Matching‐adjusted indirect comparison of efficacy and safety of bortezomib, thalidomide, and dexamethasone (VTd) as per label compared with modified VTd dosing schedules in patients with newly diagnosed multiple myeloma who are transplant eligible

Abstract Background The combination of bortezomib, thalidomide, and dexamethasone (VTd) is a standard of care for transplant‐eligible patients with newly diagnosed multiple myeloma (NDMM). Although approved labeling for VTd includes an escalating thalidomide dose up to 200 mg daily (VTd‐label), a lower fixed dose of thalidomide (100 mg daily; VTd‐mod) has become commonplace in clinical practice. To date, no clinical trials comparing VTd‐mod with VTd‐label have been performed. Here, we compared outcomes for VTd‐mod with VTd‐label using a matching‐adjusted indirect comparison. Methods VTd‐mod data were from NCT02541383 (CASSIOPEIA; phase III) and NCT00531453 (phase II); VTd‐label data were from NCT00461747 (PETHEMA/GEM; phase III). To adjust for heterogeneity, baseline characteristics from VTd‐label were weighted to match VTd‐mod. Outcomes included overall survival (OS), progression‐free survival (PFS), postinduction and posttransplant responses, and safety. Results VTd‐mod was noninferior to VTd‐label for OS, postinduction overall response rate (ORR), and very good partial response or better (≥VGPR). VTd‐mod was significantly better than VTd‐label for PFS, posttransplant ORR, and ≥VGPR. VTd‐mod was noninferior to VTd‐label for safety outcomes, and inferior to VTd‐label for postinduction and posttransplant complete response or better. Conclusions Our analysis supports the continued use of VTd‐mod in clinical practice in transplant‐eligible NDMM patients.

was noninferior to VTd-label for safety outcomes, and inferior to VTd-label for postinduction and posttransplant complete response or better. To mitigate the potentially toxic effects of high-dose thalidomide exposure (such as thrombosis and peripheral neuropathy) [3], a reduced thalidomide dosing regimen (VTd-mod; 100 mg daily) is commonly used in clinical practice. Despite routine clinical use, a clinical trial assessing efficacy and safety of VTd-label versus VTd-mod has not been performed. In the absence of head-to-head trials, indirect treatment comparisons may be done to assess the relative effectiveness of both regimens [4,5]. Our objective was to use the matching-adjusted indirect comparison (MAIC), which adjusts for heterogeneity in the baseline prognostic variables, to determine relative effectiveness of VTd-mod versus VTd-label treatment regimens in transplant-eligible patients with NDMM [6][7][8].  Comparison of VTd-mod with VTd-label was performed stepwise via a naïve, unadjusted, indirect comparison followed by MAIC. Baseline characteristics from patients in the VTd-label arm from the PETHEMA/GEM study were utilized to weight IPD in the VTd-label group and then matched with those in the pooled VTd-mod group. Hazard ratios (HRs) with two-sided 95% confidence intervals (CIs) for time to event outcomes (OS and PFS) were determined using a weighted Cox regression analysis model that was fitted to the treatment arm; a logrank test was used to determine P-values for HRs and Kaplan-Meier curves. Absolute rate difference and odds ratios (ORs), calculated with two-sided 95% CIs, were used to determine treatment response and safety outcomes; P-values for ORs were determined using the twosided Fisher's exact test. Based on recent oncology studies, results that did not achieve statistical significance (5%) were interpreted with the use of noninferiority margins [11,12]. Noninferiority margins for response, safety, PFS, and OS were identified as 13% (rate difference), 13% (rate difference), 1.333 (HR), and 1.298 (HR), respectively [11].

RESULTS
Median duration of follow-up was 35.9 months for VTd-label, 33.3 months for NCT00531453 (VTd-mod), and 18.8 months for CASSIOPEIA (VTd-mod).

Efficacy
Baseline characteristics used for matching the VTd-mod (n = 591;  Table 1. Prior to weighting, imbalances between groups were observed for ECOG status, myeloma type, ISS, and creatinine clearance; however, after weighting, baseline characteristics were similar across groups, and the cohorts were considered viable for comparison of efficacy.

OS and PFS comparisons
For OS, VTd-mod was noninferior to VTd-label for both the naïve com-
Characteristics were similar after weighting (

DISCUSSION
The triple-drug combination regimen, VTd, is regarded in clinical practice as a standard of care for patients with NDMM who are eligible for transplant [1,13,14]. Higher doses of thalidomide have been associated with AEs such as peripheral neuropathy, so clinicians have introduced a lower, fixed-dose regimen (100 mg) to offset these toxic effects [3,7,15,16]. This lower-dose regimen was administered to NDMM patients before and after ASCT in the NCT00531453 study and the phase III CASSIOPEIA study [6,7], but no direct comparisons between the higher-and lower-dose thalidomide regimens have been  [8,9] than that in CASSIOPEIA (18.8 months) [7]. In addition, per design of CASSIOPEIA, patients with at least a partial response were re-randomized (100 days post-ASCT) to either observation or daratumumab monotherapy (every 8 weeks) for a further 2 years. Thus, fewer patients in CASSIOPEIA received maintenance treatment, so this analysis likely underestimates the survival benefit of VTd-mod versus VTd-label and should be considered conservative.
In terms of safety, the lack of superiority of VTd-mod compared with VTd-label was also unexpected, but reporting rules for PETHEMA/GEM may not have been as rigorous as those for

CONCLUSIONS
Based on this MAIC analysis, VTd-mod was shown to be noninferior to

ACKNOWLEDGMENTS
The authors thank the patients who participated in this study, the staff members at the study sites, the data and safety monitoring committee, and the staff members involved in data collection and analyses.  Ingress-Health. TK, SC, JH, and AL: Employment from Janssen. PM: Consultancy for AbbVie, Amgen, Celgene, Janssen, Takeda; honoraria from AbbVie, Amgen, Celgene, Janssen, Takeda.

DATA SHARING AGREEMENT
The data sharing policy of Janssen