Rituximab in combination with glucocorticoids for treating anti-neutrophil cytoplasmic antibody-associated vasculitis

The committee considered the evidence presented by the manufacturer on the clinical effectiveness of rituximab. It noted that the evidence was primarily from the RAVE study and this was complemented by the RITUXVAS study. The committee reviewed the suitability of the clinical trial evidence and noted that only RAVE used the regimen recommended in the marketing authorisation for ANCA-associated vasculitis. Overall, the committee concluded that the studies provided adequate evidence for assessing rituximab for inducing remission of ANCA-associated vasculitis and were generalisable to UK clinical practice.

The committee discussed the clinical effectiveness of rituximab compared with cyclophosphamide as induction therapy in people with severe ANCA-associated vasculitis. The committee accepted that the RAVE results showed rituximab was non-inferior to cyclophosphamide in inducing complete remission in the full study population at 6 months, but was uncertain if the treatment benefit persisted because of the short duration of RAVE. In response to consultation, the manufacturer provided 18-month follow-up data from RAVE. The committee acknowledged that rituximab was non-inferior to cyclophosphamide in inducing complete remission at 6, 12, and 18 months. In response to consultation, the manufacturer stated that patients in RAVE had severe disease, meaning the disease threatened the function of the affected organ and had the potential to cause permanent organ damage or to threaten the patient's life unless effective therapy was implemented quickly. The committee concluded that rituximab was not less effective than cyclophosphamide as an induction treatment for people with severe ANCA-associated vasculitis.

The committee discussed the need for maintenance treatment after rituximab induction therapy. The 2013 draft British Society for Rheumatology guidelines include 4 options for maintenance treatment (see section 3.65). The committee was aware of a difference of opinion between clinical specialists about the use of maintenance treatment. Some specialists stated that maintenance treatments such as azathioprine would be given after rituximab induction therapy, whereas others stated that azathioprine would not normally be given and is not supported by clinical trial evidence. In response to consultation, several specialists stated that rituximab would be used as maintenance treatment. The committee recalled that the marketing authorisation was specifically for inducing remission (with a recommended dosage of 375 mg/m² administered as an intravenous infusion once weekly for 4 weeks) and did not include rituximab being used as maintenance treatment. It further noted that the summary of product characteristics for rituximab states that the efficacy and safety of rituximab as maintenance treatment have not been established. The committee concluded that maintenance treatment with rituximab was outside the scope of the appraisal.

The committee reviewed the subgroups presented by the manufacturer to identify which people were likely to experience a greater treatment benefit. The committee was aware that, at 6-month follow-up from RAVE, the complete remission rate for the subgroup with relapsed disease was statistically significantly higher in patients who received rituximab compared with patients who received cyclophosphamide. The committee noted that the 18-month follow-up results for this subgroup, submitted in response to consultation, showed no significant difference in remission rates between the treatment groups. The committee observed that, at 6-month follow-up for the subgroup with newly diagnosed disease, there was no significant difference in remission rates between the treatment groups. The committee concluded that, over a period of 18 months, rituximab and cyclophosphamide have similar effectiveness in inducing remission in both newly diagnosed and relapsed patients.

The committee considered whether there were additional patient subgroups who might experience a greater treatment benefit. The committee heard from the clinical specialists that there may be a small subgroup of people who would benefit from avoiding cyclophosphamide. In response to consultation, the manufacturer defined this subgroup (see section 3.21). The committee noted that the manufacturer's definition is broadly in agreement with draft guidelines from the British Society for Rheumatology. Both the manufacturer and the British Society for Rheumatology stated that patients at risk of infection would benefit from avoiding cyclophosphamide. However, the committee observed that the summary of product characteristics states that rituximab should not be used for patients with active, severe infection. In response to the second consultation, clinical specialists advised that there is evidence from case series to support the use of rituximab for people who cannot have cyclophosphamide. The committee concluded that, for the purposes of this guidance, 'people who cannot have cyclophosphamide' refers to people:

• for whom cyclophosphamide is contraindicated (as defined in the summary of product characteristics) or not tolerated; or

• who have not completed their family and whose fertility may be materially affected by treatment with cyclophosphamide; or

• with disease that has remained active or progressed despite a course of cyclophosphamide lasting 3–6 months; or

• with a previous uroepithelial malignancy.

The committee discussed the safety of rituximab compared with cyclophosphamide. It was aware that intravenous administration of cyclophosphamide is associated with a more favourable adverse-event profile than oral administration. The committee noted that the frequency and severity of short-term adverse events were broadly comparable for rituximab and cyclophosphamide in RAVE (in which cyclophosphamide was administered orally) and RITUXVAS (in which cyclophosphamide was administered intravenously). The committee noted that there were long-term adverse events associated with cyclophosphamide (such as bladder cancer and loss of fertility), but that it was not possible to form any conclusions on the long-term safety profile of rituximab because the data in the manufacturer's submission only extended to a maximum of 18 months. In response to consultation, the manufacturer submitted evidence of the long-term safety of rituximab as a treatment for rheumatoid arthritis, and evidence that rituximab does not prevent women from conceiving children. The committee concluded that the safety profiles of rituximab and cyclophosphamide seemed broadly similar in the short term, and there was uncertainty about any long-term safety benefits of rituximab because of a lack of data from patients with ANCA-associated vasculitis.

The committee discussed potential advantages associated with rituximab that were not related to its efficacy or safety. It heard from the clinical specialists and patient experts that induction treatment with rituximab was 4 weeks instead of up to 6 months with cyclophosphamide, which was more convenient for patients. The committee concluded that this benefit was important to patients.

The committee discussed the manufacturer's approach to developing its economic model. It noted that the ERG considered the manufacturer's approach to be generally in line with the NICE reference case, but that the manufacturer's decision problem did not match the final NICE scope in all areas (notably excluding some comparators and some end points). The committee concluded that the outlined economic analysis was acceptable for assessing the cost effectiveness of rituximab in treating ANCA-associated vasculitis.

The committee considered the comparators included in the manufacturer's economic analysis. The clinical specialists, and responses to consultation, confirmed that cyclophosphamide is the standard of care for inducing remission in people who can have cyclophosphamide; typically administered intravenously. The committee noted that there was a lack of consensus about the appropriate comparator for people who cannot have cyclophosphamide. The committee recalled that the ERG's exploratory analyses for this subgroup, based on the manufacturer's original model, used a comparator of best supportive care. The committee was uncertain whether best supportive care was a realistic comparator. The committee was aware that the manufacturer's updated model for this subgroup used a comparator of either mycophenolate mofetil or methotrexate. Clinical specialists at the meeting advised that neither of these drugs is a treatment of choice for people with severe disease, and methotrexate is unsuitable for people with renal disease. Also, the British Society for Rheumatology draft guidelines recommend mycophenolate mofetil or methotrexate for patients with low disease activity who are not at risk of organ damage. The committee heard from the manufacturer that the clinical specialists it consulted advised that mycophenolate mofetil or methotrexate would be used as an induction treatment for people who cannot have cyclophosphamide. The committee concluded that it was appropriate to include intravenous cyclophosphamide as the comparator in the economic analysis for people who can have cyclophosphamide, and that there was uncertainty about the appropriate comparator for people who cannot have cyclophosphamide.

The committee evaluated the treatment sequences used in the manufacturer's original economic analysis. It considered the treatment sequences to be incomplete and unsuitable because they did not enable fully incremental analyses for all populations of interest. Also, the committee learned from clinical specialists that the manufacturer's assumption that patients who had not responded to a first course of rituximab would then receive a second course did not reflect UK clinical practice. The committee agreed that the treatment sequences used by the ERG in its exploratory analyses using the manufacturer's original model were more comprehensive and therefore more appropriate. However, the committee agreed that it needed additional analyses for all possible treatment sequences for the different subgroup populations, with ICERs presented in a fully incremental analysis and as pairwise comparisons. The committee then evaluated the treatment sequences used in the manufacturer's updated economic analysis, submitted in response to consultation. The committee observed that the updated model did not consider all treatment sequences and assumed that all patients received a second course of rituximab. The results were not presented in a fully incremental analysis. The committee concluded that these issues with the manufacturer's updated economic analysis added considerable uncertainty to the cost-effectiveness estimates.

The committee discussed the uncontrolled disease health state in the manufacturer's original economic model. It noted the ERG's concerns that patients in the model spent 60–70% of their average lifespan in the uncontrolled disease state and heard from the clinical specialists that this was not realistic. The committee was aware that this health state was associated with a low utility value and understood from the clinical specialists that patients would be expected to have some disease control with treatments other than cyclophosphamide. It noted the ERG's opinion that the costs for this health state had been overestimated and was advised by the clinical specialists that the number of outpatient appointments was not plausible. The committee agreed that the utility value had been underestimated and costs had been overestimated for the uncontrolled disease health state in the manufacturer's original model. The committee noted that, in response to consultation, the manufacturer submitted an updated model with a higher utility value and lower costs in the uncontrolled disease health state. It heard from the manufacturer and the ERG that the revised utility value was based on extrapolation from the utility values in the remission and non-remission health states. The committee noted that the utility value could have been estimated using data from patients in RAVE who had not entered remission during the trial, but this analysis had not been presented. The committee concluded that the revised utility value in the uncontrolled disease health state was more plausible than the value in the original model, but was still a source of some uncertainty.

The committee discussed how adverse events and disease consequences had been incorporated into the manufacturer's original model. It noted that disutilities for cyclophosphamide's cumulative long-term toxicity had not been included in the analyses by the manufacturer, and that the costs of managing long-term toxicity could be substantial (for example, treating uroepithelial cancer or fertility problems). The committee noted that the long-term toxicity of rituximab also had not been modelled and was not fully established. It was aware that the manufacturer's model did not include inpatient costs (such as treating infections) or the costs of disease consequences (for example, managing renal disease). The manufacturer's updated model, submitted in response to consultation, did not include disutilities for long-term toxicity, inpatient costs, or the costs of disease consequences. The committee concluded that the manufacturer's original and updated models had not captured all relevant costs and disutilities, which added some uncertainty to the cost-effectiveness estimates.

The committee reviewed how the manufacturer had estimated relapse rates in its original economic model and noted that the model assumed that both minor and severe relapses would need induction treatment. The committee noted from the manufacturer's submission that, when possible, minor relapses in RAVE were managed by increasing the glucocorticoid dose. It understood from the clinical specialists that this would generally be the first approach in UK clinical practice (unless, for example, it was considered that there was a high risk of progression to a severe relapse). The committee was aware that the manufacturer had used summary statistics rather than individual patient-level data, and noted the poor fit of the exponential distributions to the Kaplan–Meier relapse curves. It agreed with the ERG's opinion that the relapse rates derived from RAVE had been poorly estimated. In response to consultation, the manufacturer submitted an updated model which assumed that only severe relapses would need induction treatment. The committee continued to have concerns about the manufacturer's use of summary statistics and concluded that the relapse rates in the manufacturer's updated model were a source of uncertainty.

The committee then considered the manufacturer's updated models submitted in response to the first consultation (see sections 3.54 to 3.62). The committee noted that the manufacturer had not provided all the analyses requested at consultation. The manufacturer's response did not include all treatment sequences, pairwise and incremental comparisons, incorporate the costs and disutility of the cumulative long-term toxicity of cyclophosphamide, or include inpatient costs associated with non-remission. The ERG identified several errors in the manufacturer's models (see section 3.63). The committee then considered the manufacturer's weighted-average threshold analysis submitted in response to the second consultation (see section 3.75). It was aware that one of the reasons the manufacturer used this analysis was that another C

The committee considered the ERG's exploratory analyses using the manufacturer's updated model for people who can have cyclophosphamide (see section 3.73). The committee noted that the ERG had corrected the errors identified in the manufacturer's updated model. The committee also noted the ERG's exploratory analysis allowed re-treatment of patients who responded to rituximab rather than re-treatment of all patients in the manufacturer's updated model. It also considered the treatment sequence, which was 1 course of cyclophosphamide followed by 2 courses of rituximab compared with 2 courses of cyclophosphamide in the comparator arm. The committee noted that there was no incremental analysis of rituximab in different places in the treatment pathway. Therefore, the ICER of £20,900 per QALY gained from the ERG's exploratory analyses did not reflect the true cost effectiveness of rituximab given after 1 course of cyclophosphamide compared with cyclophosphamide. The committee then discussed the ERG's exploratory analyses using the manufacturer's original model (see section 3.48), because these incremental analyses explored the use of rituximab in different places in the treatment pathway. The committee was aware that a treatment sequence of 2 courses of cyclophosphamide followed by 1 course of rituximab compared with 2 courses of cyclophosphamide resulted in an ICER of £12,100 per QALY gained. It noted that using rituximab after 1 course of cyclophosphamide (compared with using it after 2 courses of cyclophosphamide) or as a first-line treatment (compared with using rituximab after 1 course of cyclophosphamide), resulted in ICERs of £69,700 and £127,500 per QALY gained respectively (see section 3.48). The committee agreed that the ICERs for rituximab after 1 course of cyclophosphamide or as a first-line treatment were outside the range normally considered a cost-effective use of NHS resources. The committee noted that these exploratory analyses included maintenance treatment with azathioprine after rituximab, which may not reflect UK clinical practice (see section 4.6). The committee heard from the ERG that, based on previous exploratory analyses (see section 3.73), including maintenance treatment with azathioprine was likely to have a small impact on the ICER. The committee concluded that the most plausible ICER on which to base its decision for people who can have cyclophosphamide was £12,100 per QALY gained, provided by the comparison of 2 courses of cyclophosphamide followed by 1 course of rituximab with 2 courses of cyclophosphamide.

The committee considered the cumulative dose provided by 2 courses of intravenous cyclophosphamide. Based on the manufacturer's response to consultation, the committee was persuaded that 2 courses of intravenous cyclophosphamide provides a cumulative dose of approximately 23 g on average, which is within the limit of 25 g advised by draft guidelines from the British Society for Rheumatology. The committee further noted that approximately 23 g cyclophosphamide represented 10 infusions (the maximum number that would be administered per course of treatment) and that, according to the clinical specialists, some patients would respond with fewer infusions per cycle. The committee noted that when possible, giving 2 courses of cyclophosphamide before rituximab would represent a more cost-effective option than giving 1 course of cyclophosphamide before rituximab. The committee concluded that rituximab could be recommended as a cost-effective use of NHS resources in people with severe ANCA-associated vasculitis who can have cyclophosphamide, only if further treatment with cyclophosphamide would exceed the maximum cumulative dose (25 g) of cyclophosphamide.

The committee discussed the ERG's exploratory analyses using the manufacturer's updated model for people who cannot have cyclophosphamide (see section 3.74). The treatment sequence included 2 courses of rituximab compared with 1 course of either mycophenolate mofetil or methotrexate and the ERG's exploratory analysis of the manufacturer's updated model gave an ICER of £60,600 per QALY gained. The committee was aware of substantial uncertainty about the assumptions in the model, such as the utility of the remission health state, the cost and disutility associated with glucocorticoids, and the remission and relapse rates. Therefore, the committee agreed that the ICER of £60,600 per QALY gained was not plausible. The committee considered the ERG's illustrative analyses, submitted in response to the second consultation (see section 3.76). The ERG's illustrative analyses changed some assumptions in the model and gave an ICER of £26,400 per QALY gained for the comparison of 2 courses of rituximab with 1 course of mycophenolate mofetil. The committee heard from the ERG that the analyses illustrate the uncertainty in the estimates of cost effectiveness for people who cannot have cyclophosphamide. The committee was aware that the clinical specialists did not agree about the use of mycophenolate mofetil or methotrexate as an induction treatment in people who cannot have cyclophosphamide (see section 4.12). The committee then discussed the ERG's exploratory analyses using the manufacturer's original model for people who cannot have cyclophosphamide (see section 3.52), because these analyses included an alternative comparator. The committee noted that 1 course of rituximab compared with best supportive care gave an ICER of £11,300 per QALY gained. The committee agreed that there was a lack of consensus about the appropriate comparator for people who cannot have cyclophosphamide. The committee concluded there was substantial uncertainty about the cost effectiveness of rituximab for people who cannot have cyclophosphamide, but on balance the ICER was likely to be lower than £30,000 per QALY gained.

The committee discussed whether rituximab was innovative in its potential to make a significant and substantial impact on health-related benefits. The committee was aware that, in response to the second consultation, clinical specialists and patient experts stated that rituximab was 'scene-changing' in the treatment of ANCA-associated vasculitis. Consultees also advised that rituximab was the first new effective treatment since the introduction of cyclophosphamide in the 1970s, and rituximab may be the first of a new generation of treatments. In addition, consultees advised that people who cannot have cyclophosphamide have the highest unmet need because no alternative treatments are as effective as rituximab. The manufacturer noted that cyclophosphamide reduces fertility in men and women, and stated that the benefit of maintaining fertility while treating the disease effectively cannot be captured in the QALY. The committee was aware that the manufacturer had provided evidence that rituximab does not prevent women from conceiving children. The committee concluded that rituximab was an innovative treatment.

In summary, for people who cannot have cyclophosphamide, the committee considered the manufacturer's original and updated analyses, the ERG's exploratory and illustrative analyses, and comments received during consultation. The committee took into account the estimates of cost effectiveness and noted the uncertainty associated with them. The committee also recognised that rituximab is an innovative treatment and the high unmet need for treatment options for people who cannot have cyclophosphamide. Having taken into account all of the evidence submitted and the comments received during consultation and noting the NICE Social Value Judgements, the committee concluded that rituximab was a cost-effective use of NHS resources for treating people with severe ANCA-associated vasculitis who cannot have cyclophosphamide, as defined in section 4.8.

The committee considered whether its recommendations were associated with any issues related to the equality legislation and the requirement for fairness. The committee noted that the manufacturer stated that cyclophosphamide reduces fertility in both men and women. The committee was also aware that the manufacturer had provided evidence that rituximab does not prevent women from conceiving children and that no evidence was presented regarding the effect of rituximab on male fertility. Based on the available evidence, the committee considered that it was appropriate to accept that rituximab was likely to have a less detrimental effect on male fertility than cyclophosphamide. The committee considered that, in this context, guidance that only recommended rituximab for women who had not completed their family would potentially constitute unlawful sex discrimination. The committee concluded that it was appropriate to recommend rituximab for men and women who have not completed their family and whose fertility may be materially affected by treatment with cyclophosphamide.

The committee further considered issues related to the equality legislation. Considering that the guidance in section 1.1 recommends rituximab for people who have not completed their family and whose fertility may be materially affected by treatment with cyclophosphamide, the committee was aware that this recommendation would affect access for post-menopausal women whereas younger women and men of all ages could potentially receive rituximab. The committee discussed whether this could be regarded as indirect discrimination. The committee noted that any differential treatment of post-menopausal women arises from the different physiological features of fertility in men and women. The committee noted that rituximab and cyclophosphamide have similar effectiveness as induction treatments for severe ANCA-associated vasculitis (see section 4.7), so an effective induction treatment will also be available for post-menopausal women. Therefore, the committee agreed that its recommendations do not constitute detrimental treatment of post-menopausal women. The committee noted that the safety profiles of rituximab and cyclophosphamide are broadly similar in the short term, and there was uncertainty about any long-term safety benefits of rituximab compared with cyclophosphamide (see section 4.9). The committee concluded that the guidance would permit an effective induction treatment for all groups of people, and there was no evidence that some groups would experience more adverse effects of treatment than other groups, and therefore there was no unfairness.

In considering the potential equalities issues, the committee took into account the size and characteristics of the overall population of people with ANCA-associated vasculitis and the subgroup of people who would be affected by the recommendation relating to fertility. The committee was aware that around 1200 people are diagnosed with ANCA-associated vasculitis each year in England and Wales and the peak age of onset is between 60 and 70 years. Therefore, the committee concluded that the number of people with ANCA-associated vasculitis who have not completed their family is likely to be small.

The committee further discussed issues related to the equality legislation. Consultees suggested that children should be included in the population, but the marketing authorisation specifies 'adults' so this is not an equality issue that falls within the remit of a NICE technology appraisal. The committee concluded that its decision on the use of rituximab would not have a disproportionate impact on any group with a protected characteristic that cannot be objectively justified, and that therefore there was no need to alter or add to its recommendations.