Diagnostic yield after pre-biopsy steroid administration in primary CNS high-grade B-cell lymphoma.

Abstract

7056 Background: Primary central nervous system high-grade B-cell lymphoma (PCNSL) is typically diagnosed via brain biopsy. Corticosteroids are often administered pre-biopsy to reduce vasogenic edema, but are hypothesized to decrease diagnostic yield by inducing lymphocyte apoptosis. This study aims to evaluate whether steroid administration prior to biopsy affects the likelihood of obtaining a diagnostic result in pts with PCNSL. Methods: A retrospective chart review of Brown University Health pts (2015-2024) with PCNSL, including primary CNS diffuse large B cell lymphoma (DLBCL), was conducted. Collected variables included demographics (age, sex, ECOG), steroid dose, biopsy type (excisional, incisional, or stereotactic needle biopsy) and outcome of biopsy (diagnostic vs. undiagnostic). Statistical analysis comparing pre-biopsy steroid (PBS) group with steroid naive group was conducted using Chi-squared, Wilcoxon rank sum and Fisher’s Exact test. Results: A total of 59 pts were found to have biopsy proven PCNSL and were included in statistical analysis. 53 pts had DLBCL and 6 had high grade PCNSL. Baseline demographics and biopsy types did not significantly differ between PBS group and steroid-naive pts (p>0.05). Among pts that received PBS, 35/35 (100%) had a diagnostic biopsy compared with 22/24 (92%) in pts who did not receive steroids. Pre-biopsy steroid use was not associated with a statistically significant difference in diagnostic yield, (Fisher’s Exact p=0.16). As there were no non-diagnostic biopsies in the PBS group, the odds ratio was infinite and the upper confidence bound could not be estimated. Median dose of steroids was 213mg (prednisone equivalents; Q1 93mg, Q3 307mg). As all pts in the PBS group had diagnostic biopsies no analysis based on steroid dose could be conducted. All pts who received steroids had documented concern for vasogenic edema and steroids were not used for empiric treatment. Conclusions: In our cohort, pre-biopsy corticosteroid administration did not appear to affect diagnostic yield. There was no difference in the rate of diagnostic first biopsies between the PBS and steroid naive group. Among pts receiving steroids, all biopsies were diagnostic regardless of timing relative to biopsy or total steroid dose. Due to the small sample size and the universal diagnostic success, statistical comparisons were limited. These findings suggest that steroids may not compromise the diagnostic yield of PCNSL biopsies. Clinical characteristics and biopsy outcomes. Steroid NaiveN = 24 Steroids before Biopsy N = 35 p-value Age 69 (Q1 63; Q3 73) 66 (Q1 57; Q3 75) 0.3 Female 15 (63%) 19 (54%) 0.5 Normal LDH 16 (94%) 12 (67%) 0.088 ECOG 2-4 4 (16.6%) 9 (27.3%) 0.6 DLBCL 21 (88%) 32 (91%) 0.33 Total steroid dose 100-700mg (in Prednisone Equivalents) _ 23 (70%) Steroids within 24 hours of biopsy _ 14 (40%) Stereotactic needle biopsy 18 (75%) 17 (52%) 0.3 Diagnostic Biopsy 21 (91%) 33 (100%) 0.2