Pegylated interferon alpha 2b and ribavirin in HIV/hepatitis C virus-co-infected non-responders and relapsers to IFN-based therapy
AIDS: Volume 18(1) 2 January 2004 pp 59-66
Ballesteros, Ángel Luis; Franco, Sandraa; Fuster, Daniel; Planas, Ramónb; Martínez, Miguel Ángel; Acosta, Lesly; Sirera, Guillem; Salas, Anna; Tor, Jordi; Rey-Joly, Celestino; Clotet, Bonaventuraa; Tural, Cristina
From the HIV Clinical Unit, aIrsiCaixa Foundation, and the bHepatology and Gastroenterology Department. University Hospital Germans Trias i Pujol, Universitat Autónoma de Barcelona (UAB), Barcelona, Spain.
Chronic hepatitis C virus (HCV) infection in HIV-positive patients is a frequent and emerging health problem. Its prevalence ranges from 28 to 46% depending on the geographic location and the route of infection, increasing to 90% among injecting drug users in Southern Europe. The prolonged AIDS-free survival in highly active antiretroviral therapy-treated patients has increased the morbidity and mortality from HCV-related end stage liver disease within this population. Thus, there is an urgent need for an efficacious HCV antiviral therapy.
The primary goal of HCV treatment is to achieve a sustained virological response in order to stop fibrosis progression. Pegylated-interferon (Peg-IFN) plus ribavirin (RBV) combination therapy has obtained the highest virological response rates to date in HCV mono-infected patients. However, experience in HIV co-infection is still scarce. Limited series seem to indicate a lower response rate than in HCV mono-infected patients. Early HCV dynamics have been associated with sustained virological response (SVR), but there is insufficient information to determine its applicability in HIV-infected patients. Viral kinetics can provide new approaches to improve the management of therapy against chronic hepatitis C. As compliance to IFN-based therapies is of critical importance for SVR, an early prediction of virological response should avoid unnecessary adverse effects. A combination of directly observed treatment (DOT) strategies and early stopping rules, based on reliable predictors of efficacy, could lead to more efficient HCV implementation of therapy.
Objectives: To describe the 28-day hepatitis C virus (HCV) kinetics under Pegylated-interferon (Peg-IFN) + ribavirin (RBV) therapy in HIV/HCV co-infected patients. To evaluate the predictive value of early virological response (EVR) of achieving a sustained virological response (SVR). To investigate the baseline mutations in the interferon sensitivity determining region (ISDR)2209-2248 in the non-structural 5A protein of HCV according to genotype.
Methods: Open, prospective trial including 28 co-infected patients with directly observed treatment with Peg-IFN + RBV. We assessed the predictive values of EVR (≥ 2 log10 of HCV decay or a negative qualitative test) at days 1, 7, 28 and in week 12 of the SVR.
Results: The SVR in an intention-to-treat analysis was 28.6% (genotype 1, 1/13; genotype 3, 6/10; genotype 4, 1/5). Patients who reached SVR presented a significantly faster HCV plasma viral load reduction compared to non-responders from the first 24 h [-1.06 log10 (interquartile range, -1.7 to -0.4) versus -0.05 log10 (interquartile range, -0.4 to +0.14) respectively; P = 0.002]. The median HCV viral load at week 12 was significantly different from that at baseline in responder and transient responders but not in non-responder patients. The positive predictive value was 100% within the first month and the best negative predictive value was 92% and 88.8% at weeks 4 and 12 respectively. The only genotype 1 responder patient had eight mutations in ISDR2209-2248.
Conclusions: A very early HCV viral decay is observed in responder patients. An early virological response assessment at week 4 and 12 might be a useful tool in the clinical management of the co-infected population.
Editorial note from Jules Levin: Bear in mind this Spanish study is small, only 28 patients, and is open-label. A stopping rule in HCV monoinfected patients has been well studied and fairlu well documented finding that it is justified to syop therapy at week 12 if HCV RNA is not <200 copies/ml or a 2 log reduction has not been seen. Studies show only about 3% of patients achieve clearance of HCV RNA if they do not see this 12 week response. Nonetheless, some think continuation past 12 weeks is advisable just in case. In HCV/HIV coinfection, fewer studies with less patients have been conducted. Some studies find the 12 week rule applied effectively but the studies were small. Other studies in coinfection found that the 12 week study did not apply. HCV RNA decline has been observed to be slower in coinfected patients, thus suggesting that the 12 week stop rule is not necessarily applicable. This suggests continuing therapy for 16-24 weeks may be a better way to evaluate response, evaluation should be on an individual basis, considering the patient’s willingness to continue therapy. Bear this in mind as you read these study results and the author’s opinions. Additionally, genotype 3 patients in this study received only 24 weeks HCV treatment. In monoinfected patients 6 months therapy has been studied and may be adequate for patients with genotype 2/3. 6 months therapy has not been studied on coinfected patients, and may be inadequate. Until proven otherwise 12 months therapy for all genotypes in coinfection appears less risky.
This was an open, prospective pilot trial. The primary objective was to describe the early kinetics of HCV replication during the first 28 days of Peg-IFN + RBV therapy administered by the DOT approach in HIV/HCV co-infected patients. The secondary objective was the evaluation of the predictive value of early virological response (EVR) at different time points on the SVR. EVR is defined as a negative HCV qualitative test or a quantitative decrease of HCV viral load ≥ 2 log10 from the initiation of Peg-IFN + RBV.
We also assessed the presence of baseline mutations in the interferon sensitivity determining region (ISDR)2209-2248 in the non-structural (NS) 5A protein of HCV in the different HCV genotypes.
All patients attended the AIDS outpatient clinic at the Hospital Universitari Germans Trias i Pujol, which takes care of 2300 patients, with 43% of them HCV co-infected.
All patients in the trial were given Peg-interferon a2b 1.5 ug/kg per week (Peg-Intron; Schering-Plough, Kenilworth, New Jersey, USA) in a DOT administration, in combination with RBV 400 mg twice daily, which was above the recommended dose of 10.6 mg/kg per day for all patients (Rebetol; Schering-Plough). The length of therapy was 24 weeks for genotype 3 and 48 weeks for genotypes 1 and 4, according to the most recently accepted consensus statement.
(EDITORIAL NOTE FROM JULES LEVIN: again, there is no evidence to support that 6 months in coinfected patients with genotypes 2/3 will yield same results as 6 months therapy in monoinfected patients. Studies will be needed to evaluate this question. Until studies are conducted I think 12 months in coinfected genotype 2/3 is generally accepted.
The characteristics of the 28 HIV patients included in the study are summarized: 64% men; age 37 yrs; BMI: 22.07 kg/m2; mean CD4 count 554; 82% <200 copies/ml; 27/28 IVDU risk behavior; 46% genotype 1; 35% genotype 3; 18% genotype 4HCV viral load 5.4log IU/l. Most of them (85.7%) were on antiretroviral therapy. All possible combinations of nucleoside reverse transcriptase inhibitors were allowed. Most of the patients had a controlled HIV replication.
Patients were on average fairly well advanced with their liver disease. Twenty-one liver biopsies were performed before entry into the study. The histological activity was evaluated by a Knodell score. The median inflammatory degree was 5 (IQR, 3-7) and the distribution of fibrosis stage was as follows: 13 patients presented stage 3; three patients presented stage 1; four patients presented stage 0. No sample was obtained in one patient.
Twenty-seven patients were former injecting drug users, but only one was on methadone maintenance therapy.
There were no significant differences in the median baseline HCV viral load level between genotypes [genotype 1, 5.96 log10 (IQR, 5.13-6.35); genotype 4, 5.43 log10 (IQR, 5.29-5.86); genotype 3, 5.79 log10 (IQR, 3.97-6.03)]. We found that a higher proportion of patients with genotype 1 HCV had RNA loads > 850 000 IU/ml (8/13) than those with genotype 4 (1/5) or genotype 3 (4/10).
Early virological dynamics according to the achievement of SVR
We obtained an end of treatment response rate of 35.7% by an intent-to-treat analysis. The SVR rate was 28.6% (8/28) by an intent-to-treat and 38.1% (8/21) by an on-treatment approach. We classified the patients into two groups according to the achievement of SVR regardless of EVR. There were no significant differences in baseline HCV viral load between the two groups [baseline HCV viral load in responders and non-responders of 4.80 log10 (IQR, 3.23-5.82) and 5.60 log10 (IQR, 5.22-6.21) respectively; P = 0.121]. A significant difference in HCV plasma viral load reduction between responders and non-responders was observed within the first 24 h [-1.06 log10 (IQR, -1.7 to -0.4) versus -0.05 log10 (IQR, -0.4 to +0.14) respectively; P = 0.002]. This difference was also observed at all time points during the study period. There were significant differences in the HCV RNA viral load between baseline and the multiple time points studied during the first 28 days under Peg-IFN + RBV in the responder patients. When this analysis was performed within the non-responder patients, these differences were also found in patients with EVR who did not achieve SVR [transient responders, five patients with genotype 3 (two patients), 1 (two patients) and 4 (one patient)] but not in the non-responders without EVR. Interestingly, the median reduction of HCV viral load at week 4 and 12 in the transient responders was 1.62 log10 (IQR, 0.73-3.04) and 3.0 log10 (IQR, 2.26-3.51) respectively.
Early virological dynamics according to HCV genotypes
The distribution of SVR by an intent-to-treat analysis according to genotypes was as follows: genotype 1 (1/13), genotype 3 (6/10) and genotype 4 (1/5). We found a decrease in HCV viral load in the first day of treatment in all cases with genotype 3 (median decrease of -1.20 log10; IQR, -0.40 to -1.80, minimum decrease of -0.26 log10). In genotypes 1 and 4, the median decrease in HCV RNA load at 24 h was -0.06 log10 (IQR, -0.84 to 0.063) and -0.37 log10 (IQR, -0.57 to 0.43) respectively. These differences between genotypes achieved statistical significance (P = 0.022) There were also significant differences in the median decrease of HCV RNA load at days 4, 7, 9 and 14 between genotypes but not at 21 and 28 days [median decrease of HCV RNA load in genotypes 3, 1 and 4 at day 28, -1.76 log10 (IQR, -3.08 to -0.60 log10), -0.50 log10 (-1.32 to 0.045) and -0.88 (IQR, -2.67 to 0.27) respectively; P = 0.097]. There was a statistical difference in HCV plasma viral load between baseline and 12 weeks only for genotype 3 [median baseline and 12 week values, 5.79 log10 (IQR, 3.97 to 6.03) and 2.78 (IQR, 2.78 to 2.80) respectively; P = 0.008].
Predictive values of EVR on SVR
The two patients with an EVR at 24 h harboured genotype 3 and achieved SVR. The predictive values at 24 h and 7 days did not change. At week 4, the NPV of EVR was 92.9% and its PPV was 100%. The NPV was 88.8% at week 12 due to the delayed response in one genotype 3 patient who did not achieve EVR either at week 4 nor week 12. The PPV decreased to 58.3% at week 12 because the five cases who reached EVR later than week 4 finally relapsed (two genotype 3, two genotype 1 and one genotype 4). The only genotype 1 responder patient had the lowest baseline HCV RNA levels (3.37 log10) in his group. The only genotype 4 responder patient had a baseline HCV of 5.43 log10. Both presented an EVR at 4 and 12 weeks. The two non-responder genotype 3 patients presented a HCV viral load of > 850 000 IU/ml (one presented a breakthrough and the other relapsed).
Prognostic relevance of the NS5A gene ISDR
In the present study, the carboxyl-terminal region of NS5A protein (PKR-bd2209-2274) of HCV from the 28 patients was investigated. The ISDR could be amplified and sequenced in only two out of five patients harbouring HCV genotype 4c/4d. Both amino acid sequences obtained were different and could not be compared with an available reference sequence in the data bank to establish the number of amino acid mutations. Five out of seven patients harbouring HCV genotype 1 who did not respond had fewer than four mutations and one case could not be determined; in the only genotype 1 responder patient the ISDR had eight mutations.
Our study describes the different kinetics of HCV during the first 28 days of HCV antiviral therapy according to genotypes and the achievement of SVR in HIV co-infected patients. A reduction in HCV plasma viral load of < 2 log10 or a failure to achieve a negative qualitative test (< 50 IU/ml) after 4 and 12 weeks under Peg-IFN + RBV is associated with a lack of SVR. We have also observed that EVR assessment in the first 4 weeks of therapy is the best predictor of SVR.
Adherence to therapy is a serious concern as it is one of the most important factors in optimizing SVR. Manns et al. established the 80-80-80 rule and recommended a minimum of 80% Peg-IFN and RBV doses during at least 80% of the time prescribed. We used Peg-IFN in a DOT protocol to exclude poor adherence as a possible cause of low efficacy. The rate of premature discontinuations (28.6%) in our study is similar to that reported by other groups.
We obtained an SVR rate of 28.6% by an intent-to-treat analysis (38.1% in an on-treatment basis). The results for patients with genotype 3 are encouraging, with 60% of SVR (75% in an on-treatment basis), in contrast with only an 11.1% efficacy rate in patients with genotypes 1 and 4 (15.3% in an on-treatment basis). Although the dose of RBV could account for the low efficacy in patients with these HCV genotypes, all patients received more than the recommended dose of 10.6 mg/kg per day.
Some authors have described the predictive value of HCV kinetics at 24 h in HCV mono-infection. In our experience, the effect of HCV antiviral therapy in co-infected patients also appears as early as 24 h in responders. The rapid exponential decline in RNA levels at 24 h is explained by the effect of Peg-IFN on viral production and release. In responder patients, a continuous but slower HCV viral load decay was observed after the subsequent Peg-IFN doses in comparison to the slope observed at the first 24 h. Taking into account these two patterns, our study could suggest the existence of a second phase in co-infected patients, which could be related to the killing or clearance of infected cells, that is not well measured with the quantification of HCV plasma RNA load. Zeuzem regarded this phase as a better predictor for the likelihood of sustained response. A longer free virus half-life (7.0 h) in HCV/HIV co-infected patients compared to HCV mono-infection has been reported, but it refers only to genotypes 1 and 4. Baseline HCV RNA load has been shown to have a predictive value of sustained response. In our study, the responder patients presented a lower baseline HCV viral load than non-responder patients. Although this difference did not reach statistical significance it must be taken into account for the adequate interpretation of our results.
The majority of our patients who achieved SVR harboured HCV genotype 3 (6/8). Our data show a significant and more pronounced slope of HCV viral decrease in patients infected with genotype 3 as compared with those infected with genotype 1 and 4, similar to the data of Pawlotsky in HCV mono-infection. Perez-Olmeda et al. have reported a 30% relapse rate in genotype 3 patients, all of whom had high HCV viral loads, and suggested a slower clearance than in mono-infected patients. In our experience, the two non-responder patients harbouring genotype 3 also presented a HCV viral load of > 850 000 IU/ml. Whether or not the efficacy of Peg-IFN + RBV would improve prolonging therapy in the HIV population must be assessed, mainly in transient responders.
The standard duration therapy of 48 weeks in the difficult to treat genotypes 1 and 4 may lead to the accumulation of toxicity. Thus, a prompt recognition of no response is strongly needed. Early stopping rules might avoid toxicity and reduce costs. A reduction of ≥ 2 log10 of HCV plasma viral load from baseline, or the achievement of a negative HCV qualitative test (EVR), has been regarded as a reliable predictor of SVR. However, data about its applicability in HIV/HCV co-infection is scarce. Our results show that not achieving EVR already at week 4 has an NPV of 92.0%, as only one patient harbouring genotype 3 presented SVR but not EVR. It is important to note that the best PPV of EVR on SVR was within the first month of therapy. This finding suggests that in HIV and HCV co-infected patients the final outcome might be predicted as early as week 4, which could help in the widespread use of HCV antiviral therapy in this population. These conclusions should be explored in larger clinical trials. Early stopping rules should be applied in combination with induction strategies to minimize premature discontinuations due to toxicity, mainly in HIV-positive patients. We could hypothesize that different IFN dosing schedules might improve the initial decline of HCV plasma viral load in patients with baseline predictors of poor response.
As most of the patients harbouring genotype 1 and 4 do not present an EVR and none of them achieved an SVR, it can be assumed that those more difficult to treat patients without EVR could discontinue therapy if virological success is the only end-point considered. A tailored decision taking into account baseline liver histology and individual tolerance to therapy is recommended regarding the improvement in the necroinflammatory activity, the decrease in fibrosis progression and the lower rate of progression to end stage liver disease with IFN.
We found that the only genotype 1 responder patient presented more than four point mutations in the ISDR region, in agreement with the situation reported in HCV mono-infection. Due to the low number of patients included no conclusions about its association with SVR can be drawn. Although slow viral dynamics have also been described in HCV genotype 4 infection, it is not known if genotypes 1 and 4 share the same mechanisms of IFN resistance. We have amplified the ISDR region in only two of the five cases with genotype 4, but they could not be evaluated because of unavailable reference sequences. Given the high prevalence of genotype 4 in certain geographical areas, it could be important to investigate the mechanisms of resistance in this genotype.
In summary, the early HCV dynamics on Peg-IFN + RBV are crucial to achieve sustained response in an HCV/HIV co-infected patients. The low rate of response in this population urges the investigation of other therapeutic approaches in difficult-to-treat patients. The lack of response to Peg-IFN may be properly predicted at weeks four and twelve, particularly if adherence can be guaranteed by a DOT strategy.