Chronic hepatitis C genotype 1 patients with insulin resistance treated with pioglitazone and peginterferon alpha-2a plus ribavirin

Authors


  • Financial disclosures and potential conflicts of interest: Dr. Harrison advises and is on the speakers' bureau of Merck. He is also on the speakers' bureau of Bristol-Myers Squibb. Dr. Pandya advises and is on the speakers' bureau of Onyx. He is on the speakers' bureau of Vertex and received grants from Merck. Dr. Vierling advises, consults for, and received grants from Abbott, Bristol-Myers Squibb, Excalenz, Gilead, GlobeImmune, Hyperion, Roche, Merck, Sundise, Ocera, and Vertex. He advises and consults for Salix, HepaLife Tehcnologies, Herbalife, and Ocera. He received grants from Conatus, Idenix-Novartis, Ikara, Intercept, Mochida, Novartis, Pfizer, Pharmasset, and Zymogenetics. Dr. Sheikh is on the speakers' bureau of Merck and Salix. He received grants from Roche and Vertex.

Abstract

Patients with chronic hepatitis C and insulin resistance are less likely to respond to anti-hepatitis C virus (HCV) therapy and are at risk for more rapid fibrosis progression. Coadministration of pioglitazone with peginterferon/ribavirin improves insulin sensitivity and increases virologic response rates in insulin-resistant HCV genotype 4 patients, but it is unclear whether this finding applies to genotype 1 patients. For this reason we randomized treatment-naive HCV genotype 1 patients with insulin resistance to receive either standard care (peginterferon alpha-2a plus ribavirin for 48 weeks, n = 73) or pioglitazone 30-45 mg/day plus standard care (n = 77) in an open-label multicenter trial. Patients randomized to pioglitazone received the drug during a 16-week run-in phase, the 48-week standard-care phase, and the 24-week untreated follow-up phase. Pioglitazone treatment improved hemoglobin A1c (HbA1c), plasma glucose, insulin levels, and homeostasis model assessment of insulin resistance score and increased serum adiponectin levels during the 16-week run-in phase and maintained these improvements during the standard-care phase. However, we observed no statistically significant difference between the two groups in the primary efficacy endpoint, the decrease from baseline to Week 12 of peginterferon alpha-2a/ribavirin treatment in mean log10 HCV RNA titer (−3.5 ± 1.71 and −3.7 ± 1.62 IU/mL in the pioglitazone and standard-care groups, respectively, Δ = 0.21 IU/mL, P = 0.4394). Conclusion: Treatment with pioglitazone before and during treatment with peginterferon alpha-2a plus ribavirin improved several indices of glycemic control in patients with chronic hepatitis C and insulin resistance, but did not improve virologic response rates compared with peginterferon alpha-2a plus ribavirin alone. (HEPATOLOGY 2012)

Infection with hepatitis C virus (HCV) is often associated with changes in lipid and carbohydrate metabolism.1 These changes result in insulin resistance and have been associated with a higher prevalence of type 2 diabetes mellitus, accelerated progression of hepatic fibrosis, and increased carotid artery intima thickness.2-4

The association between HCV infection and impaired glucose metabolism is strong. Diabetes is more prevalent in patients with chronic HCV-related cirrhosis than in patients with cirrhosis attributed to a cause other than HCV infection.5 Conversely, the prevalence of HCV infection is higher among patients with diabetes than in the general population.6, 7 The exact prevalence of impaired glucose metabolism in patients with HCV infection is not known, but ranges from 30% to 70% and varies with HCV genotype.1, 8

There is growing evidence that metabolic perturbations associated with HCV infection may result from interactions between viral and host proteins.9-14 HCV core protein reduces expression of insulin receptor substrates (IRS) 1 and 2 and interferes with the cellular response to insulin. Eradication of HCV with peginterferon plus ribavirin therapy restores the expression of IRS-1 and IRS-2 and improves insulin resistance.15 HCV NS5A induces oxidative stress and increases the levels of inflammatory cytokines, which contributes indirectly to insulin resistance by stimulating nuclear factor-κB (NF-κB).9 Insulin resistance and increasing homeostasis model assessment of insulin resistance (HOMA-IR) scores have been associated with high serum HCV RNA levels.16-18

Patients with chronic hepatitis C and insulin resistance are less likely to respond to anti-HCV therapy and are at risk for more rapid fibrosis progression.19-23 Indeed, it is estimated that sustained virologic response (SVR) rates may be reduced by as much as 50% in HCV genotype 1 patients with insulin resistance.19 This suggests that measures that restore or improve insulin sensitivity might lead to better virologic response rates.

Pioglitazone reduces insulin resistance in the liver and peripheral tissues by stimulating the nuclear receptors peroxisome proliferator-activated receptor (PPAR)-γ and -α that control expression of insulin-sensitive genes. Based on these molecular mechanisms, we hypothesized that correcting insulin resistance prior to initiating anti-HCV therapy might improve the virologic response to peginterferon plus ribavirin therapy. Subsequent to initiating this study, it has been reported that the addition of pioglitazone to peginterferon plus ribavirin therapy improved both on-treatment virologic response rates and SVR rates in patients with HCV genotype 4 and insulin resistance.24 It remains to be determined whether treatment with pioglitazone can produce a similar effect in patients infected with HCV genotype 1. For this reason, the randomized, multicenter SENSITIZE study was designed to evaluate the impact of pioglitazone on virologic and metabolic outcomes in insulin-resistant genotype 1 patients receiving peginterferon alpha-2a plus ribavirin.

Abbreviations

HbA1c, hemoglobin A1c; HCV, hepatitis C virus; HOMA-IR, homeostasis model assessment of insulin resistance; IRS, insulin receptor substrates; NF-κB, nuclear factor kappaB; PPAR, peroxisome proliferator-activated receptor; SVR, sustained virologic response.

Patients and Methods

Patients.

Patients eligible for the randomized, multicenter, open-label SENSITIZE study were at least 18 years old, had chronic HCV genotype 1 infection, insulin resistance, and a fasting plasma glucose <240 mg/dL, and had not received insulin within the previous 2 weeks or prior treatment with peginterferon plus ribavirin therapy at any time. Insulin resistance was defined as one of the following: a HOMA-IR score >2, fasting blood glucose >100 mg/dL, or fasting blood insulin >10 μU/mL. Patients were required to have had either a liver biopsy within the previous 2 years that showed an absence of cirrhosis (fibrosis stage F0-3 by Metavir, F0-4 by Ishak, or <4 by Batts/Ludwig) or a FIBROSpect California serodiagnostic test with an index score ≤70 obtained within the previous 2 years. Patients receiving treatment with oral hypoglycemic agents were eligible for the trial with the exception of those who had received a thiazolidinedione (pioglitazone or rosiglitazone) within the previous 16 weeks.

Patients were excluded if they had cirrhosis, type 1 diabetes mellitus, hepatocellular carcinoma, or other comorbid liver disease, or if they were infected with hepatitis A or B viruses or human immunodeficiency virus. Patients with serious concomitant chronic diseases or laboratory abnormalities were also excluded.

Pregnant and breastfeeding women and male partners of pregnant and breastfeeding women were excluded. All nonpregnant fertile women were required to obtain a negative pregnancy test within 24 hours of the first dose of any study drug, and all fertile female patients and male patients with fertile partners were required to use two forms of effective contraception during treatment and for 24 weeks after completion of treatment.

All patients provided written informed consent, the study was conducted in accordance with provisions of the Declaration of Helsinki, and the study was approved by the Institutional Review Boards of the enrolling institutions.

Study Design and Treatment.

Patients were randomized in a 1:1 ratio, stratified by HOMA-IR score (2 to <4 or ≥4), to receive either pioglitazone (Actos, Takeda Pharmaceuticals, Deerfield, IL) or standard care. The centralized computer-generated randomization list was generated and maintained in a central repository by the sponsor. Randomization numbers were communicated to investigators by an interactive voice response system.

Patients randomized to pioglitazone entered a pioglitazone run-in phase during which they received oral pioglitazone for 16 weeks (30 mg/day for 8 weeks, then 45 mg/day for 8 weeks). After the run-in phase, patients received pioglitazone in combination with subcutaneous peginterferon alpha-2a (Pegasys, Roche, Nutley, NJ) plus oral ribavirin (Copegus, Roche) for 48 weeks and then pioglitazone alone during a 24-week follow-up period (Fig. 1). Patients randomized to standard care received peginterferon alpha-2a plus ribavirin for 48 weeks followed by a 24-week untreated follow-up period. For both groups the dose of peginterferon alpha-2a was 180 μg/week, and the dose of ribavirin was 1,000 mg/day for patients weighing <75 kg, 1,200 mg/day for patients weighing 75 to <85 kg, 1,400 mg/day for those weighing 85 to 95 kg, and 1,600 mg/day for those weighing >95 kg. The dose of pioglitazone could be held at 30 mg/day, reduced, or discontinued in the event of an adverse event.

Figure 1.

Study design.

In the event of adverse events or laboratory abnormalities, the dose of peginterferon alpha-2a could be reduced in a stepwise manner to 135, 90, and 45 μg/week, and the dose of ribavirin could be reduced in gradual 200-mg decrements. Ribavirin could be either temporarily or permanently withdrawn. Patients who discontinued prematurely were encouraged to return for follow-up HCV RNA determinations.

Treatment was discontinued if a patient did not achieve an early virologic response, defined as a ≥2-log10 decrease or undetectable HCV RNA after 12 weeks of peginterferon alpha-2a plus ribavirin treatment, or if a patient had detectable HCV RNA after 24 weeks of peginterferon alpha-2a plus ribavirin treatment.

Study Endpoints.

The primary efficacy endpoint was the change in serum HCV RNA titer from baseline (start of treatment with peginterferon alpha-2a plus ribavirin) to Week 12 of treatment with peginterferon alpha-2a plus ribavirin. Serum HCV RNA levels were determined using the COBAS TaqMan HCV Test, v. 2.0 For Use With The High Pure System (Roche Molecular Diagnostics, Pleasanton, CA).

Secondary efficacy endpoints included the change in serum HCV RNA to Week 24 and Week 48 of treatment with peginterferon alpha-2a plus ribavirin and the proportion of patients with undetectable HCV RNA (<28 IU/mL) at each timepoint. Secondary efficacy endpoints included (1) the proportion of patients with an SVR, defined as undetectable HCV RNA at the end of the 24-week follow-up period after completion of treatment with peginterferon alpha-2a plus ribavirin (Week 72); (2) the proportion of patients with virologic relapse, defined as detectable HCV RNA after discontinuation of peginterferon alpha-2a plus ribavirin up to Week 72 in a patient who had undetectable HCV RNA at Week 48; and (3) measures of glycemic control (fasting plasma glucose, insulin, and HbA1c concentrations, and HOMA-IR scores), lipid control (fasting serum triglycerides, total cholesterol, and low- and high-density lipoprotein cholesterol [LDL-C and HDL-C]), and adiponectin levels at each timepoint assessed. Safety was assessed by clinical adverse events, laboratory tests, and dosage modifications.

Statistical Considerations.

The planned enrollment was estimated assuming a common standard deviation of 1.37 log10 in the change in HCV RNA from randomization to Week 12. On the basis of this assumption, a planned enrollment of 240 patients (120 patients per arm) would provide at least 80% power to detect a 0.5 log10 difference between the two treatment groups using a two-sided Student's t test at alpha = 0.05.

The changes from baseline to each timepoint in HCV RNA titer were summarized by treatment arm using the last observation carried forward (LOCF) technique for missing assessments. A mixed-model repeated-measures analysis (with covariates baseline HCV RNA and HOMA-IR score <4 and ≥4) was used to compare the two arms up to Week 12 (intent-to-treat [ITT] population), with the best-fit variance covariance structure selected for the final model and treatment-by-time interaction effect used for testing. Estimated treatment difference was provided with standard error, 95% confidence interval (CI), and P-value based on the normal approximation of the binomial.

Results

A total of 155 patients were enrolled and randomized to either standard care (n = 76) or pioglitazone plus standard care (n = 79) at 70 sites in the United States and Puerto Rico.

The first patient was screened on January 7, 2008, and the last patient completed follow-up on December 30, 2010. The disposition of patients is shown in Fig. 2. A total of 62 of 79 patients (78.5%) randomized to pioglitazone completed the 16-week pioglitazone run-in phase. The baseline characteristics of patients enrolled in the trial are shown in Table 1. There was no statistically significant difference between the two groups in the primary efficacy endpoint, the decrease from baseline to Week 12 of peginterferon alpha-2a plus ribavirin treatment in mean log10 HCV RNA titer (−3.5 ± 1.71 and −3.7 ± 1.62 IU/mL in patients in the pioglitazone and standard-care groups, respectively, Δ = 0.21 IU/mL, P = 0.4394).

Figure 2.

Patient disposition.

Table 1. Baseline Characteristics of Patients Included in the Intention-to-Treat Population
 Intent-to-Treat PopulationPatients Completing the 16-Week Pioglitazone Run-In Phase
Standard Care (n = 73)Pioglitazone + Standard Care (n = 77)PaPioglitazone + Standard Care (n = 62)Pb
  • ALT, alanine aminotransferase; BMI, body mass index; HOMA-IR, homeostasis model assessment of insulin resistance; IQR, interquartile range; ULN, upper limit of normal; ALT quotient, ALT value divided by the upper limit of normal of the local laboratory. Values are medians (interquartile range) unless otherwise specified. Values for sex, race and ethnicity were compared by Fisher's Exact test. Other variables were compared by the Wilcoxon rank sum test.

  • a

    Standard care versus pioglitazone plus standard care (intention-to treat population).

  • b

    Standard care (intention-to treat population) versus pioglitazone plus standard care (patients who completed the run-in phase).

Male sex, n (%)45 (61.6%)50 (64.9%)0.735840 (64.5%)0.8582
Race, n (%)  0.1751 0.0543
White62 (84.9%)56 (72.7%) 45 (72.6%) 
Black10 (13.7%)19 (24.7%) 17 (27.4%) 
American Indian/Alaskan Native1 (1.4%)2 (2.6%) 0 (0%) 
Hispanic ethnicity8 (11.0%)10 (13.0%)0.80377 (11.3%)1.0000
Age (years)51.0 (47.0-54.0)52.0(45.0-55.0)0.700952.0(45.0-56.0)0.7553
Age >40 years, n (%)65 (89.0%)71 (92.2%) 56 (90.3%) 
BMI, kg/m230.1 (27.3-33.4)30.6 (26.7-36.8)0.597330.7 (26.7-37.6)0.5135
BMI > 30 kg/m2, n (%)38 (52.1%)42 (54.5%) 35 (56.5%) 
ALT quotient1.6 (1.1-2.4)1.7 (1.1-2.6)0.66131.8 (1.1-2.7)0.4090
ALT quotient ≤3 times ULN, n (%)62 (84.9%)60 (77.9%) 48 (77.4%) 
ALT quotient >3 times ULN, n (%)11 (15.1%)17 (22.1%) 14 (22.6%) 
HCV RNA titer, log10 IU/mL6.5 (6.0-6.9)6.6 (6.1-6.9)0.43643.2 (1.1-8.1)0.6225
HCV RNA titer, IU/mL x 106 IU/mL2.8 (0.9-7.2)3.6 (1.3-8.1)0.43649 (14.5%) 
HCV RNA <400,000 IU/mL, n (%)10 (13.7%)11 (14.3%) 2 (3.2%) 
HCV RNA 400,000-<800,000 IU/mL, n (%)7 (9.6%)2 (2.6%) 51 (82.3%) 
HCV RNA ≥800,000 IU/mL, n (%)56 (76.7%)64 (83.1%) 6.5 (6.1-6.9)0.6225
HOMA-IR score3.7 (2.5-6.4)3.8 (2.7-5.7)0.78983.9 (2.7-5.7)0.7057
HOMA-IR <4, n (%)39 (53.4%)40 (51.9%) 31 (50.0%) 
HOMA-IR ≥4, n (%)34 (46.6%)37 (48.1%) 31 (50.0%) 
Fasting HbA1c fraction0.06 (0.05-0.06)0.06 (0.05-0.06)0.93940.06 (0.05-0.06)0.9554
Fasting insulin, pmol/L105.6 (67.4-174.3)107.6 (72.9-164.6)0.9311118.4 (88.9-170.2)0.4489
Fasting plasma/serum glucose, mmol/L5.7 (5.1-6.7)5.9 (5.1-6.7)0.81565.9 (5.4-6.8)0.5091
Triglycerides, mmol/L1.5 (1.0-1.9)1.3 (1.0-1.9)0.80841.3 (1.0-1.9)0.9086
Total cholesterol, mmol/L4.3 (3.8-5.0)4.4 (3.9-4.9)0.93564.4 (3.8-4.9)0.6829
LDL-C, mmol/L2.6 (2.0-3.1)2.6 (2.2-2.9)0.96702.6 (2.0-2.8)0.6588
HDL-C, mmol/L1.1 (0.9-1.4)1.0 (0.9-1.3)0.78791.0 (0.9-1.3)0.7371
Adiponectin, μg/mL7.0 (5.0-10.0)6.0 (4.5-10.5)0.90456.0 (4.0-9.0)0.7474
Hemoglobin, g/dL15.8 (15.2-16.8)15.6 (14.8-16.3)0.094515.6 (14.8-16.3)0.1621
Neutrophil count x 109/L3.9 (3.2-4.8)3.8 (2.8-4.5)0.23553.7 (2.8-4.6)0.2238
Platelet count x 109/L248 (214-289)257 (173-298)0.6104260.5 (170-300)0.5292

The mean change in serum HCV RNA levels at weeks 4, 12, 24, and 48 of treatment with peginterferon alpha-2a plus ribavirin was similar in the two treatment groups (Fig. 3A); however, at each timepoint the proportion of patients with undetectable HCV RNA was consistently and numerically higher among patients in the standard-care group than in the pioglitazone group (Fig. 3B). Final SVR rates were 26.0% (20/77) and 38.4% (28/73) among patients in the pioglitazone and standard-care groups, respectively (P = 0.1021), and relapse rates were 15.6% and 19.2%, respectively (P = 0.5615).

Figure 3.

(A) Change in serum HCV RNA levels during treatment with peginterferon alpha-2a plus ribavirin (ITT population). (B) Patients with undetectable HCV RNA during treatment and at the end of follow-up with peginterferon alpha-2a plus ribavirin (ITT population).

The numerical difference in SVR rates remained consistent when African American and Latino patients were excluded from the analysis: 12 of 49 (24.5%) recipients of pioglitazone and 25 of 55 (45.5%) patients in the standard-care group achieved an SVR.

When patients who withdrew during the pioglitazone run-in phase were excluded, the virologic response rates in the pioglitazone treatment group at weeks 4, 12, 24, and 48 were 8.1% (5/62), 41.9% (26/62), 58.1% (36/62), and 48.4% (30/62), respectively, and the SVR rate was 32.3% (20/62).

Among patients randomized to pioglitazone, there was consistent improvement in glycemic variables, including HbA1c, plasma glucose, insulin level, and HOMA-IR score during the 16-week run-in phase (Table 2). Glycemic variables improved in both treatment groups during 48 weeks of treatment with peginterferon alpha-2a plus ribavirin and during 24 weeks of follow-up (Table 3), with greater improvements in plasma glucose, insulin level, and HOMA-IR score among recipients of pioglitazone and greater improvements in HbA1c among recipients of standard care. Pretreatment with pioglitazone had a minimal effect on serum lipid variables (Table 3). In fact, numerically greater improvements in total cholesterol, LDL-C, and HDL-C were seen with standard care compared with pioglitazone (Table 3). Changes in serum lipids were generally similar in patients with and without an SVR (on-line Supporting Table).

Table 2. Change in Metabolic Variables During the 16-Week Pioglitazone Lead-In Phase in Patients Randomized to Pioglitazone Plus Peginterferon Alpha-2a and Ribavirin (ITT Population)
 Pioglitazone + Standard Care
 nMedian Value at the End of Pioglitazone Lead-In Phase (IQR)Median Change from Randomization to the End of Pioglitazone Lead-In Phase (IQR)
  1. HbA1c, hemoglobin A1c; HOMA-IR, homeostasis model assessment of insulin resistance; IQR, interquartile range.

HbA1c, %625.4 (5.2, 6.0)−0.2 (−0.5, 0.1)
Plasma glucose, mmol/L635.4 (4.9, 6.0)−0.3 (−0.9, 0.0)
Insulin, pmol/L6475.0 (42.0, 125.0)−28.8 (−70.1, 3.5)
HOMA-IR score602.6 (1.7, 4.9)−1.5 (−3.0, 0.3)
Adiponectin, μg/mL6016.0 (11.0, 24.0)9.5 (5.5, 17.5)
Table 3. Change in Metabolic Variables from Baseline to End of Treatment with Peginterferon Alpha-2a Plus Ribavirin (Week 48) and End of Follow-up (Week 72)
 Standard CarePioglitazone + Standard Care
Median Change From Baseline (IQR)nTo End of Treatment (Week 48)nTo End of Follow-up (Week 72)nTo End of Treatment (Week 48)nTo End of Follow-up (Week 72)
  1. HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; IQR, interquartile range, LDL-C, low-density lipoprotein cholesterol.

HbA1c, %46−1.1 (−1.8, −0.4)380.0 (−0.50, −0.20)31−0.7 (−1.4, −0.3)29−0.1 (−0.5, 0.3)
Plasma glucose, mmol/L46−0.2 (−0.7, −0.4)38−0.3 (−0.9, 0.3)34−0.5 (−1.3, −0.5)30−0.3 (−1.1, 0.2)
Insulin, pmol/L46−18.1 (−63.2, −24.3)40−21.5 (−53.1, −4.5)32−42.0 (−100.4, −11.8)30−41.7 (−107.6, −20.1)
HOMA score44−0.6 (−2.7, −1.1)35−0.9 (−6.0, 1.1)30−2.1 (−3.5, −0.4)27−2.0 (−3.4, −1.0)
Triglycerides, mmol/L460.2 (−0.1, 1.1)390.1 (−0.3, 0.8)340.2 (−0.2, 0.6)30−0.1 (−0.6, 0.4)
Total cholesterol, mmol/L46−0.8 (−1.1, −0.2)390.0 (−0.4, 0.3)34−0.3 (−0.8, 0.0)300.3 (−0.2, 0.9)
LDL-C, mmol/L44−0.6 (−1.0, −0.2)38−0.1 (−0.4, 0.3)34−0.3 (−0.6, 0.1)300.2 (−1.0, 0.6)
HDL-C, mmol/L46−0.2 (−0.3, −0.1)390.1 (−0.1, 0.2)34−0.1 (−0.3, 0.0)300.0 (−0.1, 0.3)
Adiponectin, μg/mL440.0 (−2.0, 2.0)360.0 (−1.0, 1.0)318.0 (5.0, 16.0)294.0 (0.0, 14.0)

Median serum adiponectin levels increased from 6.0 μg/mL at baseline to 16.0 μg/mL at the end of the 16-week pioglitazone run-in phase. Moreover, the median change from the start of treatment with peginterferon alpha-2a plus ribavirin was greater in patients treated with pioglitazone plus peginterferon and ribavirin (8.0, interquartile range [IQR] 5.0 to 6.0 μg/mL) than with peginterferon and ribavirin alone (0.0, IQR −2.0 to 2.0). Changes in serum adiponectin levels were generally similar in patients with and without an SVR (on-line Supporting Table).

Safety.

The most frequent adverse events during treatment with peginterferon alpha-2a plus ribavirin were fatigue, anemia, nausea, headache, and depression, with a significantly higher incidence of depression among those receiving standard care without pioglitazone (31% versus 12% among recipients of pioglitazone plus standard care, P = 0.0051, Table 4). Seven patients reported weight gain and/or edema during treatment with pioglitazone before and after initiation of peginterferon alpha-2a plus ribavirin. The incidence of serious adverse events was similar among patients randomized to pioglitazone or standard care (13% versus 12%). Among patients who received pioglitazone, the incidence of serious adverse events was 4% (n = 3) during the run-in phase and 13% (n = 8) during treatment with peginterferon alpha-2a plus ribavirin. Serious adverse events that occurred before the start of peginterferon alpha-2a plus ribavirin included osteoarthritis (n = 1), lower limb fracture (n = 1), and acute hepatitis and acute pancreatitis (n = 1).

Table 4. Adverse Events Laboratory Abnormalities and Dose Modifications Occurring During the Run-In, Treatment with Peginterferon Alpha-2a Plus Ribavirin or Follow-Up
 Standard Care (n = 75)Pioglitazone + Standard Care (n = 77)Pa
  • a

    Fisher's Exact test.

  • b

    Adverse events that were reported by ≥25% of patients in at least one treatment group.

Common adverse events,b n (%)   
 Fatigue35 (47)30 (39)0.4126
 Anemia24 (32)28 (36)0.6107
 Nausea21 (28)26 (34)0.4855
 Insomnia23 (31)21 (27)0.7215
 Headache20 (27)14 (18)0.2453
 Depression23 (31)9 (12)0.0051
 Rash19 (25)11 (14)0.1046
Serious adverse events, n (%)9 (12)10 (13)1.0000
Discontinuation because of adverse events5 (7)10 (13)0.2772
Patients with ≥1 dose modification for adverse events or laboratory abnormalities, n (%)
 Peginterferon alpha-2a14 (18.7)20 (26.0)0.3323
 Ribavirin34 (45.3)28 (36.4)0.3222
Laboratory abnormalities during peginterferon alpha-2a plus ribavirin treatment
 Neutrophils <0.75 x 109/L16/73 (21.9)19/62 (24.3)0.3247
 Platelets <50 x 109/L2/73 (2.7)4/62 (6.5)0.4129
 Hemoglobin <10 g/dL21/73 (28.8)30/77 (39.0)0.2282
 Hemoglobin <8.5 g/dL3/73 (4.1)11/77 (14.3)0.0473

The incidence of suicidal ideation was higher among recipients of pioglitazone (n = 6, 8%) versus standard care (n = 1, 1%), although all cases of suicidal ideation occurred after the initiation of therapy with peginterferon alpha-2a plus ribavirin. More patients randomized to pioglitazone discontinued treatment prematurely (13% versus 7% with regular care), although the incidence of withdrawal from treatment was similar during the 48-week peginterferon alpha-2a plus ribavirin treatment phase. There were no deaths reported during the study.

Discussion

Our study clearly demonstrates that improvement in insulin resistance with pioglitazone before, during, and after 48 weeks of treatment with peginterferon alpha-2a plus ribavirin is not accompanied by an improved virologic response among patients infected with HCV genotype 1 with underlying insulin resistance. Although the magnitude of the reductions in HCV RNA at Week 12 of treatment with peginterferon alpha-2a plus ribavirin were similar in both groups, there was a trend toward higher virologic response rates and lower relapse rates in patients who did not receive pioglitazone in the ITT population. However, when patients who withdrew during the pioglitazone run-in phase were excluded, the apparent differences in virologic response rates between the groups were smaller.

Our findings are consistent with those of a previous study that failed to show any improvement in virological response by concomitant treatment with pioglitazone and peginterferon/ribavirin.25 Importantly, our study addressed several deficiencies in the earlier study by using a higher therapeutic dosage of pioglitazone and by starting pioglitazone before the introduction of peginterferon/ribavirin.

Our findings conflict with those in a study that showed that improved insulin sensitivity was associated with improved virologic response rates in patients infected with HCV genotype 4.24 In that study, patients treated with pioglitazone 30 mg/day in combination with peginterferon plus ribavirin for 48 weeks had significantly higher rates of rapid virologic response (27% versus 6% in patients treated with peginterferon plus ribavirin, P = 0.006) and SVR rates (60% versus 39%, respectively, P = 0.04).24

Consistent with our findings, HOMA-IR score improved significantly during treatment with pioglitazone in the study by Khattab et al.,24 and lower baseline HOMA-IR was an independent predictor of SVR.

Successful eradication of HCV has also been reported in a patient with HCV genotype 3 infection who failed to respond to a first course of peginterferon/ribavirin, but achieved an SVR after being treated with pioglitazone 45 mg/day for 5 months prior to the reintroduction of peginterferon plus ribavirin.26

The reason for the stark difference in the impact of pioglitazone on virologic outcomes between our study and that by Khattab et al. is unclear. However, there are well-established genotype-specific associations between the metabolic effects of HCV infection and treatment with peginterferon plus ribavirin. For example, hepatic steatosis is more common in patients infected with HCV genotype 3 and is significantly improved after viral eradication.27 Viral clearance is also associated with improved insulin resistance in patients infected with HCV genotypes 1 and 4, but not genotypes 2 or 3.8, 24

Metformin has also been evaluated as an adjuvant to peginterferon alpha-2a plus ribavirin in patients with HCV genotype 1 infection and insulin resistance.28 Metformin significantly improved HOMA-IR scores compared with placebo but there was no significant difference in overall SVR rates between the two treatment groups.

At the end of treatment with peginterferon alpha-2a plus ribavirin, median improvements in indicators of glycemic control (fasting glucose, insulin levels, and HOMA-IR) were greater in recipients of pioglitazone with the exception of HbA1c, which improved to a numerically greater extent in patients who did not receive pioglitazone. Perhaps the most significant metabolic improvement was the increase in adiponectin levels during the 16-week pioglitazone run-in phase. Adiponectin has antiinflammatory, antiapoptotic, and antifibrotic effects,29 and pioglitazone-induced increases in adiponectin are associated with decreased steatosis and inflammation in patients with nonalcoholic steatohepatitis (NASH).30, 31 In patients with chronic hepatitis C, there is an inverse correlation between adiponectin levels and steatosis grade, histologic activity index, and fibrosis stage.32 Low adiponectin levels are also correlated with the absence of a virologic response to peginterferon plus ribavirin therapy.33 In contrast, we found no correlation between increased adiponectin levels and virologic response. Consistent with our findings, the Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis (HALT-C) trial was also unable to show a correlation between HCV clearance, improvement in HOMA-IR score, and adiponectin levels.34

Since this trial was conceived and initiated, it has been discovered that one of the most important baseline predictors of virologic response is host IL28B genotype.35 Interestingly, this marker is also associated with metabolic phenomena. Patients with the favorable CC genotype have higher total cholesterol, LDL-cholesterol, and apolipoprotein-B levels, and the prevalence of steatosis is higher in patients with the less favorable non-CC genotypes.36, 37 It is tempting to speculate that insulin resistance may be associated with the less favorable genotype, which might explain why improved insulin resistance with pioglitazone does not translate into improved virologic response rates. However, a retrospective analysis of a large database of patients with HCV genotype 1 infection found that HOMA-IR score was not significantly associated with host IL28B genotype.38 Unfortunately, we did not determine the host genotype in our population and so are unable to comment on the impact of pioglitazone on insulin resistance and treatment outcomes by IL28B genotype.

Although there is reproducible evidence that insulin resistance impairs the response to antiviral therapy in patients with chronic hepatitis C, there is conflicting evidence to show that pharmacologic interventions that improve insulin sensitivity lead to an increase in antiviral response rates. Experimental data suggest that hyperinsulinemia up-regulates suppressor of cytokine signaling (SOCS)-3 (a known inhibitor of interferon signaling) and activates phosphatidylinositol-3-kinase, which in turn inhibits phosphorylation of signal transducers and activators of transcription-1, leading to further impairment of interferon signaling.39 However, in addition to the local indirect effects of interferon that induce an antiviral state within an infected hepatocyte, the systemic immunomodulatory effects of interferon must be considered. Interferon activates natural killer (NK) cells within 6 to 48 hours of injection, and the degree of NK-cell activation correlates with early virologic response.40 This observation is important in the context of our finding that antiviral therapy was less effective during pioglitazone therapy because there is evidence to suggest that PPAR-γ ligands alter NK-cell function. For example, exposure to a natural PPAR-γ ligand (15d-PGJ2) or a synthetic PPAR-γ ligand (ciglitazone) reduced interferon (IFN)-γ protein levels and IFN-γ gene expression in NK cells isolated from healthy donors.41 More recently, pioglitazone has been shown to reduce circulating NK cell populations in patients with type 2 diabetes.42 These data suggest that although pioglitazone reduces insulin resistance, its pleiotropic effects on immune function may impair immune-mediated clearance of infected hepatocytes.

Expression of insulin receptor substrate 1 (IRS-1) is significantly reduced in Huh-7 cells that express core protein of HCV genotype 3a or 1b.13 However, the mechanism by which core protein affects IRS-1 differs by genotype. Genotype 3a core protein promotes degradation of IRS-1 through down-regulation of PPAR-γ and by up-regulation of SOCS-7. In contrast, genotype 1b core protein activates the mammalian target of rapamycin (mTOR). Treatment with a PPAR-γ agonist (rosiglitazone) attenuated the effects of genotype 3a core protein but not genotype 1b core protein in this model.13 These results provide further insight as to why pioglitazone has differential effects in patients infected with different HCV genotypes.

This trial has certain limitations. There is an apparent imbalance between the two treatment groups, with a higher proportion of black patients and a higher proportion of patients with high HCV RNA levels in the pioglitazone group; however, these differences were not statistically significant. We failed to reach the planned enrollment because a series of large phase III trials of protease inhibitors were recruiting genotype 1 patients. The potential benefits of a higher SVR rate with protease inhibitor-based triple therapy made it difficult for investigators to recommend enrolling in this study and led to early termination of this study.

In conclusion, treatment with pioglitazone before and during treatment with peginterferon alpha-2a plus ribavirin improved several indices of glycemic control in patients with chronic hepatitis C and insulin resistance, but did not improve virologic response rates compared with peginterferon alpha-2a plus ribavirin alone.

Acknowledgements

Support for third-party writing assistance for this article, furnished by Blair Jarvis and Sue Currie, was provided by Genentech Inc. and F. Hoffmann-La Roche Ltd. Disclaimer: The opinion or assertions contained herein are the private views of the authors and are not to be construed as official or reflecting the view of the US Department of the Army or the US Department of Defense.

Ancillary