Bortezomib-Thalidomide-Dexamethasone Versus Thalidomide-Dexamethasone before and after Double Autologous Stem Cell Transplantation for Newly Diagnosed Multiple Myeloma: Final Analysis of Phase 3 Gimema-MMY-3006 Study and Prognostic Score for Survival Outcomes

Introduction: The phase 3 GIMEMA-MMY-3006 trial comparing bortezomib-thalidomide-dexamethasone (VTD) versus thalidomide-dexamethasone (TD) as induction therapy before, and consolidation after, double autologous stem cell transplantation (ASCT) for newly diagnosed multiple myeloma (MM) provided the first demonstration of increased CR rate, the primary study endpoint, and prolonged PFS with VTD (Cavo M et al, Lancet 2010). However, updating trial results with longer follow-up than earlier reported is needed to assess the effects of treatment interventions on OS and to identify factors predicting for favorable long term outcomes.

Aims: We performed a post-hoc analysis of that study to evaluate long term results and construct a prognostic index of survival.

Methods: 474 patients were enrolled, 236 randomized to VTD and 238 to TD. Median follow-up for surviving patients was 124 months (IQR: 117-131). Analyses were performed on an intention-to-treat basis. Semi-parametric Cox regression analysis was used to construct the prognostic index. To assess the evolution of prognosis over time, conditional survival CS(t|s) estimate for PFS was calculated as the probability of surviving without progression a further 2 (t) years (yrs) after having already survived s yrs.

Results: Estimates of PFS and OS at 10 yrs for the VTD arm were 34% (HR=0.62; 95% CI=0.50-0.77; p<0.001) and 60% (HR=0.68, 95% CI=0.51-0.90; p=0.007), respectively, compared with TD (corresponding values, 17% and 46%), representing a 38-32% reduction in the risk of progression and death with VTD. Outcome benefits with VTD were seen for patients with high-risk cytogenetic abnormalities (HCRA), including t(4;14) and/or del(17p) by FISH, (PFS: 17% vs 3% at 10 yrs, HR=0.45, 95% CI=0.30-0.69; p<0.001; OS: 42% vs 22% at 10 yrs, HR=0.54, 95% CI=0.34-0.88; p=0.011) and lacking HRCA (PFS: 40% vs 20%, HR=0.60, 95% CI=0.46-0.79; p<0.001; OS: 67% vs 52%, HR=0.66, 95% CI=0.46-0.95; p=0.025). On multivariate Cox regression analysis, randomization to VTD predicted for both prolonged PFS (HR=0.60, 95% CI=0.48-0.76; p<0.001) and OS (HR=0.68, 95% CI=0.50-0.91; p=0.010). Specific multivariate regression analysis not including therapy revealed that the leading factors adversely affecting PFS were the presence of HRCA (HR=1.86, 95% CI=1.45-2.38; p<0.001), ISS stage II+III (HR=1.38, 95% CI=1.10-1.74; p=0.006), and failure to achieve CR as time-dependent variable (HR=2.01, 95% CI=1.59-2.53; p<0.001). The three variables were used to build a scoring system that stratified patients into three risk groups with divergent clinical outcomes: low-risk (LR) (22%, none of the 3 adverse variables), intermediate-risk (IR) (39%, 1 adverse variable), and high-risk (HR) (39%, 2 or 3 adverse variables). Estimated 10-yr PFS rates were 44% for patients in LR, 28% for IR, and 9% for HR (p<0.001). Estimated 10-yr OS rates were 76%, 58%, and 32%, respectively (p<0.001). Consensually, the prognostic score identified three groups with statistically different PFS and OS within the TD and VTD arm (p<0.001). On VTD, the 10-yr PFS and OS rates were 51% and 79% for LR, 41% and 62% for IR, 13% and 43% for HR, respectively. Randomization to receive VTD was associated with longer PFS for the IR (41% vs 15% at 10 yrs, HR=0.50, 95% CI=0.34-0.73; p<0.001) and HR (13% vs 7% at 10 yrs, HR=0.66, 95% CI=0.47-0.92; p=0.015) subgroups compared with TD. Moreover, HR patients assigned to VTD had significantly longer OS in comparison with the same group of patients on TD (43% vs 23% at 10 yrs, HR=0.65, 95% CI=0.43-0.97; p=0.033). Assessment of conditional survival revealed that the probability of surviving without progression a further 2 yrs improved progressively after 36 months, being 65% and reaching the 91% at 96 months (p value for trend=0.009). The conditional PFS became superimposable after 78 months for the LR and IR (87% and 86%, respectively), while resulted significantly lower in the HR (62.5%) (p=0.008).

Conclusions: With a follow up of 10 yrs, the final analysis of the GYMEMA MMY-3006 trial comparing VTD versus TD showed a persistent PFS benefit translating into extended OS for the VTD arm. A prognostic model based on cytogenetic, ISS stage and achievement of CR, identified three risk groups with statistically different long-term survival probabilities. Both IR and HR groups significantly benefited from VTD. A PFS time of 78 months predicted for long term survival outcomes in the LR and IR groups.