CD30 Expression in Diffuse Large B-Cell Lymphoma

Abstract

AbstractAbstract 1558▪▪This icon denotes a clinically relevant abstract Background:Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous disease with a variable clinical course. The addition of rituximab (R) to CHOP combination chemotherapy has improved overall (OS) and progression free survival (PFS) but some patients progress despite treatment and alternative therapies are needed. Brentuximab vedotin is an antibody-drug conjugate that targets CD30. It is efficacious in the treatment of relapsed Hodgkin lymphoma and relapsed systemic anaplastic large cell lymphoma, two lymphomas associated with CD30 expression. The association between DLBCL and CD30 expression has not been well described. The aim of this study was to examine CD30 expression in DLBCL. Design:395 cases of formalin-fixed paraffin-embedded DLBCL (excluding PMBCL) in a tissue microarray were independently evaluated by two pathologists for expression of CD30, CD10, BCL6, and MUM1 by immunohistochemistry (IHC) and EBV RNA (EBER) by in situ hybridization. CD30 expression was correlated with cell of origin (COO) phenotype, OS and PFS, EBV infection, International Prognostic Index (IPI) score and CD30 mRNA expression. The COO phenotype, germinal center B-cell like (GCB) or non-GCB, was determined by IHC using the Hans algorithm. CD30 mRNA expression and COO genotype by gene expression profiling (GEP) were determined using Affymetrix U133 2.0 Plus arrays (n=170). Outcome analysis only included patients treated with R-CHOP chemotherapy. CD30 was considered positive by IHC if any malignant cells exhibited membranous staining. The threshold for calling higher CD30 expression by GEP was determined using X-Tile software. Results:25% (95/385) of DLBCL cases expressed CD30 by IHC with excellent concordance between two observers (r = 0.94). CD30 expression trended towards a non-GCB phenotype but was not significantly different (p=0.067). CD30 expression was not associated with PFS or OS in all R-CHOP treated cases (n=313); however, it was associated with a prolonged PFS in GCB-DLBCL (n=147) (p=0.019). In GCB-DLBCL CD30 expression remained an independent predictor of PFS in a multivariate analysis with IPI (p=0.038). CD30 expression by IHC was significantly associated with higher levels of CD30 mRNA (p=0.002). ABC-DLBCL exhibited significantly higher expression levels of CD30 mRNA (p=0.037). Higher CD30 mRNA expression was associated with a prolonged PFS in all R-CHOP treated DLBCL (p=0.012) as well as in DLBCL with a GCB-genotype (p=0.008), but not ABC or U-genotypes. Higher CD30 mRNA expression was also associated with a prolonged OS in the GCB-genotype (p=0.022). In the GCB-genotype higher CD30 mRNA expression remained an independent predictor of PFS, but not OS, in a multivariate analysis with IPI (p=0.037). EBV was identified in 3% of DLBCL (11/391), all of which exhibited a non-GCB phenotype (p=0.001) and were almost exclusively positive for CD30 expression (10/11)(p=<0.001). Conclusions:CD30 is expressed in approximately 25% of DLBCL and brentuximab vedotin could be considered for study in combination with traditional front-line therapies or as an alternative therapy in the relapsed or refractory disease. CD30 immunohistochemistry may be useful as a prognostic marker in R-CHOP treated GCB-DLBCL and the significant association of CD30 with EBV-positive non-GCB DLBCL suggests a distinct pathobiology for these cases. Disclosures:Slack:Seattle Genetics: Research Funding. Steidl:Seattle Genetics: Research Funding. Gascoyne:Seattle Genetics: Research Funding.