Secondary analysis of the ENDEAVOR study: carfilzomib-dexamethasone vs SC or IV bortezomib-dexamethasone, in RRMM

The ENDEAVOR study was the first head to head trial of two proteasome inhibitors (PIs) in combination with dexamethasone – carfilzomib versus bortezomib, in patients with Relapsed and Refractory Multiple Myeloma (RRMM). The initial results of this study, an updated analysis via cytogenetic risk status and an overall survival (OS) analysis, were previously covered by the MM Hub. In the latest analysis, a secondary assessment was carried out to specifically compare carfilzomib-dexamethasone with two different means of bortezomib administration: either subcutaneous (SC) or intravenous (IV). Previous analyses have suggested that both routes of administration are comparable, but that SC gives more favorable safety outcomes.

The latest analysis was published in the September edition of Leukemia and Lymphoma, by Hartmut Goldschmidt from the University Hospital of Heidelberg, Germany, and international colleagues.

Key Findings: 

• Patients (pts) = 929; randomized to receive either:
  • carfilzomib-dexamethasone (Kd) n=464 or
  • bortezomib-dexamethasone (Vd) n = 465 (30 were excluded from treatment)
• Of the 435 Vd-treated pts: SC = 360 and IV = 75
• Of the 464 Kd-treated pts: SC = 356 (termed Kd [SC Vd]), IV (termed Kd [IV Vd]) = 108
• PFS was longer for patients treated with Kd compared with SC Vd or IV Vd
• Median PFS: Kd [SC Vd] = not reached vs SC Vd = 9.5 months; HR = 0.58 (95% CI, 0.46–0.72); P<0.0001
• PFS: Kd [SC Vd] vs SC Vd in pts with
  • prior bortezomib exposure: HR = 0.66 (95% CI, 0.50–0.87); P = 0.0018
  • bortezomib-naïve patients: HR = 0.48; 95% CI, 0.34–0.70; P < 0.0001
• Median PFS: Kd [IV Vd] = 22.2 vs IV Vd = 8.5 months; HR = 0.38 (95% CI 0.24–0.61); P < 0.0001
• Median PFS IV pt cohorts: 
  • prior bortezomib exposure: HR = 0.32 (95% CI, 0.16–0.62; P = 0.0002 in favor of Kd (IV Vd)
  • bortezomib-naïve pts: HR = 0.47 (95% CI, 0.25–0.88); P = 0.0084 in favor of Kd (IV Vd)
• 1- and 2- year PFS were higher for Kd vs Vd treatment arm, irrespective of IV or SC bortezomib administration or prior exposure
• ORR:
  • Kd [SC Vd] = 75.8%, SC Vd = 64.4%, Kd [IV Vd] = 80.6%, IV Vd = 58.7%
  • prior bortezomib exposure: Kd [SC Vd] = 70.4%, SC Vd = 62.1%, Kd [IV Vd] = 74.1%, IV Vd = 57.6%
  • bortezomib-naïve pts: Kd [SC Vd] = 82.5%, SC Vd = 67.5%, Kd [IV Vd] = 87.0%, IV Vd = 59.5%
• Rates of complete response or better: Kd = 11.8–14.8%; SC Vd = 6.1%, IV Vd = 6.7%
• Median DORs: Kd [SC Vd] = not reached, SC Vd = 11.1 months, Kd [IV Vd] = 21.3, IV Vd = 10.0 months
• Rate of grade ≥2 PN: Kd [SC Vd] = 6.5%, SC Vd = 33.3%, Kd [IV Vd] = 4.6%, IV Vd = 21.3%
• Rates of any grade herpes zoster: Kd [SC Vd] = 0.3%, SC Vd = 4.2%, Kd [IV Vd] = 0.9%, IV Vd = 1.3%; bortezomib exposed pts = 0–0.5% for Kd vs 4.4% (SC Vd) or 3.0% (IV Vd)
• Any grade ≥3 AE:  Kd = 69.4–74.4%; Vd = 64.0–67.5%
• Grade ≥3 hypertension: Kd = 8.7–12.0%, Vd = 1.3–2.8%
• Grade ≥3 cardiac failure: Kd = 4.6–4.8%, Vd = 1.3–1.7%
• Grade ≥3 acute renal failure: Kd = 3.7–4.2%, Vd = 2.2–2.7%
• Discontinuation due to adverse events (AEs) similar for Kd vs Vd, although IV Vd-treated patients experienced the lowest rate

In conclusion, carfilzomib treatment (Kd) led to a greater PFS and ORR compared with Vd treatment, irrespective of the means of bortezomib administration or prior exposure. However, whilst lower rates of PN were observed with Kd treatment, it did result in higher levels of hypertension, as well as cardiac and renal failure. Whilst hypertension is manageable, the benefits for a given patient need to be balanced against the risks of more severe AEs when choosing this regimen.

Although the study was not originally powered to compare SC Vd vs IV Vd, and Kd [SC Vd] vs Kd [IV Vd], SC Vd-treated patients and Kd-treated patients pre-selected for SC Vd had a longer PFS compared with IV Vd-treated patients and Kd-treated patients who were preselected for IV Vd. This was true regardless of patient age or number of prior treatments. However, the authors caution that this study was exploratory and not statistically powered, and therefore such variations could be due to random variation.