 |
|
|
| |
Factors That Predict Response to Pegasys + Ribavirin (Copegus) in HCV/HIV Co-infection in the APRICOT Study
|
| |
| |
|
|
| |
| |
 |
|
| |
| |
Reported by Jules Levin
"HCV- but not HIV-related factorsat baselinepredict theresponse to treatment with peginterferon alfa-2a (40kd) (PEGASYS) plusribavirin (COPEGUS) in patientswith HIV/HCV co-infection: predictor analysis from the AIDS PEGASYS Ribavirin International Co-infection Trial (APRICOT"
authors: David Cooper, Greg Dore, Francisco Torriani, Maribel Rodriguez-Torres, N Brau, M Sulkowski, R Sola, C Tural, N Clumeck, MR Nelson, MV Correa, EW Godofsky, Ddieterich, F Duff, S Passe, and E Lissen
Following is the poster information reported by the authors and Roche for their study results presented at XV Intl AIDS Conference in Bangkok, July 2004.
In APRICOT the combination of peginterferon alfa-2a (40kd) (PEGASYS) plus ribavirin produced (P<0.0001) higher sustained virological response rates (SVR) in patients with HIV plus hepatitis C virus (HCV) (40%) than either PEGASYS monotherapy (20%) or the combination of conventional alfa-2a plusribavirin (RBV) (12%). The medication doses used in the study were Pegasys 180 ug/week by subcutaneous injection, ribavirin (COPEGUS) 800 mg/day. Patients received one of three randomized therapy regimens for 48 weeks followed by a 24 week followed period. This is standard for HCV therapy.
In patients with HCV mono-infection, SVR rates are influenced primarily by HCV genotype and baseline HCV RNA, and to a lesser extent by factors such as age, race, gender, histological status, and ALT levels.
The impact of HIV status on SVR rates in patients with HIV/HCV co-infection is not well understood. This study aimed to identify baseline characteristics predictive of an SVR in patients treated with PEGASYS plus RBV enrolled in Apricot.
Immediately below are the data results from this study followed by a report of the patient baseline characteristics in APRICOT, study methods, study design, and the logistic regression analysis.
SUMMARY: Based on the prospectively defined multivariate regression analysis, infection with HCV non-1genotypes and an HCV RNA titer <800,000 IU/mL had the greatest impact on the odds of achieving an SVR in patients with HIV/HCV co-infection treated with combination of Pegasys plus RBV. Younger age, a higher qualifying ALT quotient and the absence of bridging fibrosis/cirrhosis had a lesser influence. These results are consistent with observations in patients with HCV mono-infection.
HIV-disease related factors, CD4 count and use of ART were not significant predictors of an SVR, nor did HIV RNAlevels appear to be aprognostic factor according to the exploratory analysis of HIV-disease related factors in the overall populationtreated with Pegasys plusRBV. The authors conclude that HCV but not HIV related factors at baseline predict the response to Pegasys plus RBV in patientstreated with HIV/HCV co-infection. Given the overall SVR rate of 40% in APRICOT, the study authors recommend treatment of HCV in patients with HCV/HIV co-infection should be considered at the earliest opportunity.
PREDICTORS OF RESPONSE-ADDITIONAL ANALYSIS:
HCV genotypes other than 1, a pre-treatment HCV RNA level <800,000 IU/mL, higher qualifying ALT quotient (in 1-unit increments), the absence of cirrhosis/bridging fibrosis and lower age (in 10-year decrements) were associated with a higher odds of achieving an SVR.
INDEPENDENT FACTORS ASSOCIATED WITH A SUSTAINED VIRAL RESPONSE IN PATIENTS TREATED WITH PEGASYS/RBV
Low and high HCV RNA are defined as <800,000 and>800,000 IU/mL, respectively.
ODDS RATIO
ALT quotient (per 1-unit increment): 1.17
Age (per 10-year decrement): 1.28
Histology (non-cirrhotic vs cirrhotic): 1.96
HCV genotype (non-1 vs 1): 3.37
HCV RNA (low vs high): 3.56
HIV --RELATED FACTORS AND SVR
Consistent with the logistic regression analysis, baseline CD4 count and the use of ART did not influence SVR rates. HIV RNA level and the type of ART regimen used also did not influence SVR rates in patients treated with Pegasys/RBV.
Table 1. BASELINE HIV-RELATED VARIABLES AND SVR RATES IN 289 PATIENTS TREATED WITH PEGASYS PLUS RIBAVIRIN
| Variable | N | Patients with SVR(%) | | CD4 COUNT | | <200 CD4s | 17 | 8 (47%) | | <350 CD4s | 72 | 26 (36%) | | >350 CD4s | 216 | 90 (42%) | | HIV RNA LEVEL | | <50 c/ml | 173 | 72 (42%) | | 50-5000 c/ml | 66 | 23 (35%) | | >5000 c/ml | 49 | 21 (43%) |
Table 2. ANTIRETROVIRAL TREATMENT REGIMEN AT BASELINE AND SUSTAINED VIROLOGICAL RESPONSE (SVR) RATES IN 289 PATIENTS TREATED WITH PEGASYS PLUS RBV
| Variable | N | Patients with SVR(%) | | No ART at baseline | 45 | 20 (44%) | | Any ART combination | 244 | 96 (39%) | | Specific combinations: | | -NRTI+PI | 107 | 35 (33%) | | -NRTI+NNRTI | 87 | 36 (41%) | | -NRTIs only | 34 | 17 (50%) | | -NRTI+NNRTI+PI | 15 | 7 (47%) | | -PIs only | 1 | 1 |
BASELINE CHARACTERISTICS OF 289 PATIENTS TREATED WITH PEGASYS 180 ug/week PLUS RBV 800 mg/day (ITT)
STUDY METHODS
PATIENTS
Eligible patientshad detectable HIV RNA or HIV antobidies in serum and stable HIV disease with a CD4 count ?100. The use of a stable HAART regimen was allowed, but not required. Patients were also required to be HCV-treatment naïve, HCV antibody positive, and to have detectable serum HCV RNA, elevated ALT levels, and compensated liver disease.
STUDY DESIGN
Thisanalysis isrestricted to patients randomized to Pegasys plus RBV and who received at least one dose of study medication (ITT). The primary efficacy factor was SVR, defined in this analysis as asingle undetectable (<50 IU/mL) HCV RNA determination at the end of followup (week 72) by qualitative PCR (COBAS AMPLICOR HCV Test v2.0).
LOGISTIC REGRESSION ANALYSIS
The methods & factors included in this stepwise logistic regression analysis were prospectively defined in the study protocol. Baseline disease and demographic factors considered in the analysis were age, gender, (male vs female), race (Caucasian vs non-Caucasian), body surface area, log HCV RNAtiter, HCV genotype (1 vs non-1), histologic activity index (HAI) score, histologic diagnosis (cirrhosis/bridging fibrosis vs no cirrhosis/bridging fibrosis), ALT quotient, CD4 count and use of antiretroviral therapy (ART, yes or no).
In the stepwise model building process, a variable was added to the model if the adjusted chi-square statistic was significant at the 0.25 level and a variablewas deleted from the model if the Wald chi-square statistic was not significant at the 0.20 level.
The assumption of linearity in the logit was examined for each continuous variable entered into the model (age, body surface area, log HCV RNA titer, ALT quotient and Cd4 count). A descriptive method was used whereby values for a variable were grouped by deciles and the empirical logit [loge p/(1-p); p was the proportion of patients with an SVR] was calculated for each group and plotted versus the midpoint of the interval. If the assumption of linearity in the logit was evident, the values fell on approximately a straight line.
In addition to the logistic regression analysis, SVR rates were calculated, without controlling for othercovariates, according to baseline CD4 count, HIV RNA level and ART regimen.
LOGISTIC REGRESSION MODEL
Because the total HAI score did not meet the criterion for entry into the model and the sample size was reduced because of missing data, it was dropped from the list of potential variables for the subsequent analysis.
The assumption of linearity in the logit was examined for continuous variables in the final model. The relationship between HCV RNA and SVR rates is inversely proportional below about 800,000 IU/mL. Above thisconcentration the relationship 'flattens' out. For this reason, and because of similarity to results obtained in previous studies in HCV monoinfected patients, HCV RNA level was treated as a dichotomous variable in the final model using a value of 800,000 IU/mL.
The final model included HCV genotype, pre-treatment HCV RNAlevel, qualifying ALT quotient, histologic diagnosis and age.
|
| |
|
|
|
|
|
