Peginterferon alfa-2a (40 kd) and ribavirin for black American patients with chronic HCV genotype 1

Authors


Abstract

Black Americans (blacks) have a high prevalence of chronic hepatitis C virus (HCV) infection and respond poorly to therapy with interferon alfa-based regimens, but they have been underrepresented in clinical trials. The aim of this study was to assess the rate of sustained virological response (SVR) to peginterferon alfa-2a (40 kd) in combination with ribavirin in black patients chronically infected with HCV genotype 1. In a prospective, multicenter, open-label trial, 78 black and 28 white American interferon-naïve patients were enrolled to receive once weekly subcutaneous injections of 180 μg peginterferon alfa-2a plus oral ribavirin (1000 mg/d for patients weighing less than 75 kg and 1200 mg/d for patients weighing 75 kg or more) for 48 weeks. Pre- and post-treatment liver biopsies were evaluated for necroinflammation and fibrosis. SVR, defined as undetectable (<50 IU/mL) HCV RNA, was 26% in the black group and 39% in the white group. Although the SVR rate was lower in blacks than in whites, the SVR of 26% represents an improvement over previously reported SVR rates from smaller, retrospective studies of black patients. We also observed improvement in fibrosis in 25% of the black patients. No unexpected adverse events occurred. In conclusion, this prospective study evaluating responses of black patients with chronic hepatitis C to peginterferon alfa-2a/ribavirin has demonstrated that treatment can be safely offered to such individuals with reasonable antiviral and histological benefit. (HEPATOLOGY 2004;39:1702–1708.)

Large racial disparities in the prevalence of hepatitis C virus (HCV) infection exist in the United States. Antibodies to HCV are 2 to 3 times more common among black Americans (blacks) than among white Americans (whites),1 as is hepatocellular carcinoma, a serious complication of chronic HCV infection.2, 3 In addition, the distribution of HCV genotypes differs among racial groups, with genotype 1 virus responsible for approximately 90% of HCV infections in blacks, compared with only 67% in whites.1

Meaningful assessment of the efficacy and safety of antiviral therapy in blacks has been hampered by the underrepresentation of these patients in prospective clinical trials.4 Nevertheless, the limited data from retrospective analyses indicate that the response to interferon, alone and in combination with ribavirin, is lower among blacks than among whites.5–7 Pegylated interferons plus ribavirin are now the standard of care for the treatment of patients with chronic hepatitis C.8 Two recent large clinical studies of treatment with pegylated interferons plus ribavirin showed overall sustained virological response (SVR) rates of 54% and 56%.9, 10 For patients with HCV genotype 1, the SVR rates were 42% and 46% in the respective studies.9, 10 The primary objective of the present study was to investigate the efficacy and safety of peginterferon alfa-2a (40 kd) plus ribavirin in a population of non-Hispanic black patients with HCV genotype-1 infection.

Abbreviations

HCV, hepatitis C virus; SVR, sustained virological response; ALT, alanine aminotransferase; ULN, upper limit of the normal range; EVR, early virological response; HAI, histological activity index; AE, adverse event; ITT, intent-to-treat.

Patients and Methods

Patient Selection.

Eligible patients included male and female outpatients aged 18 years or older with documented chronic hepatitis C as evidenced by anti-HCV and by a polymerase chain reaction assay for HCV RNA. In addition, patients had to be non-Hispanic black or white, be naïve to treatment with any form of interferon or ribavirin, and have HCV genotype 1 as determined by automated DNA sequencing. Other inclusion criteria were the same as those described previously10; these included a liver biopsy within the previous 24 months consistent with chronic HCV infection but without evidence of cirrhosis; serum HCV RNA greater than or equal to 600 IU/mL; and serum alanine aminotransferase (ALT) above the upper limit of normal (ULN) on at least 2 occasions within the previous 12 months and at least once during the 6 months before initiation of study drug. Patients with clinically significant concomitant systemic illnesses were excluded. A negative pregnancy test for all women of childbearing age and effective contraception by all patients and their partners of childbearing age during the study period were also required, as described previously.10

Study Design.

This was an open-label, noncomparative, multicenter study investigating the efficacy of peginterferon alfa-2a plus ribavirin as initial treatment of non-Hispanic black patients with chronic HCV genotype 1 infection. The study was conducted at 11 clinical sites in the United States. Informed consent was obtained from each patient. The study protocol was reviewed and approved according to ethical guidelines of the 1975 Declaration of Helsinki by institutional review committees at the respective sites.

Patients were screened within 35 days of their first dose of peginterferon alfa-2a and ribavirin. Patients who fulfilled the inclusion and exclusion criteria received 180 μg peginterferon alfa-2a subcutaneously once weekly. In addition, patients weighing less than 75 kg received 1,000 mg/d ribavirin (400 mg in the morning and 600 mg in the evening); patients weighing 75 kg or more received 1,200 mg/d ribavirin (600 mg each in the morning and evening). Patients were treated for 48 weeks and followed up for an additional 24 weeks. Stepwise reductions in peginterferon alfa-2a or ribavirin dose were allowed to manage adverse events or laboratory abnormalities.10

Assessments.

The primary efficacy variable in this study was the proportion of patients with SVR, defined as an undetectable (<50 IU/mL) level of serum HCV RNA, as measured by the AMPLICOR HCV Test (Roche Diagnostics, Branchburg, NJ, version 2.0), at the end of the follow-up period (week 72). All virological testing was done by the central lab (LabCorp, Raritan, NJ). SVR rates were calculated separately for black and white patients.

Secondary efficacy variables included virological response over time, defined as less than 50 IU/mL serum HCV RNA at each visit during the treatment and follow-up periods. The utility of early virological response (EVR) in predicting SVR was also assessed. EVR was defined as either an undetectable HCV RNA (<50 IU/mL) or a minimum 2-log10 decrease from baseline in HCV RNA at week 12. EVR was then cross-tabulated with SVR in order to assess its predictive value.

Slides of liver biopsy specimens obtained before the study and approximately 24 weeks after completion of treatment were coded and read by a central pathologist (Dr. Sugantha Govindarajan, Liver Research Laboratory, Downey, CA), who was blinded regarding patient identity and date of biopsy. A fibrosis response was defined as a decrease of greater than or equal to 1 point and worsening was defined as an increase of greater than or equal to 1 point on the fibrosis component of the Histology Activity Index (HAI) scale.11

Safety was assessed by recording adverse events (AEs) as reported by patients, by clinical laboratory test results (including hematology, blood chemistry, thyroid function tests, and urinalysis), and by vital signs.

Data Analysis.

The trial was designed to enroll a sufficient number of black patients to estimate SVR so that the width of the 95% CI would be less than 20%. A small number of white patients was enrolled as a reference group to ensure that SVR values for whites were consistent with historical controls in previous clinical studies. As such, the ratio of black to white patients was 3:1. Efficacy analyses were performed using an intent-to-treat (ITT) population, defined as all enrolled patients who received at least 1 dose of any study medication. Two-sided 95% CIs were calculated for each efficacy variable, expressed as percentages of patients. Continuous efficacy variables were summarized using descriptive statistics.

Statistical analyses of safety were performed on all enrolled patients who received at least 1 dose of study medication and had at least 1 postbaseline safety assessment. Safety variables were summarized using proportions, frequency distributions, and descriptive statistics as appropriate.

Results

Patient Characteristics.

A total of 108 patients, 80 blacks and 28 whites, were enrolled in the study. Two patients, both blacks, did not receive any study drug, leaving 78 black and 28 white patients in the ITT and safety populations. The characteristics of the black and white groups were generally similar (Table 1), except that the mean weight of the black group was about 6 kg greater than that of the white group, and the proportion of males was higher in the black group (72%) than in the white group (61%). Due to the small number of whites in this study, none of these differences was statistically significant. Mean baseline HAI activity and fibrosis scores were comparable for the 2 groups. Only 1 patient in each group had cirrhosis (fibrosis score of F4), as reported by the study pathologist upon review of the submitted material.

Table 1. Patient Characteristics
CharacteristicBlack (n = 78)White (n = 28)
  • *

    ALT quotient represents ALT value/ULN.

Age, y  
 Mean ± SE46.3 ± 0.744.7 ± 1.4
 Minimum–maximum23–6629–62
Sex  
 Male, n (%)56 (72%)17 (61%)
 Female, n (%)22 (28%)11 (39%)
Weight, kg  
 Mean ± SE90.7 ± 1.984.7 ± 4.4
 Minimum–maximum54.3–136.447.3–128.2
Serum HCV RNA  
 >8.0 × 105 IU/mL, n (%)51 (65%)17 (61%)
 ≤8.0 × 105 IU/mL, n (%)27 (35%)11 (39%)
ALT  
 Mean ± SE (U/L)62.8 ± 5.7864.2 ± 12.5
 ALT quotient*2.12.1
Genotype, n (%)  
 1a10 (13%)5 (18%)
 1a variant31 (40%)12 (43%)
 1b37 (47%)11 (39%)
Baseline Fibrosis Scores, n (%)  
 F02 (2.6%)1 (3.6%)
 F144 (56%)16 (57%)
 F330 (38%)10 (35%)
 F41 (1.3%)1 (3.6%)

Of the 78 black patients, 62 (80%) completed 48 weeks of treatment and 60 (77%) returned at week 72. In comparison, 22 of the 28 white patients (79%) completed 48 weeks of treatment and 17 (61%) returned at week 72.

Virological Responses.

ITT analysis revealed that 20 of 78 black patients (26%; 95% CI, 16%-35%) and 11 of 28 white patients (39%; 95% CI, 21%-57%) achieved SVR. When only patients who completed 48 weeks of treatment were analyzed, 20 of 62 blacks (32%; 95% CI, 21%-44%) and 11 of 22 whites (50%; 95% CI, 30%-71%) achieved SVR. Thus, there was a trend toward a lower SVR rate for the black patients.

Virological response rates in both groups peaked at week 24 (Fig. 1), although the response rate was lower in the black group than in the white group at all study visits. EVR was 60% (47/78) for black patients, and the negative predictive value of the week-12 virological response was 100% for both blacks and whites (Table 2). Only 20 of 47 (43%) black patients with EVR went on to SVR. Further analysis showed that 22 black patients had undetectable levels of HCV RNA at week 12 and that 16 (73%) of these patients went on to SVR. Use of the HCV RNA cutoff of 30,000 IU/mL12 instead of a minimum 2-log10 decrease in viral load as a criterion for EVR did not increase the positive predictive value of EVR; in fact, it decreased the number of patients who could be excluded from treatment with a 100% negative predictive value (data not shown).

Figure 1.

Virological response over time. ITT analysis of serum samples from 78 black patients and 28 white patients. Blood was collected from patients at baseline and at weeks 1, 12, 24, 48, 68, and 72, and serum HCV RNA was measured using the Roche AMPLICOR HCV Test, version 2.0. The response rate at each time point was defined as the percentage of patients in the ITT population in each group having undetectable (<50 IU/mL) HCV RNA.

Table 2. Predictive Value of Week-12 Early Virological Response (EVR)
ParametersBlack (n = 78)White (n = 28)
SVR (n)No SVR (n)SVR (n)No SVR (n)
  • NOTE. EVR at week 12 is defined as undetectable (<50 IU/mL) HCV RNA or a decrease in viral load from baseline of >2-log10; SVR is defined as undetectable HCV RNA at week 72.

  • *

    Positive predictive value is the frequency with which patients achieve SVR if they have an EVR; negative predictive value is the frequency with which patients fail to achieve SVR if they fail to achieve EVR.

EVR at week 122027118
No EVR at week 1203109
Predictive value of EVR at week 12*    
 Positive predictive value43%58%
 Negative predictive value100%100%

Of the 45 black patients with high baseline HCV RNA titers (>8 × 105 IU/mL), 9 (20%) achieved SVR.

Independent Factors Associated With SVR.

Both univariate and multivariate analyses to identify predictors of SVR in the black patient group were performed. Factors entered into the final stepwise regression analysis included age (>40 vs. ≤40), sex (female vs. male), weight (>85 kg vs. ≤85 kg), baseline ALT (>3 times ULN vs. ≤3 times ULN), baseline HCV RNA (>800,000 IU/mL vs. ≤800,000 IU/mL), and baseline total HAI score.Univariate analysis identified baseline HCV RNA less than 800,000 IU/mL as the only significant predictor of SVR (odds ratio [OR] 0.31; 95% CI, 0.11–0.89; P = .03). On multivariate analysis, age less than or equal to 40 (OR 0.13; 95% CI, 0.02–0.82; P = .03), HCV RNA less than 800,000 IU/mL (OR 0.21; 95% CI, 0.06–0.66; P = .008), and ALT less than or equal to 3 times ULN (OR 0.06; 95% CI, 0.00–-0.84; P = .037) were significant baseline predictors of SVR. Weight (≤85 kg) was not a significant predictor in the multivariate analysis (OR 0.50; 95% CI, 0.18–1.40; P = 0.186).

Histological Response.

Paired biopsies from 53 of 78 (68%) patients in the black group and from 16 of 28 (57%) patients in the white group were obtained and reviewed. Evaluation of the paired biopsies from black patients revealed that improvements in fibrosis score occurred in 13 of 53 (25%) patients. Although the difference was not statistically significant (P = .16), only 1 of 16 (6%) in the white group showed improvement (Table 3). Over 90% of the patients in both groups showed either improvement or stabilization of fibrosis.

Table 3. Change in Fibrosis Score
Fibrosis Score (Knodell, Part IV)*Black (n = 53)White (n = 16)
  • NOTE. All patients with paired biopsies.

  • *

    Improved represents score decrease of ≥1, stable represents no change in score, worsened represents score increase of ≥1.

Improved, n (%)13 (25)1 (6)
Stable, n (%)36 (68)14 (88)
Worsened, n (%)4 (8)1 (6)

When combined virological and fibrosis responses were evaluated, a substantial group of black patients (8/36, 22%) showed fibrosis improvement without achieving SVR. These included 5 of 14 (36%) patients who relapsed and 3 of 22 (14%) nonresponders. In contrast, 5 of 17 (29%) black patients who achieved SVR also demonstrated fibrosis improvement.

Safety.

Incidence rates for most AEs were generally higher among whites than among blacks, with injection-site erythema, vomiting, alopecia, dry skin, and sinusitis having a threefold greater frequency in whites than in blacks. The most frequent AEs during treatment are delineated in Table 4. There were no unexpected AEs or unusual patterns of AEs reported during this study. One black patient died as a result of an acute anterior myocardial infarction about 180 days after the last dose of study drug, but this was not considered drug-related.

Table 4. Adverse Events Reported by 15% or More of Patients in Each Group
Adverse EventBlack, n = 78 n (%)White, n = 28 n (%)
Fatigue47 (60)20 (71)
Headache42 (54)23 (82)
Rigors27 (35)9 (32)
Insomnia21 (27)14 (50)
Rash20 (26)5 (18)
Nausea18 (23)15 (54)
Arthralgia18 (23)6 (21)
Myalgia17 (22)9 (32)
Pyrexia15 (19)7 (25)
Anorexia15 (19)3 (11)
Pain14 (18)6 (21)
Pruritus14 (18)6 (21)
Back pain14 (18)5 (18)
Depression13 (17)8 (29)
Influenza-like illness13 (17)2 (7)
Cough12 (15)3 (11)
Dizziness (excluding vertigo)10 (13)8 (29)
Dyspnea9 (12)7 (25)
Abdominal pain (upper)9 (12)6 (21)
Diarrhea9 (12)6 (21)
Irritability8 (10)8 (29)
Pharyngolaryngeal pain8 (10)6 (21)
Alopecia5 (6)9 (32)
Dry mouth5 (6)5 (18)
Injection-site erythema4 (5)8 (29)
Dry skin3 (4)7 (25)
Vomiting3 (4)6 (21)
Sinusitis2 (3)5 (18)

Dose modifications, growth factor administration, and treatment discontinuations are summarized in Table 5. Decreases in hemoglobin levels to less than 8.5 g/dL were observed in 4 black (5%) and 2 white (7%) patients, but there were no discontinuations for anemia. Neutropenia was the most common reason for modifying peginterferon alfa-2a dose, occurring in 29 of 78 blacks (37%) and 5 of 28 whites (18%); anemia was the most frequent reason for modifying ribavirin dose, occurring in 19 of 78 blacks (24%) and 9 of 28 whites (32%). Thrombocytopenia, defined as platelet counts less than 50 × 109/L, was evident in 3 black (4%) patients and 1 white (4%) patient. There were no clear patterns of dose reduction relative to any clinical parameter, and most dose reductions were transient, with patients returning to their assigned doses later in the course of treatment. No relationship between administration of erythropoietin and granulocyte colony stimulating factor and virological response was evident.

Table 5. Dose Modifications, Treatment Discontinuations, and Growth Factor Administration in Each Group
GroupDose ModificationsGrowth Factor AdministrationTreatment Discontinuations
Peginterferon alfa-2aRibavirinEPOG-CSF
  1. Abbreviations: EPO, erythropoietin; G-CSF, granulocyte colony stimulating factor.

Black36/78 (40%)31/78 (40%)3/78 (4%)3/78 (4%)18/78 (23%)
White8/28 (29%)13/78 (46%)4/28 (14%)2/28 (7%)11/28 (39%)

Premature discontinuation by the investigators for laboratory abnormalities or AEs occurred in 4 (5%) black and 5 (18%) white patients. Of the 4 black patients in this group, 2 had neutropenia, and 2 had serious infections (1 cellulitis, 1 appendicitis). The cellulitis occurred in a patient with a history of recurrent cellulitis, and the event was at week 70. Of the 5 white patients in the group, 1 had thrombocytopenia, 1 had abnormal liver function tests, and 3 had serious infections (1 cellulitis, 1 appendicitis, and 1 gram-positive sepsis). The appendicitis was diagnosed in a patient who related a history of epigastric pain, the cellulitis in a veterinary worker as a result of an animal bite, and the sepsis occurred at week 72 after an auto accident. None of the serious infections were associated with severe (grade 3/4) neutropenia, defined as less than 0.75 x 109 neutrophils/L, at any point during the study. Most of the patients that withdrew from this study either failed to return for follow-up or withdrew consent. In addition, black patients in this study were as compliant as patients in a previous trial with the same regimen, in which 22% of patients discontinued therapy.10

Discussion

We have shown in this prospective study that treatment with peginterferon alfa-2a plus ribavirin was modestly effective in black Americans chronically infected with HCV genotype 1. At the end of the follow-up period (week 72), the SVR rate was 26% in the black group, which was lower than the SVR rate (39%) in the white reference group. In a larger trial of peginterferon alfa-2a plus ribavirin, an SVR rate of 46% was reported for 298 predominantly white patients infected with HCV genotype 1.10 The lower SVR rate for whites in this study may have been due to the relatively small cohort of patients, the relatively high rate of premature discontinuations, and failure to return for the 72-week follow-up visit. In addition, response rates in nonregistration trials or clinical practice settings are often lower overall.

While parallel comparisons between the results of this trial and others must be interpreted with caution, our finding of a lower SVR in blacks than in whites is consistent with previous findings that blacks have lower rates of response. For example, in a retrospective analysis, the SVR in response to interferon monotherapy or combination therapy with interferon plus ribavirin was 11% (5/53) in blacks, compared with 27% (432/1600) in whites (P = .01).6 The results presented here also confirm those of 3 previous trials, in which it was shown that the SVR rates to interferon monotherapy were lower in blacks than in whites, even when the 2 groups were otherwise well matched in factors such as HCV genotype, markers of prior hepatitis B infection, and prevalence of diabetes.5, 13, 14

Although the reasons for apparent racial differences in the virological response to interferon therapies for chronic HCV are not understood, HCV genotype 1, which has been consistently associated with a poorer response to interferon,15–18 is more prevalent among blacks than among whites.1, 6 The genotype prevalence cannot explain the difference in this study, since all patients in this trial were infected with HCV genotype 1. The presence of higher HCV RNA titers has also been associated with a poorer treatment response.8, 10, 19 In this study, the number of patients with high viral load is too small to permit interpretation.

A recent report measured the “effectiveness” of interferon therapy, defined as the percent inhibition of virus production during the first 24 hours of therapy, in blacks and whites.19 Racial differences in therapeutic outcomes with standard interferon were attributed to differences in the effectiveness, estimated at 98% for white patients and 87% for black patients (P = .005). Only 7 of 19 blacks (37%) exhibited effectiveness greater than 90%, compared with 13 of 15 whites (87%; P = .003). These differences in viral clearance may be due to variations in interferon pharmacokinetics, signal transduction pathways, or immunologic factors.

The baseline factors identified by multivariate analysis as significant predictors of SVR were age (≤40 years), viral load (≤800,000 IU/mL), and ALT (≤3 times ULN), similar to those from a larger clinical trial of peginterferon alfa-2a plus ribavirin in which age of 40 or less, HCV genotype other than 1, and body weight of 75 kg or less were significant predictors of SVR.10 Comparisons between that trial and the results shown here are difficult to draw, since there were too few black patients in the earlier trial to make a separate calculation.

Other factors that may lead to racial disparity are socioeconomic or environmental factors, including diet, and genetic factors, including the higher levels of serum testosterone and higher rates of aberrant immune response in blacks.1, 13, 19 During acute infection with HCV, men have been shown to have a less robust response than women, and heavier individuals a less vigorous response than lighter individuals.8–10 The low SVR rate for the black patients in this trial may be related to the higher proportion of males and the greater average weight of patients in the black group. A larger clinical trial, the ViraHep-C Study (National Institute of Diabetes & Digestive & Kidney Diseases), designed to compare the effectiveness of peginterferon alfa-2a plus ribavirin in black and white American patients infected with hepatitis C genotype 1 is currently addressing the basis of ethnic differences in response rates.

We found that patients who did not experience an EVR at week 12 uniformly failed to achieve SVR at week 72. These findings are in agreement with those of previous studies, showing that, in white patients with chronic HCV, the lack of an EVR predicts failure to achieve SVR.21, 22 In contrast, the positive predictive value of an EVR at week 12 for SVR was only 43% in the black group. As in other analyses, undetectable HCV RNA at week 12 was superior as a positive predictor, with a 73% likelihood of achieving SVR. Nonetheless, there were a few patients who achieved SVR (16%) despite having detectable HCV RNA at week 12. Thus, while the positive predictive value is higher for patients with undetectable HCV RNA, these data support current practice that all genotype 1 patients with EVR, regardless of race, should continue to receive 48 weeks of treatment with peginterferon alfa-2a plus ribavirin.

Although response to therapy of chronic hepatitis C is evaluated primarily by virological criteria, the effect on fibrosis is also important in assessing efficacy of treatment. Several studies in HCV-infected patients have shown that interferon-based therapy improved the degree of fibrosis, although this effect has been most prominent in patients who achieved a virological response.9, 22–27 Of interest in this study is that 13 of 53 black patients, but only 1 of 16 white patients, showed improved fibrosis scores. Five of the black patients also had SVRs, 5 were relapsers, and 3 were nonresponders. We recognize that any improvement that occurs in virological relapsers or nonresponders may be transient, and fibrosis progression may resume after cessation of treatment. Although intriguing, these histological results did not reach statistical significance due to the small number of patients with paired biopsies. Furthermore, because the number of patients is small, sampling error may have contributed importantly to the observed differences.28 Future larger studies to evaluate the effect of treatment on fibrosis in the black American population will be important in understanding the overall benefit of treatment.

Combination therapy with peginterferon alfa-2a plus ribavirin was reasonably well tolerated by both groups of patients. Severe neutropenia was more frequent in blacks, with neutrophil counts below 0.5 × 109/L during treatment observed only among blacks, and more blacks (37%) than whites (18%) had their doses of peginterferon alfa-2a reduced for neutropenia. Healthy blacks have been observed to have lower neutrophil counts than healthy whites,29, 30 suggesting an increased prevalence of constitutional neutropenia among blacks chronically infected with HCV. Constitutional neutropenia in HCV-infected blacks, however, does not appear to increase the risk for serious infections.17, 31 Our findings suggest that peginterferon alfa-2a can be used to treat blacks, even those with constitutional neutropenia, and that it may be reasonable to lower thresholds for dose modification due to neutropenia. The 5 serious infections that did occur in this study are a matter of concern. However, 2 of these occurred at weeks 70 and 72, well after treatment completion, and their relationship to treatment is doubtful. Furthermore, of the remaining 3 infections, 1 occurred in a patient with a history of recurrent cellulitis and 1 as a result of an animal bite. Only 1 serious infection (appendicitis) was considered possibly related to treatment.

In summary, we have shown that 48 weeks of therapy with peginterferon alfa-2a plus ribavirin results in an SVR in 26% of blacks chronically infected with HCV genotype 1. To date, this is the highest response rate to treatment observed in a black population. Our observation of improvement in fibrosis score in the black population requires confirmation in a larger trial.

Acknowledgements

The authors thank Dr. George Harb for his contributions to the conduct and analysis of this study, Dr. Sugantha Govindarajan for reading of paired biopsies, Dr. Ellen Lentz for statistical analysis, Dr. Richard Alexander for medical writing and editing assistance, and Drs. Ray Koff and Janet S. Lee for their critical review of the manuscript. Investigators participating in this study also included Samuel Daniel, MD, North General Hospital, New York, NY; Michael Dragutsky, MD, GI Center of the MidSouth, PC, Memphis, TN; Michael Fallon, MD, The Kirklin Clinic, Birmingham, AL; Gabriel Garcia, MD, Stanford University Medical Center, Palo Alto, CA; Curt Hagedorn, MD, Emory University School of Medicine, Atlanta, GA; Daryl Lau, MD, University of Texas Medical Center, Galveston, TX; Christopher O Brien, MD, University of Miami School of Medicine, Miami, FL; and Thelma Wiley, MD, University of Illinois at Chicago, Chicago, IL.

Ancillary