Health-related quality of life in genotype 1 treatment-naïve chronic hepatitis C patients receiving telaprevir combination treatment in the ADVANCE study

Authors


Correspondence to:

Dr M. Vera-Llonch, Vertex Pharmaceuticals Incorporated, 130 Waverly Street, Cambridge, MA 02139, USA.

E-mail: montserrat_vera-llonch@vrtx.com

Summary

Background

Chronic hepatitis C virus (HCV) infection and its treatment impact patients' health-related quality of life (HRQL).

Aim

To report on treatment impact and predictors of HRQL among treatment-naïve patients with genotype 1 chronic HCV infection who received 12-week telaprevir (T) with 24 (T12PR24) or 48 weeks (T12PR48) peginterferon alpha-2a/ribavirin (PR), or 48 weeks of PR in the ADVANCE study.

Methods

The EQ-5D-3L (EQ-5D) questionnaire (index range: 0–1) was completed at baseline and weeks 4, 12, 24, 36, 48 and 72. Patients indicated their health state on five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression. Descriptive statistics for the EQ-5D index and descriptive system and area under the curve from baseline to week 12 were calculated. Predictors of EQ-5D index were identified using multivariate analyses.

Results

Data from 722 patients were included. The mean EQ-5D index decreased during the first 12 weeks and returned to baseline by week 72 (T12PR24 by week 36) across treatments. In multivariate analysis, sustained virological response (SVR) at week 72 was associated (P < 0.0001) with improved EQ-5D index [mean; SVR+ (0.90), SVR− (0.86)], a 4% difference, within the published range of minimal clinically important difference.

Conclusions

Post hoc analyses of data from ADVANCE suggested that HRQL worsened during the first 12 weeks of therapy and returned to baseline by week 72 across treatments. Improvements were observed early following completion of a 24-week treatment (T12PR24). Telaprevir combination therapy was associated with slightly higher reductions in HRQL during the first 12 weeks (vs. PR). SVR was a statistically significant and meaningful predictor of HRQL at week 72.

Introduction

Chronic hepatitis C virus (HCV) infection affects approximately 130–170 million individuals worldwide with estimated disease prevalence of 5 million in the US.[1] Complications of chronic infection may include cirrhosis, liver decompensation, hepatocellular carcinoma and the need for liver transplantation.[2-7]

Chronic HCV infection has been reported to be associated with significant impacts on key dimensions of health-related quality of life (HRQL)[8-14] particularly in patients with advanced fibrosis and those co-infected with the human immunodeficiency virus and also in those without severe disease.[8, 9, 12, 15-17] Many symptoms of chronic HCV infection including fatigue, irritability, nausea, anorexia, muscle ache, headache and abdominal discomfort, have been shown to adversely impact patients' health status, psychological well-being and self-perceived health.[9, 18-22] Nonviral factors such as psychiatric comorbidity associated with acquisition of HCV, intravenous drug use, or the severity of liver disease cannot fully explain the impact on HRQL.[16] Unexplained mild, yet measurable cognitive deficits have also been reported,[1, 23, 24] suggesting that chronic HCV infection may be associated with the release of inflammatory cytokines and altered serotonergic and dopaminergic neurotransmission.[16, 25] Health-related quality of life has also been reported to be influenced by the patient's awareness of a serious illness.[16, 26, 27]

Antiviral therapy with interferon (IFN) and ribavirin (RBV) has well-documented side effects that compromise HRQL. Prior studies have suggested a potential advantage of treatment of HCV infection with pegylated IFN over standard IFN regarding vitality and pain.[28] The addition of RBV to the treatment regimen contributes to fatigue and decline in physical functioning.[29] Evaluations conducted in two large studies[8, 18] suggest a pattern of decline of HRQL throughout the course of HCV treatment with a return to initial levels or even an increase once treatment is stopped. Impairment in HRQL during HCV treatment potentially contributes to early treatment discontinuation;[30] the percentage of patients who discontinue therapy due to intolerance to IFN and RBV has been reported to range from 4% to 27%.[30-32]

In a systematic review of 32 studies,[33] sustained virological response (SVR) was associated with improvement in HRQL and reported to be maintained beyond the duration of follow-up in clinical trials.[9, 18, 34] In examination of the content of HRQL measures, such improvement has been translated into working without limitations and lower rates of discouragement, tiredness and worry about hepatitis. As an SVR is considered tantamount to virological cure, its associated HRQL benefits support the value of highly effective HCV therapies.

To our knowledge, no studies to date have assessed the impact of direct-acting-antiviral (DAA)-based triple combination treatment regimens on HRQL. Telaprevir is a DAA (an NS3/4A HCV protease inhibitor) that in combination with Peg-IFN alfa and RBV (PR) was approved in the United States, Canada, Europe and Japan in 2011, for treatment of adult genotype 1 chronic HCV-infected patients with compensated liver disease. In the ADVANCE study,[35] 79% of treatment-naive patients who received 12-week telaprevir (T) with either 24 (T12PR24) or 48 weeks (T12PR48) of PR achieved SVR compared with 46% of patients who had 48 weeks of PR alone.[36] In Phase 3 studies, the most common adverse events that were observed with greater than 5% incidence in T/PR vs. PR were rash, fatigue, pruritus, nausea, anaemia, diarrhoea, vomiting, haemorrhoids, anorectal discomfort, dysgeusia and anal pruritus.[36]

In this post hoc analysis, we report on treatment impact and predictors of HRQL among patients who participated in the ADVANCE study.

Materials and Methods

ADVANCE was a Phase 3 randomised (1:1:1), placebo-controlled, double-blind, multicentre study conducted in adult treatment-naïve patients with chronic genotype 1 HCV infection. The study was designed to compare 8-week vs. 12-week durations of T in combination with PR, with 48 weeks of PR alone[35] using a response-guided treatment paradigm. Among patients treated with T combination therapy, those who had undetectable HCV RNA at weeks 4 and 12 (extended rapid virological response) received a total of 24 weeks of treatment; those who did not, received a total of 48 weeks treatment. (Findings from the 8-week telaprevir combination treatment arm are not included here.)[36]

Patient-reported outcomes including the EQ-5D-3L (EQ-5D) (EuroQoL Group)[38] were tertiary (exploratory) end points in ADVANCE. The EQ-5D questionnaire was completed by patients at baseline and weeks 4, 12, 24, 36, 48 and 72; the questionnaire was completed before disclosure of HCV RNA results at the end of the week 24 visit and subsequent visits. Both the descriptive and the visual analogue scale (findings not reported here) were administered. In the EQ-5D descriptive system, patients were asked to indicate which statements best described their health state on the day of the assessment in each of five dimensions (mobility, self-care, usual activities, pain/discomfort, and anxiety and depression) with three possible levels [no problems (level 1), some problems (level 2) and extreme problems (level 3)]. The EQ-5D valuation (or utility) index (‘EQ-5D index’) can be calculated based on this information. A higher EQ-5D index indicates a better health state; the index ranges from 0 (equivalent to death) to 1 (full health). The EQ-5D questionnaire has been previously used in studies of HCV treatment.[13, 15]

Analyses

For the mean EQ-5D descripive system, the percentages of patients indicating alternative severity levels were tabulated by study visit for each dimension. EQ-5D health states defined by the system were converted into a single summary index by assigning US-specific valuation weights to each of the levels in each dimension.[39] If one or more dimensions of the descriptive system were left unanswered, the index was not calculated. Consistent with EuroQol Group recommendations, the results of the EQ-5D descriptive system were also reported as a health profile. Levels for each dimension were dichotomized into ‘no problems’ (i.e. level 1) and ‘problems’ (i.e. levels 2 and 3), thereby changing the profile into frequencies of reported problems.

Descriptive statistics of the EQ-5D index were calculated by study visit (95% CIs) and treatment group. Corresponding dimension level responses and changes from baseline were also calculated. Analyses were conducted using the observed data (no imputation methods were employed). Descriptive analyses were threefold and addressed HRQL burden: (i) at treatment initiation; (ii) during treatment; and (iii) at end of study follow-up and impact of SVR.

Multivariate regression analyses were employed to identify predictors of the EQ-5D index at selected study visits adjusting for covariates as appropriate. Covariates included patient age, gender, race, prevalence of advanced fibrosis, number of comorbidities, employment status, baseline EQ-5D index, presence of anaemia (as reported by the investigator) and rash, respectively, during the first 12 weeks of treatment, treatment assignment, treatment discontinuation during the first 12 weeks, number of adverse events and SVR. In addition, the area under the curve (AUC) was calculated from baseline to week 12 to assess the EQ-5D index impact throughout the telaprevir treatment phase.

Results

Seven hundred and twenty-two patients were included in the analyses. Mean (s.d.) patient age was 46.6 (10.4) years, 58.9% of patients were male, the mean (s.d.) body mass index was 26.8 (5.1); 20.2% of patients had evidence of bridging fibrosis or cirrhosis. No significant differences between treatment groups for demographic or clinical characteristics were observed (Table 1).

Table 1. Baseline demographic and clinical characteristics, by treatment group
CharacteristicAll patients (n = 722)Treatment group
T12PR24 (n = 210)T12PR48 (n = 153)PR (n = 359)P-valuea
  1. Data are mean (±s.d.) or n (%) as indicated.

  2. a

    Chi-square or one-way anova.

Age (years)46.6 (±10.4)45.9 (±10.4)47.3 (±11.2)48.8 (±10.0)0.387
Gender (male)425 (59%)128 (61%)86 (56%)211 (59%)0.662
Race (white)641 (89%)191 (91%)134 (88%)316 (88%)0.087
Body mass index (kg/m2)26.8 (±5.1)26.4 (±4.7)26.3 (±5.5)27.2 (±5.1)0.074
Viral load (million IU/mL)4.7 (±5.9)4.3 (±5.0)4.7 (±4.8)5.0 (±6.8)0.394
Bridging fibrosis/cirrhosis146 (20%)38 (18%)35 (23%)73 (20%)0.533
Number of comorbidities4.1 (± 4.2)3.9 (±4.2)4.4 (±4.1)4.0 (±4.2)0.550
Unemployed157 (22%)49 (23%)29 (19%)79 (22%)0.599

Burden of Chronic Hepatitis C Infection in HRQL

At treatment initiation, the age- and gender-adjusted mean (95% CI) EQ-5D index (pooled across treatments) was 0.91 (0.90, 0.92), which is higher than the published US population norm [mean (S.E.) index = 0.87 (0.01)] for individuals aged 45–54.[40] The EQ-5D index value was represented evenly in each treatment group [T12PR24: 0.92 (0.91, 0.94); T12PR48: 0.90 (0.88, 0.93); and PR: 0.91 (0.90, 0.92)]. The percentages of patients reporting problems in each of the EQ-5D dimensions at baseline were 25.7% (pain/discomfort), 25.6% (anxiety/depression), 12.9% (usual activities), 8.2% (mobility) and 2.0% (self-care).

Burden of HCV Treatment in HRQL

The mean EQ-5D index decreased during the first 12 weeks following treatment initiation and returned to baseline levels by week 72 across all treatment groups with more rapid improvement observed among patients receiving the shorter duration of telaprevir combination therapy (T12PR24 by week 36) (Figure 1) (Table 2). Mean (95% CI) EQ-5D index during the first 12 weeks of therapy (weeks 4 and 12) was 0.80 (0.78, 0.82) (pooled T12PR) and 0.83 (0.81, 0.84) (PR). The adjusted mean (95% CI) AUC for the EQ-5D index from baseline to week 12 was 6.0 (5.3, 6.6) for T12PR and 6.6 (5.9, 7.2) for PR. At week 24, differences between treatment groups were not statistically significant. At week 48, values were 0.93 (0.92, 0.95) (T12PR24), 0.83 (0.79, 0.86) (T12PR48) and 0.84 (0.82, 0.86) (PR).

Table 2. EQ-5D Index by treatment group and study visit
VisitTreatment group
T12PR24T12PR48PR
n Mean (95% CI) n Mean (95% CI) n Mean (95% CI)
Baseline2030.92 (0.91, 0.94)1440.90 (0.88, 0.93)3460.91 (0.90, 0.92)
Week 41980.82 (0.80, 0.84)1220.81 (0.79, 0.84)3340.84 (0.82, 0.85)
Week 121940.80 (0.77, 0.82)1150.78 (0.75, 0.82)3280.82 (0.81, 0.84)
Week 241890.83 (0.81, 0.85)1090.81 (0.78, 0.85)3130.80 (0.78, 0.82)
Week 361790.91 (0.89, 0.93)1120.85 (0.82, 0.88)2960.83 (0.81, 0.85)
Week 481750.93 (0.92,0.95)1140.83 (0.79, 0.86)2780.84 (0.82, 0.86)
Week 721710.94 (0.92, 0.95)1160.90 (0.88, 0.93)2960.89 (0.88, 0.91)
Figure 1.

EQ-5D Index over time by treatment group.

The percentage of patients reporting any problems at week 4 were: pain/discomfort 61% (T12PR), 56% (PR), usual activities 51% (T12PR), 46% (PR), anxiety/depression 38% (T12PR), 38% (PR) and mobility 26% (T12PR), 20% (PR). Corresponding estimates at week 12 were: pain/discomfort 60% (T12PR), 63% (PR), usual activities 56% (T12PR), 50% (PR), anxiety/depression 51% (T12PR), 42% (PR) and mobility 29% (T12PR), 25% (PR). Fewer than 10% of patients reported problems with self-care. The percentage change from baseline by dimension and study visit is shown in Figures S1–S5.

HRQL at End-of-Study Follow-up and Benefit of SVR

The EQ-5D index [mean (95% CI)] returned to baseline levels by week 72 [0.90 (0.88, 0.93) (T12PR48), 0.89 (0.88, 0.91) (PR)], and was higher than baseline in the T12PR24 group [0.94 (0.92, 0.95)]. The EQ-5D index was available at week 72 for 382 patients who achieved SVR and for 201 patients who did not achieve SVR. The age- and gender-adjusted estimate [mean (95% CI)] was higher among patients who achieved SVR (across all treatment groups) compared with those who did not: 0.90 (0.88, 0.92) (SVR+) and 0.86 (0.83, 0.88) (SVR−), or a 0.04 difference, which is within the published range of the minimal clinically important difference (MCID) for the EQ-5D index.[41] The change from baseline in the percentages of patients reporting problems at week 72 by dimension and SVR is shown in Figure 2 with improvements observed across most dimensions among patients who achieved SVR vs. those who did not.

Figure 2.

Percentage of patients reporting problems for each EQ-5D dimension by SVR (week 72).

Predictors of EQ-5D Index at Selected Study Visits

Findings from multivariate regression analyses are reported in Table 3. At treatment initiation, being unemployed (P < 0.001), the number of comorbidities (P < 0.001) and the presence of bridging fibrosis or cirrhosis (P = 0.02) were adversely associated with HRQL. Area under the curve analyses suggested that for the first 12 weeks of therapy, a low baseline EQ-5D index value (P < 0.0001), older age (P = 0.01), receipt of T12PR (P = 0.01) and treatment discontinuation for any reason (P < 0.0001) were adversely related to HRQL. At week 12, anaemia following treatment initiation (regardless of treatment group) was a negative predictor of the EQ-5D index (P = 0.04). At week 48, a high baseline EQ-5D index and receipt of T12PR24 (P < 0.001) were positively associated with increased HRQL, while the number of adverse events during treatment had a negative impact (P < 0.001). At week 72, and adjusting for EQ-5D index at baseline, age, gender, black/nonblack race, advanced fibrosis, the number of comorbidities at baseline and the number of adverse events during treatment, SVR was associated (P < 0.001) with higher EQ-5D index values.

Table 3. Predictors of EQ-5D Index at selected study visits
VariableVisit
BaselineWeek 12Week 48Week 72
β (±S.E.)P-valueβ (±S.E.)P-valueβ (±S.E.)P-valueβ (±S.E.)P-value
  1. a

    Coefficients listed as 0.000 are rounded for presentation purposes and are actually either slightly greater than or less than zero as indicated by the preceding directional sign (+/−). Actual values for β coefficient (±S.E.) are as follows – baseline: age 0.0001(±0.0005); week 48: age 0.0002 (±0.0006); week 72: age −0.00003 (±0.00046).

Intercept0.944 (±0.022)<0.0010.420 (±0.055)<0.0010.533 (±0.055)<0.0010.479 (±0.045)<0.001
Age0.000 (±0.001)a0.765−0.001 (±0.001)0.2800.000 (±0.001)a0.793−0.000 (±0.000)a0.941
Gender (Male)0.014 (±0.009)0.1420.019 (±0.012)0.115−0.005 (±0.012)0.678−0.011 (±0.009)0.255
Race (Black)0.001 (±0.018)0.9790.006 (±0.022)0.788−0.017 (±0.023)0.467−0.061 (±0.019)0.001
Bridging fibrosis/cirrhosis−0.027 (±0.012)0.0200.008 (±0.015)0.576−0.004 (±0.015)0.777−0.017 (±0.012)0.160
No. of comorbidities−0.007 (±0.001)<0.001−0.002 (±0.001)0.127−0.001 (±0.002)0.526−0.003 (±0.001)0.024
Unemployed at baseline−0.070 (±0.011)<0.001
EQ-5D Index at baseline0.483 (±0.051)<0.0010.413 (±0.049)<0.0010.501 (±0.039)<0.001
T12PR−0.014 (±0.012)0.257
Anaemia during first 12 weeks−0.029 (±0.014)0.039
Rash during first 12 weeks−0.022 (±0.012)0.072
Discontinued in first 12 weeks−0.042 (±0.044)0.336
No. of adverse events on treatment−0.005 (±0.001)<0.001−0.002 (±0.001)0.001
T12PR240.083 (±0.013)<0.001
SVR0.041 (±0.010)<0.001

Discussion

To the best of our knowledge, this is the first report of findings from a large clinical trial on the estimated HRQL impact of DAA combination therapy in adult genotype 1 treatment-naïve chronic hepatitis C patients. Among patients in the ADVANCE study, HRQL decreased during the first 12 weeks of treatment and returned to baseline levels by week 72 across all treatment groups. Telaprevir combination treatment was associated with higher transient reductions in HRQL during the first 12 weeks of therapy than treatment with PR. For patients completing a total duration of 24-weeks of telaprevir combination treatment, improvements in HRQL were observed earlier (by week 36), and were also higher among these patients at week 72 than for those receiving T12PR48 and PR48 respectively. SVR was a statistically significant and meaningful predictor of improved HRQL at week 72.

Our study identified predictors of HRQL before, during and after HCV treatment. At treatment initiation, the prevalence of physical and mental comorbidity (approximately one-fourth of patients reported having problems with anxiety or depression and pain or discomfort respectively), the presence of bridging fibrosis or cirrhosis, and unemployment were independent predictors of HRQL impairment. Presence of these factors in genotype 1 chronic HCV-infected patients who are being considered for treatment may inform caregivers of potential treatment challenges.

The EQ-5D index at baseline among study patients was somewhat higher than published US population norms [mean (S.E.): EQ-5D index = 0.87 (0.01)] for individuals 45–54 years of age[40] and also higher than other published estimates of health state utilities.[13, 15, 19, 28, 43-46] Observed differences in the EQ-5D index may be attributable to differences in sociodemographic characteristics between populations, possibly less comorbidity among patients in the ADVANCE study, other unmeasured characteristics of patients enrolled in the study (vs. those of patients in typical clinical practice), and the fact that US population norms maybe slightly dated (derived in 2005–2006).

Both telaprevir combination treatment and PR alone adversely impacted HRQL during the first 12 weeks of therapy, most likely due to adverse events; receipt of telaprevir combination treatment (vs. PR) was associated with greater reductions during this period. At the week 12 assessment, the baseline EQ-5D index value and experience of anaemia since treatment initiation (regardless of treatment group) were statistically significant predictors of HRQL. Close monitoring of haemoglobin and management of anaemia early during treatment by means of RBV dose reductions should be encouraged to potentially minimise treatment burden in particular during the first 12 weeks of combination therapy. In pooled retrospective analyses of clinical trials data, the initial RBV dose was reduced in 68% of patients during the overall treatment phase with 90% of these patients reaching a minimum dose of 600 mg/day or less.[42] In these analyses, no substantial differences in the proportion of patients who achieved SVR were observed between patients who reduced the dose of RBV to 600 mg/day and those who did not. In addition, the prevalence of self-reported symptoms of anxiety and depression increased early during treatment compared with baseline (25%) in both treatment arms. The percentage of patients reporting such symptoms was 38% at week 4 (both treatments) and slightly higher among telaprevir-treated patients at week 12 (51% T12PR vs. 42% PR).

The relationship between treatment discontinuation and worse HRQL during the first 12 weeks of therapy was also significant in multivariate AUC analyses, suggesting that quality of life during treatment might have impacted medication compliance due to adverse effects.

Among patients who were eligible for shorter duration of telaprevir-based therapy due to early on-treatment virological response, HRQL improvements were observed after treatment completion (week 24) with values reaching baseline levels on average by week 36. These findings support the value of response-guided therapy[47] in potentially reducing the overall burden (magnitude and duration) of HCV treatment in eligible patients. By week 72 and across all treatments, HRQL returned to baseline levels consistent with published data on IFN-based regimens;[8, 18, 29, 47] slightly greater improvements were observed in the shorter telaprevir-based regimen. The extent to which HRQL improvements observed at week 72 among those patients receiving the shorter telaprevir treatment regimen and those who experienced SVR (vs. those who did not experience SVR) may be related to a longer treatment-free period (compared with 48-week treatment regimens), unknown lasting effects of HCV treatment, achievement of SVR, and/or reduced concern about infection in those who responded to treatment cannot be addressed with the data.

Our findings are also supportive of the previously reported clinical significance of viral clearance following treatment of chronic HCV infection.[16, 23] SVR, which has been linked to disease progression and survival outcomes, is also considered a relevant clinical anchor for HRQL.[33] Normalisation of the chronic low-level (cytokine-mediated) inflammatory state and reductions in metabolites and basal ganglia ratios have been suggested as reasons for improvements in cerebral immune activation and cognitive function among patients who clear the virus.[16, 23] Associated improvements in verbal learning, memory and visuospatial memory have not been reported in nonresponders.[16, 23] In our analysis, the positive impact of SVR at week 72 was prominent in the usual activities, pain/discomfort, mobility and anxiety and depression dimensions of the EQ-5D. The average utility gain of approximately 4% among patients who achieved SVR in our study is consistent with other published estimates[15] and is within the range of the MCID reported for the EQ-5D index.[41] This gain of approximately 4% also approximates the MCID for the SF-36 vitality scale, reported to be most relevant in patients with HCV infection.[33]

Several limitations should be noted in our analysis. First, the EQ-5D was an exploratory measure of treatment impact in the ADVANCE study; thus, the analyses reported here are post hoc in nature. Comparative studies that employ prespecified hypotheses on the impact of therapy and SVR respectively would likely be more robust. Second, treatment discontinuation during the first 12 weeks was associated with lower HRQL. The extent to which HRQL assessments later in the study among such patients would have further impacted the findings cannot be ascertained. Third, anaemia, as reported by the investigators, was examined as a predictor; we did not use laboratory findings in this analysis. Fourth, in light of the available published data, the patient population participating in ADVANCE was likely not representative of treatment-naïve patients in typical clinical settings who may be sicker and at risk of worse HRQL outcomes during treatment. Fifth, since patients were informed of their HCV RNA status starting at week 24, knowledge of HCV RNA values might have influenced their ratings in subsequent visits. Sixth, the degree of missing data ranged from 4% to 25% across all study visits and treatment groups, with a slightly higher prevalence (additional 5%) in the telaprevir treatment arms. Finally, sites from multiple countries participated in the ADVANCE study (approximately 60% of patients were from North America), yet US-specific valuation weights were applied to patients' response levels to derive the EQ-5D index; the extent to which use of country-specific weights could have differentially affected our estimates was not examined.

In conclusion, these results from the ADVANCE study suggest that HRQL decreased during the first 12 weeks of therapy and returned to baseline by week 72 across all treatment groups in adult treatment naïve patients with genotype 1 chronic HCV infection. Compared with PR treatment alone, greater transient reductions should be expected during the first 12 weeks among patients receiving telaprevir combination therapy. Improvements in HRQL are likely to be observed early following completion of a total duration of 24-weeks of telaprevir-based treatment among eligible patients. In this analysis, SVR was a statistically significant and meaningful predictor of HRQL. Our results are consistent with published data on the HRQL impacts of IFN-based regimens in this patient population and supportive of the value of shorter treatment duration (when clinically appropriate) and highly effective HCV therapies from a patient-reported outcomes perspective.

Authorship

Guarantor of the article: Zobair Younossi, MD.

Author contributions: MVL was involved in the data analysis plan, data analysis and synthesis, manuscript development and critical revisions to the manuscript. MM was involved in data analysis and synthesis, manuscript development and critical revisions to the manuscript. JA was involved in the data analysis plan, data analysis and synthesis, manuscript development and critical revisions to the manuscript. MD was involved in data analysis and synthesis, manuscript development and critical revisions to the manuscript. MB was involved in manuscript development and critical revisions to the manuscript. TG was involved in the data analysis plan, data analysis and synthesis, manuscript development and critical revisions to the manuscript. ZY provided background to current treatment experience for patients with chronic hepatitis C and was involved in data analysis and synthesis, manuscript development and critical revisions to the manuscript. All authors approved the final version of the manuscript.

Acknowledgments

The authors thank Leif Bengtsson for insightful discussion in early study planning; Brenda Castiglione, PharmD, Camilla S. Graham MD, MPH for critical review of the manuscript; and Rebecca Jones for medical writing assistance.

Declaration of personal interests: LB, BC and RJ are employees of Vertex Pharmaceuticals Incorporated and may or may not own stock options in that Company. CSG is a former employee of Vertex Pharmaceuticals Incorporated and may own or may have owned stock or options in that Company at the time this research was performed. MVL is an employee of Vertex Pharmaceuticals Incorporated and may or may not own stock options. MD and MM are former employees of Vertex Pharmaceuticals Incorporated and may own or may have owned stock or options in that Company at the time this research was performed. ZY has been a Consultant for Gilead, Enterome and Coneatus and has been on Advisory Committees or Review Panels for Vertex, Salix, GlaxoSmithKline, Merck and Tibotec. MB has received consulting fees from Vertex Pharmaceuticals Incorporated. JA and GT are employees of Boston Healthcare Associates, Inc.

Declaration of funding interests: This study was sponsored by Vertex Pharmaceuticals Incorporated.

Ancillary