Outcome with Lenalidomide Plus Dexamethasone Followed By Early Autologous Stem Cell Transplantation in the ECOG-ACRIN E4A03 Randomized Clinical Trial: Long-Term Follow-up

Background:  Immunomodulatory agents and proteasome inhibitors have significantly improved outcomes in the management of patients with multiple myeloma (MM). As a consequence, the role of upfront autologous stem cell transplant (ASCT) has become more controversial. The ECOG-ACRIN Cancer Research Group trial E4A03 demonstrated the superiority of lenalidomide (LEN) and low-dose dexamethasone (DEX; Ld) over LEN and high dose DEX (LD) in newly diagnosed MM (Rajkumar SV, Lancet Oncol, 2010). This led to the universal acceptance of Ld as a standard induction regimen.  More recently, the FIRST study demonstrated the superiority of Ld as continuous therapy vs Ld given x 18 cycles and melphalan-thalidomide-prednisone x 12 cycles proving continuous Ld should be considered standard of care (Benboubker L, NEJM, 2014).  Upon completion of four cycles of therapy, patients enrolled in E4A03 had the option of proceeding to ASCT or continuing on their assigned Ld or LD.  A prior post-hoc retrospective landmark analysis (Siegel DS, Abstract, Blood 2010) comparing outcomes of early ASCT vs. no early ASCT showed that in patients <65, 65-70, and >70, the survival with early ASCT at 3 years appeared higher, supporting the role of early consolidative ASCT in newly diagnosed patients. This analysis is the long-term follow-up of the aforementioned study.  

Methods: A post hoc, retrospective landmark analysis was performed evaluating overall survival (OS) by early ASCT vs. no early ASCT overall and within age subgroups. The Kaplan-Meier method was used to estimate survival distributions and Cox regression for hazard ratio (HR) estimates. The landmark analysis included only patients surviving the first four cycles of therapy.

Results: 21% of patients opted for early ASCT (n=90). At baseline (four cycles prior), early ASCT patients were younger (median 57.5y vs 66y), more fit (ECOG PS 0 45% vs 56%) and with less aggressive disease (ISS Stage 3 12% vs 28%). Median treatment duration was 7.6m for no early ASCT patients. The 1, 2, 3, 4, and 5-year survival probability estimates were higher for early ASCT vs. no early ASCT at 99%, 93%, 91%, 85%, and 80% vs. 94%, 84%, 75%, 65%, and 57%. The median OS in the early ASCT vs. no early ASCT was not reached vs. 5.8 years, respectively (Table 1). The survival hazard ratio (HR) for early ASCT was 0.55 [95%CI (0.38-0.80)]. In a further subgroup analysis of patients </= 65, 5-yr survival probability of those who had early ASCT vs. no early ASCT was 79% [95%CI (67-87)] vs. 62% [95%CI (54-69)], respectively.  In patients </= 65, median OS in the early ASCT vs. no early ASCT was not reached vs. 7.4 years and the survival HR was 0.74 [95%CI (0.48-1.15)]. For patients >65, 5-yr survival probability of those who had early ASCT vs. no early ASCT was 83% [95%CI (56-94)] vs. 52% [95%CI (45-59)], respectively. In patients >65, median OS in the early ASCT vs. no early ASCT was not reached versus 5.2 years and the survival HR was 0.38 [95%CI (0.17-0.87)]. In all age groups, overall response rate was similar prior to the decision of proceeding with ASCT vs. no early ASCT. In a multivariable model adjusting for baseline prognostic factors, early ASCT retained marginal significance.  

Conclusions: The landmark analysis demonstrated that patients opting for ASCT after induction Ld/LD had a higher 1, 2, 3, 4, and 5- year survival probability and improvement in median OS regardless of DEX dose density. Results were stronger within the older age cohort and upheld in adjusted models. Limitations include lack of randomization and data on salvage therapy. At a time when lenalidomide and dexamethasone as induction is gaining worldwide acceptance in all age groups, early ASCT is safe and efficacious and should remain a standard of care for newly diagnosed MM even in the era of novel therapies.  

Table 1: Survival Probabilities (Time from Registration) for all Patients Included in the Landmark Analysis

Median

1 year

2-year

3-year

4-year

5-year

N

Events

Median (Yrs) (95% CI)

Events

Survival Probability (95% CI)

Events

Survival Probability (95% CI)

Events

Survival Probability (95% CI)

Events

Survival Probability (95% CI)

Events

Survival Probability (95% CI)

No Early ASCT

All

341

195

5.78 (5.11 – 6.59)

22

0.94 (0.90 – 0.96)

53

0.84 (0.80-0.88)

84

0.75 (0.70-0.79)

119

0.65 (0.59-0.69)

144

0.57 (0.51 – 0.62)

Early ASCT

All

90

34

NR

1

0.99 (0.92 – 1.00)

6

0.93 (0.86 – 0.97)

8

0.91 (0.83 – 0.95)

13

0.85 (0.76 – 0.91)

18

0.80 (0.69 – 0.87)

Legend:
NR  = not reached