CD30 Expression Is Not Acquired at Time of Relapse in Diffuse Large B-Cell Lymphoma

Introduction: Prior series have identified CD30 expression by immunohistochemistry (IHC) is associated with improved overall survival (OS) in patients with newly diagnosed diffuse large B-cell lymphoma (DLBCL; 5-year OS: 79% vs 59% in CD30(-) patients, Hu et al, Blood 2013). It is unclear, however, whether CD30(+) status is retained throughout the course of the disease. Conversely, it is also unknown whether patients with CD30(-) tumors at diagnosis may present with CD30(+) disease upon relapse. As there is a currently approved antibody-drug conjugate targeting CD30, brentuximab vedotin, an improved understanding of CD30 expression in DLBCL may inform therapy options for relapsed (and potentially newly diagnosed) patients. Here, we evaluated patients with relapsed DLBCL with available tissue samples, including those with paired tissue samples from the time of diagnosis to assess for CD30 status for the duration of their disease course.

Methods: This cohort included patients ≥ 18 years old with relapsed DLBCL for whom biopsy samples and clinical data were available. Tissue samples at diagnosis and from time of relapse were collected from our institution's pathology archive, and IHC-staining for CD30 expression was performed on all available involved tissue. CD30 status was assessed using a comprehensive form including assessment of percentage of CD30(+) cells and distribution of staining within each cell. Both neoplastic and surrounding non-neoplastic cells were evaluated. All assessment of CD30 staining was completed by one hematopathologist. We also collected comprehensive clinical, demographic and pathologic data for each patient.

Results: We identified 25 patients with relapsed/refractory DLBCL with available tissue samples from the time of relapse, including 12 patients with available paired diagnostic tissue. Among all patients, the median age at diagnosis was 58 years (range 34-76), 48% were male, 56% were stage III/IV, and 62% presented with B-symptoms. Eighty-eight percent of patients received R-CHOP as frontline therapy. Cell of origin by the Hans algorithm was germinal center B-cell-like (GCB) for 9 patients, non-GCB for 5 patients, and unknown for 11 patients. After pathologic review, all 25 samples were CD30-negative at relapse, including all 12 paired samples which were CD30-negative at diagnosis and relapse, suggesting that CD30 expression does not appear to be acquired at the time of relapse in DLBCL patients who present with CD30-negative disease.

Conclusions: This retrospective, single-center cohort analysis suggests that patients with newly diagnosed DLBCL with tumors negative for CD30 expression retain CD30-negative status at relapse. Thus, assaying biopsies at relapse for CD30 positivity by IHC to investigate candidacy for salvage treatment with brentuximab vedotin in a patient with a tumor previously CD30(-) may be of limited value. Alternative computer-aided methods to assess CD30 expression in samples that are considered negative by conventional IHC may better identify the presence of CD30 among patients with relapsed DLBCL.