Sofosbuvir for treating chronic hepatitis C

  • Technology appraisal guidance
  • TA330
  • Published: 
Download guidance (PDF)

4 Consideration of the evidence

The Appraisal Committee reviewed the data available on the clinical and cost effectiveness of sofosbuvir, having considered evidence on the nature of chronic hepatitis C and the value placed on the benefits of sofosbuvir by people with the condition, those who represent them, and clinical experts. It also took into account the effective use of NHS resources.

4.1

The Committee heard from the clinical experts that chronic hepatitis C is often clinically asymptomatic, and that it is estimated to be undiagnosed in approximately 50% of people with the condition in England. However, when the condition progresses and cirrhosis occurs, it has a significant daily effect on the person with the virus and their carers. The Committee acknowledged the concerns of the patient experts that there is a stigma attached to having chronic hepatitis C, because of its link to injectable drug use. In addition, there is a reluctance to treat chronic hepatitis C in people who use injectable drugs, partly because of mistaken beliefs that they do not adhere to treatment and often become re‑infected. The Committee heard from the patient experts that the availability of sofosbuvir and other new treatments that are expected to become available over the next 5 years will encourage more people with chronic hepatitis C to seek diagnosis and treatment. In addition, people who use injectable drugs whose chronic hepatitis C is successfully treated may go on to address their drug use, leading to broader societal benefits that are not captured in the company's evidence submission. The Committee recognised the effect of chronic hepatitis C on the lives of people with the virus. It concluded that treatments that give a sustained virological response (which is considered equivalent to a cure), and that consequently help reduce the rate of HCV transmission and the stigma associated with having chronic hepatitis C, are of significant importance.

4.2

The Committee discussed the clinical management of chronic hepatitis C in adults. It heard from the clinical experts that different treatment options can have varied results, depending on the person's HCV genotype, level of liver damage, comorbidities and previous treatment history. For people with genotype 1 chronic hepatitis C, the Committee heard that boceprevir plus peginterferon alfa and ribavirin or telaprevir plus peginterferon alfa and ribavirin (see the NICE technology appraisal guidance on boceprevir for the treatment of genotype 1 chronic hepatitis C and telaprevir for the treatment of genotype 1 chronic hepatitis C) are commonly used, and that for people with genotypes 2 to 6 HCV, peginterferon alfa plus ribavirin or watchful waiting (closely monitoring the condition but not giving any treatment) are currently the main treatment options. The clinical experts highlighted that interferon‑based treatment can be associated with side effects such as chronic fatigue, neuropsychological effects and flu‑like symptoms, which can be a barrier to people wanting to start treatment, or taking their treatment for the recommended duration. The Committee also heard from the patient experts that interferon‑based treatment may cause chronic side effects, such as autoimmune responses and thyroid problems, which need additional long‑term management and therefore pose another barrier to people starting and completing treatment. The Committee acknowledged that the marketing authorisation for sofosbuvir offers people the option to have shortened courses of peginterferon alfa and ribavirin, or in some circumstances to have treatment without peginterferon alfa, thereby reducing potential adverse effects associated with interferon‑based therapy. The Committee agreed with the clinical experts and patient experts that the option to have a shortened course of interferon‑based therapy with sofosbuvir, or the possibility of sofosbuvir being used without peginterferon alfa in some circumstances, would make it a valuable treatment option for people with chronic hepatitis C.

4.3

The Committee acknowledged that the marketing authorisation in the UK for sofosbuvir licenses it to be used in adults with chronic hepatitis C in all genotypes. It heard from the clinical experts that in England, most people with chronic hepatitis C have genotypes 1 or 3 HCV (46% and 43% respectively), with genotype 1 HCV being associated with a poor response to antiviral therapy and an increased rate of progression to severe chronic liver disease. The Committee noted that the marketing authorisation also allows sofosbuvir to be used in people who have or have not had previous treatment for chronic hepatitis C. The Committee also noted that the marketing authorisation allows sofosbuvir in combination with ribavirin to be used in people with genotypes 1, 4, 5 or 6 HCV who could be considered interferon intolerant or ineligible and who are in urgent need of treatment. It heard from company representatives that people would be considered interferon intolerant or ineligible if interferon treatment was contraindicated (as described in the summary of product characteristics for peginterferon) or in people whose disease did not have an adequate response to previous interferon treatment. The Committee asked the clinicians and commissioners for a definition of who would be considered interferon intolerant or ineligible, but did not receive a clear response. The Committee heard from the clinical experts that sofosbuvir is an important new treatment that will address an unmet need, particularly in people who have previously been treated but did not have a sustained virological response, in people whose condition has relapsed, or in people who have become re‑infected after treatment. The Committee was aware that the marketing authorisation specifies that sofosbuvir treatment should be 'initiated and monitored by a physician experienced in the management of patients with chronic hepatitis C'. It agreed with comments received during consultation that treatment should be focused in specialist centres and that treatment decisions, such as determining whether someone is interferon intolerant or ineligible for interferon treatment, should be made preferably by a multidisciplinary team. The Committee concluded that most people with chronic hepatitis C are likely to have at least some benefit from adding sofosbuvir to their treatment regimen and that the condition should be treated in an appropriate setting, as specified in the sofosbuvir marketing authorisation.

Clinical effectiveness

4.4

The Committee considered the clinical effectiveness of sofosbuvir plus ribavirin, with or without peginterferon alfa, for people with genotypes 1 to 6 chronic hepatitis C. It noted the concerns of the Evidence Review Group (ERG) that because of the lack of head‑to‑head studies comparing sofosbuvir with current standard of care treatments, most of the evidence provided by the company did not directly address the decision problem. The Committee acknowledged that the company was able to provide evidence from only 1 head‑to‑head trial (FISSION, in people with genotype 2 or 3 treatment‑naive HCV, for whom interferon therapy is suitable; see sections 3.8 and 3.9) that was consistent with the decision problem. The Committee was aware that the direct comparison with standard of care treatment was further limited to people with genotype 2 HCV only because the marketing authorisation recommends 24 weeks of sofosbuvir and ribavirin treatment for people with genotype 3 HCV, but people with genotype 3 HCV in the FISSION study had 12 weeks of sofosbuvir and ribavirin. In addition, the Committee expressed concern about the robustness of the estimates of the clinical effectiveness of sofosbuvir across the different subgroups for whom it is licensed when stratified (grouped) by treatment history, presence or absence of cirrhosis, and interferon eligibility, given that most trials were single‑arm and open‑label with historical controls that only included relatively small patient numbers and provided short‑term data. The Committee heard from the clinical experts that the current standard of care has been used for many years in the UK, and has been supported by numerous trials; therefore it was not unreasonable to use historical controls. The clinical experts also commented that hepatitis C trials are often open label because some people realise they are taking an interferon‑based regimen, potentially making blinding difficult. The Committee was aware that the number of people with cirrhosis in the clinical trials was relatively small, although it reflected the proportion of people with cirrhosis seen in clinical practice, and the exclusion criteria meant that people who use injectable drugs were not included in any studies. The Committee acknowledged the limitations of carrying out trials for hepatitis C, and concluded that there was considerable uncertainty surrounding the evidence base presented by the company. Therefore the true magnitude of the effect of sofosbuvir in each subgroup could not be robustly estimated.

4.5

The Committee acknowledged that in the NEUTRINO trial, people with genotype 1 (89% of people in the trial), 4, 5 or 6 treatment‑naive HCV who had sofosbuvir plus peginterferon alfa and ribavirin had a high sustained virological response (91%) 12 weeks after treatment compared with the historical control of 60% that was presented by the company. The Committee concluded that sofosbuvir plus peginterferon alfa and ribavirin was clinically more effective than peginterferon alfa and ribavirin alone in inducing a sustained virological response in people with treatment‑naive genotype 1, 4, 5 or 6 HCV.

4.6

The Committee considered the clinical effectiveness of sofosbuvir plus peginterferon alfa and ribavirin in people with treatment‑experienced genotype 1, 4, 5 or 6 HCV. No trial data were available on the clinical effectiveness of sofosbuvir in people with these genotypes who had previously had treatment for HCV. The Committee heard from the clinical experts that there was no reason to expect a different response in treatment‑experienced HCV than in treatment‑naive HCV. The Committee also heard from clinical experts that it was unlikely that further studies of sofosbuvir plus peginterferon alfa and ribavirin in people with treatment‑experienced HCV would be started because new interferon‑free regimens are rapidly replacing older interferon‑based regimens. The Committee was aware of the evidence from the company that the US Food and Drug Administration had accepted that the increase in sustained virological response in people with genotype 1 treatment‑naive HCV from 50% to 89% in NEUTRINO (subsequently recalculated as 91%) represented an efficacy of 78% for sofosbuvir plus peginterferon alfa and ribavirin in those people who would not have a sustained virological response with peginterferon alfa and ribavirin alone (see section 3.69). The Committee also considered interim results from an ongoing open‑label, single‑arm study by Pol et al. (2013) on the efficacy of sofosbuvir plus peginterferon alfa and ribavirin in people with genotype 1 treatment‑experienced HCV, which the company provided during consultation. These interim data suggested that 74% of patients who did not previously have a sustained virological response with peginterferon alfa plus ribavirin plus another direct‑acting antiviral (ledipasvir or tegobuvir) had a sustained virological response 12 weeks after treatment with sofosbuvir plus peginterferon alfa and ribavirin. The Committee also considered that in the small numbers of people with genotype 4, 5 or 6 HCV in the NEUTRINO study (in people with treatment‑naive genotype 1, 4, 5 and 6 HCV), the sustained virological responses 12 weeks after sofosbuvir treatment were approximately 97%, which was similar to those in people with genotype 1 HCV (see section 3.3). The Committee concluded that although there was uncertainty about the robustness of the evidence base in people with HCV genotype 1, 4, 5 or 6 who have had HCV treatment before, there was sufficient evidence for the Committee to make a recommendation on the use of sofosbuvir in people with genotype 1, 4, 5 or 6 treatment‑experienced HCV.

4.7

The Committee discussed the design of the clinical trials for sofosbuvir plus ribavirin in people with genotype 2 and 3 HCV. It noted that the main evidence came from 4 trials (FISSION, [treatment‑naive HCV, interferon‑eligible], FUSION, [treatment‑experienced HCV], POSITRON [treatment‑naive and treatment‑experienced HCV, people who were ineligible for interferon or intolerant to it or unwilling to have it] and VALENCE [treatment‑naive and treatment‑experienced]. The Committee acknowledged that FISSION was the only trial with an active comparator (peginterferon alfa‑2a and ribavirin treatment for 24 weeks) but noted that it was an open‑label study, which was susceptible to the introduction of selection bias and that when broken down by genotype, treatment history, interferon eligibility and cirrhosis status, the results were based on small patient numbers. The Committee was also aware that sustained virological response in the combined study population (FISSION) was 67% in both the sofosbuvir plus ribavirin 12 week treatment arm and in the peginterferon alfa‑2a and ribavirin 24 week treatment arm. The Committee noted that when stratified by genotype, people with genotype 2 HCV had a higher sustained virological response with 12 weeks of sofosbuvir plus ribavirin (97%) than people having peginterferon alfa and ribavirin alone (78%). The Committee noted that people with genotype 3 HCV having 12 weeks of sofosbuvir plus ribavirin had a lower sustained virological response rate 12 weeks after treatment (56%) than people receiving peginterferon alfa and ribavirin alone for 24 weeks (63%). The Committee was aware that all 4 trials in people with genotype 2 and 3 HCV had small patient numbers in each stratified subgroup (by genotype, treatment history, interferon eligibility and cirrhosis) and different designs, and concluded that these factors introduced uncertainty around the clinical effectiveness of sofosbuvir. On balance, the Committee concluded that sofosbuvir plus ribavirin was clinically more effective than peginterferon alfa and ribavirin alone in inducing a sustained virological response at 12 weeks after treatment in people with genotype 2 HCV.

4.8

The Committee further considered the results of the VALENCE study, which showed that longer treatment with sofosbuvir plus ribavirin was needed for people with genotype 3 treatment‑naive HCV (24 weeks rather than 12 weeks) to obtain a comparable sustained virological response at 12 weeks after treatment to that seen in people with genotype 2 HCV (data not reported here; academic‑in‑confidence). This was also supported by the results of FISSION, which showed that the sustained virological response for 12 weeks treatment with sofosbuvir and ribavirin in people with genotype 3 HCV was consistently lower than that seen in people with genotype 3 HCV who had peginterferon alfa‑2a and ribavirin for 24 weeks (see section 3.9). The Committee also discussed the clinical effectiveness of sofosbuvir in people with genotype 2 and 3 treatment‑experienced HCV, noting that the evidence for this subgroup came from FUSION (which compared sofosbuvir plus ribavirin for 12 weeks [plus placebo for an extra 4 weeks] with sofosbuvir and ribavirin for 16 weeks), and from subpopulations in VALENCE. The Committee noted that sustained virological response was consistently higher for people with genotype 2 HCV (86% and 94% in the 12 week and 16 week treatment groups in FUSION; 93% after 12 weeks treatment in VALENCE) than for people with genotype 3 HCV, who needed longer treatment with sofosbuvir and ribavirin (16 weeks and 24 weeks) for a similar response to be shown. The Committee noted that in the studies people with cirrhosis also generally had a lower response than those without cirrhosis (irrespective of genotype). The Committee considered that treatment with sofosbuvir plus ribavirin was likely to lead to a better sustained virological response in people with genotype 3 HCV compared with the current standard of care (24 weeks of peginterferon alfa and ribavirin treatment), but only when sofosbuvir plus ribavirin treatment was extended to 24 weeks. The Committee concluded that taking into account the limitations of the trial designs and the use of historical controls there was considerable uncertainty around the true magnitude of benefit of sofosbuvir plus ribavirin compared with peginterferon alfa and ribavirin for 24 weeks in people with genotype 3 HCV.

4.9

The Committee considered the available evidence for sofosbuvir plus peginterferon alfa and ribavirin in people with genotype 3 HCV. The Committee was aware that the European public assessment report for sofosbuvir stated that because peginterferon alfa and ribavirin alone had a higher historical efficacy in people with genotype 3 HCV than in people with genotype 1 HCV, it could be inferred that a similar improvement in efficacy seen in people with genotype 1 HCV would be expected in people with genotype 3 HCV when sofosbuvir was added to peginterferon alfa and ribavirin. This was supported by the relevant results from PROTON and ELECTRON, which showed that the sustained virological responses 12 weeks after the end of treatment were 90% and 100%, respectively in people with genotype 3 HCV. The Committee was aware that these results were from open‑label studies in small numbers of people, and that there was considerable uncertainty around the true magnitude of benefit of sofosbuvir plus peginterferon alfa and ribavirin in people with genotype 3 HCV. On balance, however, the Committee concluded that 12 weeks of sofosbuvir plus peginterferon alfa and ribavirin was clinically more effective than peginterferon alfa and ribavirin alone for 24 weeks in inducing a sustained virological response in people with genotype 3 treatment‑naive HCV.

4.10

The Committee considered the available evidence for sofosbuvir plus ribavirin in people co‑infected with chronic hepatitis C and HIV. It noted that the interim analysis presented in the company's original submission and the regulatory submission was from an ongoing open‑label study with sofosbuvir and ribavirin (PHOTON‑1), which included people with genotype 1, 2 or 3 HCV and HIV who had not had treatment for hepatitis C, and people with genotype 2 or 3 HCV and HIV who had been treated before. The Committee subsequently considered the evidence provided by the company during consultation from the 1910 study (Rodriguez‑Torres et al. [2013]), which compared sofosbuvir plus peginterferon alfa and ribavirin treatment with peginterferon alfa and ribavirin alone in people with genotype 1 HCV and HIV. The Committee was aware that the interim results of both studies suggested that sustained virological responses in people with HCV and HIV‑co‑infection were similar to those seen in people with HCV mono‑infection. The Committee understood that the summary of product characteristics states that people with HCV and HIV co‑infection should have the same sofosbuvir treatment schedule as people with HCV mono‑infection, and concluded that this was appropriate.

4.11

The Committee considered the adverse reactions associated with sofosbuvir plus ribavirin with and without peginterferon alfa. It noted that the adverse events reported in the main sofosbuvir clinical studies (NEUTRINO, FISSION, FUSION, POSITRON and VALENCE) were generally consistent with those reported in other studies for hepatitis C treatments. It heard from the clinical experts that sofosbuvir is considered to have a better safety profile than peginterferon alfa and ribavirin, and most adverse events reported in the trials were likely to be related to treatment with peginterferon alfa and ribavirin rather than sofosbuvir. The Committee concluded that the adverse reactions associated with sofosbuvir plus ribavirin with or without peginterferon alfa were generally tolerable and that sofosbuvir was not likely to cause additional adverse reactions compared with existing treatment regimens.

4.12

The Committee discussed the company's mixed treatment comparison. It heard from the company that a network could not be formed for all the relevant populations and a comparison could be performed only for genotypes 1, 2 and 3 treatment‑naive HCV because of data limitations. Therefore, results from the mixed treatment comparison were not used to inform the economic model. The Committee noted that instead, the company adopted what they described as a conservative approach and used trial data that reported the highest sustained virological response for the comparators, including naive comparisons with boceprevir and telaprevir plus peginterferon alfa and ribavirin for people with genotype 1 HCV. The Committee agreed with the ERG's view that the company's mixed treatment comparison was not robust. Therefore the Committee concluded that it was reasonable for the company not to use the mixed treatment comparison to inform its cost‑effectiveness analyses.

Cost effectiveness

4.13

The Committee considered the company's original economic model provided in the company's submission, the assumptions underlying the values of the parameters, and the critique and exploratory analyses carried out by the ERG. The Committee also considered the revised base‑case model submitted by the company in response to the additional analyses requested by the Committee. The Committee noted that the company's model structure differed slightly from that used in previous technology appraisals for hepatitis C, in that people with mild and moderate chronic hepatitis C were considered collectively as a population without cirrhosis, and therefore the model distinguished only between people with and without cirrhosis. The Committee heard from the clinical experts that this approach was reasonable and consistent with how people are currently diagnosed in clinical practice. It heard from the clinical experts that previously, people had invasive liver biopsies and as a result their disease was classified as mild, moderate or severe. However, current practice involves the use of less invasive diagnostic tests that do not differentiate between mild and moderate disease and can distinguish only between cirrhosis and non‑cirrhosis. The Committee also noted that the company's model incorporated the assumption that all people who had cirrhosis were candidates for liver transplant, and that pre‑transplant patients were therefore included in the modelling presented. The Committee concluded the approach taken by the company was appropriate.

4.14

The Committee acknowledged that the ICERs from the company's original economic model were for treatment for a combined cohort of people with and without cirrhosis (hereafter referred to as the 'combined cohort'). The Committee heard from clinical experts that it was standard clinical practice for people with and without cirrhosis to be considered as separate subgroups, because cirrhosis affects a person's likelihood of a sustained virological response. The Committee considered individual ICERs presented by the company for each genotype by treatment history, interferon eligibility and cirrhosis status (where available) and noted that the ICERs were consistently much lower in the subgroups of people with cirrhosis than in the subgroups of people without cirrhosis. The Committee also noted that patient numbers underpinning the clinical evidence used in the economic model were very small for the groups of people with cirrhosis, and that the sustained virological responses were in some cases as high as in people without cirrhosis. The Committee heard from clinical experts that sustained virological responses were historically lower in people with cirrhosis across all HCV genotypes than in people without cirrhosis. In addition, the Committee was aware that the summary of product characteristics states that consideration should be given to extending sofosbuvir treatment from 12 to 24 weeks in people who have 1 or more factors historically associated with lower response rates to interferon‑based therapies and that 1 of the factors listed is cirrhosis. The Committee considered that the high sustained virological responses that were generated from the small numbers of patients in the subgroup of people with cirrhosis (which resulted in very low ICERs) in each stratified subgroup should be interpreted with caution. The Committee noted that the ICERs from the combined cohort appeared artificially low, considering that most of the group would not have cirrhosis. The Committee concluded that the consideration of the cost effectiveness of sofosbuvir for each genotype should take into consideration both the combined cohort ICER and also the estimated ICERs for treatment in people with and without cirrhosis.

4.15

The Committee acknowledged that, in response to consultation, the company presented a revised base‑case model for HCV genotypes 1, 3, 4, 5 and 6 that incorporated most of its preferred assumptions (see section 3.64). The company explained and justified deviations from the Committee's preferred assumptions, which were included in the revised model.

4.16

The Committee noted that the revised model included a transition probability from the sustained virological response cirrhotic health state to the hepatocellular carcinoma health state (0.0128) using data from Cardoso et al. as requested by the Committee. In addition, the company also updated the transition probability from the health state for people with cirrhosis who have not had a sustained virological response to the hepatocellular carcinoma health state (0.0631; also from Cardoso et al.) rather than using the transition probability estimate (0.014) from Fattovich et al. (1997) that was used in the original model. The Committee heard from the clinical experts that using both transition probabilities from the Cardoso et al. study also had face validity because it would allow the modelling of a relative reduction in the probability that a patient would progress to hepatocellular carcinoma after having a sustained virological response. The Committee also heard from the clinical experts that the Cardoso et al. evidence that a person with cirrhosis who has a sustained virological response is 4 to 5 times less likely to later have hepatocellular carcinoma is consistent with the progression to hepatocellular carcinoma seen in clinical practice. The Committee noted that the Cardoso et al. transition probabilities were based on a population whose baseline characteristics were closer to the population seen in clinical practice in England. However, the Committee also heard from clinical experts that exploring alternative sources for transition probabilities, such as Fattovich et al. was appropriate. The Committee concluded that although there is significant uncertainty about the absolute reduction in the probability of progression to hepatocellular carcinoma between the sustained virological response with cirrhosis health state and the health state of cirrhosis without a sustained virological response, Cardoso et al. was an acceptable source for transition probabilities for the company's revised base‑case model. However, the Committee also concluded that it was plausible that the transition probability for people without a sustained virological response may lie somewhere between the Cardoso et al. and Fattovich et al. estimates.

4.17

The Committee considered the impact of using alternative sustained virological responses for peginterferon alfa and ribavirin in genotype 1 HCV on the results from the revised economic model. The Committee noted that the company preferred the sustained virological responses for peginterferon alfa and ribavirin treatment in people with genotype 1 HCV from McHutchison et al. because it was a larger study and the baseline characteristics of patients were better matched to the patients in the pivotal NEUTRINO trial. It also noted that the ERG considered the estimates from Hadziyannis et al. (2004) to be more relevant because they were most generalisable to patients with HCV in England. This was because the study included people with the genotypes most relevant to the UK population, that is, genotypes 1 and 3 HCV. The Committee heard from the clinical experts that there is a wide variation in sustained virological response in clinical practice and that the baseline characteristics of patients included in each study differed. This had an impact on the absolute sustained virological responses in these studies. The clinical experts noted that it was important to consider a range of alternative sustained virological responses from the evidence base rather than arbitrarily choosing a single rate from a particular study. The Committee noted that the sensitivity analyses subsequently presented by the company showed that the ICERs for sofosbuvir plus peginterferon alfa and ribavirin compared with peginterferon alfa and ribavirin alone in people with genotype 1 treatment‑naive HCV were £25,000 per QALY gained using the estimates from Hadziyannis et al. compared with £17,500 per QALY gained using the estimates from McHutchison et al. On balance, the Committee concluded that the sustained virological responses from McHutchison et al. were an acceptable source for including in its base‑case model, but noted that the sustained virological responses could lie between those provided by the McHutchison and Hadziyannis data sets.

4.18

The Committee considered the cost of ribavirin used in the company's model. The Committee noted that the company used the cost of ribavirin from the BNF (June 2013) in the original model (Copegus; £246.65) and asked that the company explore the impact of using the price of generic ribavirin paid by the NHS (£42.05 based on the Department of Health Commercial Medicines Unit Electronic market information tool) which is available nationally through contracts negotiated by the NHS Commercial Medicines Unit. In response to consultation the company included the generic cost of ribavirin in its revised base‑case analysis, but noted that the Medicines and Healthcare products Regulatory Agency stated that generic ribavirin should only be used in combination with interferon alfa‑2b, which only has 3% of the market share in the UK. The Committee concluded that the generic cost of ribavirin had a small effect on the ICER as demonstrated by the ERG analysis, but that sensitivity analyses around the generic costs of comparator treatment were appropriate.

4.19

The Committee discussed the utility values used in the company's model. It acknowledged that health‑related quality of life was largely assessed in the clinical trials for sofosbuvir using the SF‑36 questionnaire and that none of the clinical trials collected data using the EQ‑5D quality‑of‑life measure. The Committee understood that the company obtained SF‑36 health‑related quality‑of‑life data at various time points, including 24 weeks after the end of treatment in some trials. The Committee appreciated that the company tried to be pragmatic in its approach to modelling the effects of treatment by applying a utility increment of 0.05 (from Wright et al.) after sustained virological response in the company's base‑case analysis. However, it asked that the company present a revised base‑case model that explored the use of different utility estimates including more up‑to‑date estimates from the literature such as Vera‑Llonch et al. (2013) and estimates from the pivotal clinical trials. The Committee noted that the company stated it was unable to incorporate the estimates from the pivotal clinical trials because the data were not available, but provided a revised base‑case model incorporating an alternative utility increment (0.041; Vera‑Llonch et al.) after a sustained virological response. The Committee noted that using this utility increment increased the company's base‑case ICERs slightly. The Committee concluded that although alternative utility estimates from the pivotal studies would have been preferred, using the utility increment from Vera‑Llonch et al. in its revised base case was acceptable.

4.20

The Committee discussed the discount rate used in the company's model and considered whether this appraisal met the criteria for using a non‑reference case discount rate for costs and health benefits that can be applied in situations when treatment restores people who would otherwise die or have a very severely impaired life to full or near full health, and when this is sustained over a very long period (normally at least 30 years), as described in NICE's guide to the methods of technology appraisal. The Committee noted that the company's base‑case analysis used a discount rate of 3.5% for costs and health benefits in line with the NICE reference case and that the deterministic sensitivity analysis presented by the company suggested that the ICERs were particularly sensitive to the discount rate used. The Committee heard from the clinical experts that a person who does not have cirrhosis and has a sustained virological response could be considered cured. However, the Committee was aware that no data are available beyond the follow‑up period from the trials; therefore evidence supporting the long‑term durability of a sustained virological response is lacking. The Committee also noted that people with cirrhosis who experience a sustained virological response would not have their health fully restored. Therefore, the Committee concluded that sofosbuvir did not meet the criteria for using non‑reference case discount rates, and agreed that the company's approach to using the standard discount rate of 3.5% was appropriate.

4.21

The Committee considered whether the cost effectiveness of sofosbuvir for treating hepatitis C was better assessed for the population as a whole (that is, using the 'global' ICERs presented by the company in response to consultation, which are weighted by genotype, treatment history and the presence or absence or cirrhosis, see section 3.66) or separately for each genotype. The Committee was unconvinced by the global ICER approach put forward by the company because the evidence from the clinical experts suggested that in clinical practice treatment is stratified by genotype, treatment history and other characteristics, including cirrhosis status. This is because the capacity to benefit from treatment for chronic hepatitis C differs depending on the patient's characteristics. The Committee therefore concluded that it was more appropriate to consider the clinical and cost effectiveness for each relevant subgroup of patients separately in the company's base‑case analyses.

Genotype 1

Treatment-naive, interferon eligible

4.22

The Committee considered the cost effectiveness of sofosbuvir plus peginterferon alfa and ribavirin for people with genotype 1 treatment‑naive HCV who are eligible for interferon treatment. The Committee noted that the ICER for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa and ribavirin alone for 48 weeks was less than £17,500 per QALY gained. The Committee noted that the ICERs for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with response‑guided treatment with boceprevir plus peginterferon alfa and ribavirin and telaprevir plus peginterferon alfa and ribavirin were £10,300 and £15,400 per QALY gained, respectively. The Committee noted that when stratified by the presence or absence of cirrhosis, the ICERs for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa and ribavirin for 48 weeks were £5,400 and £25,200 per QALY gained, respectively. The ICERs for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with boceprevir plus peginterferon alfa, when stratified by the presence or absence of cirrhosis, were £2,800 and £14,300 per QALY gained, respectively. The ICERs for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with telaprevir plus peginterferon alfa, when stratified by the presence or absence of cirrhosis, were £4,200 and £22,300 per QALY gained, respectively. The Committee also considered the ERG's exploratory analyses (see section 3.76). The ERG's resulting ICER for the combined cohort for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa and ribavirin alone for 48 weeks was just over £30,000 per QALY gained. The Committee believed that this ICER represented the upper limit of what could be considered to be plausible, but that the ICERs for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with response‑guided boceprevir and telaprevir plus peginterferon alfa and ribavirin treatment, which are the standard of care in the NHS, were £12,200 and £18,700 per QALY gained. The Committee concluded that sofosbuvir plus peginterferon alfa and ribavirin was cost effective for people with treatment‑naive genotype 1 HCV.

Treatment-experienced, interferon eligible

4.23

The Committee considered the cost effectiveness of sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with standard of care in people with genotype 1 treatment‑experienced HCV for whom interferon is suitable. The Committee acknowledged the uncertainty in the ICER for the population who have treatment‑experienced HCV in the light of the lack of clinical evidence, but noted that there are very few treatment options for these patients, who have a high unmet need. The Committee noted that the estimate of sustained virological response in the treatment‑experienced population provided by the company was accepted by the European Medicines Agency and clinical experts. The Committee noted that the ICERs for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa and ribavirin for 48 weeks (£12,600 per QALY gained), boceprevir plus peginterferon alfa and ribavirin (£700 per QALY gained), and telaprevir plus peginterferon alfa and ribavirin (£8,200 per QALY gained) for the combined cohort of people with and without cirrhosis could be considered cost effective although ICERs stratified by cirrhosis status were not available. The Committee considered that if the relative proportion of people with and without cirrhosis was similar to that observed in the group with treatment‑naive HCV, then it would be likely that the stratified ICERs for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa and ribavirin alone for 48 weeks would be cost effective for people with and without cirrhosis even when taking into account the assumptions in the ERG's exploratory analyses, which would increase the ICERs further. The Committee also noted that the stratified ICERs for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with response‑guided boceprevir and telaprevir plus peginterferon alfa and ribavirin treatment would be even lower for these groups, and that these 2 treatment regimens are the standard of care in the NHS. The Committee concluded that sofosbuvir plus peginterferon alfa and ribavirin is a cost‑effective treatment option for people with genotype 1 treatment‑experienced HCV who are eligible for interferon treatment.

Treatment-naive, interferon ineligible

4.24

The Committee considered the cost effectiveness of sofosbuvir plus ribavirin for 24 weeks compared with standard of care (no treatment) in people with genotype 1 treatment‑naive HCV for whom interferon is unsuitable. It noted that the ICER for sofosbuvir and ribavirin compared with no treatment for this population was £47,600 per QALY gained. In response to consultation, the company stated that although it is necessary to have options for this subgroup of patients for whom interferon treatment is unsuitable and who have a high unmet need, it is anticipated that the number of people in this group having 24 weeks of sofosbuvir plus ribavirin would be extremely low. The Committee also heard from the company that it was not expecting people with genotype 1 HCV who are interferon eligible to be given the option of the 24 week interferon‑free sofosbuvir regimen. The Committee concluded that although the number of people with genotype 1 treatment‑naive HCV for whom interferon is unsuitable is potentially small, the high ICER for sofosbuvir plus ribavirin alone compared with no treatment for this population does not represent a cost‑effective use of NHS resources and could not be recommended.

Treatment-experienced, intolerant to or ineligible for interferon treatment

4.25

The Committee considered the lack of evidence for sofosbuvir plus ribavirin for 24 weeks compared with the standard of care (no treatment) in the subgroup of people with genotype 1 treatment‑experienced HCV who are intolerant to or ineligible for interferon treatment. However, considering the Committee had accepted the ICERs generated using the sustained virological responses recognised by the US Food and Drug Administration (FDA) for the genotype 1 treatment‑experienced HCV population who are eligible for interferon, the Committee took a pragmatic view on how to establish an estimated ICER for this population. The starting point for the Committee was the ICER of £47,600 per QALY gained (that is the ICER for people with genotype 1 treatment‑naive HCV, for whom interferon is unsuitable). Assuming that the relative difference between the ICERs in the treatment‑naive and treatment‑experienced HCV groups seen in other genotypes also applies to genotype 1 HCV, the Committee would expect that the ICERs for the genotype 1 treatment‑experienced HCV group would likely be slightly lower than the ICER for people in the genotype 1 treatment‑naive HCV group. When stratified by the presence or absence of cirrhosis, the ICERs would be likely to increase in the subgroup without cirrhosis and decrease in the subgroup with cirrhosis, in a similar proportion to that seen in the subgroup of people with treatment‑naive genotype 1 HCV for whom interferon is unsuitable. However, the ICERs would still remain high. The Committee noted that if the assumptions used in the ERG's exploratory analyses were applied, the ICERs would increase in the combined cohort as well as in the subgroups with and without cirrhosis. The Committee was aware that people with genotype 1 treatment‑experienced HCV for whom interferon is unsuitable are a group with a high unmet need. However, the Committee concluded that based on the very uncertain evidence presented and the high ICERs, treatment with sofosbuvir plus ribavirin for 24 weeks does not represent a cost‑effective use of NHS resources for people with genotype 1 treatment‑experienced HCV who are intolerant to or ineligible for interferon treatment and therefore could not be recommended in this group.

Genotype 2

4.26

The Committee considered the cost effectiveness of sofosbuvir plus ribavirin compared with peginterferon alfa and ribavirin for 24 weeks in people with genotype 2 HCV who are eligible for interferon treatment, or no treatment in people who are intolerant or ineligible for treatment with interferon. The Committee noted that sofosbuvir plus peginterferon alfa and ribavirin does not have a marketing authorisation for treating genotype 2 HCV. The Committee noted that the ICER from the company's original base‑case model for sofosbuvir and ribavirin compared with peginterferon alfa and ribavirin alone was approximately £46,300 per QALY gained in people who are eligible for treatment with interferon and who have treatment‑naive HCV, and £12,500 per QALY gained in people who have treatment‑experienced HCV and are eligible for treatment with interferon. The ICER for sofosbuvir and ribavirin compared with no treatment for people who are intolerant to or ineligible for interferon was £8,200 per QALY gained for people with treatment‑naive HCV, and £8,600 per QALY gained for people with treatment‑experienced HCV. The Committee concluded that sofosbuvir plus ribavirin was not a cost‑effective use of NHS resources in adults with genotype 2 treatment‑naive HCV who are eligible for treatment with interferon. However, the Committee concluded that sofosbuvir plus ribavirin was a cost‑effective use of NHS resources for adults with genotype 2 HCV who are eligible for treatment with interferon and who have had treatment for HCV, and for adults with genotype 2 HCV who are intolerant or ineligible for interferon treatment, regardless of their treatment history.

Genotype 3

Treatment-naive, interferon eligible

4.27

The Committee considered the cost effectiveness of sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa and ribavirin for 48 weeks in people with genotype 3 treatment‑naive HCV who are eligible for treatment with interferon. The Committee noted from the company's revised base case that the combined cohort ICER (with and without cirrhosis) for sofosbuvir plus peginterferon alfa and ribavirin compared with peginterferon alfa and ribavirin alone for this population was approximately £21,900 per QALY gained. The Committee noted that when stratified by cirrhosis status, the ICER for people with treatment‑naive HCV without cirrhosis who are eligible for treatment with interferon was approximately £40,600 per QALY gained, whereas the ICER for people with cirrhosis was approximately £6,600 per QALY gained. The Committee noted that the ICERs for the subgroups of patients with or without cirrhosis were highly uncertain due to the small patient numbers included in the studies. The Committee noted that the effect of using the combined cohort analysis, which includes a larger subgroup without cirrhosis and a small subgroup with cirrhosis, resulted in a combined cohort ICER that was artificially low (£21,900 per QALY gained). The Committee considered that despite this uncertainty there was more confidence around the ICER for the subgroup with cirrhosis because the treatment remained cost effective despite using a variety of assumptions including those suggested by the ERG in its exploratory analyses. The Committee also acknowledged that people with cirrhosis are in greater need of treatment than those without cirrhosis. Therefore, the Committee concluded that sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks could be considered a cost‑effective use of NHS resources in people with genotype 3 treatment‑naive HCV who are eligible for interferon treatment and who have cirrhosis, but it was not a cost‑effective use of NHS resources in people who do not have cirrhosis.

4.28

The Committee noted that the marketing authorisation for sofosbuvir allows 24 weeks dual therapy with sofosbuvir plus ribavirin as an alternative to 12 weeks with sofosbuvir plus peginterferon alfa and ribavirin for people with genotype 3 HCV who are eligible for treatment with interferon. The Committee considered the exploratory analyses carried out by the company, modelling the effect on the revised base‑case ICERs of varying the proportion of people with genotype 3 HCV receiving sofosbuvir plus ribavirin for 24 weeks in people eligible for treatment with interferon (see section 3.75). The Committee noted that the scenario analysis presented by the company, in which 100% of people with treatment‑naive genotype 3 HCV received sofosbuvir plus ribavirin for 24 weeks compared with peginterferon plus ribavirin for 24 weeks, resulted in an ICER of £47,000 per QALY gained. The ICER for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks treatment compared with peginterferon and ribavirin treatment for 24 weeks increased from approximately £21,900 per QALY gained to approximately £22,400 and £27,100 per QALY gained when it was assumed that 2% and 20% had sofosbuvir and ribavirin for 24 weeks treatment in the population with treatment‑naive HCV who are eligible for interferon. The Committee noted that when stratified by cirrhosis status, the revised base‑case ICER for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks treatment compared with peginterferon and ribavirin treatment for 24 weeks increased in the subgroup without cirrhosis to £41,700 and £51,300 per QALY gained when it was assumed that 2% and 20% had sofosbuvir and ribavirin in the treatment‑naive population. In the subgroup with cirrhosis, the ICER increased to £6,800 and £8,400 per QALY gained using the same assumptions. The Committee concluded that the duration of treatment with sofosbuvir had a considerable effect on the ICERs in people with genotype 3 HCV, although it heard from the company, clinical experts and commissioners that sofosbuvir and ribavirin treatment for 24 weeks would only be appropriate for people ineligible for interferon therapy.

Treatment-experienced, interferon eligible

4.29

The Committee considered the cost effectiveness of sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa and ribavirin for 48 weeks in people with genotype 3 treatment‑experienced HCV who are eligible for treatment with interferon. The Committee noted that the company's revised base‑case ICER for the combined cohort was £13,900 per QALY gained. When stratified by cirrhosis status, the ICER for people without cirrhosis was £18,600 per QALY gained, whereas the ICER for people with cirrhosis was £6,300 per QALY gained. The Committee was aware that these ICERs were also uncertain, due to small patient numbers included in the studies, and that sustained virological responses were identical for people in the subgroups with and without cirrhosis, which is clinically unlikely due to the poorer sustained virological responses usually seen in people with cirrhosis. The Committee was willing to accept this uncertainty because the ICERs were within the range it could consider a technology might be cost‑effective in the group without cirrhosis and even lower in the group with cirrhosis. The ICERs remained in this range when the ERG's exploratory assumptions were used. The Committee acknowledged that this subgroup also has no further treatment options and can be considered to have a high unmet need. The Committee therefore concluded that sofosbuvir plus peginterferon alfa and ribavirin was a cost‑effective use of NHS resources in people with genotype 3 treatment‑experienced HCV who were eligible for interferon treatment.

4.30

As with the treatment‑naive, interferon‑eligible group with genotype 3 HCV, the Committee considered the exploratory analyses carried out by the company that modelled the effect on the revised base‑case ICERs of increasing the proportion of people with treatment‑experienced genotype 3 HCV receiving sofosbuvir plus ribavirin for 24 weeks (see section 3.75). The ICER for sofosbuvir plus ribavirin for 24 weeks compared with peginterferon plus ribavirin for 48 weeks in people with treatment‑experienced genotype 3 HCV who are eligible for interferon treatment increased to £48,300 per QALY gained. The ICER for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa and ribavirin for 48 weeks also increased for the population for whom interferon therapy is suitable and who had treatment‑experienced HCV, from approximately £13,900 per QALY gained to approximately £14,500 and £19,900 per QALY gained when it was assumed that 2% and 20% had sofosbuvir and ribavirin. The Committee concluded that the duration of treatment with sofosbuvir had a considerable effect on the ICERs in people with genotype 3 HCV, and it agreed with the company, clinical experts and commissioners that sofosbuvir and ribavirin treatment for 24 weeks would only be appropriate for people ineligible for interferon therapy.

Treatment-naive, interferon ineligible

4.31

The Committee considered the cost effectiveness of sofosbuvir and ribavirin for 24 weeks compared with no treatment in people with genotype 3 treatment‑naive HCV who are ineligible for treatment with interferon. The Committee noted that the company's revised base‑case ICER for this population was £21,000 per QALY gained, which was calculated based on sustained virological responses seen in VALENCE. The Committee noted that the VALENCE study was unblinded when treatment was extended for all people with genotype 3 HCV. Therefore it was of poor quality and open to potential bias. The Committee noted that when the population was stratified by cirrhosis status, the ICER for sofosbuvir plus ribavirin was £28,000 per QALY gained for people without cirrhosis (which increased to £32,000 per QALY gained using the ERG assumptions) and £10,500 per QALY gained for people with cirrhosis. The Committee concluded that given the uncertainty around the ICER in the group without cirrhosis and the possibility that the ICER may be over £32,000 per QALY gained, it could not recommend sofosbuvir plus ribavirin treatment in people with genotype 3 treatment‑naive HCV without cirrhosis who are ineligible for interferon treatment. Because the ICER in the subgroup of people with cirrhosis remained low (£15,100 per QALY gained), even when using the ERG's exploratory assumptions, the Committee concluded that sofosbuvir plus ribavirin is cost effective for people with genotype 3 treatment‑naive HCV who have cirrhosis.

Treatment-experienced, interferon ineligible

4.32

The Committee considered the cost effectiveness of sofosbuvir and ribavirin for 24 weeks compared with no treatment in people with treatment‑experienced genotype 3 HCV who are intolerant to or ineligible for treatment with interferon. The Committee considered that this group would represent a very small number of patients in the NHS. The Committee considered the company's revised base‑case ICER of approximately £27,500 per QALY gained for the combined cohort of people with and without cirrhosis. The Committee noted that this ICER was also based on the sustained virological response rates observed in VALENCE, a study that the Committee considered to be of low quality and open to potential bias (see sections 3.16 and 4.32). When the ICERs were stratified by cirrhosis status, the company's revised base‑case ICER for the subgroup without cirrhosis was £31,400 per QALY gained (£35,000 per QALY gained using the assumptions from the ERG exploratory analyses). Due to the uncertainty around the ICER and the possibility that the ICER was over £35,000 per QALY gained, the Committee concluded that sofosbuvir plus ribavirin was not a cost‑effective use of NHS resources in people with treatment‑experienced genotype 3 HCV without cirrhosis who are intolerant to or ineligible for interferon treatment. The Committee noted that the company's revised base‑case ICER for sofosbuvir plus ribavirin for 24 weeks compared with no treatment for people with cirrhosis was £19,200 per QALY gained (£29,700 per QALY gained when using the assumptions from the ERG exploratory analyses). The Committee considered the high unmet need of this subgroup for whom there are currently no other licensed treatment options. The Committee recognised the uncertainty in the evidence base for people with treatment‑experienced genotype 3 HCV who have cirrhosis and are intolerant to or ineligible for interferon. However, on balance, it concluded that it would be consistent with its other recommendations for people with genotype 3 HCV to recommend sofosbuvir plus ribavirin for 24 weeks for people with cirrhosis and that this could be considered a cost‑effective use of NHS resources, because the true ICER was likely to be between the company's revised base case and the ERG's exploratory estimates.

Genotypes 4, 5 and 6

Treatment-naive, interferon eligible

4.33

The Committee considered the company's original base‑case ICER for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa and ribavirin for 48 weeks of approximately £26,800 per QALY gained in people with genotype 4, 5 or 6 treatment‑naive HCV, for whom interferon is suitable. The Committee noted that the ICER was based on a naive comparison of the sustained virological responses 12 weeks after the end of treatment for sofosbuvir plus peginterferon alfa‑2a and ribavirin observed in NEUTRINO and the sustained virological responses 24 weeks after the end of treatment with peginterferon alfa‑2b and ribavirin observed in Manns et al. (2001). The Committee heard from clinical experts and the company that peginterferon alfa 2a and peginterferon alfa‑2b were assumed to be equally efficacious and that sustained virological response 24 weeks after the end of treatment was essentially equivalent to sustained virological response 12 weeks after the end of treatment. The Committee noted that of the 35 people with genotype 4, 5 or 6 in the NEUTRINO study, 100% of the 33 people without cirrhosis achieved a sustained virological response compared with 50% of the 2 people with cirrhosis. However in Manns et al., the subgroup without cirrhosis had a sustained virological response of 50% (as calculated by the company, based on the relative difference in sustained virological response between people without cirrhosis and people with cirrhosis in the studies in genotype 1, 2 and 3 HCV) and the subgroup with cirrhosis had a sustained virological response of 38.6%. The Committee noted that the difference in sustained virological responses between the sofosbuvir plus peginterferon alfa and ribavirin arm and the peginterferon alfa plus ribavirin alone arm of the model was a key driver of the ICER.

4.34

During consultation, the company presented the Committee with additional analyses for people with genotype 4, 5 or 6 HCV, which included alternative sustained virological responses for peginterferon alfa plus ribavirin from studies identified in a systematic review of studies in genotypes 4, 5 and 6 HCV and the impact of using different sustained virological responses on the ICER. The Committee noted that the studies were exclusively European (because the company considered the patient characteristics to be more relevant to patients in the UK) whereas the ERG had considered studies from the Middle East and Egypt to be an important source of data because these studies included larger numbers of patients. The Committee noted that the sustained virological responses for peginterferon alfa plus ribavirin in the European studies ranged from 33% (Zeuzem et al. [2005]) to 77% (Fried et al. [2002]), which spanned the range of sustained virological responses seen in the studies identified by the ERG. The Committee considered the revised base‑case ICER of £27,500 per QALY gained presented by the company for sofosbuvir plus peginterferon alfa and ribavirin compared with peginterferon alfa and ribavirin (using the sustained virological responses from Manns et al. [2001] for the latter) in people with genotype 4, 5 or 6 HCV. The Committee noted that the company varied the sustained virological response for peginterferon alfa and ribavirin using Zeuzem et al. and Fried et al. but noted that sustained virological responses were not available by cirrhosis status. Therefore the company applied the same sustained virological response for people with and without cirrhosis. The ICERs for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa plus ribavirin for 48 weeks using Zeuzem et al. and Fried et al. were £19,148 and £244,387 per QALY gained, respectively. The Committee considered that a sustained virological response of 77% for 48 weeks of treatment with peginterferon alfa and ribavirin was improbable. The Committee considered the ERG's exploratory analyses, in which the sustained virological responses in the peginterferon alfa and ribavirin arm were based on a weighted average of the responses reported in the individual studies. The ERG used a sustained virological response of 54.8% from the European studies (which included the studies by Manns et al. and Lindsay et al.) which led to an ICER of approximately £39,100 per QALY gained. The Committee considered this to be the most relevant ICER because it was based on studies with populations that were most similar to patients in England and was generated using the Committee's preferred assumptions (see section 4.15). The Committee noted that the sustained virological responses for peginterferon alfa plus ribavirin for 48 weeks in people with genotype 4, 5 or 6 HCV were higher than those reported for people with genotype 1 HCV, providing indirect evidence that people with genotype 4 HCV are not more difficult to treat. The Committee concluded that the ICER of £39,100 per QALY gained for sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks compared with peginterferon alfa plus ribavirin for 48 weeks using the ERG's calculated sustained virological response was the most plausible, although there remained considerable uncertainty about the ICER, because of the small number of people included in the NEUTRINO study.

4.35

The Committee considered comments received during consultation that recommending sofosbuvir plus peginterferon alfa and ribavirin only for a proportion of people with genotype 1, 2 or 3 HCV, but not for anyone with genotype 4, 5 or 6 HCV could potentially be interpreted as indirect discrimination. It heard from consultees that this was because a larger proportion of minority ethnic groups, people with HIV co‑infection and haemophilia are represented in the genotype 4, 5 and 6 HCV population. In light of NICE's legal obligation to promote equality, the Committee considered the additional evidence provided by the company that included family origin by HCV genotype, and the prevalence of HIV and HCV co‑infection and HCV infection in people with haemophilia. The Committee noted that the family origin evidence was self‑reported (and could therefore not be verified), and used broad categories. The Committee therefore considered this evidence to be uncertain, although it noted the anecdotal evidence provided by other consultees that minority ethnic groups are more highly represented in the genotype 4, 5 and 6 HCV population. The Committee considered the commercial‑in‑confidence evidence presented by the company about the genotype distribution of HCV in people with HCV and HIV co‑infection and agreed that a disproportionate number of people had genotype 4 HCV and HIV co‑infection compared with the overall population of people with HCV in England. The Committee noted that the evidence presented by the company suggested that 96% of people with haemophilia and HCV had genotype 1, 2 or 3 HCV, and 4% had genotype 4 or 5 because no patients were identified with genotype 6 HCV and haemophilia. The Committee noted that the distribution of HCV genotypes in people with haemophilia presented by the company was actually similar to the overall population of people with HCV in England. The Committee concluded that there did not appear to be a disproportionate percentage of people with haemophilia who had genotype 4 HCV in England. The Committee noted that the ICERs for sofosbuvir for people with genotype 4, 5 or 6 HCV for the combined cohort (people with and without cirrhosis) were very high. However, it agreed that, in the light of evidence on the higher representation of minority ethnic groups and HIV co‑infection in these genotypes, further consideration should be given to whether anything could be done to remove or reduce the disproportionate impact for the protected groups.

4.36

The Committee noted that unlike genotype 1 HCV, people with genotype 4, 5 or 6 HCV currently only have peginterferon alfa and ribavirin for 48 weeks as a treatment option. The Committee considered that the people with the highest unmet need within this population are those with cirrhosis, because their disease is less likely to respond to treatment with peginterferon alfa and ribavirin for 48 weeks. Although the sustained virological response seen in people with genotype 4, 5 or 6 HCV with cirrhosis was 50% in NEUTRINO, the Committee noted that this was based on 2 patients; 1 who had a sustained virological response and 1 who did not. As with some of the other genotypes, the Committee used a pragmatic approach in estimating an ICER for sofosbuvir for the group of people with genotype 4, 5 or 6 with cirrhosis. Using the starting point for the ICER (calculated by the ERG) as £39,100 per QALY gained, the Committee considered whether the ICER for genotypes 4, 5 and 6 responded in a similar manner as for other genotypes, that is, whether it would be significantly lower for treatment in people with cirrhosis than in people without cirrhosis. The Committee considered that it is plausible that the ICER for treatment in people with genotypes 4, 5 or 6 treatment‑naive HCV with cirrhosis could be within the range that is normally accepted as being cost effective, that is between £20,000 and £30,000 per QALY gained, and that the ICER for treatment in people with genotypes 4, 5 or 6 treatment‑naive HCV without cirrhosis is likely to be greatly in excess of the £39,100 per QALY gained estimated by the ERG for the combined cohort. Therefore, taking into consideration the potential equality issues raised about genotypes 4, 5 and 6 HCV, the high unmet need and the lack of treatment options for people with cirrhosis, the Committee considered it was reasonable to conclude that sofosbuvir plus peginterferon alfa and ribavirin for treating people with genotype 4, 5 or 6 treatment‑naive HCV who have cirrhosis was a cost‑effective use of NHS resources.

Treatment-experienced, interferon eligible

4.37

The Committee considered cost‑effectiveness evidence presented for people with genotypes 4, 5 or 6 treatment‑experienced HCV, for whom interferon is suitable. The Committee noted that the company did not provide an ICER for treatment in this group and that an estimate for this ICER could only be based on the ICER in people who are eligible for interferon who have not had treatment before. The Committee acknowledged that there is even more uncertainty in the ICER for treatment in the population who have treatment‑experienced HCV in the light of the lack of clinical evidence, but noted that there are very few treatment options for these patients, who have an even higher unmet need than people who have never been treated before. Considering the uncertainty in the evidence, but also taking into consideration the potential equality issues raised for people with genotype 4, 5 or 6 HCV, the Committee took a pragmatic view on how to establish an estimated ICER for this population. The Committee noted that sofosbuvir plus peginterferon alfa and ribavirin was recommended for genotype 1 treatment‑experienced HCV on the basis of unmet need and approval from the FDA based on a sustained virological response that was calculated from the population who had not had treatment before. Therefore, the Committee concluded that a similar approach could be used for people with genotype 4, 5 and 6 treatment‑experienced HCV. Although sustained virological responses are typically lower in the treatment‑experienced populations, the costs associated with disease progression and a comparison with no treatment mean that the ICERs for sofosbuvir in treatment‑experienced people are consistently lower than the ICERs for sofosbuvir in people who have not had treatment. Using the starting point for the ICER (calculated by the ERG) as £39,100 per QALY gained for people with genotype 4, 5 or 6 treatment‑naive HCV, the Committee considered that it was plausible that the ICER in genotypes 4, 5 or 6 treatment‑experienced HCV with cirrhosis could also be within the range that is normally accepted as being cost effective, that is between £20,000 and £30,000 per QALY gained. The Committee concluded that for all these reasons, sofosbuvir plus peginterferon alfa and ribavirin for treating people with genotype 4, 5 or 6 treatment‑experienced HCV who have cirrhosis could be considered a cost‑effective use of NHS resources.

Interferon unsuitable

4.38

The Committee considered the group of people with genotype 4, 5 or 6 HCV, for whom interferon therapy is unsuitable. The Committee noted that the company did not provide an ICER for sofosbuvir plus ribavirin in this population. The starting point for the Committee was the only cost‑effectiveness evidence provided for this population, namely the ICER of £39,100 per QALY gained for people with genotype 4, 5 or 6 HCV, for whom interferon therapy is suitable. The Committee noted that the ICER for sofosbuvir plus ribavirin compared with no treatment in people with treatment‑naive genotype 1 HCV who were not eligible for interferon was more than double the ICER for sofosbuvir plus peginterferon and ribavirin compared with peginterferon and ribavirin in people who were interferon eligible. The Committee anticipated that the ICERs for sofosbuvir plus ribavirin in people with genotype 4, 5 or 6 HCV, for whom interferon therapy is unsuitable, would increase significantly due to the fact that treatment would be offered for 24 weeks instead of 12 weeks and this was likely to increase further using the ERG's exploratory assumptions. The Committee considered that, based on the lack of evidence provided for these genotypes, it was necessary to make a value judgment. Although people with genotype 4, 5 or 6 HCV represent a small proportion of the total HCV population in England, it is still an important group with a high unmet need. The Committee reflected on the quality of evidence and the level of uncertainty, in addition to the very high costs associated with sofosbuvir plus ribavirin treatment for 24 weeks. The Committee concluded that treatment with sofosbuvir plus ribavirin for 24 weeks was not a cost‑effective use of NHS resources and could not be recommended for adults with genotype 4, 5 or 6 HCV, for whom interferon therapy is not suitable.

4.39

The Committee noted the company presented separate economic analyses for people co‑infected with HCV and HIV based on interim results from the PHOTON‑1 study and the 1910 study. The Committee was aware the PHOTON‑1 study provided results for people with genotypes 1, 2 or 3 HCV treated for 12 or 24 weeks with sofosbuvir and ribavirin and the 1910 study provided results for people with genotypes 1 or 3 HCV treated with sofosbuvir plus peginterferon alfa and ribavirin for 12 weeks. The Committee was aware that, other than incorporating higher transition probabilities from the non‑cirrhotic to the compensated cirrhosis state, the modelling did not differ for the mono‑infected and co‑infected populations. The Committee noted the ERG comment that there were differences in patient characteristics and outcomes that were not taken into account in the company's model. On balance, the Committee concluded that, based on the evidence presented and considered for this population, it was reasonable to include the group of people co‑infected with HCV and HIV in the recommendations for the mono‑infected group. However, the Committee agreed with the ERG that there were legitimate concerns about the modelling for the HIV and HCV co‑infected group, and that future economic analyses should be presented separately for this population.

4.40

The Committee considered the concern expressed in comments received during consultation that some inexperienced clinicians may want to offer sofosbuvir and ribavirin for 24 weeks to people with genotypes 1, 4, 5 or 6 HCV who are interferon eligible in order to avoid the possible adverse effects associated with interferon treatment. The Committee heard from the company that the sustained virological responses with sofosbuvir plus peginterferon alfa and ribavirin in these genotypes are superior to those achieved using sofosbuvir and ribavirin alone. It also heard from company representatives that sofosbuvir plus ribavirin alone was only licensed for people in urgent need of care with genotype 1, 4, 5 or 6 HCV in whom interferon was contraindicated (as described in the summary of product characteristics) or whose disease did not have an adequate response to interferon treatment. The company representatives agreed that sofosbuvir and ribavirin would be regarded as a second‑line option and the decision to use dual therapy should only be made by clinicians experienced in treating hepatitis C, preferably after discussion by a multidisciplinary team. However, the Committee concluded that the concerns expressed during consultation were no longer relevant in light of the Committee's decision that sofosbuvir, in combination with ribavirin, should not be recommended for treating adults with genotype 1, 4, 5 or 6 chronic hepatitis C who are interferon intolerant or ineligible (see sections 1.1 and 4.25, 4.26 and 4.39).

4.41

The Committee noted a comment received from a member of the public during the second consultation stating that it was potentially more cost effective to treat hepatitis C in people with haemophilia than people without haemophilia due to the large expense associated with treating haemophilia and the additional expenses due to monitoring liver damage in that group. The Committee noted that the clinical trials excluded patients with haemophilia and no clinical evidence or cost‑effectiveness analysis had been presented specifically for people with haemophilia and HCV. Therefore the Committee concluded that no evidence‑based decision or modelling would be possible, and therefore no separate recommendation could be made specifically for this patient group.

4.42

The Committee discussed comments from the patient experts indicating that in practice the availability of treatment for people with chronic hepatitis C who use injectable drugs was limited, which could represent a potential equality consideration. The Committee heard from the clinical experts that treatment for these people is considered on an individual basis because of concerns about safety and treatment adherence, but that clinicians would like to offer sofosbuvir to people using injectable drugs, taking into account any precautions in the summary of product characteristics. The Committee acknowledged that access to treatment for this patient group was an issue related to implementation and could not be addressed through technology appraisal recommendations. However, the Committee concluded that although people who use injectable drugs were not represented in the pivotal clinical trials for sofosbuvir, based on the current evidence available, there was no reason to deny them access to treatment; therefore any recommendations on the use of sofosbuvir would be irrespective of injectable drug use.

4.43

The Committee discussed whether sofosbuvir could be considered an innovative treatment, providing a step change in the treatment of chronic hepatitis C. The Committee agreed that sofosbuvir offers the possibility of shortened interferon‑based treatment regimens, or treatment without interferon therapy in some circumstances, which is particularly important and a major development in the current clinical management of chronic hepatitis C. The Committee therefore accepted that sofosbuvir is a valuable new therapy for treating chronic hepatitis C. The Committee agreed with the clinical experts and patient experts that there were other benefits to patients (such as relief of loss of cognitive ability in people with HCV) and public health benefits (such as reduced transmission of HCV) that were not captured in the QALY calculation and that, if taken into account, would decrease the ICERs.

4.44

During consultation, although a number of consultees noted the urgent need for guidance on the use of sofosbuvir, a comment was also received from NHS England, who is currently responsible for the commissioning of hepatitis C treatment, that it would not be possible to implement the recommendations in this guidance within 3 months. The Committee were in agreement that there may be increased demand for treatment following positive recommendations, but were not presented with evidence on the likely magnitude of this and considered that some patients may prefer to wait for NICE guidance on other interferon‑free treatments for chronic hepatitis C before they seek treatment. The Committee highlighted that it would be reasonable for NICE to reflect on whether the standard 3‑month implementation period is appropriate. The Committee noted that consultees were interested in the relative cost effectiveness of sofosbuvir compared with other agents in guidance development (although this did not fall within the scope of this appraisal). It therefore concluded that a 1 year review date for the guidance would be appropriate.