Ixazomib is an oral proteasome inhibitor used in combination with lenalidomide plus dexamethasone (IXA-LEN-DEX) and licensed for relapsed or refractory multiple myeloma. As part of a single technology appraisal (ID807) undertaken by the National Institute of Health and Care Excellence, the Evidence Review Group, Warwick Evidence was invited to independently review the evidence submitted by the manufacturer of ixazomib, Takeda UK Ltd. The main source of clinical effectiveness data about IXA-LEN-DEX came from the Tourmaline-MM1 randomized controlled trial in which 771 patients with relapsed or refractory multiple myeloma received either IXA-LEN-DEX or placebo-LEN-DEX as their second-, third-, or fourth-line treatment. Takeda estimated the cost effectiveness of IXA-LEN-DEX using a de-novo partitioned-survival model with three health states (pre-progression, post-progression, and dead). In their first submission, this model was used to estimate the cost effectiveness of IXA-LEN-DEX vs. bortezomib plus dexamethasone (BORT-DEX) in second-line treatment, and of IXA-LEN-DEX vs. LEN-DEX in third-line treatment. To estimate the relative clinical performance of IXA-LEN-DEX vs. BORT-DEX, Takeda conducted network meta-analyses for important outcomes. The network meta-analysis for overall survival was found to be flawed in several respects, but mainly because a hazard ratio input for one of the studies in the network had been inverted, resulting in a large inflation of the claimed superiority of IXA-LEN-DEX over BORT-DEX and a considerable overestimation of its cost effectiveness. In subsequent submissions, Takeda withdrew second-line treatment as an option for IXA-LEN-DEX. The manufacturer's first submission comparing IXA-LEN-DEX with LEN-DEX for third-line therapy employed Tourmaline-MM1 data from third- and fourth-line patients as proxy for a third-line population. The appraisal committee did not consider this reasonable because randomization in Tourmaline-MM1 was stratified according to one previous treatment and two or more previous treatments. A further deficiency was considered to be the manufacturer's use of interim survival data rather than the most mature data available. A second submission from the company focussed on IXA-LEN-DEX vs. LEN-DEX as third- or fourth-line treatment (the two or more previous lines population) and a new patient access scheme was introduced. Covariate modeling of survival outcomes was proposed using the most mature survival data. The Evidence Review Group's main criticisms of the new evidence included: the utility associated with the pre-progression health state was overestimated, treatment costs of ixazomib were underestimated, survival models were still associated with great uncertainty, leading to clinically implausible anomalies and highly variable incremental cost-effectiveness ratio estimates, and the company had not explored a strong assumption that the survival benefit of IXA-LEN-DEX over LEN-DEX would be fully maintained for a further 22 years beyond the observed data, which encompassed only approximately 2.5 years of observation. The appraisal committee remained unconvinced that ixazomib represented a cost-effective use of National Health Service resources. Takeda's third submission offered new base-case parametric models for survival outcomes, a new analysis of utilities, and proposed a commercial access agreement. In a brief critique of the third submission, the Evidence Review Group agreed that the selection of appropriate survival models was problematic and at the request of the National Institute for Health Care and Excellence investigated external sources of evidence regarding survival outcomes. The Evidence Review Group considered that some cost and utility estimates in the submission may have remained biased in favor of ixazomib. As a result of their third appraisal meeting, the committee judged that for the two to three prior therapies population, and at the price agreed in a commercial access agreement, ixazomib had the potential to be cost effective. It was referred to the Cancer Drugs Fund so that further data could accrue with the aim of diminishing the clinical uncertainties.