ASH 2018 | Primary analysis of the phase III COMPLEMENT A+B trial: Ofatumumab plus bendamustine for indolent NHL

On Sunday 2 December 2018, during the 60^th Annual Meeting of the American Society of Hematology (ASH), San Diego, CA, Mathias Rummel from Justus-Liebig University, Giessen, DE, presented the primary analysis of the phase III trial COMPLEMENT A+B (Abstract #450; Oral Session #623).

COMPLEMENT A+B is an open-label, global, multicenter, randomized, phase III study that compared the efficacy and safety of ofatumumab plus bendamustine versus bendamustine alone, in adult patients with indolent non-Hodgkin lymphoma (NHL). The primary endpoint of the study was progression-free survival (PFS) as assessed by Independent Review Committee (IRC). Secondary endpoints included, response rates and overall survival (OS) in all patients and in patients with follicular lymphoma (FL), as well as safety and tolerability.

Study design

• N = 346 adult patients, refractory to rituximab with ≥ 2 clearly demarcated lesions (long axis: > 1.5 cm, short axis: ≥ 1 cm) with one of the following indolent NHL subtypes:
  • FL grade 1−3a
  • Marginal zone lymphoma (MZL)
  • Lymphoblastic lymphoma (LL)
  • Small lymphocytic lymphoma (SLL)
• Patients were stratified by type of last rituximab (RTX) therapy (RTX + chemotherapy or RTX alone) and by prior exposure to bendamustine (yes or no)
• Patients were then randomized 1:1 to either:
  • Bendamustine + ofatumumab (Benda + OFA; n = 173): 90 mg/m² bendamustine on Days 1 and 2 of each cycle for up to eight (21-day) cycles. Ofatumumab was administered at 1000 mg on Day 1 of eight cycles of bendamustine and then every 28 days (four doses; total 12 doses)
  • Bendamustine (Benda; n = 173): 120 mg/m² bendamustine on Days 1 and 2 of each 21-day cycle for up to eight cycles
• Response assessments were performed on Days 84, 168 and 252, then quarterly for 18 months and then yearly until 54 months.
• Primary analysis was considered until ~215 PFS events, as determined by the IRC
• Bendamustine administration surpassing one year not resulting in at least a partial response (PR) lasting for ≥ 6 months, was not allowed
• Baseline characteristics were well matched between the two arms
  • Benda + OFA:
    • Median age (range): 61 (29−87)
    • Sex: 42% females
  • Benda:
    • Median age (range): 63 (21−86)
    • Sex: 41% females

Results

• Median follow-up (range): 61 (20−88) months
• Median treatment duration (range)
  • Benda + OFA: 260 (1−367) days
  • Benda: 135 (2−250) days
• Number of patients completing treatment:
  • Benda + OFA: 58% of patients completed OFA treatment and 65% completed Benda treatment. Three, six, and twelve treatment cycles were completed by 94%, 81%, and 59% of patients, respectively
  • Benda: 43% of patients completed treatment. Three, six, and twelve treatment cycles were completed by 91%, 65%, and 42% of patients, respectively
• PFS rates within 24 months:
  • Benda + OFA: 51% (44% of patients progressed and 8% died)
  • Benda: 58% (50% of patients progressed and 9% died)
• Median PFS after 217 events:
  • Benda + OFA: 16.7 months
  • Benda: 13.8 months
  • Comparison: HR = 0.82; (95% CI, 062−07); P = 0.1390
• Median PFS in FL patients:
  • Benda + OFA: 16.6 months
  • Benda: 12.1 months
  • Comparison: HR = 0.76; (95% CI, 0.55−08); P = 0.1076
• Median OS in all patients:
  • Benda + OFA: 58.2 months
  • Benda: 51.8 months
  • Comparison: HR = 0.89; (95% CI, 0.63−25); P = 0.497
• IRC-assessed response rates (Cheson criteria):
  • Benda + OFA (n = 173):
    • Complete response (CR): 21%
    • PR: 53%
    • Stable disease (SD): 13%
    • Progressive disease (PD): 6%
    • Not evaluable (NE): 6%
    • Not available (NA): 1%
    • Missing: 1%
  • Benda (n = 173):
    • CR: 17%
    • PR: 58%
    • SD: 9%
    • PD: 8%
    • NE: 7%
    • NA: 1%
    • Missing: 1%
  • Statistical difference in response rates: -1.2 (95% CI, -10.4−1)
• Estimated median time to progression:
  • Benda + OFA: 11.3 months
  • Benda: 11.5 months
• Median time to response (range):
  • Benda + OFA: 2.9 (2.8−9) months (127 events)
  • Benda: 2.9 (2.8−9) months (129 events)
• Duration of response (range):
  • Benda + OFA: 17.7 (13.0−3) months (78 events)
  • Benda: 14.5 (10.8−8) months (87 events)

Safety

• The frequency of death was similar between the two arms:
  • Benda + OFA: 38% of patients died (7% on treatment)
  • Benda: 41% of patients died (9% on treatment)
• The majority of deaths were treatment-unrelated
• A slightly higher incidence of any grade adverse events (AEs) was observed in the Benda arm (99% vs 96% in the Benda + OFA arm)
• The most common Grade ≥ 3 AEs than occurred in > 5% of patients were:
  • Benda + OFA (n = 172):
    • Any event: 73%
    • Neutropenia: 35%
    • Anemia: 8%
    • Leukopenia: 9%
    • Thrombocytopenia: 5%
    • Decreased neutrophil count: 6%
    • Febrile neutropenia: 7%
    • Fatigue: 3%
    • Pneumonia: 5%
    • Lymphopenia: 6%
  • Benda (n = 170):
    • Any event: 80%
    • Neutropenia: 41%
    • Anemia: 15%
    • Leukopenia: 8%
    • Thrombocytopenia: 19%
    • Decreased neutrophil count: 7%
    • Febrile neutropenia: 4%
    • Fatigue: 8%
    • Pneumonia: 6%
    • Lymphopenia: 5% 

Conclusions

• The phase III COMPLEMENT A+B trial did not find any significant difference between Benda + OFA and benda monotherapy in indolent NHL patients, who are unresponsive to RTX therapy
• The results of this trial contradict the ones from the GADOLIN phase III study, which assessed the efficacy of obinutuzumab plus bendamustine versus bendamustine monotherapy
  • This could be due to differences in treatment periods between the two studies and/or due to the fact that obinutuzumab is a type II anti-CD20 antibody and not type I, as ofatumumab
• The safety profiles were consistent with previous reports
• According to the investigations, the limitations of the study were:
  • The bendamustine monotherapy dose (120 mg/m²) could be toxic and may have contributed to the higher rates of treatment discontinuation and the higher rate of Grade ≥ 3 infections in the benda arm
  • Patients were exposed to a short period of first-line therapy prior to entering the study, as median number of prior RTX-containing lines was 1 (range, 1−8) and median time since diagnosis was 2.7 years