Abstract

Objectives

To study the association between trough ribavirin concentration (Ctrough) with sustained virological response (SVR) and haemoglobin (Hb) decrease in HIV/hepatitis C virus (HCV)-co-infected (HIV+/HCV+) patients treated with anti-HCV therapy.

Methods

HIV+/HCV+ patients treated with ribavirin and pegylated interferon were prospectively evaluated. Qualitative and quantitative HCV-RNA, Hb levels and ribavirin Ctrough were measured at baseline and weeks 2, 4, 12, 24, 36 and 48 during therapy. HCV-RNA was also measured at 24 weeks after the end of therapy. Efficacy analysis was performed on patients with a definitive virological outcome (SVR, relapser and non-responder), whereas for toxicity analysis, dropouts were considered until the last available observation.

Results

Fifty-two patients (54.7% with genotype 1 or 4) were included. Overall, no correlation between ribavirin Ctrough and early virological response (EVR) nor SVR was found. However, in patients with genotype 1 or 4, ribavirin Ctrough was independently associated with EVR (P = 0.036) and SVR (P = 0.046). A ribavirin Ctrough cut-off of 1600 ng/mL was found to be associated with both EVR (χ2 = 5.69, P = 0.028) and SVR (χ2=4.2, P = 0.04). Higher ribavirin Ctrough correlated with Hb decrease (R = −0.361, P = 0.009) and was independently associated with an Hb decrease of >4 g/dL (P = 0.009). Receiver operating characteristic (ROC) analysis indicated that a ribavirin Ctrough of >2300 ng/mL was associated with an Hb decrease of >4 g/dL (χ2 = 8.08, P = 0.01).

Conclusions

Our study confirmed a relationship between ribavirin exposure and both efficacy and toxicity. Moreover, we found ribavirin Ctrough cut-offs for both SVR in genotypes 1 and 4 and overall haematological toxicity. These findings deserve further clinical evaluation.

Introduction

Standard treatment for patients chronically infected with hepatitis C virus (HCV) is a combination of pegylated interferon and ribavirin. Comparison among international mega-trials has indirectly suggested, but not demonstrated, that higher doses of ribavirin can be associated with better outcome of anti-HCV treatment.1,2 Few studies have directly investigated the relationship between ribavirin plasma concentration and the efficacy of anti-HCV therapy.3–6 A challenging pilot study has been conducted on 10 patients with chronic hepatitis C, genotype 1 and high viral load, treated with pegylated interferon α-2a and high dose ribavirin, in order to investigate whether higher plasma concentrations of ribavirin could increase the rate of sustained virological response (SVR).7 Interestingly, this outcome was achieved by 90% of the patients.

The most common side effect of ribavirin is reversible dose-dependent haemolytic anaemia. In several studies, it was found to be dose and concentration dependent.3,4,8,9

These findings highlight the importance of determining ribavirin plasma concentration cut-off values associated with relevant decrease in haemoglobin (Hb) levels and with virological response.

In HIV co-infected subjects, the same drugs and the same doses are recommended as in HCV monoinfected patients.10–13 In this category of patients, improving the rate of SVR is important, as only 18% of those infected with high level viral load HCV genotype 1 or 4 achieved SVR. Few studies have recently investigated the relationship between plasma concentration of ribavirin, its efficacy and haematological toxicity in HIV/HCV co-infected subjects.14–17 A significant correlation with early virological response (EVR) was observed only among patients infected by HCV genotype 1 or 4.15 Cut-off values of ribavirin plasma or erythrocyte concentration associated with EVR and severe haemolytic anaemia have been reported.14–16

The aim of this study was to explore the relationship between trough ribavirin plasma concentrations (Ctrough) and the rate of SVR and haematological toxicity in a population of HIV/HCV co-infected patients treated with standard anti-HCV therapy of pegylated interferon and ribavirin.

Patients and methods

Patients

Previously untreated HIV+/HCV+ subjects attending the outpatient clinic at the Department of Infectious Diseases, University of Turin, were enrolled in this open-label observational prospective study. Inclusion criteria were as follows: age >18 years, positive anti-HIV antibody test, positive anti-HCV antibody test, detectable serum HCV-RNA and T-lymphocyte CD4+ count >200 cells/mm3. There were no restrictions on alanine aminotransferase levels, liver histology or concomitant antiretroviral therapy. Patients with other forms of liver disease, active hepatitis A or hepatitis B infection, decompensated liver disease or hepatocellular carcinoma, severe anaemia, severe depression or other psychiatric diseases, significant cardiac or renal disease, seizure disorders or pregnancy were excluded from the study.

The study was approved by the local Ethics Committee and all included patients gave informed consent.

Treatment

Patients were treated with 180 µg of pegylated interferon α-2a or 1.5 µg/kg of pegylated interferon α-2b subcutaneously once weekly, in combination with ribavirin, for 48 weeks, regardless of HCV genotype. Ribavirin was administered twice daily, orally with food as 200 mg capsules/tablets, the daily dose depending on HCV genotype and body weight, as follows: for genoype 2 or 3, a fixed dose of 800 mg daily; and for genotype 1 or 4, 1000 mg daily if <75 kg or 1200 mg daily if >75 kg.

Monitoring

Following baseline, patients were visited at 2 and 4 weeks and then monthly for the whole treatment period of 48 weeks. They were subsequently followed for 24 weeks after cessation of therapy. General monitoring included weight, full blood cell counts, Hb level, transaminases, lipid and glucose profiles, and renal and thyroid function tests. Creatinine clearance (CLCR) was calculated by the method of Cockcroft and Gault: CLCR (mL/min) = (140−age) × W/(72 × Scr) (×0.85 for women), where age is the age in years, W the body weight in kilogram and Scr the serum creatinine level in mg/dL.18 HCV genotyping was performed before inclusion by INNO LiPA HCV II. Qualitative (sensitivity <50 IU/mL, Cobas Ampliprep/Cobas Amplicor, HCV test version 2.0, Roche Molecular Systems Inc.) and quantitative (sensitivity <615 IU/mL, Versant HCV RNA 3.0 assay bDNA, Bayer) HCV RNA tests were performed at baseline and weeks 2, 4, 12, 24, 36 and 48 and then after 12 and 24 weeks of follow-up.

Measurement of plasma ribavirin concentration

At the same time-point, plasma samples were collected ∼12 h after the dose administration of ribavirin, just before the following administration (trough value). Sex, age, weight, height, concomitant medications, present dose of interferon and ribavirin, and dose changes were recorded prior to sampling. Patients were also asked about time of last dose intake. Samples were centrifugated at 3000 rpm for 10 min to separate plasma and then stored at −80°C until analysis. Ribavirin concentrations were measured by a previously validated HPLC method.19

Study endpoints

Efficacy analysis was performed on patients with a definitive virological outcome of SVR (negative qualitative HCV PCR test at 24 weeks after cessation of the treatment), relapse (positive qualitative HCV PCR test during the first 24 weeks after the cessation of a successful therapy) and therapeutic failure (positive qualitative HCV PCR test at 24 weeks of therapy, or <2 log decrease in HCV-RNA from baseline to week 12). EVR was considered as a negative qualitative HCV-RNA and/or quantitative HCV-RNA decrease >2 log at week 12. In the toxicity analysis, dropouts (for adverse events or loss to follow-up) were considered until the last available observation. Haematological toxicity was considered as a maximum decrease in Hb level >4 g/dL with respect to baseline.

Statistics

The intent to treat (ITT) last observation carried forward method was used to analyse both efficacy and toxicity.

All patients who underwent ribavirin dose reduction or who were administered erythropoietin were excluded from the analysis. Mean ribavirin Ctrough was calculated as the mean of all available Ctrough determinations for each subject along the whole study period.

Baseline qualitative characteristics were expressed by their frequency (%), and quantitative features were recorded as median and interquartile range (IQR). χ2 test was used to compare categorical variables. The relationship between potential predictive parameters and EVR, SVR or haematological toxicity was examined by univariate and multivariate linear or logistic regression analysis. Receiver operating characteristic (ROC) curve was used to calculate cut-off values. All P values were two-tailed and differences were considered significant when the P value was less than 0.05. All analyses were performed using SPSS v13.0 (SPSS Inc., Chicago, IL, USA).

Results

Fifty-two patients were included in the study. Their baseline characteristics (demographic and immunovirological) and the treatment schedules (drugs and doses) are summarized in Table 1. Median (IQR) duration of therapy was 28.8 weeks (13.7–30.28).

Table 1

Baseline characteristics of the population

Variables Results 
Demographics  
 male sex, n (%) 33 (63.5%) 
 age (years) 40 [37–42] 
 weight (kg) 66.5 [59.3–75] 
 baseline plasma creatinine (mg/dL) 0.78 [0.73–0.91] 
 creatinine clearance (mL/min) 115 [98.9–128.4] 
Immunovirological  
 genotypes 1 or 4, n (%) 29 (55.8%) 
 HCV-RNA (log10 IU/mL) 6.3 [5.6–6.5] 
 HIV-RNA (log10 copies/mL) 1.3 [1.3–4.2] 
 HIV-RNA <50 copies/mL, n (%) 27 (51.9%) 
 T-lymphocytes CD4+ (cells/mm3491 [411–620] 
 T-lymphocytes CD4+ (%) 30.6 [23.9–38.3] 
 CD4+/CD8+ ratio 0.6 [0.4–0.97] 
 Hb (g/dL) 15 [14.2–16.1] 
 ALT (IU/L) 102 [50–150] 
Treatment  
 pegylated interferon α-2a, n (%) 40 (76.9%) 
 pegylated interferon α-2b, n (%) 12 (23.1%) 
 ribavirin dose (mg), n (%) 600, 3 (5.8%) 
 800, 31 (59.6%) 
 1000, 18 (34.6%) 
 ribavirin weight-adjusted dose (mg/kg) 13 [11.8–14.3] 
 concomitant antiretroviral therapy, n (%) 11 (21.2%) 
Variables Results 
Demographics  
 male sex, n (%) 33 (63.5%) 
 age (years) 40 [37–42] 
 weight (kg) 66.5 [59.3–75] 
 baseline plasma creatinine (mg/dL) 0.78 [0.73–0.91] 
 creatinine clearance (mL/min) 115 [98.9–128.4] 
Immunovirological  
 genotypes 1 or 4, n (%) 29 (55.8%) 
 HCV-RNA (log10 IU/mL) 6.3 [5.6–6.5] 
 HIV-RNA (log10 copies/mL) 1.3 [1.3–4.2] 
 HIV-RNA <50 copies/mL, n (%) 27 (51.9%) 
 T-lymphocytes CD4+ (cells/mm3491 [411–620] 
 T-lymphocytes CD4+ (%) 30.6 [23.9–38.3] 
 CD4+/CD8+ ratio 0.6 [0.4–0.97] 
 Hb (g/dL) 15 [14.2–16.1] 
 ALT (IU/L) 102 [50–150] 
Treatment  
 pegylated interferon α-2a, n (%) 40 (76.9%) 
 pegylated interferon α-2b, n (%) 12 (23.1%) 
 ribavirin dose (mg), n (%) 600, 3 (5.8%) 
 800, 31 (59.6%) 
 1000, 18 (34.6%) 
 ribavirin weight-adjusted dose (mg/kg) 13 [11.8–14.3] 
 concomitant antiretroviral therapy, n (%) 11 (21.2%) 

Continuous variables are expressed as median and [IQR]. ALT, alanine aminotransferase.

Pharmacokinetic analysis

A total of 223 ribavirin plasma samples were collected and processed. The median (range) number of samples collected for each patient was 3 (1–12). In seven subjects, only one ribavirin Ctrough determination could be obtained. The weight-adjusted dose of ribavirin administered at baseline was 13 mg/kg (11.7–14.3). The time-course profile of plasma ribavirin Ctrough from week 2 to week 24 is shown in Figure 1. Intra- and inter-individual variabilities were 26.8% and 48.15%, respectively. Mean (±SD) ribavirin Ctrough at month 3 and at the end of therapy was 1637 ng/mL (±580) and 1699 ng/mL (±608), respectively.

Figure 1

Ribavirin Ctrough evolution over anti-HCV treatment. Medians (50th percentiles) are shown as horizontal bars. Boxes cover the IQR (25–75th percentiles) and tails show the 5th and 95th percentiles. The broken line represents mean RBV concentration at last observation. Circles represent the outliers.

Figure 1

Ribavirin Ctrough evolution over anti-HCV treatment. Medians (50th percentiles) are shown as horizontal bars. Boxes cover the IQR (25–75th percentiles) and tails show the 5th and 95th percentiles. The broken line represents mean RBV concentration at last observation. Circles represent the outliers.

Efficacy analysis

Using the ITT analysis, EVR was achieved by 32 (61.5%) patients and SVR by 25 (48.1%). Fourteen (26.9%) had therapy failure. Among the latter, 11 did not reach EVR and the treatment was stopped as they were considered as non-responders; 2 experienced a virological breakthrough after week 12 and 1 relapsed. Thirteen patients (25%) dropped out because of adverse events, mainly correlated to interferon. Four of them dropped out of the study after week 12, and all of these were early virological responders.

Efficacy analyses were performed on the 39 patients with a definitive virological outcome (SVR or non-responders). Demographic and clinical characteristics of this group were not significantly different from the whole study population (data not shown). Genotype 1 or 4 was detected in 20 (51.3%) of these patients. At week 12, a 2 log10 reduction of HCV-RNA was achieved in 28/39 (71.8%) patients, whereas undetectable HCV-RNA was seen in 24/39 (61.5%) cases. Overall, 28 patients (71.8%) achieved EVR and 25 (64.1%) SVR.

Being infected by HCV genotype 2 or 3 was associated with both EVR (χ2 = 14.555, P < 0.0001) and SVR (χ2 = 10.364, P = 0.002). In fact, EVR and SVR were observed in 19/19 (100%) and 17/19 (89.5%) subjects with genotype 2 or 3, respectively, when compared with 9/20 (45%) and 8/20 (40%) subjects with genotype 1 or 4, respectively. Results of logistic regression analysis are shown in Table 2. Baseline HCV-RNA was associated with both EVR (P = 0.021) and SVR (P = 0.044) in total population.

Table 2

Logistic regression analysis for the prediction of EVR and SVR to anti-HCV therapy

 Early virological response Sustained virological response 
 overall genotype 1 or 4 genotype 2 or 3 overall genotype 1 or 4 genotype 2 or 3 
Baseline log HCV viral load 0.021 0.1 NA 0.044 0.17 0.54 
Baseline CD4 absolute number 0.31 0.29 NA 0.64 0.42 0.6 
Type of pegylated interferon α (2a = 1) 0.51 0.76 NA 0.91 0.99 
Baseline plasmatic creatinine 0.11 0.1 NA 0.45 0.26 0.5 
Creatinine clearance 0.8 0.6 NA 0.9 0.88 0.97 
Ribavirin dose 0.34 0.37 NA 0.89 0.76 0.52 
Weight-adjusted ribavirin dose 0.71 0.89 NA 0.78 0.99 0.74 
HCV genotype (1 or 4 = 1) NA NA NA 0.004 NA NA 
Ribavirin Ctrough 0.09 0.036 NA 0.12 0.046 0.75 
 Early virological response Sustained virological response 
 overall genotype 1 or 4 genotype 2 or 3 overall genotype 1 or 4 genotype 2 or 3 
Baseline log HCV viral load 0.021 0.1 NA 0.044 0.17 0.54 
Baseline CD4 absolute number 0.31 0.29 NA 0.64 0.42 0.6 
Type of pegylated interferon α (2a = 1) 0.51 0.76 NA 0.91 0.99 
Baseline plasmatic creatinine 0.11 0.1 NA 0.45 0.26 0.5 
Creatinine clearance 0.8 0.6 NA 0.9 0.88 0.97 
Ribavirin dose 0.34 0.37 NA 0.89 0.76 0.52 
Weight-adjusted ribavirin dose 0.71 0.89 NA 0.78 0.99 0.74 
HCV genotype (1 or 4 = 1) NA NA NA 0.004 NA NA 
Ribavirin Ctrough 0.09 0.036 NA 0.12 0.046 0.75 

Results represent the P value for each parameter. P values <0.05 are marked in bold. NA, not applicable.

Overall, no association between ribavirin Ctrough and EVR (P = 0.09) or SVR (P = 0.127) was found. However, in patients infected by HCV genotype 1 or 4, logistic regression analysis showed that mean ribavirin Ctrough was independently associated with EVR (P = 0.036) and SVR (P = 0.046). By ROC analysis, a ribavirin Ctrough cut-off of 1600 ng/mL was found to be associated with both EVR (area under ROC curve; AUROC = 0.808, sensitivity = 88.9% and specificity = 63.7%) and SVR (AUROC = 0.802, sensitivity = 87.5% and specificity = 58.3%). This ribavirin Ctrough cut-off was significantly associated with EVR (χ2 = 5.69, P = 0.028) and SVR (χ2 = 4.2, P = 0.04) in patients infected with genotype 1 or 4. In fact, 8/12 (66.7%) and 7/12 (58.3%) of such subjects with a ribavirin Ctrough >1600 ng/mL had EVR and SVR, respectively, whereas both these outcomes were observed in only 1/8 (12.5%) subjects with <1600 ng/mL. As shown in Table 2, no other factors were associated with virological responses, either in the total population or according to the HCV genotype.

Toxicity analysis

Analyses of haematological toxicity were performed in 52 patients, considering data of dropouts until last observation. Dropout was mainly due to interferon side effects (flu-like syndrome, severe asthenia, thrombocytopenia and psychiatric disorders).

As a consequence of ribavirin administration, Hb levels significantly decreased from baseline to week 2 (P < 0.0001) and from week 2 to week 4 (P = 0.001). After this time-point, Hb continued to decrease from week 4 to week 12 and from week 12 to week 24, but these variations were not statistically significant (P = 0.47 and 0.64, respectively). Variation of Hb level over anti-HCV treatment is shown in Figure 2. Overall median (range) Hb decrease at last observation before stopping therapy was −1.8 g/dL (−6.1; 2.7), and the lowest Hb value reached was 12.17 g/dL (9.7–14.9). In 4/52 (7.7%) patients, Hb level decreased to below 10.0 g/dL. A decrease of >4 g/dL of Hb level was observed in 13 of 52 (25%) cases.

Figure 2

Hb evolution over anti-HCV treatment. Medians (50th percentiles) are shown as horizontal bars. Boxes cover the IQR (25–75th percentiles) and tails show the 5th and 95th percentiles. Circles represent the outliers.

Figure 2

Hb evolution over anti-HCV treatment. Medians (50th percentiles) are shown as horizontal bars. Boxes cover the IQR (25–75th percentiles) and tails show the 5th and 95th percentiles. Circles represent the outliers.

Using linear regression analysis, nadir Hb concentration was correlated with baseline Hb level (R = −0.631, P < 0.0001), ribavirin Ctrough (R = −0.361, P = 0.009), concomitant highly active antiretroviral therapy (HAART) (R = −0.288, P = 0.039) and baseline platelet count (R = 0.327, P = 0.018). At multivariate analysis, only baseline Hb level (P < 0.0001) and ribavirin Ctrough (P = 0.02) were confirmed as independent predictors of Hb nadir level. Results of the linear regression analysis are presented in Table 3.

Table 3

Linear regression analysis for the prediction of higher Hb decrease from baseline to Hb nadir and logistic regression analysis for the prediction of maximum Hb decrease >4 g/dL

 Linear regression analysis Logistic regression analysis 
 univariate multivariate univariate multivariate 
Sex −0.218 (0.12)  0.25  
Weight −0.039 (0.78)  0.52  
Age −0.053 (0.7)  0.81  
Concomitant HAART −0.288 (0.039) −0.107 (0.34) 0.08  
Type of pegylated interferon 0.067 (0.63)  0.13  
Baseline plasma creatinine −0.22 (0.11)  0.09  
Creatinine clearance 0.085 (0.57)  0.13  
Ribavirin dose −0.172 (0.22)  0.48  
Weight-adjusted ribavirin dose −0.147 (0.3)  0.13  
Baseline Hb −0.631 (<0.0001) −0.586 (<0.0001) 0.004 0.012 
Baseline platelet count 0.372 (0.018) 0.072 (0.54) 0.35  
Ribavirin Ctrough −0.361 (0.009) −0.257 (0.02) 0.009 0.043 
 Linear regression analysis Logistic regression analysis 
 univariate multivariate univariate multivariate 
Sex −0.218 (0.12)  0.25  
Weight −0.039 (0.78)  0.52  
Age −0.053 (0.7)  0.81  
Concomitant HAART −0.288 (0.039) −0.107 (0.34) 0.08  
Type of pegylated interferon 0.067 (0.63)  0.13  
Baseline plasma creatinine −0.22 (0.11)  0.09  
Creatinine clearance 0.085 (0.57)  0.13  
Ribavirin dose −0.172 (0.22)  0.48  
Weight-adjusted ribavirin dose −0.147 (0.3)  0.13  
Baseline Hb −0.631 (<0.0001) −0.586 (<0.0001) 0.004 0.012 
Baseline platelet count 0.372 (0.018) 0.072 (0.54) 0.35  
Ribavirin Ctrough −0.361 (0.009) −0.257 (0.02) 0.009 0.043 

Results represent the coefficient of regression and the P value for each parameter. P values <0.05 are marked in bold.

Using logistic regression analysis, a maximum Hb decrease above 4 g/dL was independently associated with higher baseline Hb level (P = 0.004) and higher ribavirin Ctrough (P = 0.009), and both these parameters were confirmed to be independent predictors at multivariate analysis (P = 0.012 and 0.043, respectively). Results of the logistic regression analysis are shown in Table 3.

By using ROC analysis, a cut-off of 2300 ng/mL was found to be associated with an Hb decrease of >4 g/dL (AUROC = 0.751, sensitivity = 46.2% and specificity = 89.7%).

Using this cut-off, 6/10 (60%) subjects with a ribavirin Ctrough of >2300 ng/mL showed an Hb decrease of >4 g/dL, whereas it was observed in only 7/42 (16.7%) subjects with <2300 ng/mL (χ2 = 8.08, P = 0.01).

Discussion

Response rates and the frequency of adverse events to anti-HCV combination treatment are probably dependent on the inter-individual variability of serum ribavirin concentrations. In this context, measuring ribavirin concentration in plasma could be highly helpful for the management of treated patients.

Concerning the possible correlation between ribavirin concentration and antiviral efficacy, data are lacking.3 According to Jen et al.,20 some optimal target concentrations for ribavirin have been established (10–1521 and >15 µmol/L7), but these have been tested only in small and difficult-to-treat groups of patients.

Ribavirin's major toxicity is haemolytic anaemia. Sex, age, dose of ribavirin by body weight, pre-treatment platelet count and Hb level were alternatively found to be associated with anaemia or with ribavirin dose reduction due to anaemia.22–24 Recently, two studies have addressed ribavirin plasma concentration monitoring as a tool for optimizing anti-HCV therapy in patients with expected poor tolerability.19,25

In HIV co-infected patients, a lower rate of response and a higher rate of anaemia and of ribavirin dose reduction were reported when compared with HCV monoinfected subjects. In this setting, looking for ribavirin concentration cut-off values able to predict efficacy and toxicity is particularly challenging.

The only published study found a cut-off of Ctrough ribavirin of 2700 ng/mL, being able to predict EVR with 70% sensitivity and 49% specificity (P = 0.01).14 A cut-off value of 2100 ng/mL has been recently proposed by the same authors as able to predict undetectability at week 24, with positive- and negative-predictive values of 87% and 28%, respectively.15

We previously obtained a cut-off value of 1600 ng/mL, able to predict EVR in genoype 1 HCV-infected patients,26 and this cut-off has been recently confirmed in a similar population.27

The present study was therefore the first one to investigate the relationship between ribavirin exposure and SVR. We found a significant correlation between ribavirin Ctrough and SVR in HCV genoype 1 or 4 infected patients and identified a cut-off value of 1600 ng/mL as the best predictor of both EVR and SVR. Of note, the negative-predictive value of this cut-off was 88.9% and would allow identification of patients with low ribavirin concentration, who would benefit from an increase in dose.

A toxicity cut-off value of 2800 ng/mL has been shown to be able to predict Hb drop of >2 g/dL with 73% sensitivity and 81% specificity (P < 0.001).14

In our study, we chose a more restrictive definition of haematological toxicity (>4 g/dL Hb decrease), considered as the Hb drop suffered by the higher quartile of our population. We found a cut-off value of plasma ribavirin concentration of 2300 ng/mL, being able to predict this level of haematological toxicity with 46.2% sensitivity and 89.7% specificity.

Lower cut-off values were found in our study when compared with the ones reported in previous studies.14,15 It is noteworthy that higher ribavirin doses were used in the latter study (median weight-adjusted dose of 13.9 mg/kg) than in our study (almost 60% of the patients were administered 800 mg/day, with a median weight-adjusted dose of 13 mg/kg), resulting in higher median ribavirin Ctrough (230027 and 272014 versus 1700 ng/mL, respectively). This point probably explains the difference in the cut-off values.

Our study has some limitations. First, the analysis of prediction of efficacy and toxicity was performed in two different populations (39 and 52 patients, respectively, the former being included in the latter); for this reason—even if demographic and baseline clinical characteristics were not significantly different—we could not identify a therapeutical range of ribavirin concentrations, but only two possible cut-off values that can predict the virological response and the haematological toxicity.

Moreover, we could not determine ribavirin Ctrough at all the pre-established time-points. In seven patients, only one plasma sample was collected (in three subjects, it was obtained at week 2).

Because of the accumulation of ribavirin during the first weeks of therapy (finding confirmed in 10 patients whose ribavirin concentration was measured at all time-points), total exposure to ribavirin could be underestimated.

Finally, excluding patients who required ribavirin dose reduction or who were administered erythropoietin meant that we probably failed to observe the extreme variations of Hb levels. In contrast, in this way, we could consider the mean trough concentration as a good marker of the total exposure to ribavirin, without having to apply complex adjustments for dose modifications.

In conclusion, we found that a mean ribavirin trough concentration of 1600 ng/mL is associated with higher rates of EVR and SVR in HCV genotype 1 or 4 infected patients and that a mean ribavirin trough concentration of 2300 ng/mL is associated with higher rates of severe haematological toxicity defined as >4 g/dL decrease in the Hb level. Further investigations are required to evaluate their possible use as a clinical tool.

Funding

This study was supported using internal funding.

Transparency declarations

None to declare.

References

1
Manns
MP
McHutchinson
JG
Gordon
SC
, et al.  . 
Peginterferon α-2b plus ribavirin compared with interferon α-2b plus ribavirin for initial treatment of chronic hepatitis C: a randomised trial
Lancet
 , 
2001
, vol. 
358
 (pg. 
958
-
65
)
2
Hadziyannis
SJ
Sette
H
Jr
Morgan
T
, et al.  . 
Peginterferon α-2a and ribavirin combination therapy in chronic heptitis C
Ann Intern Med
 , 
2004
, vol. 
140
 (pg. 
346
-
55
)
3
Larrat
S
Stanke-Labesque
F
Plages
A
, et al.  . 
Ribavirin quantification in combination treatment of chronic hepatitis C
Antimicrob Agents Chemother
 , 
2003
, vol. 
47
 (pg. 
124
-
9
)
4
Lindahl
K
Schvarcz
R
Bruchfeld
A
, et al.  . 
Evidence that plasma concentration rather than dose per kilogram body weight predicts ribavirin-induced anaemia
J Viral Hepat
 , 
2004
, vol. 
11
 (pg. 
84
-
7
)
5
Maynard
M
Pradat
P
Gagnieu
MC
, et al.  . 
Relevance of ribavirin plasma concentration for the prediction of treatment response
Abstracts of the Fortieth Annual Meeting of the European Association for the Study of the Liver, Paris, 2005
 
6
Souvignet
C
Stanke-Labesque
F
Bronowicki
JP
, et al.  . 
Evidence that serum ribavirin steady-state concentrations do not correlate with early virological response during peginterferon α-2a plus ribavirin combination therapy for chronic hepatitis C
Abstracts of the Fortieth Annual Meeting of the European Association for the Study of the Liver, Paris, 2005
 
7
Lindahl
K
Stahle
L
Bruchfeld
A
, et al.  . 
High-dose ribavirin in combination with standard dose peginterferon for treatment of patients with chronic hepatitis C
Hepatology
 , 
2005
, vol. 
41
 (pg. 
275
-
9
)
8
Uchida
M
Hamada
A
Yamasaki
M
, et al.  . 
Assessment of adverse reactions and pharmacokinetics of ribavirin in combination with interferon α-2b in patients with chronic hepatitis C
Drug Metab Pharmacokinet
 , 
2004
, vol. 
19
 (pg. 
438
-
43
)
9
Inoue
Y
Homma
M
Matsuzaki
Y
, et al.  . 
Erythrocyte ribavirin concentration for assessing hemoglobin reduction in interferon and ribavirin combination therapy
Hepatol Res
 , 
2006
, vol. 
34
 (pg. 
23
-
7
)
10
Chung
R
Andersen
J
Volberding
P
, et al.  . 
Peginterferon α-2a plus ribavirin versus interferon α-2a plus ribavirin for chronic hepatitis C in HIV-coinfected persons
New Engl J Med
 , 
2004
, vol. 
351
 (pg. 
451
-
9
)
11
Carrat
F
Bani-Sadr
F
Pol
S
, et al.  . 
Pegylated interferon α-2b vs standard interferon α-2b, plus ribavirin, for chronic hepatitis C in HIV-infected patients: a randomised controlled trial
JAMA
 , 
2004
, vol. 
292
 (pg. 
2839
-
48
)
12
Torriani
FJ
Rodriguez-Torres
M
Rockstroh
JK
, et al.  . 
Peginterferon α-2a plus ribavirin for chronic hepatitis C virus infection in HIV-infected patients
New Engl J Med
 , 
2004
, vol. 
351
 (pg. 
438
-
50
)
13
Laguno
M
Murillas
J
Blanco
JL
, et al.  . 
Peginterferon α-2b plus ribavirin compared with interferon α-2b plus ribavirin for treatment of HIV/HCV co-infected patients
AIDS
 , 
2004
, vol. 
18
 (pg. 
F27
-
36
)
14
Rendon
A
Nuñez
M
Romero
M
, et al.  . 
Early monitoring of ribavirin plasma concentrations may predict anemia and early virologic response in HIV/hepatitis C virus-coinfected patients
J Acquir Immune Defic Syndr
 , 
2005
, vol. 
39
 (pg. 
401
-
5
)
15
Barreiro
P
Rodriguez-Novoa
S
Labarga
P
, et al.  . 
Comparison of hepatitis C virus early kinetics in HIV/HCV co-infected patients treated with weight-based ribavirin plus either pegylated interferon α-2a or -2b
Abstracts of the Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, CA, 2007
 
Alexandria, VA, USA
Foundation for Retrovirology and Human Health
 
Abstract 901, p. 416.
16
Cassard
B
Dominguez
S
Falconer
K
, et al.  . 
Early predictive TDM of ribavirin in HIV–HCV coinfected patients
Second International Workshop on HIV and Hepatitis Co-infection, Amsterdam, 2006
  
Abstract 26
17
Rendon
A
Nuñez
M
Romero
M
, et al.  . 
Ribavirin plasma levels throughout the course of HCV therapy in HCV/HIV-coinfected patients: clinical implications
Abstracts of the Thirteenth Conference on Retroviruses and Opportunistic Infections, Denver, CO, 2006
 
Alexandria, VA, USA
Foundation for Retrovirology and Human Health
 
Abstract 581, p. 253
18
Cockroft
DW
Gault
MH
Prediction of creatinine clearance from serum creatinine
Nephron
 , 
1976
, vol. 
16
 (pg. 
31
-
41
)
19
D'Avolio
A
Ibañez
A
Sciandra
M
, et al.  . 
Validation of liquid/liquid extraction method coupled with HPLC-UV for measurement of ribavirin plasma levels in HCV-positive patients
J Chromatogr B Analyt Technol Biomed Life Sci
 , 
2006
, vol. 
835
 (pg. 
127
-
30
)
20
Jen
F
Glue
P
Gupta
S
, et al.  . 
Population pharmacokinetics and pharmacodynamic analysis of ribavirin in patients with chronic hepatitis C
Ther Drug Monit
 , 
2000
, vol. 
22
 (pg. 
555
-
65
)
21
Bruchfeld
A
Lindahl
K
Reichard
O
, et al.  . 
Pegylated interferon and ribavirin treatment for hepatitis C in haemodialysis patients
J Viral Hepat
 , 
2006
, vol. 
13
 (pg. 
316
-
21
)
22
Nomura
H
Tanimoto
H
Kajiwara
E
, et al.  . 
Factors contributing to ribavirin-induced anemia
J Gastroenterol Hepatol
 , 
2004
, vol. 
19
 (pg. 
1312
-
7
)
23
Takaki
S
Tsubota
A
Hosaka
T
, et al.  . 
Factors contributing to ribavirin dose reduction due to anemia during interferon α-2b and ribavirin combination therapy for chronic hepatitis C
J Gastroenterol
 , 
2004
, vol. 
39
 (pg. 
668
-
73
)
24
Van Vlierbergh
H
Delanghe
JR
De Vos
M
, et al.  . 
Factors influencing ribavirin-induced hemolysis
J Hepatol
 , 
2001
, vol. 
34
 (pg. 
911
-
6
)
25
Dumortier
J
Ducos
E
Scoazec
JY
, et al.  . 
Plasma ribavirin concentrations during treatment of recurrent hepatitis C with peginterferon α-2b and ribavirin combination after liver transplantation
J Viral Hepat
 , 
2006
, vol. 
13
 (pg. 
538
-
43
)
26
Gonzalez de Requena
D
Ibañez
A
Aguilar Marucco
D
, et al.  . 
Effect of ribavirin trough concentration on early virological response according to HCV genotype and on early hemoglobin decrease in HCV/HIV coinfected patients treated with ribavirin and peg-interferon
Abstracts of the Thirteenth Conference on Retroviruses and Opportunistic Infections, Denver, CO
 
Alexandria, VA, USA
Foundation for Retrovirology and Human Health
 
Abstract 857, p. 363
27
Dominguez
S
Cassard
B
Duvivier
C
, et al.  . 
Early therapeutic drug monitoring of ribavirin predicts anemia and early virological response in HIV/HCV co-infected patients
Abstracts of the Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, CA, 2007
 
Alexandria, VA, USA
Foundation for Retrovirology and Human Health
 
Abstract 893, p. 413