Carfilzomib in Relapsed or Refractory Multiple Myeloma Patients with Early or Late Relapse Following Prior Therapy: An Analysis of Overall Survival in Subgroups from the Randomized Phase 3 Aspire and Endeavor Trials

Introduction: Carfilzomib is an irreversible proteasome inhibitor with proven efficacy as a single agent and in doublet and triplet combinations (Siegel et al, Blood. 2012;120:2817-25; Dimopoulos et al, Lancet Oncol. 2017;18:1327-37; Siegel et al, J Clin Oncol. 2018;36:728-34). The phase 3 ASPIRE and ENDEAVOR trials demonstrated that treatment with carfilzomib-based regimens led to superior efficacy outcomes (progression-free survival [PFS], overall survival [OS], and overall response rate [ORR]) compared with standard regimens in patients (pts) with relapsed or refractory multiple myeloma (RRMM) (ASPIRE: carfilzomib-lenalidomide-dexamethasone [KRd] vs lenalidomide-dexamethasone [Rd]; ENDEAVOR: carfilzomib-dexamethasone [Kd] vs bortezomib-dexamethasone [Vd]). A previous subanalysis of these trials found that carfilzomib improved PFS and ORR, regardless of whether pts experienced an early or late relapse following the immediate prior therapy (Mateos et al, ASH 2017). Here we performed post hoc analyses of ASPIRE and ENDEAVOR to examine OS and updated safety in the ASPIRE and ENDEAVOR early or late relapse subgroups.

Methods: ASPIRE and ENDEAVOR enrolled pts with RRMM (1-3 prior lines of therapy). Pts in both studies received carfilzomib twice weekly in 28-day cycles. In ASPIRE, pts were randomized to KRd or Rd, and those in the KRd arm received carfilzomib 27 mg/m², which was discontinued after cycle 18. In ENDEAVOR, pts were randomized to Kd or Vd. The Kd group received carfilzomib 56 mg/m²; the Vd group received 21-day bortezomib cycles (1.3 mg/m²). In ENDEAVOR, treatment was continued until progression or unacceptable toxicity. Early relapsers were defined as pts who relapsed ≤1 year after initiating the most recent prior line of therapy (as assessed by investigator), while late relapsers were those who relapsed after >1 year. OS was summarized via Kaplan-Meier methods. In this post hoc analysis, P values were calculated for exploratory purposes. Data cutoff dates used here were April 28, 2017 for ASPIRE and July 19, 2017 for ENDEAVOR.

Results: In ASPIRE, median OS for early relapsers was 36.0 months for KRd vs 27.7 months for Rd (hazard ratio [HR]: 0.807; 95% confidence interval [CI]: 0.586-1.110; P=0.0935) (Figure 1). For late relapsers in ASPIRE, median OS was 53.2 months for KRd vs 41.2 months for Rd (HR: 0.752; 95% CI: 0.606-0.932; P=0.0046). Rates of grade ≥3 treatment-emergent adverse events (TEAEs) in ASPIRE were similar for early and late relapsers (KRd vs Rd, 89.3% vs 82.0% for early relapsers and 86.9% vs 83.0% for late relapsers). In ENDEAVOR, early relapsers had a median OS of 28.6 months for Kd vs 21.7 months for Vd (HR: 0.807; 95% CI: 0.587-1.109; P=0.0920) (Figure 2). For late relapsers in ENDEAVOR, median OS was not evaluable (NE) for Kd vs 42.3 months for Vd (HR: 0.722; 95% CI: 0.576-0.905; P=0.0023). Grade ≥3 TEAEs (Kd vs Vd) in ENDEAVOR occurred in 77.0% vs 77.0% of early relapsers and 83.6% vs 69.0% of late relapsers.

Conclusions: RRMM pts who received KRd and Kd had longer OS compared with those who received Rd and Vd, regardless of whether they had an early or late relapse following the most recent prior line of therapy. Late relapsers had a numerically greater magnitude of OS benefit with KRd and Kd compared with control arms than early relapsers. Rates of grade ≥3 AEs were consistent with those previously reported in ASPIRE and ENDEAVOR for the overall population. In conclusion, these findings underscore the impressive efficacy of carfilzomib-based therapy for the treatment of pts with RRMM.

Figure 1.

View largeDownload slide

Figure 1.

View largeDownload slide

Close modal