Background
The ANOCEF-GOELAMS Elderly trial was a phase II randomized trial comparing two non-ASCT HD-MTX–based regimens in elderly patients (age >60) with newly diagnosed PCNSL, a population unable to tolerate ASCT consolidation and representing the majority of PCNSL patients given the median age of onset. Compared regimens: MTX + temozolomide (MT, avoiding WBRT) and MTX + procarbazine + vincristine + cytarabine (MPVA). The trial addressed the unmet need for effective, tolerable PCNSL treatment in the elderly without ASCT or WBRT.
Interventions and follow up
Arm A: Methotrexate 3 g/m² IV day 1 + temozolomide 150 mg/m² PO days 1–5 (MT), q28d × 6 cycles → observation (no WBRT, no ASCT)
Arm B: Methotrexate 3 g/m² IV day 1 + procarbazine PO days 1–7 + vincristine IV + cytarabine 2 g/m² IV q12h days 2–3 (MPVA), q28d × 6 cycles → observation
Population: n=98 (49 MT, 49 MPVA); 1:1; age >60; KPS ≥50; histologically confirmed PCNSL
Primary endpoint: 1-year event-free survival; secondary: OS, CR rate, safety, QoL
mFollow up: 30 months
Arm B: Methotrexate 3 g/m² IV day 1 + procarbazine PO days 1–7 + vincristine IV + cytarabine 2 g/m² IV q12h days 2–3 (MPVA), q28d × 6 cycles → observation
Population: n=98 (49 MT, 49 MPVA); 1:1; age >60; KPS ≥50; histologically confirmed PCNSL
Primary endpoint: 1-year event-free survival; secondary: OS, CR rate, safety, QoL
mFollow up: 30 months
Results
1-yr EFS: MT 36% vs MPVA 50% (HR 0.68, P=.17, not significant)
CR rate: MT 29% vs MPVA 37% (P=.44, NS)
mOS: MT 14 months vs MPVA 31 months (HR 0.64, P=.09, trend favoring MPVA)
Tolerability: grade ≥3 toxicity and treatment delays higher with MPVA
CR rate: MT 29% vs MPVA 37% (P=.44, NS)
mOS: MT 14 months vs MPVA 31 months (HR 0.64, P=.09, trend favoring MPVA)
Tolerability: grade ≥3 toxicity and treatment delays higher with MPVA
Adverse events
Overall grade ≥3: MPVA 78% vs MT 53% (P=.01)
Hematologic grade ≥3: neutropenia MPVA 65% vs MT 29%; thrombocytopenia MPVA 44% vs MT 14%
Discontinuation/other: AE-related discontinuation MPVA 24% vs MT 10%; vincristine peripheral neuropathy (all grades) MPVA 10%; grade ≥3 MTX renal toxicity ~4% both arms
Deaths: 2 treatment-related deaths in MPVA arm vs 0 in MT arm
Hematologic grade ≥3: neutropenia MPVA 65% vs MT 29%; thrombocytopenia MPVA 44% vs MT 14%
Discontinuation/other: AE-related discontinuation MPVA 24% vs MT 10%; vincristine peripheral neuropathy (all grades) MPVA 10%; grade ≥3 MTX renal toxicity ~4% both arms
Deaths: 2 treatment-related deaths in MPVA arm vs 0 in MT arm
Conclusions
MT and MPVA showed numerically different but statistically non-significant EFS and OS in elderly PCNSL, with MPVA producing higher toxicity; neither matched outcomes seen with ASCT-based strategies in younger patients, defining the ongoing unmet need for effective, tolerable treatment in elderly PCNSL. MT provides a more tolerable backbone for the frailest patients.
Key Limitations
Phase II, small sample (n=98), underpowered for OS (P=.09 trend). Primary endpoint EFS at 1 year not significant (P=.17). MPVA's higher cytarabine intensity worsens the efficacy-toxicity balance in elderly patients. No WBRT maintenance arm (deliberately avoided for neurotoxicity), and no randomization to observation/best supportive care. Age ≥60 is heterogeneous; fitter patients aged 60–70 may tolerate MATRix. Rituximab was not part of the MT arm, reflecting limited PCNSL rituximab evidence at the time.
Clinical Context
This trial highlights the challenge of treating elderly PCNSL patients ineligible for ASCT. MT is a tolerable option for the oldest/frailest, while addition of rituximab improves outcomes. In contemporary practice, elderly PCNSL patients are stratified by fitness: fit elderly may receive modified HD-MTX-based regimens; frail elderly receive MT ± rituximab or MTX monotherapy. BTK inhibitors (ibrutinib monotherapy ~67% ORR in R/R PCNSL) and HD-MTX combinations are under study. IELSG-32, PRECIS, and ANOCEF-GOELAMS together define PCNSL treatment stratification by age and fitness.