Outcomes in Primary CNS Lymphoma Post BEAM Autologous Stem Cell Transplantation: A Single-Center Analysis.

Background Primary CNS lymphoma (PCNSL) is a variant of NHL, involving the brain, leptomeninges, spinal cord, and/or the eyes without any evidence of systemic disease. Optimal therapy in PCNSL is yet to be determined. High dose methotrexate (HD-MTx) forms the mainstay of treatment. Consolidation strategies include radiation, non-myeloablative chemotherapy or myeloablative chemotherapy followed by stem cell rescue (ASCT). Thiotepa-based conditioning is preferred for ASCT based on evidence of superiority over BEAM conditioning. These studies however, are small in size and mostly single center. Typical drug costs for thiotepa exceed $100,000 and may be 3-4 times that of BEAM. We describe our experience at the Mayo Clinic of PCNSL patients who underwent BEAM conditioning. Methods We identified patients with a diagnosis of PCNSL that received BEAM conditioning for ASCT. Baseline characteristics were collected. All the patients were diagnosed before the year 2010 (thiotepa based conditioning preferred clinical practice after). MRI brain before transplant and day +100 were reviewed, and disease status was categorized based on international PCNSL collaborative group consensus (IPCG) guidelines. Overall survival (OS) from the time of diagnosis was estimated using the Kaplan Meier method and compared using the Log-Rank test. Results 25 patients with PCNSL underwent BEAM conditioning for ASCT. The median age at diagnosis was 50 (26-67) years. There was slight male preponderance (n =15; 60%); 22 (88%) had diffuse large B-cell lymphoma with 1 case each of PTCL-NOS, HGBL-NOS, and ALK-negative ALCL. All patients received HD-MTx, and since ASCT consolidation was not standard practice at the time, some patients also received rituximab (n =9), temozolomide (n =5) and radiation therapy (n = 3). Median time to transplant from diagnosis was 6 (2-73) months. 76% patients had achieved CR (n =18) or CRU (n =1) before transplant and 24% (n = 6) had partial response. Day +100 MRI evaluation demonstrated findings consistent with progressive disease in 20% (n = 5). These were patients not in CR prior to ASCT. Median OS for the entire cohort was 144 months (Figure 1). There was a statistically significant difference in OS based on remission status prior to transplant; median OS was 164 vs 31 months (p Discussion BEAM conditioning resulted in a prolonged median OS, especially in the patients achieving CR prior to transplant. Although thiotepa-based regimens are currently preferred over BEAM conditioning for ASCT, there are significant financial implications. There remains a need for more robust clinical trial data comparing the efficacy and safety between these alternative regimens and their pharmacoeconomic impact.