Immunochemotherapy with intensive consolidation for primary CNS lymphoma: a pilot study and prognostic assessment by diffusion-weighted MRI.

PURPOSE

We evaluated a novel therapy for primary central nervous system lymphoma (PCNSL) with induction immunochemotherapy with high-dose methotrexate, temozolomide, and rituximab (MT-R) followed by intensive consolidation with infusional etoposide and high-dose cytarabine (EA). In addition, we evaluated the prognostic value of the minimum apparent diffusion coefficient (ADC(min)) derived from diffusion-weighted MRI (DW-MRI) in patients treated with this regimen.

EXPERIMENTAL DESIGN

Thirty-one patients (median age, 61 years; median Karnofsky performance score, 60) received induction with methotrexate every 14 days for 8 planned cycles. Rituximab was administered the first 6 cycles and temozolomide administered on odd-numbered cycles. Patients with responsive or stable central nervous system (CNS) disease received EA consolidation. Pretreatment DW-MRI was used to calculate the ADC(min) of contrast-enhancing lesions.

RESULTS

The complete response rate for MT-R induction was 52%. At a median follow-up of 79 months, the 2-year progression-free and overall survival were 45% and 58%, respectively. For patients receiving EA consolidation, the 2-year progression-free and overall survival were 78% and 93%, respectively. EA consolidation was also effective in an additional 3 patients who presented with synchronous CNS and systemic lymphoma. Tumor ADC(min) less than 384 × 10(-6) mm(2)/s was significantly associated with shorter progression-free and overall survival.

CONCLUSIONS

MT-R induction was effective and well tolerated. MT-R followed by EA consolidation yielded progression-free and overall survival outcomes comparable to regimens with chemotherapy followed by whole-brain radiotherapy consolidation but without evidence of neurotoxicity. Tumor ADC(min) derived from DW-MRI provided better prognostic information for PCNSL patients treated with the MTR-EA regimen than established clinical risk scores.