Treatment of chronic hepatitis C in a Canadian Aboriginal population: results from the PRAIRIE study.

BACKGROUND
The Aboriginal population of Canada is at increased risk of exposure to the hepatitis C virus (HCV). Previous data indicate that spontaneous clearance of HCV occurs more often in Aboriginals than Caucasians. Whether this enhanced response extends to antiviral therapy for chronic HCV remains to be determined.


OBJECTIVES
To document and compare the biochemical and virological responses to antiviral therapy in HCV-infected Canadian Aboriginals and Caucasians.


METHODS
A total of 101 treatment-naive adult patients (46 Aboriginal, 55 Caucasian) with chronic HCV genotype 1 infections were prospectively treated with pegylated-interferon and ribavirin and followed as per national guidelines.


RESULTS
Aboriginals had higher HCV-RNA loads at baseline (6.42log(10) versus 5.98log(10); P<0.03). Although normalization of serum aminotransferase levels, decreases in viral loads, and rapid, early and end-of-treatment virological responses were similar in the two cohorts, sustained virological responses were significantly lower in Aboriginals (35% versus 55%; P=0.047). Premature discontinuation of treatment and⁄or loss of patients to follow-up was common (Aboriginals 37%, Caucasians 27%). Treatment-related side effects were similar in the two cohorts.


CONCLUSION
Despite higher rates of spontaneous HCV clearance, the response to antiviral therapy was similar, if not lower, in Aboriginals compared with Caucasians with chronic HCV genotype 1 infections. Compliance with treatment is an issue that needs to be addressed in the management of these patients.

The study was approved by the appropriate Ethics Committees of the Universities of Manitoba (Winnipeg, Manitoba), Saskatchewan (Saskatoon, Saskatchewan) and Alberta (Edmonton, Alberta).

laboratory testing
All hematology and biochemistry testing was performed by accredited hospital laboratories at each site. Because of interlaboratory differences in the ranges of normal values, results were standardized to ratios above the upper limits of normal where appropriate.
HCV serology (anti-HCV) and HCV-RNA qualitative and quantification testing were performed using ELISAs (Amplicor and Taqman Version 1, Roche Diagnostics, Canada), respectively, at each participating centre's respective provincial laboratory. The lower limit of viral detection for the Amplicor assay is 20 IU/mL and quantification by the Taqman version 1 is 25 IU/mL.
In cases in which liver histology from within one year of treatment was available, biopsy slides were reviewed and reported (METAVIR classification system) by the local pathologist for each centre.

Statistics
Statistical analysis was performed using SAS statistical software (SAS Institute Inc, USA). Categorical variables were evaluated using χ 2 analysis. The χ 2 test of association (or F test when warranted) was used to examine differences in demographic factors, clinical variables and treatment parameters. Continuous variables were assessed using the Student's t test or ANOVA. To compare the dynamic of changes in viral load and biochemical markers between the two groups, a repeated-measures ANOVA was used. Multivariate regression analysis was used to assess whether and how patients' various demographic, clinical and viral characteristics influenced outcome variables such as rates of biochemical and virological response. Decreases in viral load and quantitative assessments (frequency) of parameters such as the number of discontinuations, withdrawals, side effects, responses, etc, were examined using Poisson regression. Logistic regression models for predicting SVR after treatment were constructed on the basis of demographic and clinical data found to be significant on univariate analysis. All data were analyzed on an intention-to-treat basis; 95% CIs were computed and statistical significance was set at P<0.05.

RESulTS
A total of 122 patients enrolled in the study. Eight individuals chose not to participate after having signed the consent form but before receiving treatment. An additional 13 were excluded due to a protocol violation (enrollment of HCV genotypes 2 or 3). Thus, the results provided are derived from a total of 101 patients: 46 Aboriginal (37 First Nations and nine Métis) and 55 Caucasian.
Baseline characteristics of the two study cohorts (Aboriginal and Caucasian) are provided in Table 1. The mean age, sex, weight, body mass index, alanine aminotransferase (ALT) and total bilirubin levels were similar in the two groups. The percentage of patients with radiological evidence of fatty infiltration of the liver was also similar (17% Aboriginal and 15% Caucasian). Although radiological evidence of cirrhosis (irregular liver margins and/or signs of portal hypertension) was more common in Caucasians (13% versus 8.7%), this difference did not reach statistical significance (P=0.74).
Of the 38 individuals (14 Aboriginal and 24 Caucasian) who had undergone a liver biopsy within one year before treatment, there were no significant differences in METAVIR grades or stages, although as with radiological imaging, a trend toward more advanced disease (stage 4) was apparent in Caucasians (21% versus 7.1%; P=0.51).
The results of serum ALT and bilirubin testing over the course of treatment are provided in Figure 1. In both cases, the decrease in values was similar in the two cohorts. The percent of Aboriginals who achieved normal ALT values at four, 12, 24 and 48 weeks were 43%, 50%, 37% and 24%, respectively, compared with 51%, 50%, 45% and 38%, respectively, in Caucasians. These differences were not statistically significant.
Treatment was discontinued in 17 (37%) Aboriginals and 15 (27%) Caucasians (P=0.30). Reasons for discontinuation of treatment are provided in Table 3. The most common reason in both cohorts was   Table 4) were similar in the two populations and consistent with those described previously for this treatment regimen.
The results of logistic regression testing are shown in Table 5. Although Aboriginal ethnicity was associated with treatment outcome on univariate analyses, only premature treatment termination, normal baseline ALT values and the presence of fatty liver on ultrasound were associated with treatment outcome following multivariate analyses.

DISCuSSIOn
Based on previous epidemiological and in vitro findings, we predicted a higher response rate to antiviral therapy in Canadian Aboriginals infected with HCV compared with Caucasians. However, the results of the present study suggest that response rates are similar, if not lower, in Aboriginals. There are a number of possible explanations for this finding. First, the in vitro findings referred to above reflect differences in the innate immune response to HCV, whereas adaptive immunity is believed to be more relevant for individuals who have progressed to and are being treated for chronic HCV infection (10). Second, baseline HCV viral loads of Aboriginals were higher than that of Caucasians, and higher viral loads are associated with lower antiviral response rates (11,12). Third, although not statistically significant, the higher rate of noncompliance among Aboriginals (twice that of Caucasians) would likely have contributed to the subsequent lower SVRs in that cohort (13). Finally, the limited number of study subjects introduces the possibility of a type 2 statistical error.
Unfortunately, there are no previous publications of prospective studies describing response rates to antiviral therapy in HCV-infected Aboriginals (relative to Caucasians) that could be used for comparative purposes. The only previous report that addressed this issue was a retrospective analysis of treatment results in HCV-infected Aboriginals and Caucasians (14). However, subjects in that report were infected with various HCV genotypes and unspecified numbers were treatment naive or treatment experienced. Moreover, no standardized treatment regimen or follow-up was used. Notwithstanding these limitations, the reported EVR, ETR and SVR rates were similar in that study's two cohorts.
A somewhat unexpected finding in the present study was the higher viral load documented at baseline in the Aboriginal cohort. Here again, the explanation for this finding remains unclear. Alcohol abuse can increase HCV viral loads and, although subjects with alcohol dependency were excluded from the study, intermittent alcohol abuse very likely occurred (15). Whether such abuse was more common in Aboriginal study participants, as has been described for Aboriginals in general, was not determined (16). The absence of immunosuppressive medications and negative serological testing for HIV infection (as per study exclusion criteria) render these additional causes of increased HCV viral loads unlikely.
There are a number of limitations to the present study that warrant consideration. First, the number of subjects enrolled was small. This stems from the study having been prematurely terminated due to ethical concerns resulting from the licencing of protease inhibitors that rendered pegylated-interferon and ribavirin therapy no longer the    standard of care. Second, the limited number of liver biopsies available for analysis precluded an accurate determination of whether advanced fibrosis or cirrhosis was present and/or more common in one study cohort. Third, study investigators did not have access to interleukin-28b phenotype testing; this marker has been identified as an important predictor of response to interferon-based therapy for HCV (17). Finally, as reflected by the high noncompliance rates, resources were not available to encourage patients to continue to participate in the trial. While presumably impacting on response rates and SVRs, this limitation also provides a more 'real-world' expectation of the results of antiviral therapy in community-based populations.

COnCluSIOn
The principal finding of the present study suggests that HCV genotype 1-infected, adult Canadian Aboriginals have lower SVR rates following treatment with pegylated-interferon and ribavirin compared with Caucasians. Although the explanation for this finding remains to be determined, differences in patient compliance need to be explored.

ACKnOWlEDGEMEnTS: This study was supported by a grant from
Roche Canada Ltd. The authors thank the following staff for their assistance; Ms Beth McHattie, Pam Thompson, Suzanne Sumner, Sylvia Skrypnyk, Carol Dupasquier, Marianne Wiebe and Caroline Blackner. They thank Ms R Vizniak for her prompt and accurate typing of the manuscript.