diabetes was not associated with increased risk of rejection, retransplantation, or death at 1 year after transplantation, and had no impact on overall survival after ~ansplantation. Co~rclusions: 1) The risk of diabetes is not increased in patients with liver cirrhosis due to cholestatic liver disease but is in patients with liver cirrhosis due to hepatitis C or alcoholic liver disease; 2) cofactors (age, sex, body mass index, hepatitis G virus coinfection, hepatitis C virus genotype, or HLA-DR3IDR4) did not explain the increased risk of diabetes in patients with hepatitis C; 3) diabetes before liver transplantation did not change the outcome at 1 year after transplantation or survival.
I
1.5% of the population in the U.S. and Western Europe is infected with this virus (1,2). The possibility of extrahepatic manifestations in the setting of hepatitis C virus infection was suggested first in 1990 (3). Since then, various extrahepatic immunologically mediated syndromes have been recognized to complicate the infection, including essential mixed cryoglobulinemia (4), porphyria cutanea tarda (5), membranous glomerulonephritis (6), Mooren cornea1 ulcer (7), autoimmune
NFECTION
with hepatitis C virus is common throughout the world. It has been estimated that 1.2% to
Received 16 February; revised 26 July; accepted 2 August I999 Corresponderzce: Nizar N. Zein, Division of Gastroenterology and Hepatology and Internal Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA. Tel: 507 284 3335. Fax: 507 284 2107. e-mail: [email protected]
Key wouds: Alcoholic liver disease; Cirrhosis; Diabetes mellitus; Disease; Endpoints, Hepatitis; Liver; Outcome; Tr~splantation.
209
N. N. Zein et al.
thyroiditis (8), Sjiigren syndrome (9), lichen planus (lo), and idiopathic pulmonary fibrosis (11). A possible link between hepatitis C infection and development of diabetes mellitus has been suggested but not completely characterized. The increased aminotransferase levels common in patients with type 2 diabetes have traditionally been attributed to fatty infiltration of the liver (12). However, most of these studies were published before the identification of hepatitis C virus. Recently, 2 studies from the United Kingdom and Spain (13,14) showed a high prevalence of hepatitis C virus antibodies in patients with type 2 diabetes, suggesting a possible link between the two disorders. Another study from the United Kingdom by Allison and colleagues (15) provided evidence for increased prevalence of type 2 diabetes in patients with hepatitis Cassociated liver cirrhosis compared with those with liver cirrhosis due to other causes, supporting the likelihood of a causal relationship. A more recent Italian study provided evidence against a potential association between the two disorders (16). Hepatitis G virus is a positive-stranded RNA virus of the Flaviviridae family (17). The genomic organization of this virus is similar to that of the hepatitis C virus, but phylogenetic studies have demonstrated that the viruses are distinct (18-20). The blood-borne nature of hepatitis G virus has been well demonstrated by analyses showing the appearance of the virus in patients after transfusion (17-20). This mechanism of transmission has also been implied by the high frequency of infection among intravenous drug users (21), hemodialysis patients (19,22), and persons with chronic hepatitis C (23). Although hepatitis G virus has not been clearly associated with any human disease, its high prevalence in patients with HCV raises the possibility of it being a cofactor indevelopment disorders that have been attributed to HCV, including diabetes mellitus. In this study, we report the prevalence of type 2 diabetes mellitus in patients in the U.S. with end-stage liver disease and explore the factors that potentially may be associated with the development of diabetes in patients with hepatitis C virus infection, including coinfection with hepatitis G virus. We also examine the impact of pretransplantation diabetes on liver transplantation outcome in patients with end-stage liver disease.
Materials and Methods Patients and controls All patients with a pretransplantation diagnosis of hepatitis C who underwent liver transplantation at our institution between 1986 and 1996 were identified (73 patients). Diagnosis of hepatitis C was based on the detection of hepatitis C virus antibodies by commercially avail-
210
able enzyme immune antibody assay (EIA 2.0, Ortho Diagnostic Systems, Raritan, NJ, USA) and confirmed by a strip immune assay (SIA 2.0, Ortho Diagnostic Systems, Raritan, NJ, USA). In patients who were evaluated for liver transplantation before the introduction of routine hepatitis C virus testing, the diagnosis was made using stored sera. Of the 73 patients, 64 had hepatitis C as the only diagnosis and nine had hepatitis C plus primary biliary cirrhosis or primary sclerosing cholangitis. The demographic data collected at the time of pretransplantation evaluation included age, sex, alcohol use, and the estimated time of acquisition of hepatitis C virus. Data on alcohol consumption in patients with HCV infection were obtained from a standard questionnaire given to all patients undergoing a pretransplantation evaluation. Patients were considered to have a significant alcohol consumption if their mean daily consumption was equal to or greater than 30 g. If accurate data were not recorded on the questionnaire, medical records were reviewed. Patients with hepatitis C were compared with two other groups of patients who had similarly advanced liver disease. The first group included all patients who underwent transplantation at our institution for end-stage liver disease due to cholestatic liver disease between 1986 and 1990 (79 patients, of whom 37 had primary biliary cirrhosis and 42 had primary sclerosing cholangitis) who had no serologic or virologic markers for hepatitis B or C. The second group included those who underwent transplantation at our institution between 1986 and 1996 for alcoholic liver disease in the absence of serologic and virologic markers of hepatitis B or C or evidence for any other liver disease (53 patients). Blood glucose values of all patients and controls were reviewed retrospectively to identify those with diabetes. The prevalence of this disease among patients with liver cirrhosis of different causes was compared with that expected in an age- and sex-matched population. The data on diabetes prevalence in a population without liver cirrhosis were based on the prevalence of diabetes among Olmsted County, Minnesota, residents, which was published recently as part of an epidemiologic study performed at our institution and included all the residents of Olmsted County (24). The diagnosis of diabetes was based on the status at the time of assessment for liver transplantation candidacy. Cases were identified using criteria recommended by the National Diabetes Data Group (25) for the diagnosis of diabetes. These included fasting blood glucose of ~140 mg/dl on 2 separate occasions in those not receiving hypoglycemic medications or corticosteroids. Medications used during the pretransplantation follow-up period were obtained from medical records. Medications were recorded during each weekly visit before transplantation on a special form to ensure accurate records of medications used by each patient. Patients who were receiving an oral hypoglycemic agent or insulin at the time of evaluation were assumed to have diabetes. Patients were excluded if they were receiving treatment with corticosteroids or interferon at the time of blood glucose determination or if 2 blood glucose determinations within a year before transplantation were not available. Although C-peptide or insulin levels were not measured to differentiate type 1 diabetes from type 2 diabetes, these patients were prospectively followed at our institution over a long period and were not prone to ketosis. Patients who had a history of diabetic ketoacidosis or those younger than 30 years presenting with a clinical requirement for insulin were excluded. Based on the clinical information, diabetic patients in this study were assumed to have type 2 diabetes mellitus. Patients who had transplantation for hepatitis B-associated liver disease were excluded because of their small number at our institution. We did not use the new criteria for diagnosis of diabetes mellitus (fasting blood glucose value of >126 mg/dl) that was proposed by the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus (26). This allowed us to compare our findings with those published to date, including the prevalence of diabetes mellitus in the general population and in patients with liver cirrhosis.
Hepatitis C virus RNA detection Viremia was confirmed in all patients with hepatitis C virus-positive antibodies by the detection of hepatitis C virus RNA using a poly-
Risk of diabetes in liver cirrhosis merase chain reaction assay developed in our laboratory (27). In summary, hepatitis C virus RNA was isolated from 100 pl of serum using a commercially available reagent (RNAzol B, Biotecx Laboratories, Houston, TX, USA) according to the manufacturer’s instructions, with 1 modification, the addition of 20 pg of glycogen to each sample as a carrier during isopropanol precipitation. Hepatitis C virus RNA was reverse transcribed and partially amplified with the polymerase chain reaction using primers specific for the 5’ untranslated region of the viral genome. Amplication products were detected by Southern blotting. Hepatitis C virus genotyping For the purpose of genotyping hepatitis C virus, polymerase chain reaction amplification with primers specific for the NS 5 region of the hepatitis C virus genome was performed according to our previously published protocols (28). To avoid possible contamination with exogenous sequences during extraction or amplification, nucleic acid extraction, amplification, and detection steps were performed in separate laboratories. Negative and positive controls were extracted, reverse transcribed, and amplified in each batch of samples tested with the polymerase chain reaction. Each fragment of approximately 401-bp (370 to 410 bp) polymerase chain reaction product was desalted before sequencing using a direct column purification method (Wizard PCR preps DNA purification system, Promega, Madison, WI, USA). Automated sequencing was performed with a standard Sanger procedure involving the incorporation of fluorescein-labeled dideoxynucleotides and detection on an acrylamide gel (ABI model No. 373 A, Applied Biosystems, Hercules, CA, USA). Nucleotide sequences were aligned and compiled with the previously reported sequences by using the Pileup program (courtesy of Wisconsin Genetic Computer Group, Madison, WI, USA [29]). Cluster analysis was performed using the unweighted pair mean group average (UPMGA) included in the program. These methods allowed comparison of a 222-bp fragment of DNA homologous to nucleotide posititions 7975 to 8196 in the prototype virus. A standardized system recently proposed for the classification of hepatitis C virus genotypes and subtypes was adopted for our study (30). According to this system, hepatitis C virus can be classified into 6 major genotypes comprising a total of 11 subtypes. Hepatitis G virus RNA extraction and detection Hepatitis G virus RNA was extracted and then reverse transcribed following the same procedure for hepatitis C virus RNA and using previously published primers (17). Polymerase chain reaction amplification was accomplished by addition of a second mixture containing 0.8XPCR buffer II (500 mM KCl, 100 mM Tris-HCl, pH 8.3, Perkin Elmer, Norwalk, CT, USA), 2.5 pl of Amplitaq DNA polymerase (Perkin Elmer, Norwalk, CT, USA), 10% (vol/vol) glycerol, 1.75 mM magnesium chloride, 190 mM dUTP, 10 mM dig dUTP (Boehringer Mannheim, Mannheim, Germany), UNG (1 pl/bl, Epicentre Technologies, Madison, WI, USA), and 0.5 PM sense primer 77F (5’-CTC TTT GTG GTA GTA GCC GAG AGA T-3’) (21). Amplifications were carried out in a DNA thermal cycler 480 as follows: 50°C for 10 min, 94°C for 4 min, then 50 cycles of 94°C for 1 min and 58°C for 1 min, followed by a final extension step of 72°C for 5 min. A 40-~1 aliquot of each reverse transcriptase polymerase chain reaction was then tested using the polymerase chain reaction ELISA DIG Detection Kit (Boehringer Mannheim, Mannheim, Germany) and capture probe 152F (S’TCG GTT ACT GAG AGC AGC TCA GAT GAG-3’) for detection of the 156-bp amplification product (21). HLA dermination A standard microlymphocytotoxicity technique was used for class I and class II typing in patients with hepatitis C virus. Several locally derived and commercially available sera were used to define each specificity, as described elsewhere (31,32). Post-transplantation outcome To compare liver transplantation outcome in patients with hepatitis C and diabetes with those without diabetes, we used rejection, re-
transplantation, or death at 1 year after transplantation as the end point. We also evaluated the effect of different variables, including diabetes, on overall survival after liver transplantation. Statistical analysis Two-sample t-tests were used to compare the continuous variables for 2 groups, and the Fisher exact test was used to compare binary outcomes. Because of the small number of patients infected with each hepatitis C virus genotype, all tests of associations between genotype and other factors were based on data that were collapsed into 4 groups: genotypes la, lb, 2a-b, and 3a-4a. The SAS statistical analysis package (SAS Institute, Cary, NC, USA) and S-PLUS (MathSoft, Seattle, WA, USA) were used for all calculations. The Cox proportional hazards model was used to assess the effect of the independent variables on survival. The expected number of patients with diabetes was calculated by applying the age- and sexspecific rates obtained from the Olmsted County, Minnesota, general population. Confidence intervals and p-values for comparing observed and expected number of diabetes cases were based on the Poisson distribution,
Results Patient characteristics and prevalence of diabetes Of 204 patients with end-stage liver disease, 28 (13.7%)
had diabetes. Diabetes was present in 16 of 64 patients (25%) with hepatitis C (not including the nine patients with hepatitis C plus either primary biliary cirrhosis or primary sclerosing cholangitis), one of 78 (1.3%) with cholestatic liver disease (primary biliary cirrhosis or primary sclerosing cholangitis), and ten of 53 (19%) with alcoholic cirrhosis. Of the nine patients with hepatitis C plus cholestatic liver disease, one (11%) had diabetes mellitus. The prevalence of diabetes among patients with hepatitis C was significantly higher than among those with cholestatic liver disease (p=O.OOOl). Similarly, diabetes was more common among those with alcoholic liver cirrhosis than cholestatic liver cirrhosis (p=O.O04). Although diabetes was more prevalent among patients with hepatitis C than among those with alcohol-induced cirrhosis, the difference was not significant (p=O.36). The mean age of patients with hepatitis C (48.72 1.2 years) was marginally higher (p=O.O5) than for those with cholestatic liver cirrhosis (45.7+ 1.1 years) but not different from that of those with alcoholic cirrhosis (52.32 1.l years). Sex distribution was also different: 25/78 (32.1%) of those with cholestatic liver disease were males compared with 44/73 (60.3%) of those with hepatitis C (~~0.001) and 40/53 (75.5%) of those with alcoholic liver cirrhosis (~~0.001). Sex distribution was not different between the hepatitis C and alcoholic cirrhosis groups. No differences were noted among the 3 groups with regard to family history of diabetes. The body mass index of patients with hepatitis C and those with alcoholic liver disease (2820.6 and 27.3kO.7, respectively) was slightly higher than that of patients with cholestatic liver disease (23.450.5). However, the former was also more likely to have detectable ascites 211
N. N. Zein et al,
(67% and 88% of patients with hepatitis C and alcoholic liver disease, respectively, had ascites, compared with 54% of patients with cholestatic liver disease), precluding the ability to compare the body mass index among the 3 groups. Risk factors for diabetes Age, sex, family history, and body mass index were not different between those with diabetes and those without diabetes for either the cholestatic or alcoholic groups {Table 1). For the hepatitis C group, the mean age of those with diabetes (53.0~2.3 years) was marginally higher (p=O.O4) than that of those without diabetes (47.321.3). The difference between the 2 groups for the number of years from exposure to hepatitis C virus to liver transplantation, based on the available data, was not significant (Table 1). Also, there were no significant differences between the 2 groups with respect to sex distribution, excessive alcohol intake, hepatitis G virus coinfection, or HLA-DR3 and HLADR4 (Table 1). Five of 17 (29%) patients with hepatitis C and diabetes had a family history of diabetes, as compared with one of 56 (1.8%) patients with hepatitis C alone (p=O.O002). These data may suggest an important role for family history as a risk factor for the development of diabetes in patients with hepatitis C cirrhosis. The hepatitis C virus genotype was determined for 12 of the 17 (70%) patients with hepatitis C and diabetes and for 47 of the 56 (83%) patients without dia-
betes (Table 1). Genotyping was not possible on the remaining samples because of poor sequencing results. The genotype distribution in patients with diabetes was la in three (25O/), lb in five (42%), and 2b in two (17%), and in patients without diabetes, la in 22 (47%), lb in 16 (34%), 2a + b in seven (15%), and 3a and 4a in two (4%). Genotype distribution was not different among those with and those without diabetes (p= 0.74). Observed versus expected number with diabetes mellitus Table 2 shows the observed number of cases of diabetes by sex for the hepatitis C, cholestatic liver disease, and alcoholic cirrhosis groups. The expected age- and sexmatched prevalence rates of diabetes were obtained from prevalence of adult-onset diabetes mellitus among Olmsted County, Minnesota, residents older than 45 years on January 1, 1990. Our samples included people younger than 45 years, Their expected number was calculated using rates for those 45 to 54 years old. This results in a higher number of expected, because the prevalence of diabetes increases with age until age 65 to 74. Thus, the results presented here are conservative. The observed number of cases of diabetes among patients with hepatitis C was significantly higher than the expected number for both men and women (6.2 times higher for men and 10.1 times higher for women, p
TABLE 1 Characteristics of patients with liver cirrhosis* Characteristic
Mean age, years Males Body mass index Family history of DM Years of hepatitis C virus? Excessive alcohol intake Hepatitis G virus coinfection Hepatitis C virus genotype la lb 2a+b 3a+4a HLA-DR3 HLA-DR4 Rejection at 1 year ~transplan~tion at 1 year Death at any time after liver transplantation
PBCX’SC
Hepatitis C
Alcohol
Without DM (n=56)
With DM (n= 17)
Without DM (n=77)
With DM (n=l)
Without DM (n=43)
With DM (n= 10)
47.32 1.3 35 (62.5) 28.120.7 1 (1.8) 24.521.3 7 (12.5) 6 (11)
53.022.3 9 (53) 27.52 1.5 5 (29) 25.423.0 5 (29) 4 (23.5)
45.421.1 25 (32) 23.5kO.S I (1.3)
57.6 0 18 0
51.721.3 33 (77) 27.5z0.8, 2 (4.7)
55121.7 7 (70) 26.5?2.0 0
26 (60.5) 5 (11.6) 13 (30)
7 (70) 1 (10) 2 (20)
22147(47) 16147(34) 7i47 (15) 2147 (4) 9 (16) 8 (14.3) 35 (62.5) 8 (14.2) 11 (19.6)
3112 (25) S/l2 (42) 202 (16.7) 2l12 (16.7) 1 (6) 4 (23.5) 8 (47) 1 (5.9) 1 (5.9)
37 (48) 6 (7.8) 22 (28.6)
0 0 1 (LOO)
DM, diabetes mellitus; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis. * Values are number of patients (%) or value. t Years from the estimated time of exposure to hepatitis C virus to transplantation (available for 52173patients).
212
Risk of diabetes in liver cirrhosis TABLE 2 Observed and expected cases by sex of diabetes mellitus in cholestatic liver disease and hepatitis C* Characteristic
Number
Hepatitis C group Males Females Totalt
Diabetes mellitus
Observed/expected
Observed
Expected
Ratio
95% CI
p-value
1.46 0.79 2.24
6.2 10.1 7.6
(2.8, 11.7) (4.4, 20.0) (4.4, 12.i)
44 29
9 8
73
17
PBCYPSC group Males Females Totalf
25 53 78
0
1 1
0.72 1.16 1.88
0.00 0.86
0.53
(0.0, 4.2) (0.02, 4.8) (0.01, 2.3)
Alcoholic cirrhosis group Males Females Total
40 13 53
7 3 10
1.51 0.39 1.90
4.6 7.7 5.3
(1.9, 9.5) (1.6, 22.5) (2.5, 9.7)
CI, confidence interval; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis. * Expected number of cases was derived from data from residents of Olmsted County, Minnesota (24). + No significant differences by sex (p=O.31). 1 No significant differences by age (p=O.32).
liver disease showed no difference from the expected number of cases in the age- and sex-matched general population without liver disease (Table 2). The observed number of diabetes cases in patients with endstage alcoholic liver disease was also si~ifi~antly higher than that expected for both males and females (4.6 times higher for men and 7.7 times higher for women, p
outcome
and overall survival
Of 28 patients with diabetes (all groups), 15 (53.6%) had at least 1 episode of rejection at 1 year after liver transplantation compared with 98 of 176 patients (55.7%) without diabetes @=O.l). No death occurred in either group at 1 year, but 19 of 176 patients (10.8%) without diabetes required retransplantation compared with two of 28 patients (7.1%) with diabetes (p=O.3) (Table 1). Indications for retransplantation included hepatic artery thrombosis (6 patients), severe cholestasislhepatitis C recurrence (4 patients), ischemic biliary complications (4 patients), primary graft dysfunction (3 patients), and others (4 patients). There were no significant differences in outcome or overall survival between patients with diabetes mellitus and those without it when similar analysis was applied on the basis of the underlying liver disease (Table 1). The overall survival of the group was evaluated. Because follow-up time was different for the groups, Cox proportional hazards models were run to determine if there were any differences in survival time. Sex was the only significant factor for survival of patients with cho-
lestatic liver disease (males at higher risk for death) (p=O.O15) but not for the hepatitis C group or alcoholic cirrhosis group (p=O.16). Neither age at the time of transplantation nor diabetes was significant for overall survival in any of the 3 groups during followup* In the hepatitis C group, a univariate model was performed for the variables specific for this group (alcohol intake, hepatitis G virus, coinfection, hepatitis C virus genotype, HLA-DR3 or HLA-DR4). None of these variables was significant for survival or for death of patients with hepatitis C.
1,000 1 800
;:
-
HCV
- - Aicot~olic -*.-a Cholestatic ---- General population
/
;
0-
0.) . .. . . . . 20-44
. ., . . yyy.~:“:-
,.e”=E:‘-______““‘“’
55-64
45.54
65-74
Age group
Fig. 1. Age-dependent prevalence of diabetes meliitus among Ulmsted County, Minnesota, residents (general population) and in those with liver cirrhosis. Alcoholic, alcoholic liver disease; HCV, hepatitis C; PBCIPSC, cholestatic liver disease (primary biliary cir~hosislpri~ary sclerosing cirrhosis).
213
N. N. Zein et al.
Discussion Impaired glucose tolerance and liver disease frequently occur concurrently (33-35). As many as 80% of patients with liver cirrhosis are glucose intolerant, and 10% to 20% of these patients have diabetes (36,37). It has been proposed that peripheral resistance to insulin as well as impairment in pancreatic insulin secretion may be the underlying mechanisms for diabetes in patients with liver cirrhosis (33-37). In our study, we found that 13.7% of patients with liver cirrhosis had diabetes, consistent with previously published results. However, the prevalence of diabetes was significantly greater among patients with hepatitis C- or alcoholrelated cirrhosis compared with those with cholestatic liver disease. In fact, patients with cholestatic liver cirrhosis had a prevalence of diabetes similar to that of an age- and sex-matched general population of Olmsted County, Minnesota, suggesting that the mechanism of diabetes in patients with liver cirrhosis is related more closely to the underlying cause of liver cirrhosis. The increased risk of diabetes has also been reported in a retrospective study of patients with chronic hepatitis C without liver cirrhosis (38), supporting the role of hepatitis C virus in the pathogenesis of diabetes. These findings may provide an explanation for the discrepancies in the literature based on the selection of the control group. de1 Olmo and colleagues (39) from Spain found no increase in the risk of diabetes in hepatitis Crelated cirrhosis compared with those with alcohol- or hepatitis B-related cirrhosis. However, those with cholestatic liver disease were not included as controls, as in our study. Although both patients with hepatitis C and patients with alcoholic liver disease had a high prevalence of diabetes, patients with both hepatitis C and heavy alcohol consumption did not have increased prevalence of diabetes in comparison with those with hepatitis C and low or no alcohol consumption. This may suggest that hepatitis C and excessive alcohol consumption are not likely to have an additive effect in relation to the risk of having diabetes in any given patient. Family history of diabetes was an important predictor for the development of diabetes in HCV patients but not in the alcoholic or cholestatic disease groups and did not appear to explain the majority of diabetes cases. After adjusting for the family history, patients with HCV continued to have a significantly higher prevalence of diabetes than patients with cholestatic liver disease or than Olmsted County residents. A less likely explanation of the high prevalence of diabetes in patients with end-stage liver disease due to HCV or alcohol is that diabetes accelerates the progression of liver disease in such patients. Our study was
214
not designed to address this issue, but future prospective studies may shed more light on the relationship between diabetes and the underlying liver disease. However, diabetes did not appear to be associated with accelerated disease in our patients after liver transplantation. One of the shortcomings of this study is the selection of patients with decompensating liver cirrhosis. This may not be an accurate representation of all patients with chronic liver disease or compensated liver cirrhosis. Future studies may need to include patients with different stages of liver disease. The small number of patients in each of the three groups of patients with liver diseases is another major shortcoming of our study and may have potentially underestimated or overestimated the prevalence of diabetes in these patients. However, a recent retrospective study that included 1117 patients with chronic viral hepatitis showed a 21% prevalence rate of diabetes in patients with chronic HCV, which is similar to the findings of our study (40). The pathogenesis of diabetes in patients with hepatitis C- or alcohol-associated cirrhosis is not well understood. Features that are present in these 2 disorders and absent from cholestatic liver disease are likely to provide clues to the pathogenesis of “hepatic diabetes.” Increased fat and iron deposition in the liver are common features among patients with hepatitis C virus infection and those with alcoholic liver disease (4143). It has been suggested that liver fat may contribute to insulin resistance (44), which may lead to loss of the restraining effect of insulin on hepatocyte production of glucose, leading to diabetes. This hypothesis is supported by the increased rate of hepatic glucose production in patients with diabetes (45). It also may provide insight into the link between diabetes and nonalcoholic steatohepatitis (46,47). The relationship between iron and the development of diabetes is most evident in hemochromatosis. The mechanism is thought to be related to damage of the exocrine pancreas by excess iron. Excessive hepatic iron deposition in patients with alcohol-related or hepatitis C liver disease is possibly associated with excessive iron deposition in the exocrine pancreas as well, but this hypothesis needs to be evaluated in future studies. Conversely, the pathogenesis of diabetes in patients with hepatitis C may be different from that of patients with alcoholic cirrhosis. We were not able to account for the increased prevalence of diabetes on the basis of family history alone. Body mass index was not different between hepatitis C patients with diabetes and
Risk of diabetes in liver cirrhosis
those without diabetes, providing evidence against it as a major factor in the development of diabetes in these patients. The risk of diabetes increases with age (24), but age alone does not seem to explain the differences found in the present study. Although the mean age of patients with hepatitis C or alcoholic cirrhosis was higher than that of those with cholestatic liver cirrhosis, the difference was marginal (approximately 3 years). Moreover, patients with hepatitis C or alcoholic cirrhosis had a much higher prevalence of diabetes than an age-matched group of patients with cholestatic liver disease or an age-matched general population in Olmsted County, Minnesota (Fig. 1). Several potentially confounding factors in patients with hepatitis C were evaluated, including the infecting hepatitis C virus genotype and coinfection with hepatitis G virus. Previous reports have suggested a relationship between specific hepatitis C virus genotypes and some of the extrahepatic manifestations of hepatitis C virus infection (48). In the present study, hepatitis C virus genotype distribution was similar among the patients with diabetes and those without, suggesting that diabetes is not a genotype-dependent manifestation of the hepatitis C virus. However, a recent study (40) suggested a potential role for HCV genotype 2a in the development of diabetes in patients with chronic HCV infection. Because only three patients were infected with HCV genotype 2a in our study, we were unable to analyze the role of this specific genotype in our patients. Similarly, coinfection with hepatitis G virus was not associated with an increased risk of diabetes in hepatitis C virus-infected patients. Currently, the role of hepatitis G virus in the pathogenesis of any human disease is unclear. The finding that excessive alcohol intake by patients with hepatitis C was not an important factor in the pathogenesis of diabetes may suggest that the effects of alcohol and hepatitis C are not additive with respect to the development of diabetes. Uncommon forms of immune-mediated diabetes have been described. These include stiff-man syndrome, an autoi~une disorder of the central nervous system associated with high titers of antibodies to glutamic acid decarboxylase and diabetes (49). Anti-insulin antibodies can cause diabetes by binding to the insulin receptor, thereby blocking the binding of insulin to its receptors in target tissues (50). We previously reported significant increase in the production of autoantibodies in autoimmune disorders in patients with hepatitis C (51). The production of autoantibodies was strongly associated with HLA-DR3, whereas autoimmune disorders were associated with HLA-DR4 (51). In the present study, patients with hepatitis C and dia-
betes were not more likely to have HLA-DR3 or HLADR4, suggesting that autoimmunity is not a likely cause of diabetes in these patients. The results of orthotopic liver transplantation in patients with diabetes are controversial. A study from Spain suggested that the prognosis of patients with posttransplantation diabetes is worse than that of those without diabetes (52). A study from the U.S. showed that transplantation could be accomplished successfully in patients with diabetes, with results similar to those of patients without diabetes, although infections and renal dysfunction were more common in the former group (53). Because liver transplantation outcome in patients with hepatitis C or alcoholic cirrhosis is less favorable than that of those with cholestatic liver disease, it is important to rule out diabetes as a negative prognostic sign for transplantation outcome, We compared outcomes, including rejection, retransplantation, and death at 1 year, as well as overall survival among those with and those without diabetes. Pretransplantation diabetes had no impact on the outcome of transplantation and did not increase the likelihood of allograft rejection. It is important to understand the limitations of these findings, because patients with complicated diabetes (presence of end-organ damage, including retinopathy, nephropathy, or vasculitis) were excluded from transplantation. Therefore, our data are applicable only to those with pretransplantation diabetes without significant diabetes-related complications. In summary, our results suggest that the underlying cause of liver cirrhosis is an important determinant of the risk of diabetes. The cause of diabetes in patients with hepatitis C or alcoholic cirrhosis is not clear. Direct viral-induced injury, insulin resistance related to excess liver fat, or excess iron deposition may have a role and need to be investigated in future studies. If the findings of this study are confirmed in future investigations, diabetes should not be included in the clinical spectrum of liver cirrhosis; however, it should be added to the list of extrahepatic manifestations of hepatitis C and alcoholic liver cirrhosis.
Acknowledgements This study was supported in part by a grant from the American Liver Foundation (NNZ). This study was presented in part at the annual meeting of the American Association for the Study of Liver Diseases, Chicago, Illinois, November 1996.
References 1. Alter MJ. Epidemiology of hepatitis C in the West. Semin Liver Dis 1995; 15: 5-14. 2. Shakil AO, Conry-Cantilena C, Alter HJ, Hayashi P, Kleiner DE, Tedeschi V, et al. Volunteer blood donors with antibody to hepa-
21.5
N. N. Zein et al.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12. 13.
14.
15.
16.
17.
18. 19. 20. 21.
22.
23.
24.
216
titis C virus: clinical, biochemical, virologic, and histologic features. The Hepatitis C Study Group. Ann Intern Med 1995; 123: 330-7. Pascual M, Perrin L, Giostra E, Schifferli JA. Hepatitis C virus in patients with cryoglobulinemia type II [letter]. J Infect Dis 1990; 162: 569. Gumber SC, Chopra S. Hepatitis C: a multifaceted disease. Review of extrahepatic manifestations. Ann Intern Med 1995; 123: 615-20. Ferri C, Greco F, Longombardo G, Palla P Moretti A, Marzo E, et al. Association between hepatitis C virus and mixed cryoglobulinemia. Clin Exp Rheumatol 1991; 9: 6214. DeCastro M, Sanchez J, Herrera JF, Chaves A, Duran R, GarciaBuey L, et al. Hepatitis C virus antibodies and liver disease in patients with porphyria cutanea tarda. Hepatology 1993; 17: 551-7. Johnson RJ, Gretch DR, Yamabe H, Hart J, Bacchi CE, Hartwell P, et al. Membranoproliferative glomerulonephritis associated with hepatitis C virus infection. N Engl J Med 1993; 328: 46570. Wilson SE, Lee WM, Murakami C, Weng J, Moninger GA. Mooren-type hepatitis C virus-associated cornea1 ulceration. Ophthalmology 1994; 101: 73645. Haddad J, Deny P Munz-Gotheil C, Ambrosini JC, Trinchet JC, Pateron D, et al. Lymphocytic sialadenitis of Sjiigren’s syndrome associated with chronic hepatitis C virus liver disease. Lancet 1992; 339: 321-3. Gruppo Italian0 Studi Epidemiologici in Dermatologia (GISED). Lichen planus and liver diseases: a multicentre case-control study. BMJ 1990; 300: 227730. Ueda T, Ohta K, Suzuki N, Yamaguchi M, Hirai K, Horiuchi T, et al. Idiopathic pulmonary fibrosis and high prevalence of serum antibodies to hepatitis C virus. Am Rev Respir Dis 1992; 146: 2668. Stone BG, Van Thiel DH. Diabetes mellitus and the liver. Semin Liver Dis 1985; 5: 8-28. Gray H, Wreghitt T, Stratton IM, Alexander GJ, Turner RC, O’Rahilly S. High prevalence of hepatitis C infection in AfroCaribbean patients with type 2 diabetes and abnormal liver function tests. Diabet Med 1995; 12: 2449. Simo R, Hernandez C, Genesca J, Jardi R, Mesa J. High prevalence of hepatitis C virus infection in diabetic patients. Diabetes Care 1996; 19: 998-1000. Allison ME, Wreghitt T, Palmer CR, Alexander GJ. Evidence for a link between hepatitis C virus infection and diabetes mellitus in a cirrhotic population. J Hepatol 1994; 21: 1135-9. Mangia A, Schiavone G, Lezzi G, Marmo R, Bruno F, Villani MR, et al. HCV and diabetes mellitus: evidence for a negative association. Am J Gastroenterol 1998; 93: 2363-7. Linnen J, Wages J Jr, ZhangKeck ZY, Fry KE, Krawczynski KZ, Alter H, et al. Molecular cloning and disease association of hepatitis G virus: a transfusion-transmissible agent. Science 1996; 271: 505-8. Pessoa MG, Wright TL. Hepatitis G: a virus in search of a disease. Hepatology 1996; 24: 461-3. Alter HJ. The cloning and clinical implications of HGV and HGBV-C (editorial). N Engl J Med 1996; 334: 1536-7. Di Bisceglie AM. Hepatitis G virus infection: a work in progress (editorial). Ann Intern Med 1996; 125: 772-3. Aikawa T, Sugai Y, Okamoto H. Hepatitis G infection in drug abusers with chronic hepatitis C [letter]. N Engl J Med 1996; 334: 195-6. Masuko K, Mitsui T, Iwano K, Yamazaki C, Okuda K, Meguro T, et al. Infection with hepatitis GB virus C in patients on maintenance hemodialysis. N Engl J Med 1996; 334: 1485-90. Tanaka E, Alter HJ, Nakatsuji Y, Shih JW, Kim JP, Matsumoto A, et al. Effect of hepatitis G virus infection on chronic hepatitis C. Ann Intern Med 1996; 125: 74&3. Leibson CL, O’Brien PC, Atkinson E, Palumbo PJ, Melton LJ III. Relative contributions of incidence and survival to increasing
25.
26.
27.
28.
29. 30.
31.
32.
33.
34. 35.
36.
37. 38.
39. 40.
41.
42.
43.
44.
45.
46.
prevalence of adult-onset diabetes mellitus: a population-based study. Am J Epidemiol 1997; 146: 12-22. National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979; 28: 1039-57. Gavin JR, Alberti KGMM, Davidson MB, DeFronzo RA, Drash A, Gabbe SG, et al. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 1998; 21 Suppl 1: S5-19. Zein NN, McGreger CG, Wendt NK, Schwab K, Mitchell PS, Persing DH, et al. Prevalence and outcome of hepatitis C infection among heart transplant recipients. J Heart Lung Transplant 1995; 14: 865-9. Zein NN, Rakela J, Krawitt EL, Reddy KR, Tominaga T, Persing DH. Hepatitis C virus genotypes in the United States: epidemiology, pathogenicity, and response to interferon therapy. Collaborative Study Group. Ann Intern Med 1996; 125: 6349. Genetics computer group. Program manual for GCG package, version 7. Madison,Wisconsin 53711. April 1991. Simmonds P, Alberti A, Alter HJ, Bonino F, Bradley DW, Brechat C, et al. A proposed system for the nomenclature of hepatitis C viral genotypes [letter]. Hepatology 1994; 19: 13214. Czaja AJ, Rakela J, Hay JE, Moore SB. Clinical and prognostic implications of HLA B8 in corticosteroid-treated severe autoimmune chronic active hepatitis. Gastroenterology 1990; 98: 158793. Czaja AJ, Strettell MD, Thomson LJ, Santrach PJ, Moore SB, Donaldson PT, et al. Associations between alleles of the major histocompatibility complex and type 1 autoimmune hepatitis. Hepatology 1997; 25: 317-23. Petrides AS, DePronzo RA. Glucose metabolism in cirrhosis: a review with some perspectives for the future. Diabetes Metab Rev 1989; 5: 691-709. Megyesi C, Samols E, Marks V Glucose tolerance and diabetes in chronic liver disease. Lancet 1967; 2: 1051-6. Proietto J, Alford FP, Dudley FJ. The mechanism of the carbohydrate intolerance of cirrhosis. J Clin Endocrinol Metab 1980; 51: 103&6. Kruszynska YT, Home PD, McIntyre N. Relationship between insulin sensitivity, insulin secretion and glucose tolerance in cirrhosis. Hepatology 1991; 14: 103-l 1. Nolte W, Hartmann H, Ramadori G. Glucose metabolism and liver cirrhosis. Exp Clin Endocrinol Diabetes 1995; 103: 63-74. Fraser GM, Harman I, Meller N, Niv Y, Porath A. Diabetes mellitus is associated with chronic hepatitis C but not chronic hepatitis B infection. Isr J Med Sci 1996; 32: 52630. de1 Olmo JA, Serra MA, Rodrigo JM. Liver cirrhosis and diabetes mellitus [letter]. J Hepatol 1996; 24: 645. Mason AL, Lau JY, Hoang N, Qian K, Alexander GJ, Xu L, et al. Association of diabetes mellitus and chronic hepatitis C virus infection. Hepatology 1999; 29: 328-33. Fiore G, Fera G, Napoli N, Vella F, Schiraldi 0. Liver steatosis and chronic hepatitis C: a spurious association? Eur J Gastroenter01 Hepatol 1996; 8: 125-9. Riggio 0, Montagnese F, Fiore P, Folino S, Giambartolomei S, Gandin C, et al. Iron overload in patients with chronic viral hepatitis: how common is it? Am J Gastroenterol 1997; 92: 1298301. Eaton S, Record CO, Bartlett K. Multiple biochemical effects in the pathogenesis of alcoholic fatty liver. Eur J Clin Invest 1997; 27: 719-22. Banerji MA, Buckley MC, Chaiken RL, Gordon D, Lebovitz HE, Kral JG. Liver fat, serum triglycerides and visceral adipose tissue in insulin-sensitive and insulin-resistant black men with NIDDM. Int J Obes Relat Metab Disord 1995; 19: 84650. Petrides AS, Vogt C, Schulze-Berge D, Matthews 4 Strohmeyer G. Pathogenesis of glucose intolerance and diabetes mellitus in cirrhosis. Hepatology 1994; 19: 61627. Ludwig J, McGill DB, Lindor KD. Review: nonalcoholic steatohepatitis. J Gastroenterol Hepatol 1997; 12: 398403.
Risk of diabetes in liver cirrhosis 47. Sheth SG, Gordon FD, Chopra ‘S.Nonalcoholic steatohepatitis. Ann Intern Med 1997; 126: 13745. 48. Chin CF, Zein NN. Extrahepatic syndromes associated with hepatitis C infection. Gastroenterol Int 1998; 11: 26-35. 49. Bingley PJ, Bonifacio E, Williams AJ, Genovese S, Bottazzo GF, Gale EA. Prediction of IDDM in the general population: strategies based on combinations of autoantibody markers. Diabetes 1997; 46: 1701-10. 50. Taylor SI. Lilly Lecture: molecular mechanisms of insulin resistance. Lessons from patients with mutations in the insulin-receptor gene. Diabetes 1992; 41: 1473-90.
51. Zein NN, Persing DH, Czaja AJ. Viral and genetic factors as determinants of autoimmune expression in chronic hepatitis C [abstract]. Gastroenterology 1996; 1lO(Supp1): A1367. 52. Navasa M, Bustamante J, Marroni C, Gonzalez E, Andreu H, Esmatjes E, et al. Diabetes mellitus after liver transplantation: prevalence and predictive factors. J Hepatol 1996; 25: 64-71. 53. Trail KC, Stratta RJ, Larsen JL, Ruby EI, Patil KD, Langnas AN, et al. Results of liver transplantation in diabetic recipients. Surgery 1993; 1 14: 650-6.