Kinetics of Response to Bortezomib/Thalidomide/Dexamethasone (VTD) in Multiple Myeloma: Implications for the Choice and Design of Pretransplantation Induction Regimens

Background: Autologous stem-cell transplantation (ASCT) is the standard of care for younger patients with multiple myeloma (MM). The degree of tumor reduction after ASCT is the crucial factor associated with a prolonged PFS and OS, being the M-protein decrease at the time of transplant the most important predictor of residual disease after ASCT. While there is an agreement that bortezomib and dexamethasone associated to a third drug is the induction of choice, which should be the third drug (thalidomide, lenalidomide, doxorubicin or cyclophosphamide) and the optimal number of cycles remain unknown. The results of phase 3 PETHEMA study GEM05menos65 showing a CR rate of 35% increasing overtime during the 6 induction cycles of VTD (Rosiñol et al, Blood 2012) prompted us to study the kinetics of response to VTD and TD in this study and in similar historic data with the VD regimen in phase 2 PETHEMA VELCA/DEXA trial (Rosiñol et al, JCO 2007).

Objective: To study the kinetics of response by cycle during the 6 cycles of induction with VTD, TD and VD using a random effects model methodology.

Patients and Methods: In GEM05menos65 study patients were randomized to receive 6 induction cycles of either VTD or TD followed by ASCT. One hundred and thirteen patients were treated with TD (thalidomide 200 mg daily; dexamethasone 40 mg on days 1-4 and 9-12) and 122 with VTD (TD at identical doses plus bortezomib 1.3 mg/m2 on days 1, 4, 8 and 11) (Rosiñol et al, Blood 2012) and had complete data set for this analysis. In VELCA/DEXA study 40 patients received 6 induction cycles of bortezomib and dexamethasone on an alternating basis (Rosiñol et al, JCO 2007). Linear random effects models were employed to analyze the tumor response kinetics using the absolute decrease value of the serum M-protein after each cycle. Because the nonlinearity in the change of the M-protein overtime, a piecewise linear model was used to estimate mean changes in M-protein in each of the 6 cycles.

Results: Three different comparisons were made: 1) the decrease of the M-protein by cycle within each treatment group, 2) the total M-protein decrease at the end of induction with VTD compared to TD and VD and 3) the decrease by cycle comparing VTD vs TD and VTD vs VD.

Concerning the M-protein decrease by cycle within each arm, statistically significant decreases versus the previous cycle were observed in the first 5 cycles of VTD, the first 3 of TD and the first 4 of VD. The serum M-protein reduction at the end of the 6 induction cycles was significantly higher with VTD when compared with TD (p<0.0001) and VD (p<0.0001). Finally, when comparing the serum M-protein decrease between VTD and TD by cycle, the M-protein reduction was significantly higher with VTD in the first 5 cycles and the same analysis between VTD and VD showed that the serum M-protein decrease was significantly higher with VTD in the first 3 cycles (Tables 1 and 2).

Conclusions: In the cycle by cycle analysis VTD continued to improve M-protein reduction significantly in the first 5 cycles. When compared with TD and VD, the M-protein decrease at the end of induction was significantly higher with VTD. Furthermore, in the cycle by cycle comparison there was a significantly higher efficacy of VTD over TD in the first 5 cycles and over VD in the first 3 cycles. Our results suggest a synergistic rather than only an additive effect between thalidomide and bortezomib supporting the use of an IMiD as the drug of choice to be combined with bortezomib and dexamethasone. Finally, our study supports an induction period beyond 3 or 4 cycles when using a bortezomib/IMiD regimen in order to maximize the induction efficacy.