Summary Clinical data on primary central nervous system (CNS) lymphoma (PCNSL) patients is mostly generated from prospective studies, and many frail real‐world patients are not included. Recently,the diagnosis and treatment of PCNSL patients was confounded by the COVID‐19 pandemic. In particular, treatment with high‐dose cytarabine was linked to increased risk of pneumonia and virus persistence. We report on outcome of the induction regimen R‐MIV (rituximab, methotrexate, ifosfamide, and vincristine) involving intensive administration of high‐dose methotrexate (3.5 g/m2) with ifosfamide, every 2 weeks and rituximab once per week for six doses. The median age and performance status (PS) for 64 patients was 58 years and 2 (PS 3; 22%) respectively. The overall response rate by magnetic resonance imaging/computed tomography (MRI/CT) was 73% (n = 46/63), with an additional 17.5% (n = 11/63) patients without measurable disease at baseline. Grade 3–4 haematological toxicity was low for R‐MIV (neutropenia: 25% and thrombocytopenia: 1%). Three patients (4.7%) died from treatment‐related toxicity. Co‐existence of SARS‐CoV‐2 infection with cytomegalovirus reactivation and the varicella‐zoster virus in two patients was fatal. Fifty patients (78%) were eligible for consolidation. Median progression‐free and overall survival were not reached (median follow‐up: 44 months). In conclusion, the R‐MIV regimen is feasible in routine practice, effective and safe, even during the COVID‐19 pandemic.

Methods: This is an investigator-initiated, single-arm phase II study, and immunocompetent patients with untreated PCNSL were enrolled.Patients were treated with pomalidomide 4 mg once per day on days 1-14, orelabrutinib 150 mg once daily continuously, and rituximab 375 mg/m 2 intravenous once on day 1 of each 21-day cycle. After four cycles, HD-MTX chemotherapy was sequentially added for two additional cycles. The primary endpoint was overall response rate (ORR) at the end of four cycles of POR. Circulating tumor DNA detection of cerebrospinal fluid was performed. This trial was registered at www.clinicaltrials.gov (#NCT 05390749). Findings:The data cut-off date was 15 February 2023, and thirteen patients were enrolled in this study. The median age was 58 years (range 34-79 years). Four patients had eye involvement and one patient had leptomeninges involvement. Eight patients performed ctDNA analysis and all had MYD88 l265p mutation. One patient discontinued the experimental treatment due to treatment-related adverse events (TRAE) and eleven patients finished four cycles of POR and were evaluable. The ORR after four cycles of POR was 90.1% and the complete response rate was 36.4%. Twelve patients who finished at least 1 cycle were evaluated for safety analysis. The most common grade 3 or 4 AEs were neutropenia (G3, n = 1; G4, n = 2), thrombocytopenia (G3, n = 2), rash (G3 n = 1, G4 n = 1), ALT increased (G3, n = 1) and anemia (G3, n = 1). With a median follow-up of 4 months, one patient relapsed and died of PCNSL. Conclusion:This is the first study to treat newly diagnosed PCNSL with a targeted therapy combination before chemotherapy. POR produced a high ORR with good tolerance. This suggested the potential of noncytotoxic first-line therapies for PCNSL.

No abstract