SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References

Elevated low-density lipoprotein (LDL) levels and statin use have been associated with higher sustained virological response (SVR) rates in patients receiving chronic hepatitis C therapy. However, these relationships have not been well characterized in randomized controlled trials. Furthermore, little is known about the relationship between high-density lipoprotein (HDL) and virological response. To determine whether baseline LDL or HDL levels and statin use affect SVR rates, we retrospectively evaluated the IDEAL (Individualized Dosing Efficacy Versus Flat Dosing to Assess Optimal Pegylated Interferon Therapy) trial, in which 3070 treatment-naive, hepatitis C virus (HCV) genotype 1–infected patients were treated for up to 48 weeks in one of the following arms: (1) peginterferon (PEG-IFN) alfa-2b at 1.5 μg/kg/week with ribavirin (RBV) at 800 to 1400 mg/day, (2) PEG-IFN alfa-2b at 1.0 μg/kg/week with RBV at 800 to 1400 mg/day, or (3) PEG-IFN alfa-2a at 180 μg/week with RBV at 1000 to 1200 mg/day. Virological responses were assessed by pretreatment statin use and baseline elevated LDL levels (≥130 mg/dL) or low HDL levels (<40 mg/dL for men and <50 mg/dL for women). In 1464 patients with baseline elevated LDL levels or low HDL levels, the SVR rate was significantly higher than that in patients with normal levels (44.9% versus 34.0%, P < 0.001). In 66 patients receiving a statin pretreatment, the SVR rate was higher than the rate of those not receiving it (53.0% versus 39.3%, P = 0.02). In a multivariate logistic regression analysis using the stepwise selection method with baseline characteristics, a high LDL level [odds ratio (OR) = 1.6, 95% confidence interval (CI) = 1.4-1.8, P < 0.001], a low HDL level (OR = 0.5, 95% CI = 0.3-0.8, P = 0.004), and statin use (OR = 2.0, 95% CI = 1.1-3.7, P = 0.02) were independently associated with SVR. Conclusion: Baseline elevated LDL levels or low HDL levels and preemptive statin usage were associated with higher SVR rates. Prospective studies may be considered to explore the biological impact of these factors on HCV RNA replication and treatment response. (HEPATOLOGY 2010;)

Chronic hepatitis C (CHC) remains an important health problem in the United States and around the world, with 130 million to 170 million people thought to be chronically infected.1 The treatment response rates have improved significantly over the past 15 years, but they remain suboptimal, especially for patients infected with hepatitis C virus (HCV) genotype 1. Sustained virological response (SVR) rates in those with HCV genotype 1 range from 40% to 46% with current standard-of-care therapy, peginterferon (PEG-IFN) and ribavirin (RBV). There are multiple factors, both host and viral, associated with poorer treatment outcomes; these include older age, male gender, obesity, insulin resistance or metabolic syndrome, African American race, steatosis or advanced fibrosis on liver biopsy, and high viral loads. Although recent findings indicate that the interleukin-28B polymorphism explains approximately half of the treatment differences between African American and white patients, other factors, including cholesterol levels and statin use, may also play a role in virological response.2 Recent evidence from small retrospective studies suggests that elevated low-density lipoprotein (LDL) levels may be associated with higher SVR rates in patients infected with HCV genotype 1.3–17

In vitro data suggest that lipid-lowering agents (3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors or statins) may inhibit viral replication. However, cell culture models of HCV infection have demonstrated that anti-HCV activity is not the same for all statins: fluvastatin, atorvastatin, simvastatin, and lovastatin inhibit viral replication to variable extents, whereas pravastatin and rosuvastatin have little to no anti-HCV activity.18–21 The underlying mechanism remains unclear; however, data suggest that HCV enters hepatocytes via several lipoprotein receptors (LDL and scavenger receptor class B type 1). At first glance, this appears to be counterintuitive because the statins increase LDL receptors on hepatocytes and in theory increase HCV infectivity. To this end, three small prospective human trials with either atorvastatin or fluvastatin have yielded conflicting results. One study suggested that high-dose fluvastatin increased HCV RNA levels after 4 weeks of treatment,22 whereas another trial using low-dose atorvastatin showed no decrease in HCV RNA at week 4 or 12.23 Alternatively, another study using high-dose fluvastatin showed a modest decrease in HCV RNA replication in 50% of patients.24

Little is known about the relationship between high-density lipoprotein (HDL) and the virological response to PEG-IFN and RBV. HDL, a scavenger receptor class B type 1 receptor ligand, may modulate the cell entry of the virus via scavenger receptor class B type 1.25 Subsequently, a low HDL level would be expected, at least in theory, to decrease cellular infectivity.

We retrospectively evaluated the IDEAL (Individualized Dosing Efficacy Versus Flat Dosing to Assess Optimal Pegylated Interferon Therapy) study database to assess the virological response to PEG-IFN and RBV treatment in patients with CHC genotype 1 infection who either had a baseline elevated LDL level or low HDL level or were taking a statin at the time of enrollment in the study.

Patients and Methods

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References

The IDEAL trial was a randomized, parallel-group, multicenter study conducted at 118 academic and community centers in the United States. This study was carried out in accordance with the Declaration of Helsinki, current guidelines on good clinical practices, and local ethical and legal requirements. All patients provided voluntary written, informed consent before trial entry. Eligible patients were 18 years old or older, had a chronic HCV genotype 1 infection and compensated liver disease, and had not been previously treated for hepatitis C. All patients were required to have an absolute neutrophil count ≥ 1500/mm3, a platelet count ≥ 80,000/mm3, and a hemoglobin level ≥ 12 g/dL for females and ≥ 13 g/dL for males. Patients were excluded if they had human immunodeficiency virus or hepatitis B coinfection, any other cause of significant liver disease, poorly controlled diabetes, weight > 125 kg, or severe depression or other psychiatric disorders or were active substance abusers. Patients were required to have undergone liver biopsy within 3 years prior to screening.

For inclusion, patients were also required to have a fasting glucose level of 70 to 140 mg/dL. Patients with levels between 116 and 140 mg/dL or those with diabetes mellitus were required to have hemoglobin A1c levels ≤ 8.5%. Patients weighing > 105 to 125 kg with a body mass index (BMI) > 30 kg/m2 were excluded if they had a history of three or more of the following risk factors: uncontrolled hypercholesterolemia, diabetes, hypertension, smoking, and a family history of coronary heart disease.

All medications used by the patients within 2 weeks prior to screening and throughout the study period were collected.

Study Design.

Patients received one of three treatment regimens: PEG-IFN alfa-2b (1.5 μg/kg/week or 1.0 μg/kg/week) in combination with oral RBV dosed by body weight (40-65 kg, 800 mg/day; >65 to 85 kg, 1000 mg/day; >85 to 105 kg, 1200 mg/day; >105 to 125 kg, 1400 mg/day) or PEG-IFN alfa-2a (180 μg/week) in combination with oral RBV at 1000 to 1200 mg/day (<75 kg, 1000 mg/day; ≥75 kg, 1200 mg/day). Treatment was discontinued because of treatment failure in patients with detectable HCV RNA and a <2-log10 decrease in HCV RNA from the baseline at week 12 and in patients with detectable HCV RNA at week 24.26, 27

Laboratory Assessments.

Fasting LDL and HDL levels were measured at screening visit 2 (baseline), in treatment weeks 12, 24, 36, and 48, and in follow-up week 24. HCV RNA levels were measured at screening visits 1 and 2 (baseline), in treatment weeks 2, 4, 12, 24, and 48, and in follow-up weeks 4, 12, and 24 with the COBAS TaqMan assay (Roche Diagnostics; lower limit of quantitation = 27 IU/mL). All values below the limit of detection were considered undetectable.

Endpoints and Statistical Analyses.

Statistical analyses included all randomized patients who received at least one dose of the study medication. SVR was defined as undetectable HCV RNA levels 24 weeks after the completion of therapy. If the 24-week posttreatment HCV RNA level was missing, the 12-week posttreatment level was used. Relapse rates, defined as undetectable HCV RNA at the end of treatment (EOT) with detectable HCV RNA levels at the end of follow-up, were also evaluated. Other virological response rates included rapid virological response (RVR; undetectable HCV RNA in week 4 of treatment), early virological response (EVR; undetectable HCV RNA in week 12 of treatment), and EOT response (undetectable HCV RNA at EOT).

Categorical variables were summarized with percentages, and continuous variables were summarized with means and standard deviations (SDs). The positive predictive value of on-treatment response was computed as the proportion of patients who had SVR among those who had an on-treatment response at the specific time points of interest (weeks 4 and 12). Two-sided chi-square tests were used to provide nominal P values for comparisons of response rates. Multivariable, stepwise logistic regression analysis was used to identify baseline factors that were predictors of SVR. P values based on the logistic regression model, odds ratios (ORs), and 95% confidence intervals (CIs) for the ORs were reported for factors that were significant (P < 0.05).

Results

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References

Patient Characteristics.

Between March 2004 and June 2006, 4469 patients were screened, and 3070 patients were randomized and treated in the IDEAL trial. Of the 3070 treated patients, 1464 (47.6%) had baseline elevated LDL levels (≥130 mg/dL) or low HDL levels (<40 mg/dL for men and <50 mg/dL for women) and were not receiving treatment with statins at the baseline. In this group, 333 patients had elevated LDL levels only, 927 had low HDL levels only, and 204 patients had both elevated LDL levels and low HDL levels. Of patients with baseline elevated LDL levels or low HDL levels, 64% were male; the mean age was 47.3 years, and the mean weight was 86.2 kg (Table 1). The numbers of patients with baseline elevated LDL levels or low HDL levels were similar among the three treatment arms. The mean baseline LDL level was 115 mg/dL for patients with elevated LDL levels or low HDL levels versus 93 mg/dL for patients with normal levels.

Table 1. Patient Baseline Demographics and Disease Characteristics
 Baseline Elevated LDL or Low HDL (n = 1464)*Baseline Normal LDL and HDL (n = 1540)*Baseline Statin Use (n = 66)No Baseline Statin Use (n = 3004)
  • Abbreviation: ULN, upper limit of normal.

  • *

    Statin users (n = 66) were excluded; nine patients with missing baseline LDL and HDL levels were included in the normal LDL and HDL group.

Treatment arm, n (%)    
 PEG-IFN alfa-2b, 1.5 μg/kg/RBV478 (33)519 (34)22 (33)997 (33)
 PEG-IFN alfa-2b, 1.0 μg/kg/RBV507 (35)487 (32)22 (33)994 (33)
 PEG-IFN alfa 2a/RBV479 (33)534 (35)22 (33)1013 (34)
Male sex, n (%)932 (64)855 (56)46 (70)1787 (59)
Race, n (%)    
 White1087 (74)1056 (69)46 (70)2143 (71)
 African American234 (16)322 (21)14 (21)556 (19)
 Hispanic106 (7)105 (7)2 (3)211 (7)
 Asian18 (1)31 (2)2 (3)49 (2)
 Other19 (1)26 (2)2 (3)45 (1)
Age (years), mean (SD)47.3 (7.8)47.6 (8.2)52.2 (6.5)47.4 (8.0)
Weight (kg), mean (SD)86.2 (16.2)80.6 (15.9)87.2 (16.6)83.3 (16.3)
BMI (kg/m2), mean (SD)28.8 (4.8)27.5 (4.8)29.5 (5.4)28.2 (4.8)
Source of infection    
 Parenteral996 (68)1042 (68)41 (62)2038 (68)
 Transfusion279 (19)312 (20)19 (29)591 (20)
 Sporadic140 (10)146 (9)6 (9)286 (10)
 Other49 (3)40 (3)074 (2)
Mean ALT level (U/L)2.19 × ULN2.08 × ULN2.14 × ULN2.13 × ULN
ALT level > ULN, n (%)1216 (83)1211 (79)52 (79)2427 (81)
Baseline HCV RNA    
 Log10 IU/mL, mean (SD)6.4 (0.7)6.3 (0.7)6.4 (0.8)6.3 (0.7)
 >600,000 IU/mL, n (%)1207 (82)1263 (82)48 (73)2470 (82)
Genotype 1 subtype, n (%)    
 1a911 (62)968 (63)40 (61)1879 (63)
 1b513 (35)547 (36)25 (38)1060 (35)
METAVIR fibrosis score of F3/F4, n (%)175 (12)149 (10)4 (6)324 (11)
Hepatic steatosis, n (%)    
 No482 (33)615 (40)20 (30)1097 (37)
 Yes912 (62)853 (55)41 (62)1765 (59)
 Missing70 (5)72 (5)5 (8)142 (5)
Baseline LDL (mg/dL), mean (SD)115 (36)93 (22)100 (34)104 (31)
Medical history of diabetes mellitus, n (%)93 (6)90 (6)23 (35)183 (6)
Medical history of hypertension, n (%)405 (28)389 (25)41 (62)794 (26)

Sixty-six patients (2.1%) were receiving a statin at the start of treatment [atorvastatin (29), pravastatin (14), simvastatin (10), rosuvastatin (6), lovastatin (5), fluvastatin (1), and ezetimibe/simvastatin (1)]. Seventy percent of statin users were male, and the statin users had a higher mean age (52.2 versus 47.4 years) and a higher mean weight (87.2 versus 83.3 kg) versus those who did not use statins (Table 1). The percentage of patients with a high baseline viral load (HCV RNA level > 600,000 IU/mL) was slightly lower among those using statins (73% versus 82%, P = 0.07). Statin users had similar mean baseline LDL levels in comparison with patients not using statins (100 versus 104 mg/dL, respectively). More patients on statins had preexisting diabetes mellitus and hypertension in comparison with those not receiving statin treatment.

Cholesterol Levels and Virological Response.

The SVR rate for patients with baseline elevated LDL levels or low HDL levels was 44.9% (95% CI = 42.3%-47.5%), which was significantly higher than the rate of 34.0% for patients with normal baseline levels (95% CI = 31.6%-36.4%, P < 0.001). In addition, the virological response rates in weeks 4, 12, and 24 and at EOT were significantly higher in the group with baseline elevated LDL levels or low HDL levels (Fig. 1). Relapse rates were similar between the groups: 24.4% (95% CI = 21.6%-27.5%) in patients with baseline elevated LDL levels or low HDL levels versus 27.3% (95% CI = 24.1%-30.8%) in those with normal LDL and HDL levels with a difference of −3.2% (95% CI = −7.5% to 1.1%). Further analysis of the data demonstrated that patients with baseline elevated LDL levels and low HDL levels had an SVR rate of 53.1% (95% CI = 46.3%-59.9%); these patients were followed by those with elevated LDL levels and normal HDL levels, who had an SVR rate of 50.9% (95% CI = 45.6%-56.2%). Patients with normal LDL levels but low HDL levels had an SVR rate of 41.7% (95% CI = 38.6%-44.9%), which was followed by an SVR rate of 34.0% (95% CI = 31.7%-36.4%) for patients with normal HDL and LDL levels.

thumbnail image

Figure 1. Percentage of patients with undetectable HCV RNA by baseline LDL/HDL levels. Abbreviation: TW, treatment week.

Download figure to PowerPoint

SVR rates in patients with elevated LDL levels or low HDL levels were compared to those in patients with normal levels for subgroups of patients defined by demographic and disease characteristics (Table 2). In nearly all of these subgroups, including those that are typically more difficult to treat, patients with baseline elevated LDL levels or low HDL levels had significantly higher SVR rates. The only three exceptions were patients 40 years of age or younger (P = 0.08), patients with normal baseline alanine aminotransferase (ALT) levels (P = 0.07), and patients with a baseline weight of 40 to 65 kg (P = 0.18). Although the lipid effect (patients with elevated LDL levels or low HDL levels at the baseline were compared to patients with normal baseline levels) was not statistically significant for these three subgroups, the SVR rates were higher in patients with elevated LDL levels or low HDL levels at the baseline. In agreement with the overall IDEAL study results, younger patients, patients with baseline fasting glucose levels < 5.6 mmol/L, patients with normal baseline ALT levels, patients without hepatic steatosis, patients with low baseline viral loads, patients without advanced fibrosis, and patients with an assigned RBV dose > 13 mg/kg/day had higher SVR rates, regardless of their baseline lipid levels.28

Table 2. Viral Response According to the Baseline and On-Treatment Factors
 Baseline Elevated LDL or Low HDL (n = 1464)Baseline Normal LDL and HDL (n = 1540)PBaseline Statin Use (n = 66)No Baseline Statin Use (n = 3004)P
  • Abbreviation: NA, not applicable.

  • *

    HCV RNA level < 27 IU/mL.

  • Statin users (n = 66) were excluded; nine patients with missing baseline LDL and HDL levels were included in the normal LDL and HDL group.

  • P values are nominal and are not adjusted for multiple hypotheses.

  • §

    Relapse rates are based on patients with undetectable HCV RNA levels at EOT and detectable HCV RNA levels at the end of follow-up.

Overall SVR, % (n/N)*44.9 (657/1464)34.0 (523/1540)<0.00153.0 (35/66)39.3 (1180/3004)0.02
Gender, % (n/N)      
 Female45 (242/532)36 (248/685)0.00170 (14/20)40 (490/1217)0.007
 Male45 (415/932)32 (275/855)<0.00146 (21/46)39 (690/1787)0.33
Age, % (n/N)      
 ≤40 years56 (130/233)48 (106/223)0.08100 (1/1)52 (236/456)0.34
 >40 years43 (527/1231)32 (417/1317)<0.00152 (34/65)37 (944/2548)0.01
Body weight, % (n/N)      
 40-65 kg46 (75/163)39 (108/274)0.1860 (3/5)42 (183/437)0.41
 >65 to <75 kg44 (95/214)35 (113/320)0.0464 (7/11)39 (208/534)0.10
 75-85 kg46 (151/325)34 (136/395)0.00135 (6/17)40 (287/720)0.70
 >85-105 kg43 (253/583)30 (132/445)<0.00156 (14/25)37 (385/1028)0.06
 >105-125 kg46 (81/177)32 (34/105)0.0363 (5/8)41 (115/282)0.22
Baseline fasting glucose, % (n/N)      
 <5.6 mmol/L48 (490/1011)38 (433/1140)<0.00163 (24/38)43 (923/2151)0.01
 ≥5.6 mmol/L37 (167/452)22 (89/399)<0.00139 (11/28)30 (256/851)0.30
ALT level, % (n/N)      
 Normal40 (98/248)32 (106/329)0.0736 (5/14)35 (204/577)0.98
 Elevated46 (559/1216)34 (417/1211)<0.00158 (30/52)40 (976/2427)0.01
Hepatic steatosis, % (n/N)      
 No52 (253/482)43 (267/615)0.00355 (11/20)47 (520/1097)0.50
 Yes41 (370/912)27 (234/853)<0.00156 (23/41)34 (604/1765)0.004
Baseline viral load, % (n/N)      
 >600,000 IU/mL40 (484/1207)29 (370/1263)<0.00144 (21/48)35 (854/2470)0.19
 ≤600,000 IU/mL67 (173/257)55 (153/277)0.00478 (14/18)61 (326/534)0.15
METAVIR fibrosis score, % (n/N)      
 0/1/247 (571/1219)36 (476/1319)<0.00153 (30/57)41 (1047/2538)0.09
 3/430 (52/175)17 (25/149)0.006100 (4/4)24 (77/324)<0.001
Assigned RBV dose, % (n/N)      
 ≤13 mg/kg/day43 (324/747)28 (182/644)<0.00142 (14/33)36 (506/1391)0.48
 >13 mg/kg/day46 (333/717)38 (341/896)0.00164 (21/33)42 (674/1613)0.01
Response rates in patients with RBV dose reduction, % (n/N)      
 EOT68 (687/1011)56 (577/1030)<0.00170 (32/46)62 (1264/2041)0.37
 Relapse§23 (151/650)26 (140/544)NA10 (3/31)24 (291/1194)NA
 SVR50 (506/1011)40 (409/1030)<0.00161 (28/46)45 (915/2041)0.08

In patients who achieved RVR (undetectable HCV RNA at week 4) and EVR (undetectable HCV RNA at week 12), SVR rates were similar for those with baseline elevated LDL levels or low HDL levels and those with normal levels (Table 3). However, more patients with baseline elevated LDL levels or low HDL levels achieved RVR [170/1464 (12%) versus 136/1540 (9%)] and EVR [677/1464 (46%) versus 526/1540 (34%)]. These differences contributed to a higher overall SVR rate in those with baseline elevated LDL levels or low HDL levels. Relapse rates increased as the time to HCV RNA being first undetectable increased similarly in both groups.

Table 3. Predictability of SVR or Relapse Based on the HCV RNA Change from the Baseline, Detectability, or Time to First Undetectability at Various Treatment Weeks
 Baseline Elevated LDL or Low HDLBaseline Normal LDL and HDLBaseline Statin UseNo Baseline Statin Use
  • *

    Per protocol, therapy was discontinued in patients with detectable HCV RNA and a less than 2-log10 IU decline at treatment week 12 and in patients with detectable HCV RNA at week 24; however, in some cases, therapy was continued on the basis of a request by an investigator. A total of 174 patients continued therapy beyond week 12 or 24 despite persistent low-level viremia. Of these patients, SVR was achieved in only 14 patients.

  • Statin users (n = 66) were excluded; nine patients with missing baseline LDL and HDL levels were included in the normal LDL and HDL group.

HCV RNA change from the baseline: SVR, % (n/N)    
 Week 4    
  <1 log104 (11/283)4 (20/461)0 (0/6)4 (31/744)
  1 to <2 log1028 (86/310)24 (94/396)22 (4/18)25 (180/706)
  2 to <3 log1055 (133/244)48 (99/208)83 (5/6)51 (232/452)
  3 to <4 log1070 (153/219)60 (92/153)71 (10/14)66 (245/372)
  ≥4 log1055 (130/238)53 (99/186)50 (5/10)54 (229/424)
  Undetectable85 (144/170)88 (119/136)92 (11/12)86 (263/306)
 Week 12    
  <1 log10*1 (1/105)0.5 (1/185)0 (0/2)0.7 (2/290)
  1 to <2 log10*0.7 (1/149)1 (3/231)0 (0/6)1 (4/380)
  2 to <3 log104 (4/91)4 (6/140)0 (0/2)4 (10/231)
  3 to <4 log1020 (25/126)21 (30/140)33 (1/3)21 (55/266)
  ≥4 log1028 (88/316)24 (77/318)18 (3/17)26 (165/634)
  Undetectable79 (538/677)77 (406/526)86 (31/36)78 (944/1203)
 Week 24    
  ≥2-log10 drop at week 12 and undetectable72 (623/866)68 (492/723)81 (35/43)70 (1115/1589)
Weeks to HCV RNA first being undetectable: SVR, % (n/N)    
 288 (60/68)92 (56/61)100 (2/2)90 (116/129)
 480 (86/107)81 (64/79)90 (9/10)81 (150/186)
 1277 (401/523)73 (290/399)80 (20/25)75 (691/922)
 2442 (104/246)42 (109/260)40 (4/10)42 (213/506)
Weeks to HCV RNA first being undetectable: relapse, % (n/N)    
 22 (1/60)2 (1/57)0 (0/2)2 (2/117)
 410 (9/93)6 (4/68)10 (1/10)8 (13/161)
 1215 (70/467)15 (52/339)13 (3/23)15 (122/806)
 2450 (101/202)50 (107/214)43 (3/7)50 (208/416)

Statin Use and Virological Response.

A significantly higher SVR rate was also seen among patients who were receiving a statin prior to antiviral therapy versus those not using statins [Fig. 2; 53.0% (95% CI = 40.3%-65.4%) versus 39.3% (95% CI = 37.5%-41.1%), P = 0.02]. Similarly, there was a significant statin effect on virological response rates in treatment weeks 4 and 12; although the response rates continued to be higher in patients who received a statin, the difference was not statistically significant in week 24 or at EOT. Relapse rates were lower among the statin users: 16.7% (7.0%-31.4%) versus 25.8% (23.6%-28.0%) with a difference of −9.1% (95% CI = −20.6% to 2.4%).

thumbnail image

Figure 2. Percentage of patients with undetectable HCV RNA by baseline statin use. Abbreviation: TW, treatment week.

Download figure to PowerPoint

The statin effect was also examined in subgroups of patients defined by demographic and baseline characteristics. For every subgroup, the SVR rate was numerically higher for patients on a statin versus those not on a statin at the start of therapy (Table 2). In many cases, the statin effect was not statistically significant, but the number of patients on statins was generally very small. Nevertheless, females were found to derive a significant benefit in the SVR rate from statin use: 70% versus 40% (P = 0.007). A nonsignificant increase in SVR rates was seen in males on pretreatment statin therapy (46% versus 39%). Patients older than 40 years had a significantly higher SVR rate: 52% versus 37% (P = 0.01). Interestingly, in patients with a baseline fasting glucose level < 5.6 mmol/L, there was a significant statin effect on SVR: 63% versus 43% (P = 0.01). The presence of hepatic steatosis was associated with a significantly higher SVR rate of 56% in statin users versus 34% in patients who were not on statins at the baseline (P = 0.004). Advanced fibrosis (METAVIR F3 and F4) was associated with a significantly higher SVR rate in the statin group (100% versus 24%, P < 0.001), although the numbers in the statin group were exceedingly low (n = 4). The statin effect on SVR was also significant for patients who received an assigned RBV dose > 13 mg/kg/day: 64% versus 42% (P = 0.01).

Although the number of patients in each group was small, more baseline statin users achieved RVR [12/66 (18%) versus 306/3004 (10%); Fig. 2]. Among statin users, SVR rates were slightly higher in those with RVR (92% and 86%, respectively; Table 3). The number of patients with a higher drop (≥2-log10 decline from the baseline) in the viral load or undetectable HCV RNA in week 4 was higher in the statin group [42/66 (64%) versus 1554/3004 (52%)]. The complete EVR rate in week 12 was higher in 55% (36/66) of the statin group versus 40% (1203/3004) in the nonstatin group and was associated with increased SVR rates of 86% versus 78%. When the two groups were evaluated according to weeks to HCV RNA first being undetectable, a greater number of patients in the statin group achieved undetectable HCV RNA in weeks 2, 4, and 12 with higher SVR rates versus the nonstatin group. Relapse rates increased with increasing time to undetectable HCV RNA, with similar relapse rates at each time point among those with and without baseline statin use.

Cholesterol Levels and Statin Use.

A multivariate logistic regression model using stepwise selection was then used to assess independent predictors of SVR (Fig. 3). LDL and HDL levels were evaluated as continuous variables; higher LDL levels resulted in an improved SVR rate (OR = 1.58, 95% CI = 1.37-1.83), as did low HDL levels (OR = 0.51, 95% CI = 0.32-0.80). Additionally, in the same model, baseline statin use was associated with an improved SVR rate (OR = 2.00, 95% CI = 1.09-3.67). LDL, HDL, and statin use were independent predictors of SVR. Other variables associated with improved SVR rates included low baseline viral loads, nonblack race, less fibrosis (stages F0-F2), absence of hepatic steatosis, elevated ALT levels, baseline fasting blood glucose levels < 5.6 mmol/L, nonsmokers or exsmokers, and low baseline hemoglobin levels.

thumbnail image

Figure 3. Multivariate logistic regression for SVR with the stepwise selection method.

Download figure to PowerPoint

Adverse Events, Discontinuations, and Dose Reductions.

No differences in deaths, serious or common adverse events, or hematological parameters were noted between the group with elevated LDL levels or low HDL levels and the group with normal LDL and HDL levels (Table 4). However, fewer patients discontinued treatment in the group with elevated LDL levels or low HDL levels (40% versus 52%). This appeared to be due to a better virological response in that group. Dose modifications were similar between the two groups for both PEG-IFN and RBV.

Table 4. Adverse Events, Discontinuations, and Dose Reductions
 Baseline Elevated LDL or Low HDL (n = 1464)*Baseline Normal LDL and HDL (n = 1540)*Baseline Statin Use (n = 66)No Baseline Statin Use (n = 3004)
  • *

    Statin users (n = 66) were excluded; nine patients with missing baseline LDL and HDL levels were included in the normal LDL and HDL group.

Deaths, n66012
Serious treatment-related adverse events, n (%)64 (4)64 (4)3 (5)128 (4)
Patients discontinuing treatment, n (%)592 (40)799 (52)25 (38)1391 (46)
 Adverse events169 (12)186 (12)7 (11)355 (12)
 Treatment failure315 (22)504 (33)11 (17)819 (27)
Hematological parameters, n (%)    
 Neutrophil count (<750/mm3 / <500/mm3)280 (19)/50 (3)361 (23)/59 (4)7 (11)/1 (2)641 (21)/109 (4)
 Hemoglobin (<10 g/dL / <8.5 g/dL)401 (27)/37 (3)448 (29)/43 (3)19 (29)/5 (8)849 (28)/80 (3)
 Erythropoietin-stimulating agent use228 (16)246 (16)8 (12)474 (16)
Dose modification, n (%)561 (38)633 (41)29 (44)1194 (40)
 PEG-IFN alfa only139 (9)151 (10)5 (8)290 (10)
 RBV only250 (17)270 (18)17 (26)520 (17)
 Both172 (12)212 (14)7 (11)384 (13)
Common adverse events (≥25% incidence), n (%)    
 Fatigue977 (67)990 (64)37 (56)1967 (65)
 Headache672 (46)730 (47)30 (45)1402 (47)
 Nausea591 (40)568 (37)28 (42)1159 (39)
 Insomnia579 (40)619 (40)20 (30)1198 (40)
 Chills491 (34)492 (32)22 (33)983 (33)
 Pyrexia462 (32)443 (29)19 (29)905 (30)
 Anemia433 (30)525 (34)28 (42)958 (32)
 Irritability399 (27)370 (24)11 (17)769 (26)
 Myalgia376 (26)389 (25)12 (18)765 (25)
 Arthralgia366 (25)348 (23)12 (18)714 (24)
 Neutropenia342 (23)426 (28)9 (14)768 (26)
 Dyspnea253 (17)254 (16)18 (27)507 (17)

Comparing the statin users to the patients who did not use statins, we found that no deaths occurred in the small statin user group, and serious adverse event rates were similar. As in the group with baseline elevated LDL levels or low HDL levels, fewer patients discontinued therapy in the statin group (38% versus 46%) because of improved virological response. A reduction of hemoglobin levels to <8.5 g/dL was slightly more common in the statin users (8% versus 3%), although the number of patients in the statin group was small. Only 12% of patients in the statin group used an erythropoietin-stimulating agent versus 16% in the nonstatin group. Dose modification of RBV was higher in the statin group: 26% versus 17%. When we assessed common adverse events, we found that fewer myalgias and less fatigue, insomnia, irritability, and neutropenia were reported in the baseline statin user group, but more anemia and dyspnea were reported.

Discussion

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References

One of the major findings in this large retrospective analysis derived from a multicenter trial was the improved SVR rate seen in patients with baseline elevated LDL levels or low HDL levels. Smaller previous studies have suggested that this might be the case for LDL,3, 4 but this is the largest trial to date to corroborate these findings and the first published study to our knowledge to show a relationship between HDL and SVR. In a small retrospective study, Gopal et al.4 found that, among patients with EVR, EOT response, and SVR, the LDL and total cholesterol levels were higher, and this was independent of age.4 Subsequently, Akuta et al.3 found that an LDL level ≥ 86 mg/dL was an independent predictor in a multivariate logistic regression analysis of EVR and SVR in 114 Japanese adults with HCV genotype 1 who were treated with PEG-IFN and RBV for 48 weeks.

The reason for SVR improvement in patients with elevated LDL levels or low HDL levels is unknown. However, the LDL receptor, a membrane glycoprotein, has been shown to be one of the proteins involved in HCV entry into hepatocytes,29 and data suggest that HCV RNA levels correlate with LDL receptor expression.28 Elevated serum concentrations of LDL may decrease the number of LDL receptors located on hepatocytes and effectively decrease overall cellular infectivity. This hypothesis would perhaps argue against the advantage of statin use in the treatment of hepatitis C. The finding of a low HDL level being associated with improved SVR supports previous work showing that HDL may modulate virus entry into the cell. Further study is necessary to delineate the role of a low HDL level in improving SVR rates.25

Another important finding of this study is the link between statin use and improved SVR. An earlier pilot study by Sezaki et al.30 suggested that improved SVR rates might be obtained with the use of statins. Statins function by inhibiting the rate-limiting enzyme in hepatocyte cholesterol synthesis, 3-hydroxy-3-methylglutaryl coenzyme A reductase. Intuitively, one might suspect that inhibiting this enzyme would result in more LDL receptors appearing on the cell membrane and thus potentially increase cell infectivity. In fact, a pilot study in patients coinfected with human immunodeficiency virus and HCV did show an increase in HCV RNA levels after 4 weeks of fluvastatin.22 Conflicting evidence suggests that the opposite may be true, with less efficient uptake of HCV pseudoparticles (34%) seen in hepatoma cells treated with mevastatin as well as less efficient (48% inhibition) internalization of HCV into cells treated with statins.20 Clearly, a better understanding of the role of the LDL receptor in HCV infectivity is required, and it is possible that the pleiotropic effects of statins may prove to be more relevant. Evidence suggests that HCV replicates within hepatocytes after an HCV replication complex is formed on lipid rafts within the endoplasmic reticulum. This complex involves an isoprenoid protein called geranylgeranyl pyrophosphate, which is attached to specific G proteins via a process called protein prenylation.31In vitro studies of various statins have shown depletion of specific nonsterol isoprenoids (particularly geranylgeranyl pyrophosphate) that are generated via mevalonate metabolism. The availability of fewer lipid rafts for replication seems to be a plausible explanation for at least some of the anti-HCV properties of statins.32, 33

The well-studied anti-inflammatory and immunomodulatory effects of statins with respect to cardiovascular and other organ systems may also explain the beneficial effects observed here in conjunction with PEG-IFN and RBV therapy. Peroxisome proliferator-activated receptor alpha (PPAR-α) is a nuclear hormone receptor important in regulating lipid, cholesterol, carbohydrate, and steroid metabolism in the liver as well as the kidneys, heart, and skeletal muscle.34 Activation of PPAR-α results in increased uptake and catabolism of fatty acids, which lead to decreased triglycerides, increased gluconeogenesis, and enhanced HDL synthesis.35 PPAR-α activation also leads to the inhibition of inflammatory response in hepatocytes via down-regulation of nuclear factor kappa B, which is well known to induce inflammatory genes and modulate the immune response.34, 35 Several studies have shown that statins increase PPAR-α expression in vitro and in vivo and induce transcriptional activity.36–38

Limiting the development of drug resistance in the era of directly acting antiviral therapies will be critical. This will likely involve synergy with other compounds demonstrating antiviral activity. The possibility that statins may play a role is intriguing. Delang et al.20 showed that statins delay or prevent development of resistance. Another finding of this study is the small number of patients with CHC who were actually on a statin. Overall, only 2.1% of the patients in our cohort were on a statin, even though 18% of the patients had baseline elevated LDL levels. The reason is likely related to provider apprehension about starting statin therapy in the setting of chronic liver disease. Provider resistance should be mitigated because recent data from several studies have demonstrated the safety of these compounds in patients with CHC.39, 40 Our data support these findings.

In agreement with previous studies evaluating the use of statins in chronic liver disease, our data suggest that statins are safe to use in patients with CHC who are undergoing therapy with PEG-IFN and RBV. No significant differences in adverse events were seen. However, there was a significantly higher number of dose modifications of RBV among the statin users.

The limitations of this study include its retrospective nature, although the numbers of patients with baseline elevated LDL levels or low HDL levels or baseline statin use were similar among the treatment arms. Additionally, only one measurement of baseline lipid levels was conducted prior to the initiation of study treatment, and this perhaps led to an overestimation of the number of patients with high LDL levels. Only a small number of patients (n = 66) were using statins at the start of treatment, and this limited a more thorough analysis of this subset of patients. Statin use was determined throughout the trial, but for the analysis of this population, only patients who were on statins at the time of the initiation of study drugs were included. There were 34 patients who began using statins during treatment. Because of the variability in the duration of statin use, these patients were not included in the cohort using statins. Because of the limitations of this study, further investigation of these factors in virological response is necessary.

In conclusion, our study suggests that patients with baseline elevated LDL levels or low HDL levels and patients on statin therapy have improved SVR rates. Our understanding of the mechanisms behind the improved response to baseline elevated LDL levels or low HDL levels is limited, and our questions cannot be answered by this study. It appears that the statins, though not effective as monotherapy, may improve SVR when they are added to standard PEG-IFN and RBV. Prospective clinical trials assessing the potential benefit of statins as adjuvant therapy for CHC may be considered.

References

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References