Phase III results on rituximab addition to high-dose methotrexate chemotherapy for newly-diagnosed PCNSL

The results of the multicenter, open-label, randomised phase III trial HOVON105/ALLG NHL24 were recently published in The Lancet Oncology by Jacoline Bromberg from the Erasmus MC Cancer Institute, Rotterdam, NL, and colleagues.

The aim of this trial was to investigate the efficacy and safety of combining rituximab with high-dose methotrexate-based chemotherapy in patients with newly-diagnosed primary central nervous system lymphoma (PCNSL). Although in the last years prognosis has improved for PCNSL patients, outcomes are still poor, with only 30−40% of patients achieving long-term survival. The primary endpoint of this trial was event-free survival (EFS). Secondary endpoints included, progression-free survival (PFS), overall survival (OS), and safety. 

Study design

• N = 199 patients were randomly assigned 1:1 to methotrexate-based chemotherapy (MBVP; n = 100) with or without rituximab (R-MBVP; n = 99) between 2010–2016
• Eligible patients were:
  • Newly-diagnosed with PCNSL, aged 18−70 years, with an Eastern Cooperative Oncology Group (ECOG) performance status of 0−3, and with histologically-proven CD20-positive diffuse large B-cell lymphoma (DLBCL)
  • Alternatively, patients were eligible if they had MRI evidence of a brain parenchymal lesion indicative of lymphoma in combination with evidence of CD20-positive large cell lymphoma in cerebrospinal fluid or vitreous fluid, or in both vitreous fluid and CSF without a brain parenchymal MRI lesion
  • Median age (range): 61 (55−67)
• Dosing:
  • MBVP: In total two 28-day cycles with 3 g per m2 methotrexate on Day 1 and 15 of each cycle. Intravenous administration of 100 mg per m2 teniposide on Day 2 and 3, 100 mg per m2 carmustine on Day 4 (intravenously), and 60 mg per m2 prednisolone on Day 1−5 (orally)
  • R-MBVP: Same as above for MBVP dosing with the addition of 375 mg per m2 rituximab on Day 0, 7, 14, and 21 in cycle one and Day 0 and 14 in cycle two
  • In cases of renal or hepatic dysfunction methotrexate was postponed and subsequent dose reduced occurred
  • Patients with persistent disease in the CSF after the first cycle of MBVP or R-MBVP intrathecal methotrexate was additionally administered
  • After two cycles of MBVP or R-MBVP, patients with a complete response (CR) or partial response (PR) received consolidation with high-dose cytarabine 2000 g per m² intravenously, twice daily on two consecutive days (between 4−6 weeks after last high-dose methotrexate administration)

Key findings

• At data cut-off (June 2017), n = 119 patients were alive (MBVP, 59; R-MBVP, 60)
• Median follow-up (range): 32.9 (23.9−51.5) months for both cohorts
  • Median follow-up for MBVP arm: 32.9 months
  • Median follow-up for R-MBVP arm: 31.9 months
• Ninety percent of patients received both MBVP or R-MBVP cycles and 81% of patients received high-dose cytarabine consolidation
• At one year, EFS was:
  • MBVP: 49% (95% CI, 39−58)
  • R-MBVP: 52% (95% CI, 42−61)
  • Comparison: HR = 1.00; 95% CI, 0.70−43; P = 0.99
• Median EFS was:
  • MBVP: 10.8 months (95% CI, 5.9−3)
  • R-MBVP: 14.9 months (95% CI, 7.0−4)
• At one year, PFS was:
  • MBVP: 58% (95% CI, 48−67)
  • R-MBVP: 65% (95% CI, 55−74)
  • Comparison: HR = 0.77; 95% CI, 0.52−13; P = 0.18
• The number of patients achieving a response after induction chemotherapy was:
  • MBVP: 86%
  • R-MBVP: 86%
• Number of patients achieving CR or unconfirmed CR (CRu) after consolidation with cytarabine:
  • MBVP: 66%
  • R-MBVP: 68%
  • Comparison: odd ratio = 1.08; 95% CI, 0.59−98; P = 0.81
• Within the first year, 21% of patients from each arm died
• Median OS rates were:
  • MBVP: 56.7 months (95% CI, 37.7−not reached)
  • R-MBVP: not reached
  • Comparison: HR = 0.93; 95% CI, 0.59−44; P = 0.74
• One-year OS rates were:
  • MBVP: 79% (95% CI, 69−86)
  • R-MBVP: 79% (95% CI, 69−85)
• Two-year OS rates were:
  • MBVP: 65% (95% CI, 55−74)
  • R-MBVP: 71% (95% CI, 60−79)
• Three-year OS rates were:
  • MBVP: 61% (95% CI, 51−71)
  • R-MBVP: 58% (95% CI, 46−68)

Safety

• Toxicity was similar between the two arms
• Grade ≥ 3 adverse events (AEs) occurred in:
  • MBVP: 58%
  • R-MBVP: 63%
• Infections and infestations were the most frequent Grade 3−5 AE, occurring in:
  • MBVP: 24%
  • R-MBVP: 21%
• Other common Grade ≥ 3 AEs were:
  • Hematological toxicity (MBVP, 15% vs R-MBVP, 12%)
  • Nervous system disorders (MBVP, 10% vs R-MBVP, 15%)
  • Febrile neutropenia (MBVP, 12% vs R-MBVP, 11%)
  • Electrolyte disturbances (MBVP, 9% vs R-MBVP, 11%)
• Serious AEs were reported in:
  • MBVP: 48% (fatal or life threatening for 12% of patients)
  • R-MBVP: 54% (fatal or life threatening for 10% of patients)
• The most common serious AE was infection, with other common ones being:
  • Nervous system disorders (seizures, neurological deterioration)
  • Febrile neutropenia
  • Gastrointestinal disorders
• Sixteen patients stopped protocol treatment due to toxicity (5% MBVP and 11% R-MBVP) 

Conclusions

• Rituximab addition to standard MBVP chemotherapy does not improve EFS, OS or PFS of newly-diagnosed PCNSL patients
• Rituximab addition to MBVP did not lead to increased toxicity