P037 | SURGICAL RESECTION VS STEREOTACTIC BIOPSY IN PRIMARY CENTRAL NERVOUS SYSTEM LYMPHOMA: A MULTI-CENTER RETROSPECTIVE ANALYSIS OF 143 PATIENTS.

Background: The aim of this study was to evaluate the effect of surgical resection and stereotactic biopsy on the overall survival (OS) of 143 patients diagnosed in 4 italian institution with primary central nervous system lymphoma (PCNSL) and to explore the predictors of selection for resection. Methods: A retrospective analysis was performed of 143 patients with PCNSL, which was diagnosed by surgical resection (n=58) or stereotactic brain biopsy (n=85) in 4 departments from January 2000 to April 2025. Clinical characteristics, imaging findings and OS were retrospectively reviewed and compared among these two groups. Results: The median age was 65.1 years, with 75 male and 69 females, there was no sex predominance and the median performance status according to Eastern Cooperative Oncology Group (ECOG) was 2 (range 0-4). The most common clinical manifestation was a motor deficit (98, 69%) followed by cognitive impairment (81, 57%) and sensory disturbances (44, 31%). No significant differences in these characteristics were found in the two groups. With a median follow-up of 92 months, the median OS of all patients was 11.1 months, with a range 0.1-266. The mean OS for patients who underwent surgical resection was 10.1 months, whereas it was 12.9 months for those who had a brain biopsy (p=0.84) (figure A). Patients submitted to surgical resection started chemotherapy treatment after a median of 23 days (range 7-347), while those submitted to stereotactic biopsy started after a median of 18 days (range 5-98). The difference is statistically significant (p=0.046) There was a trend in OS (p=0.07) when combining the location of the disease with the biopsy method, with the best life expectancy for patients with a resected superficial lesion (13 months, range 0.5-266) and the worst for patients with resected deep lesions (10 months, range 1-95) (figure B). The same trend is less evident (p=0.12) when comparing patients with superficial lesions (12 months, range 1-266) compared with deep lesions (10 months, 1-195), regardless of the surgical treatment. In univariate analysis, we identified the tumour not involving deep structures (p=0.00066) as the only predictor of selection for resection. Conclusions: In PCNSL, surgical resection tends to be performed mainly in patients with superficial lesions. Compared to stereotactic biopsy, there is a longer time between surgery and the start of therapy, but survival is the same.