Epidemiology and risk factors for hepatitis C in Alaska Natives

Health Care System and HCV Testing.

Alaska Natives are comprised of three main ethnic groups: Eskimo, Aleut, and Indian. Alaska natives receive their health care at no charge through an integrated health care system of primary community health practitioners in small villages, regional hospitals, and clinics located in medium-sized communities and at the ANMC, a referral hospital with specialty care and sophisticated diagnostic facilities located in Anchorage. It has been estimated that at least 90% of ANs receive most of their health care via this system. Since July 1990, blood donors used in this population have been screened for HCV infection using an ELISA.

Since July 1990, ANMC and Artic Investigations Program has offered at no charge testing for antibody to HCV (anti-HCV) for ANs from anywhere in Alaska. All anti-HCV tests from hospitals and clinics in Anchorage, Fairbanks, Sitka, Juneau, and Bethel are carried out at ANMC. These areas account for 79% (93,882 persons) of the AN population.12 Sera for HCV testing also are received intermittently at ANMC from the other areas of the state. The ANMC Viral Hepatitis Program staff conducts hepatology clinics 3 days a week in Anchorage, and one to three times per year in Fairbanks, Juneau, Sitka, Ketchikan, Barrow, and Bethel. Patients are referred from all areas in Alaska to ANMC for liver biopsies and evaluation for treatment, according to NIH Consensus Conference Recommendations.13

HCV Registry, Recruitment Efforts, and Enrollment.

In 1992, the ANMC Viral Hepatitis Program began a registry of ANs with HCV to improve patient care and tracking. All persons positive for anti-HCV and who have a positive confirmatory test (either recombinant immunoblot assay [RIBA] or polymerase chain reaction [PCR] for HCV RNA) are enrolled in this registry. In 1994, a formal study of HCV outcomes was begun by recruiting patients from this registry. The Alaska Area Native Investigational Review Board, the Indian Health Service National, and three regional AN health boards approved this study. All persons identified to be anti-HCV positive, or referred to one of the hepatology clinics, are invited by letter or phone call to participate in the long-term study. Those enrolled in the long-term follow-up study provide written and oral informed consent. Since 1997, patients not responding to the letters have had a note placed in their medical charts asking that they be referred to one of the hepatology clinics. Patients hospitalized at ANMC who are identified as anti-HCV positive are visited in the hospital and are invited to participate in the study.

On enrollment, a history, physical examination, and laboratory evaluations are carried out for each participant. Participants complete an in-person standard interview administered by one of the study nurses to obtain demographic data and information about risk exposures and behaviors associated with HCV infection. The two principal risk exposures that we asked patients were a history of intravenous drug use (IDU) and history of blood transfusion or blood products exposure before July 1992. Other behaviors of interest included household member history of HCV, history of tattooing, history of intranasal cocaine use, >10 lifetime sexual partners, and a sexual partner with a history of IDU. For purposes of analysis, persons with IDU were assigned to that risk exposure category. Persons who were not intravenous drug users but had received a blood transfusion were assigned to the blood transfusion risk exposure category. Persons without a history of IDU and who had not received a blood transfusion were assigned to the “other” category. Individual medical records of participants are reviewed for relevant medical history, including history of liver disease, previous liver biopsy, and alcohol use.

Retrospective Testing of Stored Sera and Estimation of Date of Infection.

The Alaska Area Native Health Services has a serum bank containing more than 600,000 sera collected from previous studies over the past 30 years that is maintained by the CDC Arctic Investigations Program in Anchorage, Alaska, and many HCV long-term study enrollees have sera stored here. When participants are enrolled in the study for HCV follow-up, their consent also is obtained for anti-HCV testing of stored sera. Available sera are tested in chronologic order to identify the earliest anti-HCV-positive date. The estimated date of infection is determined using the historical risk factor data and the results from tests of stored sera. The estimated date of infection is taken as the date of first intravenous drug use (IDU) or blood transfusion exposure (before July 1992) correlated with anti-HCV testing of stored sera specimen. For those persons without a history of IDU or blood transfusion, the estimated date of infection is taken as the date midway between the date of the last anti-HCV-negative sera result and the first anti-HCV-positive result.

Laboratory Testing.

Testing for anti-HCV is performed by enzyme-linked immunosorbent assay using commercial assays (Abbott Laboratories, Abbott Park, IL) at ANMC. Persons with an anti-HCV antibody-positive specimen have confirmatory testing by RIBA or PCR. If PCR was the first test performed and the result was negative for HCV RNA, then a RIBA test was performed. If RIBA was the first confirmatory test performed and it was positive, PCR was then performed. Testing for HCV RNA and HCV genotype was performed at the University of Washington. HCV viral RNA levels were determined by the branched DNA assay version 2.0 (Bayer Corporation, Tarrytown, NY) and by quantitative reverse transcription PCR. The limit of detection of the branched DNA assay is 200,000 genome equivalents/mL (g equivalents/mL). These results were converted to international units (IU/mL) using the manufacturer's recommended conversion factor of 5.8 g equivalents/mL to 1 IU/mL. For samples negative below this limit, an endpoint dilution assay using Roche Amplicor (Roche Diagnostic Systems, Branchburg, NY) was used to characterize further low-level viremia (limit, 100 copies/mL). HCV genotype was performed by restriction fragment length polymorphism analysis of the 5′ noncoding region as previously described.14 All participants are screened for antibody to hepatitis B core antigen, and those who are positive also are tested for hepatitis B surface antigen and antibody to hepatitis B surface antigen.

Statistical Analysis.

Minimum prevalence estimates were calculated as of June 30, 2002, using population denominators estimated from the 2000 U.S. census data. Comparisons of proportions were made using the χ2 test or logistic regression in univariable analyses. Continuous variables were compared between groups by use of the Wilcoxon rank sum test. Comparisons of genotype distribution were made with a likelihood ratio test using the software StatXact 4.0 (Cytel Software Corporation, Cambridge, MA). Exact P values or Monte Carlo P values for the likelihood ratio test were reported. Multivariable analyses of the proportion of participants who were RNA positive and with RNA concentration levels ≥ 2 million were conducted by use of logistic regression. Variables with a univariable P value less than 0.25 were considered in the multivariable models. Forward selection was used to evaluate main effects. Variables were considered confounders and remained in the model if their exclusion changed the value of the coefficient(s) of interest by more than 15%. We tested for statistical significance of two-way interactions after selection of main effects. P values were considered statistically significant at the 0.05 level.

Participant Demographics.

A total of 981 ANs and AIs were found to be anti-HCV-positive and had a positive confirmatory test between July 1992 and June 2002. The minimum estimated prevalence of HCV in the AN population was found to be 0.82%, with the highest prevalence, 2.63%, found in those aged 40 to 59 years (Table 1). Of these, 759 (77.4%) are enrolled in the study. The mean age of the enrollees was 39.7 years (range, 1–82 years). Of those enrolled, 45.7% were male and 67.2% resided in an urban area (Anchorage, Fairbanks, or Juneau). No significant differences in age, gender, and residence (urban, rural) were seen between those persons who were enrolled and those not enrolled.

Distribution of Risk Factors.

A risk exposure and behavior history was obtained for 750 (98.8%) study participants. Overall, 451 (60.1%) reported a history of IDU and 105 (14.0%) had a blood transfusion as their risk exposure. Other behaviors of interest were reported in 20.3% (n = 152); these included 39 (25.7%) with a history of tattooing, 105 (69.1%) who used intranasal cocaine, 95 (62.5%) who had >10 lifetime sexual partners, 22 (14.5%) who had lived with a household member with HCV, and 46 (30.3%) who had a sexual partner with a history of IDU. Forty-two participants (5.6%) did not report any risk exposures or behaviors of interest. Females did not report any risk exposures or behaviors of interest (7.9%) more often than males (2.9%; P < 0.01). There was no difference in the distribution of risk exposures or other behaviors of interest between urban and rural residents. Twenty-five enrollees (3.3 %) were known to be human immunodeficiency virus positive. Twenty-two percent (n = 167) of enrollees were positive for antibody to hepatitis B core antigen, of whom 11 (1.4%) were coinfected with hepatitis B virus (hepatitis B surface antigen positive).

Estimated Date of HCV Infection.

The year of HCV infection was estimated for 611 (80.5%) of the 759 participants using historical risk factor data and the results from anti-HCV tests on stored sera. The median estimated length of HCV infection at consent was 13.0 years (range, 0–49 years; Fig. 1). There was no difference in length of infection by residence; however, for males, the median length of infection (16.0 years) was higher than for females (11.0 years; P < 0.01). The median length of infection was higher for blood transfusion recipients (16.0 years) and those with a history of IDU (15.0 years) than for those without these two risk factors (7.0 years). The median age at the estimated date of infection was 24.0 years (range, 0–65 years). For those with a blood transfusion as their risk exposure, the median age at infection was 27.0 years (range, 0–65 years) compared with the median age for IDU of 22.0 years (range, 8–57; P < 0.01). There was no difference in median age at estimated date of infection by gender or residence.

HCV Genotype.

Genotype was determined within 1 year after enrollment for 498 (97.1%) of the 513 HCV RNA-positive participants. For 15 persons, HCV RNA was isolated, but we were unable to determine the genotype. The HCV genotype distribution by age group found in the AN population is shown in Table 2. The percentage of persons with HCV genotype 1 increased along with age group (<40 years [58%], 40–50 years [67%], ≥50 years [70%]; P = 0.03), and the percentage with HCV genotype 3 decreased with age group (P < 0.01; Table 2). We found no association between genotype and gender, residence (rural or urban), age at estimated date of HCV acquisition (<20 years, 20–39 years, ≥40 years), risk exposure (IDU, blood transfusion, or neither), and estimated decade of HCV infection (before 1970, 1970s, 1980s, after 1989).

HCV RNA Testing.

HCV RNA testing using a branched DNA assay or reverse transcription PCR was positive for 513 (73.2%) of the 701 who were tested within 1 year of enrollment (Table 3). The percentage of participants who were RNA positive did not differ by the year enrolled into the long-term study. Males were more likely to be RNA positive than females (odds ratio [OR], 1.51; 95% confidence interval [CI], 1.07–2.12; Table 3). Urban and rural enrollees had similar proportions of persons who were RNA positive. Of human immunodeficiency virus positive patients and hepatitis B virus carriers, 80% (20/25) and 50% (5/10), respectively, were RNA positive. Persons who had a blood transfusion as their risk factor were more likely to be RNA positive than those with a history of IDU as their risk or those with no risk exposure history (OR, 2.66; 95% CI, 1.44–4.89). No significant difference was found in the proportion who were RNA positive between those with IDU as a risk factor and those with neither IDU nor blood transfusion. Those who acquired their HCV infection at age 40 years or older tended to be more likely to be RNA positive than those infected when younger than age 40 years (OR, 2.19; 95% CI, 0.90–5.34). The RNA -positivity rate differed by estimated decade of infection (P < 0.01). This was the result of a higher frequency among those infected with HCV before 1970 compared with those infected after 1969 (OR, 3.83; 95% CI, 1.49–9.80). In a multivariable analysis that considered gender, age at infection (<20 years, 20–29 years, 30–39 years, ≥40 years), residence, risk factor, and decade of infection, only blood transfusion as a risk factor (vs. all others, P < 0.01; OR, 2.87; 95% CI, 1.51–5.45) and decade of infection (<1970 vs. after 1969, P < 0.01; OR, 3.47; 95% CI, 1.34–8.96) were independent factors associated with HCV RNA positivity. Controlling for age at infection (P = 0.58), the odds ratio for blood transfusion as a risk exposure for RNA positivity is 2.63 (95% CI, 1.37–5.04) and for decade of infection is 3.56 (95% CI, 1.34–9.47).

HCV RNA Concentration.

HCV RNA concentration was determined for 492 (95.9%) of the 513 RNA-positive participants. The median RNA concentration was 3.84 million copies/mL (662,069 IU/mL) and 61% (299/492) of participants had a concentration ≥2 million (344,828 IU/mL; Table 4). Males were more likely to have an RNA concentration ≥2 million genome equivalents than females (OR, 2.18; 95% CI, 1.51–3.17). No differences were observed in the proportion of participants with an RNA concentration ≥2 million by residence (urban or rural), risk factor, or age when infected. However, RNA levels increased along with the estimated duration of HCV infection. For each 5-year increase in the duration of infection, the odds of an RNA level ≥2 million copies/mL increased by 1.16 times (95% CI, 1.04–1.29). The percentage of participants with RNA levels ≥2 million varied significantly by genotype. RNA concentrations ≥2 million were seen most commonly among persons infected with genotypes 1a (66%; 132/200) and 2b (79%; 55/70). A multivariable analysis that considered gender, residence, risk factor, age at infection, duration of infection, and genotype found that gender (males vs. females, OR, 1.94; 95% CI, 1.24–3.03; P < 0.01) and genotype (P < 0.01 overall; 1a vs. 3a: OR, 1.92; 95% CI, 1.02–3.64; 2b vs. 3a: OR, 3.17; 95% CI, 1.36–7.39) were independently associated with having an RNA concentration level ≥2 million genome equivalents. Although in this model the duration of HCV infection was not predictive of RNA level (P = 0.12), we adjusted for this variable as a confounder in the final model.