First-line therapy for non-transplant eligible patients with multiple myeloma: direct and adjusted indirect comparison of treatment regimens on the existing market in Germany.

Objectives: The purpose of this study was to compare approved first-line therapies for patients with multiple myeloma.
Methods: A systematic literature search for phase III randomized controlled trials (RCTs) comparing first-line chemotherapies approved in Germany and recommended by guidelines at the time of study design was conducted. Random-effects meta-analysis (MA) was used for direct and the Bucher method for adjusted indirect treatment comparison.
Results: One RCT comparing melphalan and prednisone plus bortezomib (VMP) vs. melphalan and prednisone (MP) and six RCTs comparing MP plus thalidomide (MPT) vs. MP were analysed. For MPT vs. MP, an individual patient data (IPD) MA was used for sensitivity analyses. VMP and MPT were superior to MP regarding efficacy endpoints (VMP vs. MP, overall survival (OS): hazard ratio (HR) 0.70, 95 % confidence interval (CI) 0.57-0.86; progression-free survival (PFS): HR 0.56, 0.39-0.79; complete response (CR), risk-ratio (RR) for non-response: 0.70, 0.65-0.75; MPT vs.
Mp, Os: HR 0.83, 0.66-1.03; PFS: HR 0.67, 0.56-0.81; CR, RR for non-response 0.92, 0.88-0.95); but had a higher risk of developing any grade 3-4 adverse events (AEs) (VMP vs. MP: RR 1.13, 1.06-1.20; MPT vs. MP: RR 2.06, 1.43-2.98). The indirect comparison of VMP vs. MPT via MP showed a statistically not significant advantage for VMP regarding survival outcomes (OS: HR 0.85, 0.63-1.14; PFS: HR 0.83, 0.56-1.23) and a significant advantage regarding CR (RR for non-response 0.76, 0.70-0.83) and AEs (RR 0.55, 0.38-0.80). Treatment comparisons using results of IPD MA yielded similar effect sizes.
Conclusions: VMP and MPT seem more effective than MP, VMP was superior to MPT regarding response criteria and AEs. Our results may best be confirmed by a head-to-head trial of VMP vs. MPT.