Stereotactic Radiosurgery as an Effective and Safe Treatment Modality in Central Nervous System Lymphoma, a Single Institution Experience

Abstract

CNS lymphoma (CNSL) is an aggressive liquid cancer with the potential for rapid progression. Standard treatment regimens for CNSL have included chemotherapy and whole brain radiotherapy (WBRT). WBRT carries a high risk of leukoencephalopathy (LE) manifesting as cognitive decline and is identifiable on brain MRI as fluid attenuation inversion recovery (FLAIR) changes. Stereotactic radiosurgery (SRS) can precisely deliver higher biologically effective doses of radiation, while minimizing damage to surrounding healthy brain tissue. SRS offers enhanced local control in CNSL with minimal risk of LE. Patients undergoing SRS at an NCI-designated comprehensive cancer center from 2000 to 2022 were prospectively enrolled into a database. Twenty-nine consecutive patients diagnosed with relapsed primary CNSL (59%) or secondary CNSL (41%) were included in this study. Median age at diagnosis was 70 (range 23-91) years and 14 were female. Follow up MRI imaging was obtained at 1-, 3-, 6- and 12-months post-procedure. Measurement of tumor volumes was performed at treatment and each subsequent follow-up by image co-registration and tumor segmentation. Pre and post SRS brain MRIs were independently reviewed by a neuro-radiologist to score the grade of LE for each patient. Progression-free survival (PFS) and overall survival (OS) were calculated using Kaplan-Meier analyses and Log-rank tests using statistical software. Twenty-nine patients treated for 50 tumors were studied. Twenty-four patients were treated for a single tumor, and 5 for multiple tumors. Six patients underwent repeat SRS treatment for 10 out-of-field recurrences and 1 patient for an in-field recurrence. Of the 35 SRS procedures performed, 32 (88%) were single session delivery and 3 were hypofractionated. Stereotactic immobilization was achieved with frame (57%) or thermoplastic mask (43%). Median prescribed dose was 14 Gy (range 10-21) to the 50% isodose line (range 40-100). Median coverage, selectivity and gradient index were 0.97, 0.69, and 2.9 respectively. Median follow-up time was 4.5 months (range 0.5-64). Twenty-six (52%) tumors completely responded to treatment with no residual enhancement, while the rest decreased in volume. No tumor failed to respond to treatment. Median time to best volumetric response was achieved in 2.9 months (range 1-6.1) and median volume of best response was zero cc (range 0-1.19). The median OS of the cohort was 7.0 months (95% CI 5.1-8.9), while the median PFS was 5.0 months (95% CI 1.2-8.8). Median score of LE was zero, and only one patient experienced grade 3 MRI FLAIR changes. With a median time to response of 2.9 months, a 100% initial response rate and a single in-field failure, our analysis demonstrates a rapid and effective response of CNSL to SRS. Distant failures were controlled with repeat SRS. The absence of LE in the treated patients provides further evidence for safety and tolerability of SRS in CNSL.