Autologous stem cell transplantation versus bortezomib-melphalan-prednisone for newly diagnosed multiple myeloma patients: Final analysis of the phase 3 EMN02/HO95 study

Abstract

The multicenter, randomized, phase 3 EMN02/HO95 trial was aimed to prospectively compare (randomization 1, R1) standard dose intensification therapy with bortezomib-melphalan-prednisone (VMP) vs high dose melphalan plus single or double autologous stem cell transplantation (ASCT) after a bortezomib-based induction in patients (pts) with newly diagnosed multiple myeloma (NDMM). A second randomization (R2) was designed to evaluate the role of short term consolidation treatment with bortezomib-lenalidomide-dexamethasone) vs no consolidation, to be followed by lenalidomide maintenance until progression or toxicity. Progression-free survival (PFS) from R1 and R2 were the primary study end points. Herein we report the results of the final analysis of the study for R1 population. From February 2011 to April 2014, 1503 pts were enrolled; among them, 1197 resulted eligible for R1 and were randomized between VMP (n=495) and ASCT (n=702) arms. Median age was 58 years in both groups; ISS stage III was 20% in ASCT arm and 21% in VMP arm; Revised-ISS stage III was 8% in both groups. An high risk cytogenetic profile (HiR-cyto), defined by the presence of at least one among del(17p), t(4;14) and t(14;16) as detected by FISH analysis, was detected in 25% of pts in both groups. After a median follow-up (mFU) of 60 (52-68) months from R1, on an intention-to-treat basis, median PFS was 42 months in the VMP-group vs 57 months in the ASCT group (HR=0.73, CI=0.63-0.85, p=0.0001). Prespecified subgroup analyses confirmed that PFS-benefit with ASCT was not influenced by revised stage (stage II: p=0.001; stage III: p=0.001), HiR-cyto (present p=0.006; absent: p=0.001); age (>55 years: p=0.028; <0.001). The probability of achieving >= best VGPR was 85% in the ASCT group and was 78% in the VMP group (p=0.012). In a multivariate Cox regression analysis, randomization to ASCT (HR=0.67; CI=0.55-0.80; p<0.001), best >=VGPR (HR=0.38; CI=0.30-0.47; p<0.001), R-ISS I vs III (HR=0.43; CI=0.30-0.63; p<0.001), absence of HiR-cyto (HR=0.70; CI=0.56-0.87; p=0.014), were independent predictors of prolonged PFS. Overall survival rate at mFU was 72% in VMP and 75% in ASCT arm (p=0.359). In conclusion, the final analysis confirmed that in comparison with VMP, ASCT increased the rate of high quality responses and extended PFS in pts with standard-risk and high-risk features. Upfront ASCT still represent the reference treatment choice for NDMM pts in the era of bortezomib-based therapies.