CD30 Expression in Diffuse Large B-Cell Lymphoma and Its Relation to Important Clinical and Biological Disease Features

Abstract

AbstractAbstract 1592 Background:CD30 is a well-known diagnostic marker in both anaplastic large cell lymphoma (ALCL) and classical Hodgkin's lymphoma (CHL). Recently the chimeric drug brentuximab vedotin that combines an anti-CD30 monoclonal antibody with the anti-tubulin agent monomethyl auristatin E demonstrated activity in patients with relapsed ALCL and CHL. Previous observational studies have suggested that CD30 may be expressed in 10 to 20% of DLBCLs. It is possible that CD30+ DLBCLs may show different biologic behavior and be amenable to anti-CD30 therapy. The aim of this study was to determine the prevalence of CD30 expression in DLBCL by immunohistochemistry and explore possible relationships with important clinical and biologic variables of DLBCL. Methods:We retrospectively identified cases of DLBCL diagnosed between July 2003 and July 2012 at our institution. Eligible cases included patients with diagnosis of DLBCL irrespective of anatomic site or tumor stage. The diagnosis of DLBCL was based on the current WHO 2008 criteria. The following large B cell lymphoma subtypes were excluded from this analysis: post-transplant lymphoproliferative disorders with DLBCL morphology, Primary Mediastinal large cell lymphoma and the unclassifiable lymphomas with features intermediate between either DLBCL and Burkitt’s lymphoma or between DLBCL and Hodgkin’s lymphoma. Immunohistochemistry was performed as part of the routine workup of the cases (Monoclonal Mouse Anti-Human CD30, Dako) and CD30 was considered positive when ≥30% of neoplastic cells stained positive. DLBCLs were classified into germinal center (GC) or non-GC subtypes applying the Hans algorithm. Logistic regression analysis was performed to assess association between selected variables and CD30 expression. Results:A total of 333 cases of DLBCL were eligible for this study and of these 148 cases (44.4%) had CD30 results available. Selected demographic, clinical and histological characteristics were similar between cases on which CD30 IHC was performed compared to those in which CD30 IHC was not performed (data not shown), arguing against a selection bias in the performance of CD30 immunohistochemistry in these cases. Twenty-three percent (95% CI: 16.2% – 29.8%) of DLBCL tumors expressed CD30. CD30 expression was not significantly different between females and males (27.6% and 20.0% respectively, p = 0.28). The patients with CD30+ DLBCLs were 11.6 years younger than those with CD30- DLBCLs (95% CI: 5.3 – 17.9 years). The optimal cutoff age for CD30 expression was 47 years (ROC area under the curve: 0.70, p < 0.001). CD30 expression was more frequent in nodal and BCL2+ DLBCLs (Table 1). There was a non-significant difference between the expression of CD30 in non-GC type compared to GC type DLBCL with a pronounced trend for higher proportion of CD30 positivity in non-GC DLBCL (Table 1). CD30 expression was not significantly different in Epstein-Barr virus positive (EBV) compared to EBV negative DLBCLs (Table 1). Conclusions:CD30 is expressed in approximately 20% of all DLBCL and is more frequently expressed in younger patients and in BCL2+ DLBCLs. Although statistical significance was not reached, a trend towards more frequent CD30 expression in non-GC DLBCLs was observed. The increased expression of CD30 in the younger age group is intriguing, as other CD30+ neoplasms (CD30+ ALCL, CHL, and embryonal carcinoma) are also commonly observed in younger patients. The development of brentuximab vedotin and its well established effectiveness in other types of relapsed lymphomas opens the possibility of its applicability in CD30+ DLBCLs.Table 1CD30 expression by age, site, BCL2 expression, subtype of DLBCL and EBV status.Variable**nCD30+ n (%)OR (95% CI)Age4.4 (1.9–10.2)≤47 years6424 (37.5)>47 years8410 (10.3)Site3.0 (1.3–7.0)Nodal5217 (32.7)Extra nodal9617 (17.7)BCL25.2 (1.4–19.0)Positive6419 (29.7)Negative403 (7.5)Type2.5 (0.9–6.8)Non-GC5515 (28.8)GC507 (14.0)EBV1.7 (0.3–9.3)Negative446 (18)Positive182 (11)The second category listed is the reference category.Table 2Multivariate analysis of association of CD30 expression with age, site, BCL2 expression, and subtype of DLBCL.Independent variablesOR (95% CI)Age ≤47 years10.0 (2.5–40.1)Nodal DLBCL1.1 (0.3–4.0)BCL2 positive7.7 ( 1.4–43.9)Non-GC2.1 (0.6–7.9)N = 78. Dependent variable: CD30 expression. R2 = 0.35, Hosmer-Lemeshow goodness of fit p-value = 0.54. Disclosures:No relevant conflicts of interest to declare.