| |
Insulin Resistance Impairs Response to IFN/RBV
|
| |
| |
"Insulin resistance impairs sustained response rate to peginterferon plus ribavirin in chronic hepatitis C patients"
Gastroenterology
March 2005 Vol 128, Numb 3
Manuel Romero-G—mez et al
Hospital Universitario de Valme
--Patients with insulin resistance were less likely to achieve an SVR (viral response to therapy)
--In genotype 1 patients this was particularly stark
--Fibrosis and BMI were associated in this study with having insulin resistance
--Patients who cleared HCV by month 6 experienced decreased insulin resistance.... These data support a link between active HCV replication and insulin resistance...
"...insulin resistance could be a pathogenic factor in the resistance to interferon/ribavirin therapy. As such, therapeutic intervention aimed at decreasing insulin resistance may be warranted in these patients before or during peginterferon and ribavirin treatment...'
"...It is interesting to note that insulin resistance seems to be a common denominator among groups of poor-response patients, such as those with cirrhosis4 or a high BMI, African Americans, and those with HIV/HCV co-infection. Whether insulin resistance is a marker for patients who are very difficult to treat or whether it plays a role in interferon plus ribavirin resistance remains unclear... TNF levels could predict the risk of developing diabetes mellitus in patients with chronic hepatitis C..."
ABSTRACT
Background & Aims: We evaluated the effect of insulin resistance and viral factors on sustained virological response in patients with chronic hepatitis C treated with peginterferon plus ribavirin.
Methods: Patients (n = 159; 94 men; age, 41.7 ± 11.1 years) with chronic hepatitis C (genotype 1, n = 113; non-1 genotype, n = 46) received treatment with interferon plus ribavirin. Serum levels of leptin and insulin were measured, and the insulin resistance index (HOMA-IR: homeostasis model of assessment) and body mass index were calculated.
Results: A sustained virological response was associated with lower age, insulin resistance index, body mass index, and Ei-glutamyltranspeptidase and serum leptin levels. There was no association with viral load, sex, type of interferon, or cholesterol levels.
A sustained virological response was achieved in 43.4% (46/113) of genotype 1 and 89% (32/36) of genotype 2 and 3 (P = .0001) patients. Necroinflammatory activity and steatosis were not associated with the sustained virological response rate. Multivariate regression analysis indicated that the independent variables related to sustained virological response were genotype (odds ratio, 3.57; 95% confidence interval, 1.49-8.3; P = .001), insulin resistance index (odds ratio, 1.82; 95% confidence interval, 1.08-3.06; P = .012), and fibrosis (odds ratio, 1.36; 95% confidence interval, 1.01-1.84; P = .029). A sustained virological response in patients with genotype 1 and insulin resistance (HOMA-IR > 2) occurred in 23 of 70 (32.8%; 95% confidence interval, 21.9%-43.9%) patients, vs. 26 of 43 (60.5%; 95% confidence interval, 45.9%-75.1%) genotype 1 patients without insulin resistance (P = .007; odds ratio, 3.12, 95% confidence interval, 1.42-6.89).
Conclusions: Insulin resistance, fibrosis, and genotype are independent predictors of the response to antiviral therapy in chronic hepatitis C patients treated with peginterferon plus ribavirin.
INTRODUCTION
Host and viral factors influence the sustained virological response (SVR) rate to combined interferon/ribavirin therapy. Low viral load and non-1 genotype are the most favorable viral factors,1 whereas host factors, including HLA class I2 and II3, ethnicity,4 and body mass index (BMI),5 could also influence SVR.6 Indeed, overweight5 and African American7 patients have shown a characteristic resistance to combination therapy, and increased BMI has been identified as an independent variable associated with poorer response.5 Experimental and clinical studies support a role of hepatitis C virus (HCV) infection in the development of insulin resistance.8 Patients with mild chronic hepatitis have a higher homeostasis model of assessment insulin resistance index (HOMA-IR) than healthy controls matched for age and BMI.9 Also, insulin resistance has been implicated in fibrosis progression10 and steatosis development11 in chronic hepatitis C infection. It is interesting to note that insulin resistance is frequently seen in very difficult to treat chronic hepatitis C patients, such as cirrhotics, African Americans, overweight patients, and co-infected human immunodeficiency virus (HIV)/HCV patients. The aim of this study was to analyze the effect of body composition, the leptin system, and insulin resistance, together with viral factors, on SVR in patients with chronic hepatitis C treated with interferon plus ribavirin.
Insulin resistance, advanced fibrosis, and genotype 1 are independent predictors of a poor response to antiviral therapy in chronic hepatitis C. It is interesting to note that insulin resistance seems to be a common denominator among groups of poor-response patients, such as those with cirrhosis4 or a high BMI,13 African Americans,14 and those with HIV/HCV co-infection.15 Whether insulin resistance is a marker for patients who are very difficult to treat or whether it plays a role in interferon plus ribavirin resistance remains unclear. Other host factors associated with insulin resistance, such as increased BMI and hepatocyte steatosis, have also been observed to be associated with a poorer SVR rate.16 In this study, both variables were associated with impaired SVR in univariate statistical analysis. However, in the multivariate analysis, HOMA-IR remained as an independent variable, and BMI and steatosis did not enter into the statistical equation. Our data strongly suggest that HOMA-IR may assist in further refining the prediction of antiviral response in patients infected by genotype 1. HOMA-IR was higher in advanced fibrosis and in patients infected by genotype 1 and, despite these associations, these variables seemed to be independently related to SVR. Thus, HOMA-IR could be the best host marker to predict SVR to combined therapy, at least in genotype 1-infected patients. Also, because insulin resistance can be modified by changes in lifestyle habits and by insulin-sensitizer drugs, this could be a potential target for treatment in chronic hepatitis C. Baseline viral load was not associated with SVR. Genotype has been found to be a stronger factor associated with SVR than baseline HCV RNA levels.6 Indeed, in the largest pivotal study from Fried et al,1 baseline HCV RNA was associated with SVR in the entire group, but there was no difference according to low or high viral load when patients were segregated by genotype. In a multivariate analysis, baseline HCV RNA was not an independent variable related to SVR,1 and a lack of association between the baseline viral load and the chance of achieving SVR has also been previously reported.2,5
Interferon induces insulin resistance immediately after administration in healthy controls.17 In chronic hepatitis C patients treated for 2 weeks, interferon induces a decrease in glucose uptake by peripheral tissue and the liver.18 However, this effect disappears when analyzed after 3 months of treatment19 or during follow-up at the conclusion of treatment.20 Thus, interferon-induced cytokine release could explain the acute increase of insulin resistance that disappears whether treatment is maintained. Furthermore, in patients who cleared HCV RNA, insulin resistance decreased significantly at the end of follow-up, and this provides support for a role of HCV in the development of insulin resistance in this group of patients, as has been reported in an experimental model.8 Nevertheless, the underlying mechanisms are not completely understood. Tumor necrosis factor (TNF) has been implicated in the development of insulin resistance and could be up-regulated by chronic hepatitis C infection. Indeed, TNF levels could predict the risk of developing diabetes mellitus in patients with chronic hepatitis C,21 and TNF levels correlate with HOMA-IR in cirrhotic and noncirrhotic HCV patients.22 Also, TNF has been found to be a marker for a poor response to interferon therapy.23 As such, patients showing insulin resistance might display higher TNF levels and resistance to antiviral therapy. In vitro experiments with liver samples indicate that HCV infection leads to a postreceptor defect in insulin receptor substrate 1, thereby contributing to insulin resistance.24 Indeed, in a murine model that is transgenic for the HCV core gene and that shows high TNF production, HCV infection seemed to be associated with insulin resistance development, and anti-TNF treatment improved the insulin sensitivity in this model.8 Although interferons have been reported to impair glucose tolerance25 and although some types of interferon-alfa could induce alterations in lipoprotein profile,26 in this study, the metabolic changes that occur during and after treatment were not related to the type of peginterferon (alfa-2a or alfa-2b).
In this study, the clearance of the virus induced an improvement in insulin resistance, despite a potential increase in resistance associated with the drug. However, in nonresponders, HOMA-IR and insulin secretion remained unchanged, despite a decrease in BMI. These data support a link between active HCV replication and insulin resistance. Indeed, HOMA-IR varies significantly according to the presence or absence of viremia during and after treatment. As previously reported, our results confirm an association between fibrosis progression and insulin resistance.9,27,28
Finally, we observed that insulin resistance strongly influenced the rate of SVR in patients infected by genotype 1; it ranged from 60% in patients without insulin resistance to only 20% in patients with an altered HOMA-IR (>4). Thus, HOMA-IR could be introduced as a marker that is predictive of SVR in patients infected by genotype 1.
In summary, our study found insulin resistance, together with genotype and fibrosis, to be prognostic factors in relation to the rate of SVR to combined therapy in chronic hepatitis C patients. The discussion remains open, however, as to how insulin resistance could be a marker of these poor-response patients and whether its measurement should be included in the prognostic profile when the therapeutic regimen is planned. Also, insulin resistance could be a pathogenic factor in the resistance to interferon/ribavirin therapy. As such, therapeutic intervention aimed at decreasing insulin resistance may be warranted in these patients before or during peginterferon and ribavirin treatment.
AUTHOR DISCUSSION
Insulin resistance, advanced fibrosis, and genotype 1 are independent predictors of a poor response to antiviral therapy in chronic hepatitis C. It is interesting to note that insulin resistance seems to be a common denominator among groups of poor-response patients, such as those with cirrhosis4 or a high BMI,13 African Americans,14 and those with HIV/HCV co-infection.15 Whether insulin resistance is a marker for patients who are very difficult to treat or whether it plays a role in interferon plus ribavirin resistance remains unclear. Other host factors associated with insulin resistance, such as increased BMI and hepatocyte steatosis, have also been observed to be associated with a poorer SVR rate.16 In this study, both variables were associated with impaired SVR in univariate statistical analysis. However, in the multivariate analysis, HOMA-IR remained as an independent variable, and BMI and steatosis did not enter into the statistical equation. Our data strongly suggest that HOMA-IR may assist in further refining the prediction of antiviral response in patients infected by genotype 1. HOMA-IR was higher in advanced fibrosis and in patients infected by genotype 1 and, despite these associations, these variables seemed to be independently related to SVR. Thus, HOMA-IR could be the best host marker to predict SVR to combined therapy, at least in genotype 1-infected patients. Also, because insulin resistance can be modified by changes in lifestyle habits and by insulin-sensitizer drugs, this could be a potential target for treatment in chronic hepatitis C. Baseline viral load was not associated with SVR. Genotype has been found to be a stronger factor associated with SVR than baseline HCV RNA levels.6 Indeed, in the largest pivotal study from Fried et al,1 baseline HCV RNA was associated with SVR in the entire group, but there was no difference according to low or high viral load when patients were segregated by genotype. In a multivariate analysis, baseline HCV RNA was not an independent variable related to SVR,1 and a lack of association between the baseline viral load and the chance of achieving SVR has also been previously reported.2,5
Interferon induces insulin resistance immediately after administration in healthy controls.17 In chronic hepatitis C patients treated for 2 weeks, interferon induces a decrease in glucose uptake by peripheral tissue and the liver.18 However, this effect disappears when analyzed after 3 months of treatment19 or during follow-up at the conclusion of treatment.20 Thus, interferon-induced cytokine release could explain the acute increase of insulin resistance that disappears whether treatment is maintained. Furthermore, in patients who cleared HCV RNA, insulin resistance decreased significantly at the end of follow-up, and this provides support for a role of HCV in the development of insulin resistance in this group of patients, as has been reported in an experimental model.8 Nevertheless, the underlying mechanisms are not completely understood. Tumor necrosis factor (TNF) has been implicated in the development of insulin resistance and could be up-regulated by chronic hepatitis C infection. Indeed, TNF levels could predict the risk of developing diabetes mellitus in patients with chronic hepatitis C,21 and TNF levels correlate with HOMA-IR in cirrhotic and noncirrhotic HCV patients.22 Also, TNF has been found to be a marker for a poor response to interferon therapy.23 As such, patients showing insulin resistance might display higher TNF levels and resistance to antiviral therapy. In vitro experiments with liver samples indicate that HCV infection leads to a postreceptor defect in insulin receptor substrate 1, thereby contributing to insulin resistance.24 Indeed, in a murine model that is transgenic for the HCV core gene and that shows high TNF production, HCV infection seemed to be associated with insulin resistance development, and anti-TNF treatment improved the insulin sensitivity in this model.8 Although interferons have been reported to impair glucose tolerance25 and although some types of interferon-alfa could induce alterations in lipoprotein profile,26 in this study, the metabolic changes that occur during and after treatment were not related to the type of peginterferon (alfa-2a or alfa-2b).
In this study, the clearance of the virus induced an improvement in insulin resistance, despite a potential increase in resistance associated with the drug. However, in nonresponders, HOMA-IR and insulin secretion remained unchanged, despite a decrease in BMI. These data support a link between active HCV replication and insulin resistance. Indeed, HOMA-IR varies significantly according to the presence or absence of viremia during and after treatment. As previously reported, our results confirm an association between fibrosis progression and insulin resistance.9,27,28
Finally, we observed that insulin resistance strongly influenced the rate of SVR in patients infected by genotype 1; it ranged from 60% in patients without insulin resistance to only 20% in patients with an altered HOMA-IR (>4). Thus, HOMA-IR could be introduced as a marker that is predictive of SVR in patients infected by genotype 1.
In summary, our study found insulin resistance, together with genotype and fibrosis, to be prognostic factors in relation to the rate of SVR to combined therapy in chronic hepatitis C patients. The discussion remains open, however, as to how insulin resistance could be a marker of these poor-response patients and whether its measurement should be included in the prognostic profile when the therapeutic regimen is planned. Also, insulin resistance could be a pathogenic factor in the resistance to interferon/ribavirin therapy. As such, therapeutic intervention aimed at decreasing insulin resistance may be warranted in these patients before or during peginterferon and ribavirin treatment.
RESULTS
Factors associated with sustained virological response
Patients with SVR were younger (39.7 ± 10.4 years vs. 44.5 ± 10.9 years; P = .006), had a lower HOMA-IR (2.36 ± 1.85 vs. 3.76 ± 3.22; P = .001), had a lower BMI (25.4 ± 3.65 kg/m2 vs. 27.1 ± 4.89 kg/m2; P = .027), had lower Ei-glutamyltranspeptidase levels (46.8 ± 49.5 U/L vs. 106.7 ± 107.2 U/L; P = .001), and had lower serum leptin levels in men (5.1 ± 3.97 ng/mL vs. 12.4 ± 13.7 ng/mL; P = .048) and in women (13.3 ± 10.5 ng/mL vs. 20.5 ± 13.5 ng/mL; P = .016). There was no association with sex distribution, type of interferon, alcohol consumption, or cholesterol levels. The HCV viral load was similar in both groups. SVR was achieved in 43.4% (46/113) of genotype 1 patients vs. 89% (32/36) of genotype 2 and 3 patients (P = .0001). The SVR rate was 64.6% in mild fibrosis (F0-F2) and 27.8% in advanced fibrosis (F3-F4; P = .001). Also, the median fibrosis was lower between responders (1.41 ± 0.88 vs. 2.16 ± 1.39; P = .0001). Hepatocyte steatosis was associated with a poorer SVR rate in genotype 1 patients (18.2% vs. 53.7%; P = .001). However, necroinflammatory activity was not associated with the SVR rate.
With reverse stepwise logistic multivariate regression analysis, the independent variables related to SVR were genotype (odds ratio [OR], 3.57; 95% confidence interval [CI], 1.49-8.3; P = .001), HOMA-IR (OR, 1.82; 95% CI, 1.08-3.06; P = .012), and fibrosis (OR, 1.36; 95% CI, 1.01-1.84; P = .029). The SVR rate in genotype 1 patients segregated with respect to HOMA-IR was 60.5% (95% CI, 45.9%-75.1%; 26/43) if HOMA was <2, 40% (95% CI, 25.7%-54.3%; 18/45) if HOMA was between 2 and 4, and only 20% (95% CI, 4.3%-35.7%; 5/25) if HOMA was >4 (P = .004).
Factors associated with insulin resistance
Insulin resistance correlated with BMI (r = 0.51; P = .001), age (r = 0.33; P = .002), leptin levels (r = 0.41; P = .003), serum C-peptide (r = 0.86; P < .001), and fibrosis stage (r = 0.41; P = .001). Despite the BMI correlation with HOMA-IR, insulin resistance was found in 20 of 54 patients (37%) with a BMI <25 kg/m2. HOMA-IR was higher in patients infected by genotype 1 than by genotype 3a (3.26 ± 2.82 vs. 2.28 ± 1.83; P = .003), whereas no difference was seen in BMI or leptin levels with respect to genotypes. In genotype 1 patients, HOMA-IR correlated with hepatocyte steatosis (r = 0.40; P = .001). Neither necroinflammatory activity nor cholesterol levels were associated with insulin resistance. In multivariate analysis, by using reverse stepwise logistic regression, BMI (OR, 1.69; 95% CI, 1.15-2.49; P = .0006) and fibrosis (OR, 2.27; 95% CI, 1.1-5.26; P = .03) were the independent variables associated with insulin resistance.
Insulin resistance and hepatitis C virus viremia
In 50 patients, a group representative of the entire cohort (37 were infected by genotype 1, 28 received peginterferon alfa-2a, and 25 were sustained responders), the HOMA-IR was measured before treatment, at 6 months of treatment, and at the end of follow-up. Figure 2 shows the mean ± SD HOMA values segregated with respect to sustained responders (n = 25), nonresponders (n = 11), and relapsers (n = 14). The HOMA-IR index decreased during the first 6 months of treatment in patients who had achieved clearance of HCV RNA at month 6 (from 2.55 ± 2.52 to 1.85 ± 1.66; P = .09), but it remained unchanged in patients with HCV RNA positive at month 6 (3.65 ± 2.03 to 3.53 ± 1.85; P = not significant). BMI decreased in both groups from 25.6 ± 3.8 kg/m2 to 24.3 ± 4 kg/m2 (P = not significant) in HCV RNA-negative patients and from 26.7 ± 3.8 kg/m2 to 24.9 ± 4.1 kg/m2 (P = not significant) in HCV RNA-positive at month 6 of treatment. In nonresponders, treatment was stopped, and HOMA-IR remained unchanged on follow-up. In sustained responders, HOMA-IR continued to decrease, and the paired t test indicated statistical significance when comparing baseline HOMA-IR with the values at the end of follow-up (2.55 ± 2.52 vs. 1.50 ± 0.77; P < .05). Conversely, in relapsers, the HOMA-IR increased from 6 months to the end of follow-up (Figure 2). The type of peginterferon administered did not seem to influence the changes in glycemia, insulinemia, or HOMA-IR during or after treatment
|
|
| |
| |
|
|
|
