OBJECTIVES: Selinexor plus dexamethasone (Sd) is the first regimen licensed for penta-refractory multiple myeloma (PR-MM), seeking reimbursement in Europe through Health Technology Assessment (HTA). To inform cost-effectiveness analyses (CEA) for HTA, robust data in a relevant population is required, including health state utility values (HSUVs) representative of health-related quality of life (HRQoL) outcomes of interventions versus standard of care.

METHODS: A systematic review was conducted to identify economic evaluations, costs and resource use, and health utility data, in patients with relapsed and/or refractory MM (RRMM), including PR-MM.

RESULTS: Bibliographic database searching with additional handsearching identified 62 records reporting utility values mostly in earlier lines of RRMM (2L/3L/4L). Additional studies were identified reporting HRQoL data from measures mappable to EQ-5D, however, there were no published studies reporting utility values for a PR-MM population. Two records were identified from the phase 2b STORM trial of Sd, in a penta-exposed population (PE-MM) with ≥50% PR-MM, reporting HRQoL data using FACT-G/MM/MM-TOI for which mapping to EQ-5D-3L/5L is required to elicit utility values. As a potential proxy to PR-MM, a CEA of belantamab mafodotin (belamaf) versus Sd in a 5L+ triple-class refractory (TCR) population applied HSUVs from DREAMM-2 (4L+ triple-class exposed [TCE]) obtained by mapping EORTC-QLQ-C30/MY-20 data to EQ-5D-3L. These mapped utilities were referenced in a second belamaf CEA (5L+ TCE). A third CEA of treatments in heavily pre-treated RRMM applied published HSUVs from MM-003 in double-class refractory participants who had received a median of five prior lines. The pre-progression HSUV reported across all three records was 0.73.

CONCLUSIONS: Although robust data is required for CEA informing HTA of newer interventions, studies only reported HRQoL data specifically for patients with PE-MM and nothing specifically was for PR-MM. Furthermore, reported HRQoL data were derived from tools other than EQ-5D, necessitating mapping exercises to obtain HSUVs for CEA.