Prebiopsy Steroids and Diagnostic Yield in Patients With Diffuse Large B-Cell Lymphoma.

Importance: Corticosteroids are frequently used to manage symptoms in patients with diffuse large B-cell lymphoma (DLBCL), but longstanding clinical caution suggests that prebiopsy administration may reduce diagnostic accuracy. The association between steroids and biopsy outcomes in patients with systemic DLBCL remains unexamined. Objective: To determine whether corticosteroid use before biopsy is associated with diagnostic yield in patients with newly diagnosed DLBCL. Design, Setting, and Participants: Retrospective case series of adult patients newly diagnosed with DLBCL at a single institution tertiary academic medical system affiliated with Brown University between 2015 and 2024. Patients were identified through institutional hematology and pathology databases. Patients with relapsed or refractory DLBCL or primary central nervous system lymphoma were excluded. Logistic regression was used to assess the association between steroid exposure (dose, duration, and withdrawal) and diagnostic yield. Exposure: Corticosteroid administration within 30 days before tissue biopsy. Main Outcomes and Measures: Diagnostic yield of initial biopsy for DLBCL, repeat biopsy rates, and time from first negative biopsy to chemotherapy initiation. Results: Of 320 patients (mean [SD] age, 68 [14] years and 164 [51%] female), 48 (15%) received corticosteroids prebiopsy. Diagnostic yield was 83% (40 of 48 patients) in the steroid group and 81% (221 of 272 patients) in the nonsteroid group, with no statistically significant difference between the groups (prevalence ratio [PR], 1.03; 95% CI, 0.89-1.18). Among patients with negative first biopsies, time to treatment initiation did not differ significantly between groups. No significant associations were observed between diagnostic yield and total steroid dose, duration, or days off steroids before biopsy. Biopsy type was significantly associated with diagnostic yield. Compared with excisional and incisional biopsies, core needle and fine needle aspiration were 12% (PR, 0.88; 95% CI, 0.81-0.96; P = .01) and 69% (OR, 0.31; 95% CI, 0.18-0.63; P < .001) less likely to be diagnostic, respectively. Conclusions and Relevance: In this case series of 320 patients with DLBCL, prebiopsy corticosteroid administration was not associated with reduced diagnostic yield or delay in treatment in patients with newly diagnosed DLBCL. Biopsy technique was most associated with diagnostic success. These findings support a clinically pragmatic approach to steroid use during lymphoma evaluation.