Prognostic Assessment of Surgical Strategies and Extent of Resection in Methotrexate-Based Treatment for Primary Central Nervous System Lymphoma: A Propensity-Matched Analysis.

[BACKGROUND] The prognostic significance of different surgical strategies in primary central nervous system lymphoma (PCNSL) remains unclear. Moreover, the extent of resection (EOR) on clinical outcomes in PCNSL has not yet been evaluated.

[METHODS] Newly diagnosed and methotrexate-treated PCNSL patients from May 1, 2011, to December 31, 2023, were included. Propensity score matching (PSM) (1:1) was implemented to adjust confounders between the resection and biopsy cohorts. Survival outcomes were compared between cohorts using Kaplan-Meier analyses and evaluated using Cox proportional hazards models. Preoperative tumor volume and residual tumor volume were measured based on the contrast-enhanced magnetic resonance imaging to calculate EOR.

[RESULTS] A total of 149 patients in the resection cohort and 128 in the biopsy cohort were included. After PSM, 104 patients remained in each cohort. Biopsy was associated with superior progression-free survival (P < 0.05), with consistent benefit across pathologic subtypes and adjuvant treatment subgroups. Older patients (≥60 years) derived greater benefit from biopsy than younger patients, whereas no obvious advantage of biopsy was observed for superficial or solitary/limited (≤2) lesions. In the matched resection cohort, patients were stratified by the median EOR (67%; range, 24%-93%), with the EOR subcohort showing the poorest prognosis, whereas the EOR subcohort had outcomes comparable to the biopsy cohort.

[CONCLUSIONS] Overall, the biopsy cohort demonstrated superior prognosis, particularly in elderly patients and those with deep-seated lesions; however, the opposite was observed for superficial and solitary/limited lesions, where higher EOR in the surgery cohort yielded outcomes comparable or superior to biopsy.