Updated results from the phase III EMN02/HO95 study in patients with newly diagnosed multiple myeloma

The phase III EMN02/HO95 trial (NCT01208766) was designed to compare both the efficacy and safety of bortezomib (V) plus melphalan (M) and prednisolone (P; VMP) with autologous hematopoietic stem cell transplant (auto-HSCT) as an intensification therapy, and whether consolidation therapy with V, lenalidomide (Len, R) and dexamethasone (D; VRD) can deepen responses in patients with newly diagnosed multiple myeloma (NDMM).

Results from the intensification therapy (R1) and interim results from the consolidation therapy (R2) randomizations have previously been published¹, with a summary available on the Multiple Myeloma Hub. At the 62nd American Society of Hematology (ASH) Annual Meeting and Exposition, Michele Cavo presented updated analyses from the first randomization in the study with a longer follow-up, and Pieter Sonneveld reported the final analysis of the second randomization.^2,3

Study design

In total, 1,503 adult patients ≤65 years old with symptomatic NDMM were enrolled across 172 European Myeloma Network (EMN) centers. Overall, 1,212 patients were randomized to either VMP (n = 495) or auto-HSCT after high-dose melphalan (n = 702) in R1. Subsequently, patients were randomized to consolidation with VRD (n = 449) or no consolidation (n = 428) in R2, and received 10 mg lenalidomide maintenance daily on Day 1–21 of 28-day cycles until progression or unacceptable toxicity.

The primary outcomes of the study were progression-free survival from R1 (PFS-R1) and from R2 (PFS-R2), and secondary outcomes included overall survival (OS) and safety. For the updated analysis of R1, endpoints included OS, PFS from randomization to progression from the subsequent treatment (PFS2), and time to next treatment (TNT). Outcomes for patients randomized to upfront auto-HSCT and those who received auto-HSCT at the time of progression after primary randomization to VMP (delayed auto-HSCT) were also compared.

Key results

Updated analysis of R1

Table 1 shows outcomes from the longer follow-up of the first randomization. At an extended median follow-up time of 75 months (IQR, 67–84), survival was significantly better for patients randomized to auto-HSCT vs VMP. This benefit was notably the case for patients with unfavorable prognostic characteristics at baseline, such as high-risk cytogenetics. The survival advantage was most remarkable for patients with del(17p). Patients randomized to VMP had a shorter PFS2 of 73 months vs not reached in the auto-HSCT group, and TNT was also lower.

Table 1. Patient outcomes from the longer-term follow-up of R1 (all comparisons were statistically significant [p value < 0.05] in favor of the auto-HSCT arm)²

Patient demographics and baseline characteristics in the delayed ASCT group were comparable with those of patients who were randomized to upfront ASCT. After a median follow-up of 85 months in the upfront auto-HSCT group vs 51 months in the delayed auto-HSCT group, median PFS2 rate was 55% vs 32% (HR, 0.52; 95% CI, 0.40–0.66; p < 0.0001). Moreover, median OS was not reached in the upfront auto-HSCT group vs 81 months in the delayed auto-HSCT group, with OS rates of 69% vs 58% (HR, 0.68; 95% CI, 0.51–0.93; p = 0.0164).

Overall, 600 patients in the auto-HSCT group and 377 in the VMP group received maintenance therapy with lenalidomide:

• Median duration of therapy was 34.3 months (IQR, 13.3–62.8)
• Median PFS from start of maintenance was 56 vs 42 months (HR, 0.77; 95% CI, 0.65–0.91; p = 0.0017)
• OS was 74% vs 69% (HR, 0.84; 95% CI, 0.66–1.08; p = 0.1694)
• The most frequent causes of discontinuation were progressive disease (63%), treatment-related adverse events (27%), and second primary malignancies (4%).

Final analysis of R2

At the time of the final analysis, the median follow-up time from R2 was 71.3 months (IQR, 63–80), and 512 events had been reported for PFS-R2. Response status at the time of R2 was equal between patients randomized to receive VRD consolidation vs no consolidation, with 20% patients achieving a complete response or better, 67% a very good partial response or better, and 92% a partial response or better.

Outcomes after the second randomization are shown in Table 2. Longer PFS-R2 with adjustment for R1 was seen in patients randomized to VRD consolidation compared with no consolidation (HR, 0.80; 95% CI, 0.67–0.95; p = 0.013), consistent with interim analyses.

Table 2. Patient outcomes from the final analysis of R2³

Cox regression analysis revealed that a revised international scoring system (ISS) stage III (HR, 2.05; 95% CI, 1.43–2.92) and ampl1q (HR, 1.68; 95% CI, 1.38–2.06) were adverse prognostic factors at the time of diagnosis. Furthermore, the PFS benefit was observed across subgroups, including revised ISS stage I–III, standard-risk cytogenetics, and randomization to VMP or auto-HSCT in R1. Overall response rates were significantly improved in patients randomized to VRD consolidation vs no consolidation.

Median duration of maintenance with lenalidomide was 33 months (0–97+ months), with 32% of patients on treatment 5 years after maintenance initiation. A longer follow-up period is required before OS can be accurately evaluated.

Overall, Grade 4 adverse events were reported in 5% of patients randomized to VRD consolidation, of which the majority were neutropenia (2%), and thrombocytopenia (2%). Cumulative incidence of second primary malignancies 6 years from R2, excluding superficial skin cancer, was 5–6% in both randomization groups.

Conclusion

A longer follow-up of patients in this study confirmed that auto-HSCT gives better survival outcomes than the VMP treatment, and this benefit is retained in patients with unfavorable prognostic indicators. Furthermore, patients who initially received VMP but progressed to auto-HSCT showed poorer survival than those who received auto-HSCT upfront. The authors concluded that these data support the continued use of upfront auto-HSCT in NDMM. In addition, preceding lenalidomide maintenance with VRD consolidation treatment improved PFS and achieved deeper responses, with an acceptable toxicity rate.

Recruitment for this study is complete; however, follow-up continues until 10 years postenrollment, with further analyses anticipated.