4 Evidence and interpretation

The Appraisal Committee (Appendix A) considered evidence from a number of sources (see Appendix B).

4.1 Clinical effectiveness

4.1.1 The standard measurement of effectiveness of treatment of CHC is the virological response rate sustained for 6 months, called the SVR. SVR has been shown to closely reflect biopsy and ALT results taken from the same people at the same time.

4.1.2 Peginterferon alfa combination therapy versus interferon alfa combination therapy

4.1.2.1 The effectiveness of peginterferon alfa and ribavirin combination therapy, compared with interferon alfa and ribavirin combination therapy, for patients being treated with interferon alfa or peginterferon alfa for the first time has been investigated in two randomised controlled trials (RCTs) lasting 48 weeks. One trial used ribavirin and peginterferon alfa-2a (n = 1121) and the other used ribavirin and peginterferon alfa-2b (n = 1530). The results were broadly similar. For the first trial, peginterferon alfa-2a in combination with ribavirin yielded an SVR of 56% versus 44% for interferon alfa-2b in combination (95% confidence interval [CI] on the difference of 12 percentage points is 5 to 19 percentage points). In the second trial, the intention-to-treat analysis (which included patients taking ribavirin at lower than the licensed dose) of peginterferon alfa-2b in combination with ribavirin, the SVR was 54% versus 47% for interferon alfa-2b in combination (95% CI on the difference of 7 percentage points is 0.4 to 12.7 percentage points). In this arm of the study, all patients received 800 mg of ribavirin with 1.5 μg per kg body weight of peginterferon alfa-2b. The effect of ribavirin dose adjusted according to body weight was analysed in a subset of 188 of these patients. In this sub-population, the SVR for peginterferon alfa-2b in combination was 61% versus 47% for interferon alfa-2b in combination (95% CI on the difference of 14 percentage points is 5 to 22 percentage points). The licence for peginterferon alfa-2b combination therapy is based on this weight-adjusted ribavirin dosage. The Assessment Report recognises that the treatments with peginterferon alfa-2a and alfa-2b (both in combination with ribavirin) may be different and that there are differences between the trial populations. However, it shows that, if the results of the trials of these two treatments are pooled, peginterferon alfa combination therapy yields an SVR of 56% on an intention-to-treat basis, whereas the interferon alfa combination yields an SVR of 47% on the same basis. The difference (9 percentage points) has a 95% CI from 5 to 13 percentage points. A second trial of peginterferon alfa-2a combination therapy has so far been reported in abstract form only. It extends the knowledge gained from the first trial by comparing different doses of ribavirin and lengths of treatment. Broadly, it confirms the results of the first trial using peginterferon alfa-2a. (Further results are currently commercial-in-confidence.)

4.1.2.2 The SVR in each of the two fully reported trials varied with both the baseline viral load and the genotype of the HCV. When there were more than 2 million copies of the virus in each millilitre of a patient's blood, the SVR was significantly lower than when there were fewer than 2 million copies. This was true for both arms of both of the trials.

4.1.2.3 SVRs for patients infected with HCV G1 are much lower than those for G2/3, whereas SVRs for genotypes 4, 5 and 6 (when they are known) appear to be between those of the more prevalent genotypes. For G1, SVRs for peginterferon alfa-2a combination therapy were 46%, compared with 36% for interferon alfa-2a combination therapy. When peginterferon alfa-2b combination therapy and interferon alfa-2b combination therapy were compared, the SVR values were 42% and 33%, respectively, on an intention-to-treat basis. On a weight-based ribavirin dosage, they were 48% and 34%, respectively. For G2/3, the SVR for peginterferon alfa-2a combination therapy was 76%, compared with 61% for the interferon alfa-2a therapy. When the peginterferon alfa-2b and interferon alfa-2b combination therapies were compared, the SVR values were 82% and 79%, respectively, on an intention-to-treat basis. On a weight-based ribavirin dosage, they were 88% and 80%, respectively.

4.1.2.4 Patients infected with HCV G2/3 respond to combination treatment with peginterferon alfa-2a in 95% of cases or more, and in about 80% of cases the response is sustained 6 months after treatment has finished. These rates are achieved after 24 weeks of treatment and are not increased by prolonging treatment for a further 24 weeks. For G1, however, the SVR after 48 weeks of treatment is much higher than that for 24 weeks of treatment, even though it is of the order of only 40–50%. This pattern follows that of combination therapy with interferon alfa-2a and interferon alfa-2b.

4.1.2.5 After 12 weeks of treatment, the viral load in people who eventually have an SVR after 24 or 48 weeks' treatment is generally reduced by a factor of 100 or more. That is, for every 1000 copies of the virus in the blood at the beginning of treatment, there would be 10 or fewer copies at the end of 12 weeks' treatment. This is known as a 2-log reduction.

4.1.2.6 For patients infected with HCV G2/3, more than 99% will respond with a 2-log reduction at 12 weeks. About 80% will eventually have an SVR. Of the very small number of patients not responding at 12 weeks, very few (perhaps less than 0.5% of the group that started treatment) have an SVR. For genotypes 1, 4, 5 and 6 (together called G1+), only 70–80% have a 2-log reduction at 12 weeks and, of these, about 60% (40–50% of the total group) have an SVR. Of the 20–30% that are non-responders at 12 weeks, few (perhaps 0.5% of those originally treated) go on to have an SVR.

4.1.2.7 Data from a subgroup of people with cirrhosis or bridging fibrosis and G2/3 in a recent trial of peginterferon alfa-2a, details of which are still confidential until its full publication, suggest that treatment beyond 24 weeks does not result in an increase in the SVR.

4.1.3 Peginterferon alfa monotherapy versus interferon alfa monotherapy

4.1.3.1 The Assessment Report found four RCTs that compared peginterferon alfa monotherapy with interferon alfa monotherapy. Three of these trials, involving about 960 people, were conducted with peginterferon alfa-2a, and one trial, involving more than 1200 people, was conducted with peginterferon alfa-2b. SVRs were much lower than for combination therapy. Peginterferon alfa-2a yielded a 36% pooled response, compared with 14% for interferon alfa-2a, whereas the SVR values for peginterferon alfa-2b versus interferon alfa-2b were 23% and 12%, respectively. Different doses of peginterferon alfa were used in three of the four trials, which occurred at different stages of drug development. All trials were 48 weeks in duration; hence the shorter treatment possibility for G2/3 was not tested.

4.1.4 Re-treatment of non-responders

4.1.4.1 The Assessment Report found 10 RCTs, involving some 860 people, that compared interferon alfa combination therapy with interferon alfa monotherapy for the re-treatment of non-responders to interferon alfa monotherapy. Of those re-treated with monotherapy, only 7 out of 413 had a virological response at the end of the trial, whereas for combination therapy, 53 out of 449 had such a response. For studies including both failure to respond and relapses from previous monotherapy, there were 16 responses out of 323 for monotherapy, compared with 75 out of 330 for combination therapy. The differences between the success rates for monotherapy compared with combination therapy are marked, although the percentage of successes when re-treating people failing to respond to monotherapy with combination therapy is only of the order of 10%.

4.1.4.2 Data for re-treatment with peginterferon alfa combination therapy for people previously treated with interferon alfa monotherapy or combination therapy is still tentative.

4.1.5 Adherence

4.1.5.1 Three studies (one published and one unpublished study of peginterferon alfa combination therapy, and one study of peginterferon alfa monotherapy) have retrospectively examined satisfactory adherence, defined as adhering to the designated dosing pattern at least 80% of the time. All studies show that SVR is significantly higher among people with G1 who show satisfactory adherence. For people with G2/3, one of the three studies also shows that SVR is significantly higher among those with satisfactory adherence.

4.1.6 Other patient subgroups: haemophilia

4.1.6.1 Many people with haemophilia were infected by blood products, in most cases by HCV G1. Many cases of G1 did not respond to monotherapy, or relapsed within 6 months. Small studies showed that a small but significant proportion of these relapses and treatment failures responded to peginterferon alfa combination therapy.

4.1.7 Other patient subgroups: HIV comorbidity

4.1.7.1 It is not unusual for people with HCV to be co-infected with HIV, because of their common transmission routes. Several patient submissions, one manufacturer and the Assessment Report examined this set of circumstances.

4.1.7.2 In people infected with both viruses, the rate of progression of CHC is much faster.

4.1.7.3 Several small trials have been conducted, all involving interferon alfa-2b, which show that the SVRs are of the order of 30% lower (for example, 35% instead of 50%) for people co-infected with HIV than for those without HIV.

4.1.7.4 There is no evidence that interferon alfa interacts with drugs taken for HIV, but there is evidence that ribavirin could do so when taken with peginterferon alfa, and may prove toxic. Additional care is called for when monitoring people receiving medication for HIV co-infection.

4.1.8 Other patient subgroups: injecting drug users

4.1.8.1 Current injecting drug users can have high rates of discontinuation in trials, and thus do not achieve success rates in trials with interferon alfa therapy as high as those obtained by other participants. However, there is evidence that where adherence is achieved, success rates are not significantly different.

4.1.9 Other patient subgroups: people with continued alcohol consumption

4.1.9.1 Alcohol consumption of more than 7 units per week not only increases liver damage for those infected with HCV, but also adversely affects its treatment.

4.1.10 Other patient subgroups: liver transplants

4.1.10.1 People with CHC who require a liver transplant usually develop the disease in the new liver. Very limited data (six people) showed that four people responded to peginterferon alfa-2b combination therapy.

4.1.11 Other patient subgroups: age, gender and ethnicity

4.1.11.1 Some differences have been observed in the success of treatment between people of different ages, between men and women, and between people of different ethnicity. These differences are relatively small compared with those resulting from viral genotype or viral load.

4.1.12 Other patient subgroups: mild CHC and acute hepatitis C

4.1.12.1 Trials in people with mild disease have not yet reported. Treatment of mild CHC is outside of the scope of this appraisal.

4.1.12.2 Acute infection is not covered by this appraisal. One trial has shown better clearance if HCV infection is treated immediately after onset, but it may not be possible to generalise its results to most people infected with HCV.

4.2 Cost effectiveness

4.2.1 Peginterferon alfa combination therapy versus interferon alfa combination therapy

4.2.1.1 The Assessment Report shows that peginterferon alfa combination therapy is a very cost effective intervention compared with interferon alfa combination therapy. For G2/3, given the very high sustained success rates at 24 weeks, treatment is cost effective at 24 weeks but not thereafter. For G1, 48-week treatment is cost effective compared with stopping therapy after 24 weeks. See Table 1.

Table 1: Cost effectiveness of combination therapy for different HCV genotypes

Comparison

Genotype

Treatment length (weeks)

Estimated incremental cost/QALY

(1) Peginterferon alfa combination vs interferon alfa combination

1

48

£4000 to £11,000

(2) Peginterferon alfa combination vs interferon alfa combination

4–6

48

£9000

(3) Peginterferon alfa combination vs interferon alfa combination

2–3

48

£7000 to £38,000

(4) Peginterferon alfa combination (24 weeks) vs peginterferon alfa combination (48 weeks)

Not 1

24 vs 48

£69,000 to negative benefits compared with24 week treatment

(5) Peginterferon combination (24 weeks) vs peginterferon combination (48 weeks)

1

24 vs 48

£15,000 to £19,000 compared with 24 week treatment

Notes on Table 1: The estimated incremental cost/QALY figures were obtained from the Assessment Report using a modelling approach.

Rows (1) and (3): use data from the pivotal trials of peginterferon/interferon alfa-2a and of alfa-2b. The estimates differ because they are based on (a) different trials and (b) different doses of ribavirin.

Row (2): Uses data from the pivotal trial of peginterferon/interferon alfa-2b.

Rows (4) and (5): use data from an unpublished trial of peginterferon/interferon alfa-2a submitted in confidence by the manufacturer, for different doses of ribavirin. The genotype 'not 1' essentially refers to genotypes 2 and 3, as the numbers of those in genotypes 4, 5 and 6 were small.

4.2.1.2 In the Assessment Report, the estimates of incremental cost effectiveness ratios by viral genotype differ depending on thetype of peginterferon alfa or interferon alfa and the dose of ribavirin. For G1, the estimated cost per quality-adjusted life ear (QALY) gained of peginterferon alfa combination therapy compared with the corresponding interferon alfa combination therapy for 48 weeks' treatment ranges from £4000 to £11,000. For G2/3, the corresponding figures are £7000 to £38,000.

4.2.1.3 For monotherapy, all treatments are for 48 weeks (see Table 2).

Table 2: Cost effectiveness of monotherapy for different HCV

Comparison

Genotype

Estimated incremental cost/QALY

Peginterferon alfa monotherapy vs interferon alfa monotherapy

1

£19,000

Peginterferon alfa monotherapy vs interferon alfa monotherapy

2 and 3

£7000

Peginterferon alfa monotherapy vs interferon alfa monotherapy

4–6

£2000

Note on Table 2: The estimated incremental costs per QALY gained were obtained from the Assessment Report, based on a modelling approach using SVRs taken from a meta-analysis.

4.2.1.4 The manufacturers' models are similar in structure to that of the Assessment Report, and the estimates of cost effectiveness derived from them show even lower costs per QALY. In one instance, this can be explained in part by the longer time horizon (expected lifetime, as opposed to 30 years).

4.2.1.5 The Assessment Report shows that testing viral load at 12 weeks for G1+ and stopping treatment for people who do not exhibit a 2-log reduction in viral load is cost effective compared with continuing treatment. Some 20-30% of people infected with G1+ do not respond at 12 weeks, and of these, less than 2% will eventually have an SVR. The cost per QALY gained from continuing treatment for the non-responders at 12 weeks is estimated to be £227,000. This is not the case for G2/3; there are very few non-responders at 12 weeks.

4.2.1.6 The cost effectiveness of treating with peginterferon combination therapy non-responders to interferon monotherapy has been estimated to be £3000 per QALY against no treatment. For non-responders to interferon combination therapy, it is £9000 per QALY against no treatment.

4.3 Consideration of the evidence

4.3.1 The Committee reviewed the evidence available on the clinical and cost effectiveness of treatment with interferon alfa and peginterferon alfa and ribavirin in CHC, having considered evidence on the nature of the condition and the value placed by users on the benefits of interferon and peginterferon alfa and ribavirin from people with CHC, those who represent them, and clinical experts. It was also mindful of the need to take account of the effective use of NHS resources.

4.3.2 The Committee considered that peginterferon alfa combination therapy was both clinically and cost effective compared with interferon alfa combination therapy. Additionally, peginterferon alfa monotherapy was both clinically and cost effective compared with interferon alfa monotherapy. The Committee concluded that peginterferon alfa therapy should therefore supersede treatment using interferon alfa for all people unless particular side-effect considerations (neutropenia and thrombocytopenia risk) favour interferon alfa (without pegylation). The Committee also concluded that combination therapy should be used rather than monotherapy, except for people for whom ribavirin is contraindicated or cannot be tolerated.

4.3.3 The Committee gave careful consideration to the differential efficacy of treatment for patients infected with the different HCV genotypes. On the basis of the evidence reviewed, the Committee concluded that patients infected with HCV G2/3 should be considered differently from those with G1, and genotypes 4, 5 and 6 should be treated as for G1. For combination therapy, people with G2/3 should receive 24 weeks' treatment, whereas people with all other genotypes who have demonstrated a sufficient initial response should receive 48 weeks' treatment.

4.3.4 For combination therapy, the Committee discussed the requirement for testing for viral load at 12 weeks after the initiation of treatment as a means of assessing response. For G2/3, the number of non-respondents at this stage was such a small proportion that testing them to exclude further treatment was not considered cost effective. For all other genotypes, because the proportion of non-responders was much higher than for G2/3, the viral load response at 12 weeks is important to inform the need for treatment up to 48 weeks.

4.3.5 The Committee further considered the clinical and cost effectiveness of peginterferon alfa-2a versus peginterferon alfa-2b. Although it was aware that the two drugs had different dosage regimens and pharmacokinetic profiles, it considered that, in the absence of any head-to-head trials and on the basis of expert opinion received, the evidence was insufficient to recommend one of these products over the other. The Committee concluded that it would be important for clinicians to have the choice of either product in order to target different groups of people under particular clinical circumstances. The Committee, in this context, gave consideration to genotypes 2 and 3. The Committee noted that the SVRs for peginterferon alfa-2a were 15 percentage points above those for interferon alfa-2b, whereas for peginterferon alfa-2b, the SVRs were only 3 percentage points above those for interferon alfa-2b. However, the Committee considered the response of the former control group to be relatively low, whereas that of the latter control group was relatively high. It also noted that, when the weight-related dosage of ribavirin (on which the peginterferon alfa-2b licence is based) was considered, the relative efficacy of peginterferon alfa-2b compared with interferon alfa-2b was more marked. The Committee therefore considered that the apparent differences in response between the two forms of peginterferon for G2/3 should not result in a differential recommendation.

4.3.6 The Committee considered the use of peginterferon alfa for combination therapy in groups of people with HCV infection that were not represented in the pivotal clinical trials. These included people with haemophilia and people co-infected with HIV. The Committee concluded that, based on the evidence available, there was no reason to make any different provision for these groups. It did, however, note that there might be occasions where ribavirin may interact with medication for HIV, necessitating either a change in the latter or a switch to peginterferon alfa monotherapy.

4.3.7 For combination therapy, the Committee considered the differences in treatment efficacy for people of different age, gender and ethnicity, and decided that, where sufficient evidence existed, the efficacy differences were not great enough to give rise to a different recommendation for any of these subgroups.

4.3.8 The Committee heard that, although injecting drug users with HCV might, on average, seek treatment less frequently than other people with HCV, those who do seek treatment have similar adherence rates to other people with HCV. Furthermore, the evidence provided by the experts persuaded the Committee that current information indicated that HCV re-infection rates for people on interferon or peginterferon therapy were low in those who continue to inject illicit drugs. Thus, although rates of discontinuation of injecting drug users in trials have been high, the Committee was prepared to accept that in naturalistic settings, the rate of discontinuation would not be so great as to prevent the treatment being cost effective.

4.3.9 The Committee heard that continued alcohol consumption even at levels of intake much lower than the recommended maximum levels for the general population might be harmful for people with CHC-induced liver disease. This is because of the effect of alcohol on the progression of liver disease and also because alcohol reduces the efficacy of peginterferon/interferon alfa as therapy for CHC. The Committee considered that continued alcohol consumption was, however, not in itself an absolute contraindication to therapy but should be emphasised as an important factor to be taken into account in advice and information given by the clinical team.

4.3.10 The Committee carefully considered the situation of people who had already been treated with peginterferon or interferon alfa. Evidence shows that for those treated with interferon alfa monotherapy who had either not responded, or had responded but then relapsed, further treatment with combination therapy (with either peginterferon or interferon alfa) will be cost effective, although not as cost effective as for people not previously treated. The Assessment Group produced a further modelling analysis which assumed that re-treatment with peginterferon alfa combination, for those who had already undergone interferon alfa combination therapy, would yield an SVR equal to the difference between the SVRs of the two therapies. This analysis suggested that it could be cost effective to re-treat those previously treated with interferon alfa combination therapy who had relapsed or who had not responded. The Committee, after also receiving expert clinical advice on this matter and recognising the great uncertainty surrounding these estimates, decided that these groups of people should be suitable for treatment. The Committee reached the same conclusion for the few, if any, people previously treated without sustained virological response with peginterferon alfa monotherapy. However, it decided that there was no clinical or modelling evidence, or expert opinion, to support re-treatment of people who had previously been treated with peginterferon alfa combination therapy.

4.3.11 For people unable to take ribavirin, the Committee decided that peginterferon alfa monotherapy should be the treatment of choice, because it is both clinically and cost effective compared with interferon alfa monotherapy, despite lower clearance rates of the virus than for combination therapy. All people taking peginterferon alfa monotherapy should receive treatment for 48 weeks, regardless of genotype, because it was noted that there is currently no evidence for the effectiveness of a shorter period (24 weeks) of treatment. The requirement for 12 weeks' viral load testing was also considered for this group, and it was concluded that it should apply to people with every HCV genotype. Although there was no direct evidence of the cost effectiveness for this recommendation, it could reasonably be assumed that viral testing at 12 weeks would be at least as cost effective as in combination therapy, and there was no evidence to support G2/3 being treated any differently from other genotypes. The provisos for combination therapy in Section 4.3.3 (except for the 24-week treatment for G2/3), and Sections 4.3.5 to 4.3.9, also apply to treatment with monotherapy.

4.3.12 The Committee considered the treatment of people classified on the basis of liver biopsy as having mild chronic CHC. It was aware that there were two trials of people with mild disease that would shortly be reporting. The correct and cost-effective management of this group was considered very important and, although people with mild disease represent a small subgroup of the current RCT evidence base, it was decided that waiting for the current specific trials to report would provide a more robust basis on which to provide guidance to the NHS.

4.3.13 The Committee discussed the question of the need for liver biopsy at some length. It concluded that, because the basis for the original guidance (see Section 8.1) required the definition of the extent of liver disease, the requirement for biopsy before deciding on appropriate therapy should remain. It was persuaded that alternative non-invasive tests of liver function could not currently be relied upon to act as appropriate surrogates for direct histological examination. However, the Committee considered that, in due course, the result of the trials in mild disease might affect this requirement. The Committee believed that there were grounds for making exceptions for people with haemophilia and risk of bleeding or with a previous adverse reaction to liver biopsy, and for those with extra-hepatic symptoms sufficient to merit treatment.

4.3.14 The Committee considered that the effective delivery of the guidance in Section 1 would be critically dependent on the existence of a properly structured clinical environment for people with CHC. Thus, it concluded that the decision to undertake therapy should only be initiated by a physician with specialist knowledge of the treatment of CHC. Additionally it is important that a clinical team including specialist nurses is available for lifestyle advice to facilitate the informed decision of the individual to undertake treatment and to help him or her successfully complete the course of therapy.