Superior Rate of Complete Response with up-Front Velcade-Thalidomide-Dexamethasone Versus Thalidomide-Dexamethasone in Newly Diagnosed Multiple Myeloma Is Not Affected by Adverse Prognostic Factors, Including High-Risk Cytogenetic Abnormalities.

Complete response (CR) is an important objective of autologous stem-cell transplantation (ASCT) in multiple myeloma (MM). In comparison with conventional induction treatments, newer combinations of novel agents may effect increased rates of CR and near CR (nCR), a benefit potentially translating into even higher frequencies of CR/nCR after ASCT and improved clinical outcome. We designed a phase III study to detect an increase in CR+nCR rates from 10–15% with conventional Thalidomide-Dexamethasone (TD) to 20–30% with Velcade added to TD (VTD) in newly diagnosed MM. Both TD and VTD were given as three 21-day cycles in preparation for double ASCT. In the present analysis, CR+nCR rates by the two induction treatments were examined in relationship to baseline prognostic variables in 399 evaluable pts aged ≤65 years, of whom 199 randomized to VTD and 200 to TD. All analyses were intent to treat. In comparison with TD, VTD effected higher rates of CR+nCR (12% vs 33%, P<0.001) and of ≥very good partial response (VGPR) (30% vs 61%, P<0.001). By univariate analysis, superiority of VTD to TD was maintained across all sub-group analyses according to standard prognostic factors, including β2-m, albumin, stage (ISS), Hb, PLTs, bone marrow PC, M protein isotype, LDH, CRP. In particular, the rates of CR+nCR with VTD vs TD in pts with standard poor prognostic factors were as follows: ISS stage 3 (23.5% vs 6%, P=0.03), Hb<10 g/dL (24% vs 4%, P=0.002), PLTs<150.000/μL (35% vs 4%, P=0.009), bone marrow PC ≥50% (31% vs 13%, P<0.001), IgA isotype (63% vs 15%, P<0.001), LDH >190 U/L (33% vs 9%, P<0.001), CRP ≥8 mg/L (29% vs 10%, P=0.004). We next examined CR+nCRs by treatment arms in relationship to cytogenetics (FISH data available in 93% to 99% of all pts). Superior CR+nCR rates were effected by VTD vs TD in the presence of high-risk cytogenetics, including del(13) (39% vs 10%, P<0.001), t(4;14) (39.5% vs 10%, P=0.002), combined t(4;14) and del(13) (32% vs 0%, P=0.001), and del(17p) (28.5% vs 0%, P=0.03). Remarkably, when examined in the context of the VTD arm, high-quality response rates were significantly higher for pts carrying del(13) and t(4;14) vs those who lacked these abnormalities [del(13): CR+nCR:39% vs 24%, P=0.03; ≥VGPR: 71% vs 48%, P=0.001] [t(4;14): ≥VGPR:79% vs 55%, P=0.007)]. An opposite trend was noted for pts in the TD arm, whose probability to attain ≥VGPR was adversely affected by the presence of del(13) (P=0.07) and del(17p) (P=0.03). Variables associated with achievement of CR+nCR in the two arms that retained statistical significance when assessed by multivariate Cox regression analysis included randomization to VTD (P<0.001), light chain only subtype (P<0.001), IgA isotype (P<0.001) and Hb>10 g/dL (P=0.01). In the VTD arm, a positive correlation was observed with del(13) (P=0.006) and t(4;14) (P=0.02). Response to first ASCT with melphalan 200 mg/m² could be evaluated in 297 pts, of whom 145 randomized to VTD and 152 to TD. Randomization to VTD was closely associated with increased CR+nCR rates (54% vs 29% with TD, P<0.001) and remained statistically significant (P<0.001) also in the multivariate analysis. Additional factors predicting for superior post-ASCT CR+nCR rates in the multivariate setting included light chain only subtype (P<0.001) and IgA isotype (P=0.005). We conclude that randomization to up-front VTD was the strongest and independent factor associated with increased rates of CR+nCR before ASCT. Superiority of VTD to TD pertained in both low-risk and high-risk sub-groups, including the traditionally unfavorable sub-groups carrying del(13), t(4,14) and del(17p). Remarkably, in the VTD arm improved postinduction CR+nCR rates were significantly associated with the presence of del(13) and t(4;14) in the multivariate analysis. Benefit from VTD vs TD as primary induction therapy translated into significantly improved CR+nCR rates after the first ASCT and remained statistically significant when assessed by multivariate analysis.