Racial Differences in Responses to Therapy With Interferon
in Chronic Hepatitis C
K. RAJENDER REDDY,1 JAY H. HOOFNAGLE,2 MYRON J. TONG,3 WILLIAM M. LEE,4 PAUL POCKROS,5 E. JENNY HEATHCOTE,6
DONALD ALBERT,7 AND TENSHANG JOH8 FOR THE CONSENSUS INTERFERON STUDY GROUP
The likelihood of a sustained response to a course of
interferon in patients with chronic hepatitis C correlates
with several clinical and viral factors, including age, viral
genotype and initial levels of hepatitis C virus (HCV) RNA
in serum. The role of race and ethnicity has not been
assessed. We evaluated the association of race with re-
sponse to interferon in a large randomized, controlled trial
using either consensus interferon (9 mg) or interferon
alfa-2b (3 million units) given three times weekly for 24
weeks. African-American patients participating in the study
were similar to white patients in mean age (43 vs. 42 years)
and baseline levels of HCV RNA (3.6 vs. 3.0 million
copies/mL) but had lower rates of cirrhosis (5% vs. 12%)
and more frequently had viral genotype 1 (88% vs. 66%: P 5
.004). Most strikingly, the rates of end-of-treatment and
sustained virological responses were lower among the 40
African-American patients (5% and 2%) than among the 380
white patients (33% and 12%) (P 5 .04 and .07). Rates of
response among Hispanic and Asian-American patients
were not statistically different than non-Hispanic white
patients. Median viral levels decreased by week 24 of
therapy by 2.5 logs in white patients (from 3.0 to 0.012
million copies/mL) but by only 0.5 logs among African-
American patients (from 3.6 to 1.8 million copies/mL).
Thus, there are marked racial differences in virological
responses to interferon in hepatitis C that must be consid-
ered in assessing trials of interferon therapy and in counsel-
ing patients regarding treatment. The differences in re-
sponse rates are as yet unexplained. (HEPATOLOGY 1999;30:
787-793.)
Chronic hepatitis C is the major cause of chronic hepatitis
in the United States, Europe, and Japan, ranking equal to
alcohol-induced liver disease as a cause of cirrhosis, end-
stage liver disease, and hepatocellular carcinoma.1 Population-
based surveys indicate that 1.5% to 2.0% of the general
population has antibody to hepatitis C virus (HCV), the
majority of whom are thought to be chronically infected.2-4
Many cases of chronic hepatitis C are mild, asymptomatic,
and nonprogressive, but a proportion, variably estimated to
be 20% to 40%, are progressive and are likely to lead to
cirrhosis and mortality or morbidity from liver disease.5,6
At present, therapy of chronic hepatitis C is problematic.7
Several forms of interferon are approved for use in the United
States and Europe, including alfa-2a, alfa-2b, alfa-n1 (in
Europe), and consensus interferon. Prospective controlled
trials have documented that a 6-month course of interferon
leads to clearance of HCV RNA from serum in 30% to 45% of
patients during therapy and sustained loss of this viral marker
persisting thereafter in 6% to 15%.7-12 Longer courses have
yielded higher sustained response rates, averaging 15% to
25%.8,11-13 Follow-up of patients with sustained virological
responses has shown that most remain without evidence of
continued infection and that the liver disease resolves.14,15
Thus, the quality of responses to interferon therapy can be
excellent, but the response rate is low. Furthermore, inter-
feron therapy is expensive and often poorly tolerated. Modifi-
cations in the method of administration of interferon, use of
higher doses, induction regimens, and retreatment regimens
are being evaluated.16-18 Most importantly, the combination of
interferon with the oral antiviral agent ribavirin has recently
been shown to increase the response rate substantially.19,20
Indeed, the combination of alfa interferon with ribavirin for
24 weeks (for patients with genotypes 2 and 3) or 48 weeks
(for patients with genotype 1) is now considered standard
therapy of hepatitis C.21
In retrospective analyses, several host and viral factors have
been found to correlate with a higher response rate to
interferon.12,22 These factors include younger age, female
gender, lack of cirrhosis or marked fibrosis on pretreatment
liver biopsy specimens, low levels of serum HCV RNA at the
start of treatment, and viral genotypes 2 and 3 (in comparison
with genotype 1). There are also marked geographic differ-
ences in response rates with highest rates being reported from
Japan.23,24 We have evaluated the possibility that there are
racial or ethnic differences in response rates to interferon
therapy using data from a large randomized controlled trial of
interferon alfa-2b and consensus interferon in a cohort of
North American patients with chronic hepatitis C.9
MATERIALS AND METHODS
Consensus Interferon Trial Design. The consensus interferon trial
was a prospective three-armed randomized, double-blind controlled
trial conducted in 41 centers in the United States and Canada.9 A
total of 704 patients were enrolled into one of three groups to receive
Abbreviations: HCV, hepatitis C virus; PCR, polymerase chain reaction.
From the 1University of Miami, Miami, FL; 2Division of Digestive Diseases and
Nutrition, National Institute of Diabetes and Digestive and Kidney Diseases, National
Institutes of Health, Bethesda, MD; 3Huntington Memorial Hospital, Pasadena, CA;
4University of Texas, Southwestern Medical Center, Dallas, TX; 5Scripps Clinic, La Jolla,
CA; 6University of Toronto, Toronto, Ontario, Canada; and 7Amgen, Inc., Boulder, CO
and 8Thousand Oaks, CA.
Received February 25, 1999; accepted June 24, 1999.
Supported by Amgen Inc., Thousand Oaks, CA.
Address reprint requests to: Jay Hoofnagle, M.D., Bldg. 31, Room 9A23, NIH,
Bethesda, MD 20892. E-mail: HoofnagleJ@extra.niddk.nih.gov; fax: 301-496-2830.
Copyright
r
1999 by the American Association for the Study of Liver Diseases.
0270-9139/99/3003-0026$3.00/0
787

(group 1) interferon alfa-2b in a dose of 3 million units (MU), or
(group 2) consensus interferon in a dose of 9 µg, or (group 3)
consensus interferon in a dose of 3 µg, with all groups receiving the
interferon by subcutaneous injection three times weekly for 24
weeks. All patients underwent medical evaluation, virological test-
ing, and liver biopsy before therapy. Monitoring included evaluation
of symptoms and side effects, routine liver tests and hematological
tests, and HCV-RNA levels taken at 2- to 4-week intervals during
and at 4- to 8-week intervals for at least 6 months after therapy.
Compliance was monitored by counting returned vials. The results
of this trial have been reported in full form.9
Overall, the end-of-treatment (24 weeks) and sustained (48-
week) biochemical and virological response rates were similar in
patients receiving interferon alfa-2b in a dose of 3 MU and
consensus interferon in a dose of 9 µg. Response rates in patients
receiving consensus interferon in a dose of 3 µg were less than half of
those with the higher dose, and the 3-µg dose was considered
suboptimal. Side effects were similar in frequency and severity
among patients receiving 9 µg of consensus interferon and those
receiving 3 MU of interferon alfa-2b. The details of this trial were
reviewed and approved by the local institutional review boards at
each participating center, and all patients gave written, informed
consent.
Cohort for the Current Analysis. The current analysis was limited to
patients who received either interferon alfa-2b or consensus inter-
feron in a dose of 9 µg. This cohort consisted of 472 patients, of
whom 380 were non-Hispanic white patients, 40 African-American
patients, 40 Hispanic white patients, 10 Asian-American patients,
and 2 patients with unknown racial background. Racial background
information was determined by a patient questionnaire used uni-
formly at all sites; more extensive family and racial background
information was not obtained.
Virological Assays. HCV RNA was tested by quantitative polymer-
ase chain reaction (PCR) assays performed under contract at
National Genetics Institute (Los Angeles, CA). The qualitative PCR
assay is sensitive to a level of 100 viral copies per milliliter.22 Viral
genotypes were assessed by a commercial line probe assay (Lipa;
Innogenetics, Ghent, Belgium).25
Statistical Analyses. Both univariate and multivariate methods
were used in analysis. For variables with binary outcomes, univari-
ate comparisons between groups were conducted using Fisher’s
exact test. For quantitative variables that were not normally distrib-
uted (such as changes in HCV-RNA levels) the Wilcoxon 2-sample
test was used. Logistic multiple regression was used to model
demographic parameters that might have the greatest influences on
end-of-treatment and sustained HCV-RNA responses. The baseline
parameters included in the models were race, gender, baseline HCV-
RNA levels (logarithmically transformed), and genotype. Race was
modeled as a dichotomous dummy variable and HCV RNA as a
covariate that was modeled continuously. Because of the sparse
numbers of some genotypes, the genotype variable was analyzed as
either genotype 1 or not 1. A stepwise regression was conducted in
which each variable was considered in the model. A significance
level of .05 was used to determine which variables to include in the
model.
RESULTS
Pretreatment Characteristics. Among 472 patients analyzed,
380 were non-Hispanic whites, 40 African Americans, 40
Hispanic whites, and 10 Asian Americans, and 2 were of
unknown racial background. The clinical, biochemical, hema-
tological, and virological factors in the 4 racial-ethnic groups
are shown in Table 1. Patients ranged in age from 18 to 76
years (mean, 43 years), and 72% of patients were men. The
average age and sex distribution was similar in the 4 groups.
The average pretreatment levels of serum alanine and aspar-
tate transaminase, total bilirubin, albumin, platelet counts,
and prothrombin times were also not significantly different
among the 4 groups. Liver biopsy specimens were read using
a modification of the histology activity index and yielded
similar scores among the 4 groups of patients. However, the
percent of patients with cirrhosis was lower among the 40
African American patients (5%) than the 380 white patients
(12%) (P 5 .10).
All patients were initially reactive for HCV RNA by PCR,
because this was required for enrollment. Median and mean
serum HCV-RNA levels were similar among the 4 racial
groups, although they were slightly higher in the African-
American cohort (3.6 million copies/mL) than the white
cohort (3.0 million copies/mL) (P 5 .50). For analyses of
genotypes and responses, patients were categorized into 2
groups; genotype 1 (1a, 1b, and 1; n 5 317) and other (2a, 2b,
3, 6, and not classified; n 5 155). Genotype 1 was more
frequent among the African-American patients (88%) than
the white (66%), Hispanic white (69%), or Asian-American
(40%) patients (P 5 .004).
Response Rates. Responses to therapy with interferon were
classified as biochemical (alanine transaminase levels within
the range of normal) or virological (HCV-RNA negative by
PCR) and as end-of-treatment (the 2 last specimens taken
during therapy) or sustained (6 months after therapy).
Analysis of response rates by racial group are shown in Table
2. End-of-treatment and sustained biochemical and virologi-
cal responses were not statistically different for the non-
Hispanic white, Hispanic white, and Asian-American groups.
In contrast, response rates in all categories were lower for the
African American cohort. Only 2 of the 40 African-American
patients became HCV-RNA negative during therapy, and only
TABLE 1. Comparison of Baseline Features in Four Racial Groups
Feature
White
(N 5 380)
African
American
(N 5 40)
Hispanic
White
(N 5 40)
Asian
American
(N 5 10)
Mean age (yrs) 42 43 45 50
Sex (male) 74% 63% 70% 70%
History of injec-
tion drug use 48% 50% 25% 10%
Mean ALT (U/L) 132 116 118 117
Mean AST (U/L) 81 96 95 74
Mean albumin
(gm%) 4.2 4.0 4.1 4.1
Mean prothrom-
bin time
(sec) 12.5 12.9 12.9 12.7
Mean platelet
count
(109)/L 196 207 165 150
Mean histology
activity
index score 7 6 8 8
Cirrhosis 12% 5% 20% 10%
Median HCV
RNA level 3.0 3 106 3.6 3 106 1.9 3 106 1.4 3 106
Genotypes
1 2% 3% 3% 0
1a 38% 52% 40% 20%
1b 26% 32% 25% 20%
2a 4% 10% 3% 20%
2b 12% 3% 12% 30%
3 15% 0 10% 0
Other/mixed 3% 0 8% 10%
Abbreviation: ALT, alanine transaminase; AST, aspartate transaminase.
788 REDDY ET AL. HEPATOLOGY September 1999