Ixazomib with lenalidomide and dexamethasone for treating relapsed or refractory multiple myeloma

3.1 Multiple myeloma is typically incurable and is a progressive disease that affects survival and quality of life. The patient experts explained that multiple myeloma causes debilitating symptoms including bone pain and bone fractures, tiredness, infections, hypercalcaemia (too much calcium in the blood) and kidney problems. The clinical experts noted that treating the disease becomes more and more challenging with each relapse. Treatment side effects and frequent hospital visits have an impact on people with multiple myeloma and their carers. The patient expert explained that ixazomib with lenalidomide and dexamethasone (ixazomib combination) has allowed them a long period of progression-free survival and an improved quality of life with minimal side effects. The patient expert noted that they valued the oral administration of ixazomib combination. The committee considered the evidence from patient experts who have experience taking ixazomib combination and concluded that it is a valuable treatment option for people with relapsed or refractory multiple myeloma.

3.2 Treatment options for multiple myeloma depend on a person's treatment history, their response to the treatments and their preferences. After 2 previous lines of treatment, options include lenalidomide plus dexamethasone, or panobinostat plus bortezomib and dexamethasone. After 3 previous lines of treatment, options include pomalidomide plus dexamethasone, or panobinostat plus bortezomib and dexamethasone. Daratumumab alone, and isatuximab with pomalidomide and dexamethasone are also available in the Cancer Drugs Fund. Ixazomib combination can be used after at least 1 previous therapy. The committee in the original appraisal (NICE technology appraisal guidance 505, from now referred to as TA505) understood that ixazomib combination would be used in the same place in the pathway that lenalidomide and dexamethasone is currently used; that is, for people who have had 2 or 3 previous therapies. The committee concluded that the population relevant to this review was people who have had 2 or 3 previous therapies.

3.3 The clinical experts explained that oral treatment regimens, which reduce pressure on hospital outpatient units and chemotherapy day units compared with intravenous options, are very important to people with multiple myeloma. They explained that oral treatment regimens reduce the risk of acquiring infections and that many people with multiple myeloma are particularly concerned about contracting COVID-19 from hospitals. The clinical lead for the Cancer Drugs Fund explained that triple therapy is now standard at first and second line for people with multiple myeloma and that a fourth-line triplet therapy is available in the Cancer Drugs Fund. The patient expert explained that the benefits of ixazomib combination were not limited to the oral administration of the treatment, but also that it is a triplet therapy. The clinical lead for the Cancer Drugs Fund added that ixazomib combination has met the need for this preferred treatment option at third line. The committee concluded that, without ixazomib combination, there would be an unmet need for an oral triplet therapy for people who have had 2 or 3 previous therapies.

3.4 In TA505 the committee agreed that lenalidomide with dexamethasone was the relevant comparator. This was because, although the scope included bortezomib for people who have had at least 1 therapy, the committee for TA505 recalled that ixazomib combination was not expected to be widely used in people who have had 1 previous therapy. So, the committee in TA505 agreed bortezomib was not a relevant comparator. NICE guidance recommends both panobinostat plus bortezomib and dexamethasone, and lenalidomide plus dexamethasone after 2 previous lines of treatment. But the clinical experts in TA505 stated that they would always prefer to use lenalidomide before panobinostat, and therefore panobinostat is not used unless people have had 3 therapies or more. No changes to the scope are permitted in a Cancer Drugs Fund review and treatments in the Cancer Drugs Fund cannot be considered comparators or included in the treatment sequence. So, the committee concluded that after 2 or 3 lines of treatment, lenalidomide plus dexamethasone remained the most relevant comparator.

3.5 In TA505 the clinical-effectiveness evidence for ixazomib combination came from a second interim analysis data cut of TMM1, a phase 3 randomised controlled trial. The primary endpoint of the trial was progression-free survival. The median progression-free survival for ixazomib combination after 2 or 3 previous lines of treatment was 22 months compared with 13 months for lenalidomide and dexamethasone. The stratified hazard ratio was 0.62 (95% confidence interval [CI] 0.45 to 0.86, p=0.0033). This showed that ixazomib combination was associated with a statistically significant improvement in progression-free survival compared with lenalidomide and dexamethasone. The progression-free survival observations were not collected beyond the second interim analysis of TMM1 because this data was mature. So, the progression-free survival data was the same as the original appraisal. In TA505 the committee noted that the second interim analysis showed a reduced progression-free survival difference between arms compared with the first analysis, which was no longer statistically significant for the intention-to-treat population. However, it did remain significant for people who had had 2 or 3 previous therapies. The committee concluded that ixazomib combination improved progression-free survival in people who have had 2 or 3 lines of therapy.

3.6 The committee in TA505 was concerned that the clinical benefit of ixazomib combination was uncertain because the overall-survival data was too immature to be considered robust. To address the committee's concerns from the original appraisal, the company provided the clinical-effectiveness for ixazomib combination from the final data cut of TMM1. The trial included 722 people with relapsed or refractory multiple myeloma. A subgroup of 297 people, which included people who had had 2 or 3 previous lines of treatment, was relevant to this appraisal. The unadjusted median overall survival was 53 months for ixazomib combination and 43 months for lenalidomide and dexamethasone. The stratified hazard ratio was 0.85 (95% CI 0.64 to 1.11, p=0.232), showing ixazomib combination was not associated with a statistically significant improvement in overall survival compared with lenalidomide and dexamethasone. However, the clinical experts explained that overall survival is highly dependent on the choice of post-progression treatments. They further explained that the TMM1 trial had no UK centres, and the post-progression treatments would have varied depending on location and so were unlikely to be representative of UK practice. The company considered it was not appropriate to use the unadjusted overall-survival estimates from TMM1. The company explained that it had adjusted the data to remove the effect of subsequent therapies. This was because the treatment pathway in the trial did not reflect UK clinical practice and the distribution of subsequent treatments was not balanced between treatment arms (see section 3.8). The clinical experts noted that ixazomib is a proteasome inhibitor. They explained that because TMM1 was a blinded trial, some people whose disease progressed with ixazomib combination may have been given another proteasome inhibitor as a subsequent therapy. These people were likely to have been refractory to proteasome inhibitors at progression, so were given a potentially less effective treatment sequence than people in the lenalidomide and dexamethasone arm. The company did not adjust for this potential confounding, so, the committee heard that the overall-survival results were likely to be biased against ixazomib combination. The committee concluded that although the effect of ixazomib commination on overall survival in TMM1 was likely confounded by subsequent treatments, the data was acceptable for decision making.

3.7 Treatments recommended in the Cancer Drugs Fund cannot be considered comparators or subsequent treatments in NICE appraisals. NICE's guide to the methods of technology appraisal 2013 also advises that treatments not embedded in clinical practice in the NHS should not be considered as comparators or subsequent treatments. So, the company used the 2-stage method with recensoring to remove the effect of subsequent treatments from TMM1 not routinely used in the NHS, including those available via the Cancer Drugs Fund. The company assumed that the treatments removed in the 2-stage adjustment were effective and improved survival. The adjusted median overall survival was 51.4 months for ixazomib combination and 41.5 months for lenalidomide and dexamethasone. The stratified hazard ratio was 0.71 (95% CI 0.54 to 0.95, p=0.0216), showing ixazomib combination was associated with a statistically significant improvement in overall survival compared with lenalidomide and dexamethasone. The ERG explored additional analyses of the intention-to-treat population, including all 722 people, from TMM1. The ERG stated that in this population, which included people who had only had 1 previous line of treatment, the confidence intervals of overall survival crossed 1. The company highlighted that analysing this population was outside the scope of this review and it included people from a different treatment line who would have had different treatments and disease prognoses. The company also noted that a regional study of TMM1, in China, which did not face challenges of adjusting data for subsequent treatment therapies, showed a survival benefit with ixazomib combination. The committee concluded that an overall-survival benefit of ixazomib combination is likely.

3.8 After adjustment for subsequent treatments, modelled post-progression life expectancy reduced to a greater extent in the lenalidomide and dexamethasone arm (a reduction of 0.3 life years) compared with the ixazomib combination arm (a reduction of 0.03 life years). The ERG noted that there were only small differences in the proportion of treatments not routinely used in the NHS or within the Cancer Drugs Fund in both treatment arms. Therefore, it was unclear why the reduction in life years was larger in the lenalidomide plus dexamethasone arm. The company explained that more people had treatments that extend survival in the lenalidomide and dexamethasone arm. The ERG also noted that modelled post-progression life-year gain, after adjustment, was 30.5% of the total life-year gain. Whereas, modelled post-progression life-year gain before adjustment was 7.4% of total life-year gain. The ERG stated that the change in life years gained departed from clinical plausibility. The company highlighted that there was a minimal difference between the cost-effectiveness estimates for the 2-stage method with recensoring and the unadjusted analysis. The committee noted these areas of uncertainty associated with the company's overall-survival adjustment, but was aware that using the adjusted analysis with recensoring versus unadjusted overall-survival data in the model made very little difference to the cost-effectiveness results. The committee concluded that although there was uncertainty about the adjustment to overall survival for subsequent treatments, the company's approach was appropriate.

3.9 The committee noted that the company recensored the overall-survival data following its 2-stage adjustment, and that recensoring can overestimate the treatment effect. The committee considered that uncensored data should have been considered and that the cost-effectiveness estimates were lower than they would have been if the company had not recensored the data. The company highlighted that the NICE Decision Support Unit technical support document 16 states that recensoring is required to avoid bias, particularly in cases where treatment switching is likely to be associated with prognosis, which was the case in TMM1. The committee concluded that the bias from recensoring the data was likely less than the bias from not censoring the data, but that the true cost-effectiveness estimates could be higher than the ones presented by the company.

3.10 NHS England also provided data from the SACT dataset. It was collected from 2,460 people who had ixazomib combination through the Cancer Drugs Fund between December 2017 and June 2020. The clinical experts explained that the clinical experience with ixazomib combination was positive. The clinical experts added that the SACT dataset shows that disease control and disease response with ixazomib combination was encouraging. The committee noted that the adjusted median overall survival in the trial was longer (51.4 months) than in the SACT dataset (30 months). Both the clinical experts and the clinical lead for the Cancer Drugs Fund explained that people included in the SACT dataset were older and had a poorer prognosis than people in TMM1. This may have contributed to the differences in the median survival estimates. The clinical experts added that the median follow up for overall survival was also shorter than the follow up in TMM1 and so not all benefits from ixazomib combination would have been captured. The committee recalled that the SACT dataset did not provide comparative evidence versus lenalidomide and dexamethasone and so a relative treatment effect from the use of ixazomib combination in the NHS could not be estimated. The committee agreed that the evidence from TMM1 was robust and the most appropriate for decision making. The committee concluded that population differences likely contributed to the differences in treatment effects between TMM1 and the SACT dataset.

3.11 The company used data from the subgroup of people from TMM1 with relapsed and refractory multiple myeloma who had 2 or 3 previous lines of treatment. It used the generalised gamma model to estimate overall survival with ixazomib combination and lenalidomide and dexamethasone. It explained that it chose this model because its clinicians stated that the generalised gamma provided a reasonable estimation of long-term outcomes. The ERG proposed that the Weibull model was as valid on the grounds of clinical plausibility as the generalised gamma curve and was statistically better fitting than the generalised gamma model for both ixazomib combination and lenalidomide and dexamethasone. The clinical experts acknowledged that both the generalised gamma and Weibull curves estimated similar overall-survival outcomes. The clinical experts noted that survival seen in real-world practice in the NHS was much lower than the survival estimated by the parametric models. The company highlighted that there was a difference between clinical trial data and real-world data; in TMM1 around 35% of people were still alive after 8 years. The company added that people in clinical trials are often younger, have fewer comorbidities and have a better prognosis compared with people not in clinical trials. The committee concluded that although the generalised gamma curve was acceptable to extrapolate overall survival for ixazomib combination and lenalidomide and dexamethasone, the Weibull curve should also be considered.

3.12 In TA505 the company assumed that ixazomib combination's relative survival benefit in the clinical trial, compared with lenalidomide and dexamethasone, was maintained for the rest of a person's life after treatment stopped. The committee concluded in TA505 that, although it was biologically plausible for the relative treatment benefit to continue after stopping treatment, that level might not be maintained for the rest of a person's life. For this review, the company did not include a treatment waning effect in its base case. The company explained that the discontinuation of treatment had been almost completely captured within the 8-year observation time of TMM1. The company added that the hazard ratio or treatment effect in the ixazomib combination arm was a composite measure reflecting a pathway of treatments. The ERG agreed that the discontinuation and waning of treatment had been almost completely captured within the observed time of the trial, but noted that this was separate to the waning of the treatment effect. The ERG noted that over 90% of people were only observed for around 2 years following discontinuation of ixazomib combination. The ERG thought this was an insufficient time to capture any waning of treatment effect and conducted scenarios to explore the potential impact of treatment waning. The ERG did not include these scenarios in its base case. The ERG applied weighted hazards produced at each model cycle to generate an adjusted overall-survival estimate for people having ixazomib combination. Introducing a treatment waning effect increased the cost-effectiveness estimates of ixazomib combination. The committee recalled the long follow up of TMM1 and concluded that it largely captured the treatment waning and no other adjustments to treatment waning were needed.

3.13 In TA505 the committee recognised that the utility for progressed disease was higher than the UK population norms, and higher than in previous multiple myeloma appraisals. In this review the company used updated EQ-5D data from TMM1. The ERG agreed that the utilities estimated from the final analysis of TMM1 better aligned with the literature and reflected the health-related quality of life of people with relapsed or refractory multiple myeloma. The committee concluded that the utility values used in the model were acceptable.

3.14 The committee considered the advice about life-extending treatments for people with a short life expectancy in NICE's guide to the methods of technology appraisal. In TA505 the committee considered that although ixazomib combination has the potential to improve overall survival, it did not meet the criterion for extension to life. The committee concluded that ixazomib combination could not be considered as an end of life therapy. The committee in this review concluded that its view from TA505 had not changed.

3.15 The committee preferred an analysis that included:

3.16 Clinical experts noted that myeloma is twice as common in people of African-Caribbean family background. The committee considered that its recommendation applies equally, regardless of family background. It concluded that this difference in prevalence did not itself represent an equality issue in this appraisal.