4 Evidence and interpretation

The appraisal committee (appendix A) considered evidence from a number of sources (appendix B).

4.1 Clinical effectiveness

4.1.1

The assessment group performed a systematic review to identify randomised controlled trials (RCTs) evaluating the clinical effectiveness of peginterferon alfa plus ribavirin for the treatment of chronic hepatitis C in three specific groups:

  • people eligible for shortened treatment courses

  • people eligible for re-treatment following previous non-response or relapse of their condition

  • people co-infected with HCV and HIV.

    Included in the review were studies in which a standard duration of combination therapy with peginterferon alfa plus ribavirin (or peginterferon monotherapy for people with contraindications to ribavirin) was compared with courses of combination therapy of a shortened duration (24 weeks for people with HCV genotype 1 treated with either peginterferon alfa-2a or -2b; 16 weeks for people with HCV genotype 2 or 3 treated with peginterferon alfa-2a only). The review also attempted to identify studies in which treatment was compared with best supportive care for subgroups that were co-infected with HIV or were being re-treated. Clinical outcomes included measures of virological response during and after treatment, and adverse effects.

4.1.2

Six RCTs reported in 8 publications were identified, all of which reported on peginterferon alfa and ribavirin combination therapy in people eligible for shortened courses of treatment. No studies were found that compared treatment with peginterferon alfa (with or without ribavirin) with best supportive care for people co-infected with HIV and HCV or for people whose hepatitis C had not responded to previous treatment or had responded but subsequently relapsed.

Shortened treatment courses

4.1.3

Of the 6 trials, 4 evaluated the effect of shortened treatment courses with peginterferon alfa plus ribavirin for people with HCV genotype 1 infections, while the other 2 trials evaluated the effect for people with genotype 2 or 3 infections. No trials assessed the effect of shortened treatment courses for people with genotype 4 infections. Five of the trials included people with a low viral load at the start of treatment. The comparator in all included studies was the same intervention for a standard duration of treatment. The dose of peginterferon alfa-2a was 180 micrograms per week and the dose of peginterferon alfa-2b was 1.5 micrograms per kg per week in all trials. All 6 RCTs reported sustained virological response rates as the primary outcome measure. A sustained virological response was defined as undetectable serum HCV RNA at the end of 24 weeks' follow-up in 4 trials, and as undetectable serum HCV RNA at the end of treatment and at the end of follow-up in the other 2 trials. In many of the trials the sustained virological response rates were presented for subgroups of people who had been randomised and who achieved a rapid virological response. It was not reported whether these subgroups were powered to detect a statistically significant difference between trial arms. A rapid virological response was defined as an undetectable serum concentration of HCV RNA (25 IU/ml or less) in 1 trial, serum HCV RNA negative (50 IU/ml or less) in 3 trials, serum HCV RNA of 600 IU/ml or less in 1 trial, and serum HCV RNA of 650 IU/ml or less in 1 trial, all at week 4 of treatment.

4.1.4

In people with a low viral load (800,000 IU/ml or less) who attained a rapid virological response up to and including week 4 of treatment, sustained virological response rates were comparable between the group that received 48 weeks (standard duration) of treatment (range 83% to 100%) and the group that received shortened treatment courses (range 84% to 96%), with no statistically significant differences between the groups. Regardless of HCV genotype, sustained virological response rates were similar across studies, with the exception of 1 trial that reported lower rates.

4.1.5

In 2 of the trials of peginterferon alfa-2a plus ribavirin, the ribavirin dose was varied according to body weight, which is no longer consistent with its marketing authorisation. The marketing authorisation specifies that ribavirin should be given as a fixed dose of 800 milligrams in people with HCV genotype 2 or 3, and both trials assessed the effectiveness of treatment only in people with HCV genotype 2 or 3 infections. The assessment group noted that if these 2 trials had been excluded on the basis of the variable ribavirin dosages, there would have been no evidence on the impact of shortened treatment durations in people with HCV genotype 2 or 3 infections.

4.1.6

Rapid virological response rates were comparable between the groups that received the standard duration of treatment and those that received shortened courses. However, there was a large range in reported rapid virological response rates between the studies, with rates in people with HCV genotype 2 or 3 infections generally being higher than those in people with genotype 1 infections.

4.1.7

The relapse rate in a subgroup of people with a low viral load and a rapid virological response was reported in 1 trial. Relapse was defined as the re-appearance of serum HCV RNA during the 24‑week follow-up period in people whose condition initially responded to treatment. The rates of relapse were low and were not statistically significantly different between the treatment arms (3.6% for 24 weeks of treatment versus 0% for 48 weeks of treatment; p=1.00). In contrast, in people with a high viral load at start of treatment and a rapid virological response later, shortening the duration of treatment was associated with a statistically significant higher rate of relapse (23.5% for 24 weeks versus 0% for 48 weeks; p=0.045).

4.1.8

Adverse events were presented for treatment groups as a whole, and not for the subgroup of people with a low viral load who had a rapid virological response. Overall, the frequency of adverse events did not differ statistically between treatment arms, although a lower incidence of adverse events was reported in 3 trials in people treated for a shorter duration with combination therapy. The most frequently occurring adverse events included influenza-like symptoms, insomnia, anorexia, dermatological symptoms and alopecia.

4.1.9

The cumulative incidence of serious adverse events ranged from 0% to 7% over the duration of the trials, and was not considered to be substantially different between treatment arms, although levels of statistical significance were lacking in most studies. One death was reported, which was a result of re-activation of pulmonary tuberculosis at week 36 of treatment in a person who was treated with peginterferon alfa-2a plus ribavirin. One trial reported that people receiving a shortened treatment regimen were less likely to discontinue their treatment than people receiving longer courses (3% versus 10%; p=0.045).

Re-treatment after non-response or relapse

4.1.10

The assessment group did not identify any RCTs that compared treatment with peginterferon alfa (with or without ribavirin) with best supportive care for people whose hepatitis C had not responded to previous treatment or had responded but subsequently relapsed. The assessment group did identify a number of RCTs comparing treatment with peginterferon alfa (with or without ribavirin) with active treatment comparators (such as non-peginterferon alfa plus ribavirin), but these did not meet the inclusion criteria for the review (based on the decision problem) because they featured an active treatment comparator. However, the assessment group assumed in its economic model (see section 4.2) that the sustained virological response rates for people whose condition did not respond or responded but subsequently relapsed on peginterferon alfa therapy were those reported in the clinical trials that used an active comparator. The assessment group noted that these reported sustained virological response rates may not be representative of people with relapsed disease because the trial participants with HCV genotype 1 infections had less intensive initial treatment than what would be regarded as standard treatment.

HCV and HIV co-infection

4.1.11

The assessment group did not identify any RCTs that compared treatment with peginterferon alfa (with or without ribavirin) with best supportive care for people with HCV and HIV co-infection. A number of RCTs were identified that compared peginterferon alfa (with or without ribavirin) with active treatment comparators (such as non-peginterferon alfa plus ribavirin) for this subgroup, but these did not meet the inclusion criteria for the review because they featured an active treatment comparator. The assessment group noted that it is unlikely that any studies, whether randomised or not, would be available that include a non-active treatment arm, because withholding treatment would be unlikely to be considered ethical.

Summary of clinical-effectiveness evidence

4.1.12

The collective evidence for combination therapy for both peginterferon alfa-2a and peginterferon alfa-2b suggests that there are no statistically significant differences in sustained virological response rates between people receiving shortened durations of treatment with peginterferon alfa plus ribavirin of 16 weeks (HCV genotype 2 or 3; peginterferon alfa-2a) or 24 weeks (HCV genotype 1; both peginterferon alfa-2a and alfa-2b) and people receiving the standard duration of treatment. Rates of relapse following a shortened course of treatment were reported to be low and not statistically significantly different from relapse rates after a standard course of treatment in people with a rapid virological response and low viral load. For people with a rapid virological response and a high viral load, shortening the duration of treatment may be associated with a higher rate of relapse. However, the assessment group noted that analyses of sustained virological responses in people with a low viral load at the start of treatment and a rapid virological response are likely to be underpowered because they were based on subgroups of the randomised participants, and therefore the results should be interpreted with caution. No RCTs were identified that compared peginterferon alfa (with or without ribavirin) with best supportive care for people who were re-treated after non-response or relapse of their condition, or for people with HCV and HIV co-infection.

4.2 Cost effectiveness

4.2.1

The assessment group identified 2 studies examining the cost effectiveness of the treatment of people co-infected with HCV and HIV. The studies compared peginterferon alfa plus ribavirin with non-peginterferon plus ribavirin, peginterferon alfa monotherapy and no treatment (supportive care). One study also had an additional non-peginterferon alfa monotherapy arm. No economic evaluations were identified for people requiring re-treatment following previous non-response or relapse, or for people eligible for shortened courses of treatment.

4.2.2

In both studies of people co-infected with HCV and HIV, peginterferon alfa monotherapy or peginterferon alfa plus ribavirin dominated (that is, treatment was more effective and less costly) the other strategies in people with HCV genotype 1. In 1 study, peginterferon alfa monotherapy was the most favourable option in each scenario. For people with other HCV genotypes (that is, not genotype 1), peginterferon alfa plus ribavirin was the least favourable option in one study (incremental cost-effectiveness ratios [ICERs] of $300,800 to $4,000,000 per quality-adjusted life year [QALY] gained). In the second study peginterferon alfa plus ribavirin dominated the other strategies for all genotypes reported. The incremental costs per life year saved (LYS) for peginterferon alfa plus ribavirin in people with HCV genotypes other than genotype 1 were $39,300/LYS in women and $39,700/LYS in men and $70,000/LYS in women and $73,000/LYS in men in people with genotype 1).

Manufacturers' models

Roche Products
4.2.3

A health state transition model submitted by Roche Products was used to assess the cost effectiveness of treatment with peginterferon alfa-2a plus ribavirin in 3 groups: people who had been treated previously with peginterferon alfa plus ribavirin, including those whose hepatitis C did not respond to this previous treatment (by HCV genotype) and those whose condition responded but then relapsed; people with a low viral load and a rapid virological response who received shortened courses of peginterferon alfa plus ribavirin (by genotype); and people co-infected with HCV and HIV.

4.2.4

The model used clinical-effectiveness data from published RCTs, although evidence of effectiveness for shortened treatment courses was derived from subgroup analyses. For people receiving shortened treatment courses, the sustained virological response rates for both groups in the model were taken from unpublished analyses of subgroups of clinical trial participants. For the subgroups with HCV genotypes 2 and 3 the data were taken from the ACCELERATE study reported by Shiffman and colleagues, and for the genotype 1 and 4 subgroups the data were taken from a trial reported by Hadziyannis and colleagues (study NV15942).

4.2.5

A number of the clinical-effectiveness studies used by the manufacturer to obtain data for the economic model (namely for people who had been re-treated or had HIV co-infection) had active comparators, rather than comparing treatment with best supportive care (as outlined in the decision problem). For people with HCV and HIV co-infection, data on sustained virological response rates were taken from a clinical trial comparing non-peginterferon alfa-2a with peginterferon alfa-2a, which is not consistent with the scope for this appraisal. For people whose hepatitis C did not respond or relapsed on previous peginterferon therapy, data on sustained virological response rates were taken from 2 clinical trials. In the majority of situations it was assumed that people receiving best supportive care would not achieve a spontaneous sustained virological response. No discussion or critical analysis of the reliability or generalisability of the clinical-effectiveness evidence used to populate the model was provided by the manufacturer.

4.2.6

Shortening the duration of treatment was associated with a QALY loss compared with the standard treatment duration (because of a slight reduction in the sustained virological response rate), as well as with a reduction in costs. Since both costs and outcomes were lower with shortened treatment duration, the ICERs represent savings per QALY lost. For people with HCV genotype 1 or 4 the ICER was £15,472 per QALY, and for people with genotype 2 or 3 the ICER was £2,719 per QALY. The assessment group noted that when there are both lower costs and lower effectiveness, the analysis can be better understood by applying the net benefits framework. Using the net benefits framework, a treatment would be accepted when the value of the incremental benefits exceeds the incremental costs, resulting in a positive incremental net benefit. Using this framework, the incremental net monetary benefit for a shortened duration of treatment is positive for people with HCV genotype 1 or 4 over a wide range of willingness-to-pay values (below the ICER value of £15,472 per QALY). For people with HCV genotype 2 or 3 the incremental net monetary benefit is positive only at comparatively low threshold values (below the ICER value of £2,719 per QALY).

4.2.7

Re-treating people whose hepatitis C relapsed following previous peginterferon treatment was reported by the manufacturer as dominating best supportive care. This is the result of including a high sustained virological response rate observed in 1 trial that may not be generalisable to other populations of people whose condition relapses. The majority of people had HCV genotype 1 infections and their previous treatment was of shorter duration than the current standard of care (that is, 24 weeks rather than 48 weeks). The sustained virological response rates applied in the model for the re-treatment of people whose hepatitis C had not responded to previous peginterferon treatment were lower than those for people whose condition had relapsed following treatment. Treatment was associated with QALY gains and increased costs compared with best supportive care, resulting in ICERs of £3,334 per QALY gained for people with HCV genotype 1 and £809 per QALY gained for people with other genotypes. The majority of people recruited to the trial whose condition did not respond to previous peginterferon treatment had HCV genotype 1; only 29 were infected with other HCV genotypes (9% of the arm used to estimate the effectiveness of treatment in the model), and the majority (66%) of these had genotype 4 infections.

4.2.8

For people with HCV and HIV co-infection, treatment with peginterferon plus ribavirin was estimated to dominate treatment with non-peginterferon. However, this is not the comparison specified in the decision problem issued by NICE (where best supportive care was stated as the comparator). The assessment group extended the analysis by assuming that the sustained virological response rate for untreated people would be zero, and estimated a QALY gain (using the manufacturer's model) of 1.95 and an incremental cost of £1,765 for peginterferon plus ribavirin compared with best supportive care, resulting in an ICER of £903 per QALY gained.

4.2.9

The assessment group thought that the cost-effectiveness results were generally robust to variations in the limited number of parameters included in a deterministic sensitivity analysis reported in the manufacturer's submission. Probabilistic sensitivity analyses were also conducted by the manufacturer, and included the majority of parameters in the model. Although appropriate probability distributions appear to have been used for the probabilistic sensitivity analyses, the assessment group noted that limiting the distributions for some inputs does not appear to make the best use of data reported in the submission. Moreover, the assessment group felt that the manufacturer's model did not adequately consider logical relationships and potential correlations between model inputs. Rather than reporting the probability of cost effectiveness at certain willingness-to-pay thresholds, the submission identified a maximum threshold of £15,000 for all analyses. Further analyses of the manufacturer's model undertaken by the assessment group generally resulted in less favourable ICERs, but did not substantially alter the conclusions from those in the manufacturer's submission.

Schering-Plough
4.2.10

The manufacturer presented cost-effectiveness results for three populations: people who had been treated previously with peginterferon and whose hepatitis C either did not respond to treatment or relapsed after treatment (broken down further by broad HCV genotype categories: genotypes 1 and 4 combined, and genotypes 2 and 3 combined) (data obtained from the EPIC3 clinical study report); and people co-infected with HCV and HIV (using effectiveness data from a trial reported by Laguno and colleagues). The manufacturer subsequently presented an analysis of the cost effectiveness of shortened compared with standard treatment durations for people with HCV genotype 1.

4.2.11

Model-based economic evaluations were based on clinical data from the EPIC3 study (a multicentre, non-randomised open-label uncontrolled study of re-treatment in people whose hepatitis C did not respond or relapsed after treatment) and from the study of Laguno and colleagues (a phase 3 open-label trial that recruited treatment-naïve [naïve to combination therapy] people with histologically verified liver disease, who were HIV-positive with controlled disease [for people with HCV and HIV co-infection]). As the included studies do not make the comparisons specified in the decision problem (that is, anti-viral treatment compared with best supportive care), the manufacturer assumed that people receiving best supportive care would not achieve a spontaneous sustained virological response. The model includes a low probability of a spontaneous sustained virological response for people with mild chronic hepatitis C, which is applied to the best supportive care and active treatment cohorts.

4.2.12

The manufacturer's model is structurally similar to that used in a previous assessment report for NICE technology appraisal guidance 106. However, it does not distinguish between people achieving a sustained virological response from any of the treatment-eligible states (mild or moderate hepatitis C and compensated cirrhosis). The parameters reflecting natural history in the model are similar to those adopted in the previous assessment report, as are the health state utilities and health state costs (inflated from 2003/4 to 2007/8 costs using the HCHS Pay and Prices Index).

4.2.13

Re-treating people whose hepatitis C did not respond or relapsed following previous interferon treatment was estimated to result in a QALY gain of 1.03 compared with best supportive care at an incremental cost of £4,536, resulting in an ICER of £4,387 per QALY gained. These results were reported for a combined cohort of people with HCV genotypes 1 and 4 (84% of total) and people with genotypes 2 and 3. Separate results were also reported for the 2 genotype subgroups: the ICERs were £7,177 per QALY gained for people with genotypes 1 and 4, and £783 per QALY gained for people with genotypes 2 and 3. Separate subgroup analyses (not stratified by genotype) for people whose condition did not respond to treatment and for people whose condition relapsed following treatment were also presented which suggested that the QALY gain was higher for the non-response subgroup than for the relapse subgroup.

4.2.14

For a cohort of people (all HCV genotypes) co-infected with HCV and HIV, treatment with peginterferon plus ribavirin was estimated to result in a gain of 2.32 QALYs compared with best supportive care at an incremental cost of £2,502, resulting in an ICER of £1,077 per QALY gained. For people with genotypes 1 and 4 the ICER was estimated at £1,637 per QALY gained, while for people with genotypes 2 and 3 the ICER was £403 per QALY gained.

4.2.15

The manufacturer also conducted an analysis of the subgroup of people eligible for a shortened treatment duration. The analysis demonstrated that a shortened treatment duration of 24 weeks dominated the standard treatment duration of 48 weeks for people with HCV genotype 1. The manufacturer noted that these results should be treated with caution, as there was a large numerical difference in the rapid virological response rates at week 4 between the patient groups (standard versus shortened treatment) in the clinical trial, despite the fact that they had received the same treatment up to this point.

4.2.16

Deterministic sensitivity analyses provided by the manufacturer showed that the ICERs for both the re-treated and co-infected groups were sensitive to variations in the early virological response rate and the sustained virological response rate, as well as to changes in body weight (since the dosing regimens of both peginterferon alfa-2b and ribavirin were weight-based). In the re-treatment group the ICERs showed a small increase in response to changes in the distribution of disease severity within the group.

4.2.17

Probabilistic sensitivity analyses were conducted by the manufacturer which included the majority of parameters in the model. The manufacturer performed probabilistic sensitivity analyses (for overall cohort and by genotype) for each group (re-treated and co-infected). The probabilistic sensitivity analysis indicated a high probability (over 90%) that treatment with peginterferon alfa-2b plus ribavirin is cost effective for all analyses at willingness-to-pay thresholds of both £20,000 and £30,000 per QALY gained.

4.2.18

The assessment group was concerned about the manufacturer's assumption that a person would achieve a sustained virological response immediately after treatment is given, and therefore accrue health benefits before entering the model. Furthermore, the assessment group did not consider that it was appropriate to apply the same utility values for a sustained virological response irrespective of the stage of disease when the treatment was given. Further analyses of the manufacturer's model undertaken by the assessment group generally resulted in less favourable ICERs, but did not substantially alter the conclusions from those in the manufacturer's submission.

Assessment group's model

4.2.19

The assessment group adapted a previously published model to undertake an independent economic assessment of shortened treatment durations with peginterferon alfa plus ribavirin, using clinical-effectiveness data from studies included in the systematic review. The economic model was structurally similar to those developed by the manufacturers, and used similar input parameters to model disease progression, health state costs and utility. The model consists of 9 health states representing stages of chronic liver disease and death.

4.2.20

The economic model contains 3 health states representing cure of chronic hepatitis C, which are differentiated by the stage of the disease before treatment, as this is expected to have an impact on the subsequent risk of progressive liver disease, post-treatment surveillance and health-related quality of life. The health states mild hepatitis C, moderate hepatitis C, compensated cirrhosis, decompensated cirrhosis, hepatocellular carcinoma and liver transplant represent stages of progressive liver disease. People not exhibiting a sustained virological response are expected to have the same risk of disease progression as untreated people. These assumptions are all consistent with previous assessments and with other published economic evaluations of anti-viral treatment for chronic HCV infection. The model has a cycle length of 1 year and incorporates a half-cycle adjustment.

4.2.21

Baseline population data in the model were based on the results of a clinical audit undertaken at a London teaching hospital. New patients were taken to represent the population of people with HCV and HIV co-infection and the population eligible to receive shortened courses of combination therapy as applicable, and existing patients were taken to represent the population of people treated previously with peginterferon alfa and ribavirin. Patients were differentiated in terms of average age and distribution across stages of chronic liver disease (mild hepatitis C, moderate hepatitis C and compensated cirrhosis). The proportion of men in the baseline cohort was based on the assessment group's previous assessment. The majority of these assumptions do not affect response to treatment, but relate to the risk of all-cause mortality. The influence of the stage of chronic liver disease on response to treatment, and on the cost effectiveness of the intervention, was assessed in a sensitivity analysis.

4.2.22

Data on sustained virological response rates for the subgroup of people receiving shortened treatment courses were extracted by the assessment group from clinical trials included in the clinical effectiveness review. For the subgroups of people co-infected with HIV and HCV and of people whose hepatitis C did not respond to previous treatment or responded but subsequently relapsed, sustained virological response rates were taken from RCTs that included active comparators, which were not systematically reviewed by the assessment group. Data on sustained virological response rates were used in the model to estimate the probability of treatment-eligible people moving to the relevant state of a sustained virological response. Where applicable, early virological response rates were used to estimate the average duration of treatment and total drug acquisition costs for each anti-viral treatment strategy.

4.2.23

Shortening the duration of treatment with peginterferon alfa-2a plus ribavirin from 48 weeks to 24 weeks for people with HCV genotype 1, a baseline low viral load and a rapid virological response resulted in a reduction in total QALYs of between 0.08 and 0.14. This shortened treatment duration was also associated with a reduction in the total cost, resulting in ICERs that reflected 'savings per QALY lost' (ranging from £34,510 to £64,880 per QALY). For people with HCV genotypes 2 and 3, a baseline low viral load and a rapid virological response, a shortened duration of treatment of 16 weeks with peginterferon alfa-2a plus ribavirin dominated the standard 24-week treatment duration. For the subgroup of people with HCV genotype 1, a baseline low viral load and a rapid virological response, a shortened duration of treatment of 24 weeks with peginterferon alfa-2b plus ribavirin dominated the standard 48‑week treatment duration.

4.2.24

The ICER for the subgroup of people who had been treated previously with peginterferon alfa-2b plus ribavirin or peginterferon monotherapy but whose hepatitis C did not respond to treatment, or responded initially to treatment but subsequently relapsed, was £23,912 per QALY gained (compared with best supportive care) for people with HCV genotypes 1 and 4. If an early stopping rule was applied, the ICER for people with genotypes 1 and 4 was reduced to £7681 per QALY gained. For people with genotypes 2 and 3, treatment with peginterferon alfa-2b plus ribavirin dominated best supportive care (with or without an early stopping rule). Re-treatment with peginterferon alfa-2a plus ribavirin of people whose hepatitis C did not respond to previous peginterferon therapy resulted in an ICER of £52,587 per QALY gained for people with HCV genotype 1 and £10,926 per QALY gained for people with other genotypes, each compared with best supportive care. If an early stopping rule was applied at 12 weeks for people re-treated with peginterferon alfa-2a plus ribavirin who did not show an early virological response at this time, the ICERs were reduced to £9,169 per QALY gained for people with HCV genotype 1 and £2,294 per QALY gained for people with other genotypes.

4.2.25

For people co-infected with HCV and HIV, treatment with peginterferon alfa-2a plus ribavirin resulted in an ICER of £7,941 per QALY gained for people with HCV genotypes 1 and 4 compared with best supportive care. Peginterferon alfa-2a plus ribavirin dominated best supportive care for people with genotypes 2 and 3. Treatment of people co-infected with HCV and HIV with peginterferon alfa-2b plus ribavirin resulted in an ICER of £11,806 per QALY gained for people with HCV genotypes 1 and 4 and an ICER of £2,161 per QALY gained for people with genotypes 2 and 3.

Summary of cost-effectiveness evidence

4.2.26

Cost-effectiveness analyses indicate that adopting a shortened duration of treatment with peginterferon alfa-2a plus ribavirin for people with HCV genotype 1 is associated with fewer QALYs, but also with lower treatment costs, resulting in ICERs ranging from £34,000 to £65,000 of savings per QALY lost. For people with genotypes 2 and 3, a shortened treatment course with peginterferon alfa-2a plus ribavirin dominates the standard treatment duration. Similarly, for people with HCV genotype 1, a shortened duration of treatment with peginterferon alfa-2b plus ribavirin dominates the standard treatment duration. The results submitted by the manufacturers and those reported by the assessment group also indicate that treatment with peginterferon alfa plus ribavirin for people who are eligible for re-treatment following previous non-response or relapse of their condition is associated with QALY gains and an increase in costs. If an early stopping rule was applied at 12 weeks, re-treatment with peginterferon alfa plus ribavirin was associated with ICERs below £10,000 per QALY gained for people with HCV genotypes 1 and 4. For people with HCV genotypes 2 and 3, re-treatment was associated with ICERs below £2,294 per QALY gained or dominated best supportive care. For people co-infected with HCV and HIV who were treated with peginterferon alfa plus ribavirin, either all ICERs were below £12,000 per QALY gained or treatment dominated best supportive care.

4.3 Consideration of the evidence

4.3.1

The appraisal committee reviewed the data available on the clinical and cost effectiveness of peginterferon alfa (2a or 2b) plus ribavirin, having considered evidence on the nature of chronic hepatitis C and the value placed on the benefits of peginterferon alfa (2a or 2b) plus ribavirin by people with the condition, those who represent them, and clinical specialists. It also took into account the effective use of NHS resources.

4.3.2

The committee was aware from the evidence presented that there were no statistically significant differences in sustained virological response rates between subgroups receiving shortened and standard courses of treatment with peginterferon alfa plus ribavirin. The committee heard from the clinical specialist that, although some trials have reported that people who received shortened courses of treatment had (non-significant) lower sustained virological response rates, using shortened rather than standard courses of therapy was not expected to have clinical implications. Therefore the committee concluded that shortened and standard treatment durations could be viewed as clinically comparable.

4.3.3

The committee noted that no RCTs were identified that compared peginterferon alfa (with or without ribavirin) with best supportive care for people co-infected with HIV and HCV, or for people whose hepatitis C did not respond to previous treatment or responded but subsequently relapsed. The committee heard from the clinical specialist that it would be unethical to conduct trials without an active comparator arm for these groups, and therefore the committee should consider the use of sustained virological response rates from trials that had an active comparator to be appropriate for these subgroups. The committee agreed that using the results from trials that had an active comparator to derive data on clinical effectiveness for the economic model was appropriate.

4.3.4

The committee discussed the assumption that people who receive best supportive care for chronic hepatitis C will not have a sustained virological response. The clinical specialist stated that sustained virological response rates are generally extremely low for people not receiving active therapy. The committee concluded that it was therefore reasonable to assume that people who receive best supportive care do not achieve a sustained virological response.

4.3.5

The committee heard from the patient experts that many people with chronic hepatitis C would be willing to tolerate the adverse effects of treatment with peginterferon alfa, with or without ribavirin, for the longest available treatment period to increase their chance of achieving a sustained virological response or to reduce the possibility of relapse following treatment, even if they knew that a shortened course of treatment may be associated with fewer adverse events. The committee acknowledged that some people prefer to have a choice when considering their treatment options.

4.3.6

In response to comments received after consultation on the ACD, the committee noted that the definition of undetectable serum HCV RNA (which denotes clearing of HCV) varied between the clinical trials (see section 4.1.3). They acknowledged that different laboratories in the UK use different tests and set different thresholds to determine whether a virus is undetectable, and that the quality of the test used may influence treatment decisions. The committee therefore agreed that a highly sensitive test should be used to detect serum HCV RNA, to minimise the chance of false negative results – that is, to minimise the chance that the test indicates the absence of virus, leading erroneously to a shortened course of treatment, when in fact virus is present.

4.3.7

The committee discussed the cost-effectiveness data presented by the manufacturers and the assessment group. The committee noted that the assessment group adapted a previously published economic model, and that the model was structurally similar to those developed by the manufacturers, and used similar input parameters to model disease progression, health state costs and utility. The committee noted that the model submitted by Roche Products may have overestimated the QALY gains from achieving a sustained virological response, because the model did not differentiate between mild and moderate hepatitis C and because age-specific utilities were applied to the sustained virological response state but not to other health states. The committee considered that it was not appropriate for the model to apply age-specific utility values to only one health state, but concluded that the additional analysis carried out by the assessment group to correct for this issue was appropriate.

4.3.8

The committee was aware that the model from Schering-Plough for peginterferon alfa-2b assumed that a sustained virological response occurs at the beginning of a cycle rather than mid-cycle, and that the same utility values for a sustained virological response were used irrespective of the stage of disease when treatment starts. The committee was satisfied that the additional analyses carried out by the assessment group corrected for these issues and did not substantially alter the results.

4.3.9

The committee considered the cost effectiveness of peginterferon alfa and ribavirin combination therapy for the 3 populations covered by this appraisal. For people with HCV genotype 1, a baseline low viral load and a rapid virological response, the Committee noted that the shortened treatment duration of 24 weeks with peginterferon alfa-2a plus ribavirin was associated with a reduction in both QALYs and costs compared with the standard treatment duration of 48 weeks, resulting in ICERs that reflected 'savings per QALY lost' (ranging from £34,500 to £64,900 per QALY). The committee considered situations where an ICER is derived from decreased effectiveness and decreased costs, and concluded that the commonly assumed decision rule of accepting ICERs below a given threshold is reversed, and so ICERs that are above the threshold are considered acceptable in this case. The committee also noted that for people with genotype 1 who had a low viral load and a rapid virological response, the shortened treatment duration of 24 weeks with peginterferon alfa-2b plus ribavirin dominated the standard treatment duration of 48 weeks. Similarly, for people with HCV genotype 2 or 3, a baseline low viral load and a rapid virological response, the shortened treatment duration of 16 weeks with peginterferon alfa-2a plus ribavirin dominated the standard treatment duration of 24 weeks. Based on these ICERs, the committee concluded that shortened courses of peginterferon alfa (2a or 2b) and ribavirin combination therapy represent a cost-effective use of NHS resources for people eligible for this treatment regimen. The committee noted that the cost effectiveness results also indicated that standard treatment durations were not likely to be considered cost effective compared with shortened treatment durations. Since the shortened and standard treatment durations are considered to be clinically comparable (see section 4.3.2), the committee concluded that shortened treatment courses should be recommended rather than recommended as an option. However, the committee was aware that shortened treatment courses are not suitable for all people. The committee highlighted that when considering a shortened treatment course, clinicians should take into account the licensed indication of the chosen drug (peginterferon alfa-2a or peginterferon alfa-2b), the HCV genotype, the viral load at the start of treatment and the response to treatment (as indicated by the viral load).

4.3.10

The committee then discussed whether peginterferon alfa (2a or 2b) plus ribavirin should be recommended for people who have been treated previously with peginterferon alfa (2a or 2b) and ribavirin in combination, or with peginterferon alfa monotherapy, and whose condition either did not respond to treatment or responded initially to treatment but subsequently relapsed. The committee concluded that the analysis that incorporates an early stopping rule at 12 weeks for people who do not show an early virological response, as specified in the marketing authorisation, was the most appropriate scenario forming the basis of the cost-effectiveness analysis (see section 4.2.24). The committee noted that the assessment group's analysis for people re-treated with peginterferon alfa plus ribavirin resulted in ICERs below £10,000 per QALY gained for people with HCV genotypes 1 and 4, and re-treatment dominated best supportive care for people with HCV genotypes 2 and 3. The committee then considered the cost-effectiveness estimates for the subgroup of people with HIV and HCV co-infection. The committee noted that in the assessment group's analysis for people co-infected with HCV and HIV who were receiving peginterferon alfa (2a or 2b) plus ribavirin, either all ICERs were below £12,000 per QALY gained or treatment dominated best supportive care. The committee concluded that the ICERs presented in the base-case analysis by the assessment group were plausible for both the subgroup of people being re-treated and the subgroup of people co-infected with HCV and HIV. Based on these ICERs, the committee was satisfied that peginterferon alfa (2a or 2b) and ribavirin combination therapy represents a cost-effective use of NHS resources both for people who have been treated previously and whose condition either did not respond to treatment or responded initially to treatment but subsequently relapsed, and for people with HIV and HCV co-infection.

Summary of appraisal committee's key conclusions

Key conclusions

  • Section 1.1: Combination therapy with peginterferon alfa (2a or 2b) and ribavirin is recommended as a treatment option for adults with chronic hepatitis C:

    • who have been treated previously with peginterferon alfa (2a or 2b) and ribavirin in combination, or with peginterferon alfa monotherapy, and whose condition either did not respond to treatment or responded initially to treatment but subsequently relapsed or

    • who are co-infected with HIV

  • Section 1.3: Shortened courses of combination therapy with peginterferon alfa (2a or 2b) and ribavirin are recommended for the treatment of adults with chronic hepatitis C who:

    • have a rapid virological response to treatment at week 4 that is identified by a highly sensitive test and

    • are considered suitable for a shortened course of treatment.

Current practice

Clinical need of patients including the availability of alternative treatments
  • Clinical practice for the use of these technologies is described in NICE technology appraisal guidance 75 and 106. This appraisal addresses extensions to the marketing authorisations.

The technology

Adverse effects
  • Section 4.3.5: The committee heard from the patient experts that many people with chronic hepatitis C would be willing to tolerate the adverse effects of treatment with peginterferon alfa, with or without ribavirin, for the longest available treatment period to increase their chance of achieving a sustained virological response or to reduce the possibility of relapse following treatment, even if they knew that a shortened course of treatment may be associated with fewer adverse events.

Evidence for clinical effectiveness

Availability, nature and quality of evidence
  • Section 4.3.3: RCT evidence was available for shortened treatment courses. No RCTs were identified that compared peginterferon alfa (with or without ribavirin) with best supportive care for people co-infected with HIV and HCV, or for people whose hepatitis C did not respond to previous treatment or responded but subsequently relapsed. The committee agreed that it was appropriate to use results from trials that had an active comparator to derive data on clinical effectiveness for the economic model.

Relevance to general clinical practice in the NHS
  • Section 4.3.4: The clinical specialist stated that sustained virological response rates are generally extremely low for people not receiving active therapy. The committee concluded that it was therefore reasonable to assume that people who receive best supportive care do not achieve a sustained virological response.

  • Section 4.3.6: The committee noted that the definition of undetectable serum HCV RNA (which denotes clearing of HCV) varied between the clinical trials, and therefore that the quality of the test used may influence treatment decisions. The committee therefore agreed that a highly sensitive test should be used to detect serum HCV RNA, to minimise the chance of false negative results.

Uncertainties generated by the evidence
  • There were no major uncertainties generated by the evidence.

Estimate of the size of the clinical effectiveness including strength of supporting evidence
  • Section 4.3.2: The committee was aware from the evidence presented that there were no statistically significant differences in sustained virological response rates between subgroups receiving shortened and standard courses of treatment with peginterferon alfa plus ribavirin. The committee heard from the clinical specialist that, although some trials have reported that people who received shortened courses of treatment had (non-significant) lower sustained virological response rates, using shortened rather than standard courses of therapy was not expected to have clinical implications. Therefore the committee concluded that shortened and standard treatment durations could be viewed as clinically comparable.

Evidence for cost effectiveness

Availability and nature of evidence
  • Cost effectiveness data were presented by the manufacturers and the assessment group.

  • Section 4.3.7: The committee noted that the assessment group adapted a previously published economic model, and that the model was structurally similar to those developed by the manufacturers, and used similar input parameters to model disease progression, health state costs and utility.

Uncertainties around and plausibility of assumptions and inputs in the economic model
  • There were no material uncertainties around the assumptions and inputs in the economic models presented.

  • Section 4.3.7: The committee noted that the model submitted by Roche Products may have overestimated the QALY gains from achieving a sustained virological response, because the model did not differentiate between mild and moderate hepatitis C and because age-specific utilities were applied to the sustained virological response state but not to other health states. The committee considered that it was not appropriate for the model to apply age-specific utility values to only one health state, but concluded that the additional analysis carried out by the assessment group to correct for this issue was appropriate.

  • Section 4.3.8: The committee was aware that the model from Schering-Plough for peginterferon alfa-2b assumed that a sustained virological response occurs at the beginning of a cycle rather than mid-cycle, and that the same utility values for a sustained virological response were used irrespective of the stage of disease when treatment starts. The committee was satisfied that the additional analyses carried out by the assessment group corrected for these issues and did not substantially alter the results.

Have any potential significant and substantial health-related benefits been identified that were not included in the economic model, and how have they been considered?
  • No potential significant and substantial health-related benefits were identified that were not included in the economic models.

Most likely cost-effectiveness estimate (given as an ICER)
  • Section 4.3.9: For people with HCV genotype 1, a baseline low viral load and a rapid virological response, the Committee noted that the shortened treatment duration of 24 weeks with peginterferon alfa-2a plus ribavirin was associated with a reduction in both QALYs and costs compared with the standard treatment duration of 48 weeks, resulting in ICERs that reflected 'savings per QALY lost' (ranging from £34,500 to £64,900 per QALY).

  • Section 4.3.9: The committee considered situations where an ICER is derived from decreased effectiveness and decreased costs, and concluded that the commonly assumed decision rule of accepting ICERs below a given threshold is reversed, and so ICERs that are above the threshold are considered acceptable in this case.

  • Section 4.3.9: The committee also noted that for people with genotype 1 who had a low viral load and a rapid virological response, the shortened treatment duration of 24 weeks with peginterferon alfa-2b plus ribavirin dominated the standard treatment duration of 48 weeks. Similarly, for people with HCV genotype 2 or 3, a baseline low viral load and a rapid virological response, the shortened treatment duration of 16 weeks with peginterferon alfa-2a plus ribavirin dominated the standard treatment duration of 24 weeks.

  • Section 4.3.10: The committee noted that the assessment group's analysis for people re-treated with peginterferon plus ribavirin resulted in ICERs below £10,000 per QALY gained for people with HCV genotypes 1 and 4, and re-treatment dominated best supportive care for people with HCV genotypes 2 and 3.

  • Section 4.3.10: The committee noted that in the assessment group's analysis for people co-infected with HCV and HIV who were receiving peginterferon alfa (2a or 2b) plus ribavirin, either all ICERs were below £12,000 per QALY gained or treatment dominated best supportive care. The committee concluded that the ICERs presented in the base-case analysis by the assessment group were plausible for both the subgroup of people being re-treated and the subgroup of people co-infected with HCV and HIV.

Additional factors taken into account

Equalities considerations, social value judgements
  • No issues relating to equality were raised at any stage during the development of this appraisal.