Plasmablastic lymphoma was initially described as a variant of diffuse large B-cell lymphoma (DLBCL) involving the oral cavity of HIV+ patients and characterized by immunoblastic morphology and a plasma cell phenotype. However, other lymphomas may exhibit similar morphologic and immunophenotypic features. To determine the significance of plasmablastic differentiation in DLBCL and examine the heterogeneity of lymphomas with these characteristics, we examined 50 DLBCLs with low/absent CD20/CD79a and an immunophenotype indicative of terminal B-cell differentiation (MUM1/CD38/CD138/EMA-positive). We were able to define several distinct subgroups. Twenty-three tumors were classified as plasmablastic lymphoma of the oral mucosa type and showed a monomorphic population of immunoblasts with no or minimal plasmacytic differentiation. Most patients were HIV+ and EBV was positive in 74%. Eleven (48%) cases presented in the oral mucosa, but the remaining presented in other extranodal (39%) or nodal (13%) sites. Sixteen cases were classified as plasmablastic lymphoma with plasmacytic differentiation. These were composed predominantly of immunoblasts and plasmablasts, but in addition exhibited more differentiation to mature plasma cells. Only 33% were HIV+, EBV was detected in 62%, and 44% had nodal presentation. Nine cases, morphologically indistinguishable from the previous group, were secondary extramedullary plasmablastic tumors occurring in patients with prior or synchronous plasma cell neoplasms, classified as multiple myeloma in 7 of the 9. Two additional neoplasms were an HHV-8+ extracavitary variant of primary effusion lymphoma and an ALK+ DLBCL. HHV-8 was examined in 39 additional cases, and was negative in all. In conclusion, DLBCLs with plasmablastic differentiation are a heterogeneous group of neoplasms with different clinicopathological characteristics that may correspond to different entities.

The tumor microenvironment plays an important role in the biologic behavior of follicular lymphoma (FL), but the specific cell subsets involved in this regulation are unknown. To determine the impact of FOXP3-positive regulatory T cells (Tregs) in the progression and outcome of FL patients, we examined samples from 97 patients at diagnosis and 37 at first relapse with an anti-FOXP3 monoclonal antibody. Tregs were quantified using computerized image analysis. The median overall survival (OS) of the series was 9.9 years, and the FL International Prognostic Index (FLIPI) was prognostically significant. The median Treg percentage at diagnosis was 10.5%. Overall, 49 patients had more than 10% Tregs, 30 between 5% to 10%, and 19 less than 5%, with a 5-year OS of 80%, 74%, and 50%, respectively (P ‫؍‬ .001). Patients with very low numbers of Tregs (< 5%) presented more frequently with refractory disease (P ‫؍‬ .007). The prognostic significance of Treg numbers was independent of the FLIPI. Seven transformed diffuse large B-cell lymphomas (DLBCLs) had lower Treg percentages (mean: 3.3%) than FL grades 1,2 (mean: 12.1%) or 3 (mean: 9%) (P < .02).