Abstract 5582: A targeted molecular classifier of MYC activity and BCL-2 expression in aggressive B-cell lymphomas, designed for clinical practice

Abstract

Abstract Background: High MYC and BCL2 co-expression as detected by immunohistochemical staining (IHC) of fixed biopsy samples identifies a sub-group of diffuse large B-cell lymphomas (DLBCLs) with inferior outcome among patients treated with standard chemotherapy, and the differential expression of MYC and BCL2 among DLBCL subtypes provides a biological basis for the prognostic value of the ‘Cell of Origin’ classification system. Thus quantifying MYC activity and BCL2 expression in formalin-fixed paraffin embedded (FFPE) biopsy specimens could help to identify patients who might benefit from more aggressive chemotherapy. Unfortunately, IHC is not a reproducibly quantitative test due to a number of pre- and post- analytical factors. In contrast, gene expression profiling (GEP) allows for the possibility of better standardization and quantitation of biomarkers in biopsy samples, but traditional GEP has required RNA isolated from frozen tissue. Design: We sought to develop a molecular classifier of MYC activity and BCL2 expression that is applicable to FFPE biopsy samples using the ‘NanoString nCounter’ platform in a 2-stage approach: 1. Discriminate between Burkitt Lymphoma (BL) and DLBCL using a selection of genes specific for each diagnostic category. 2. Quantify MYC and BCL2 expression using statistically justified ‘MYC target’ genes as well as other genes selected with an unbiased approach, those with significant differential expression between MYC IHC High and IHC Low cases. 3. Normalize data to selected housekeeping genes and assess the tissue microenvironment. The initial gene set was developed in silico based on the whole genome gene expression of 56 carefully selected de novo DLBCL that had companion MYC immunostaining. Results: An initial gene set comprising 200 genes was tested on a discovery cohort of FFPE biopsy samples of 42 aggressive B-cell lymphomas (12 Burkitt Lymphoma [BL] and 30 DLBCL). Differential analysis and prediction models were used to construct a classifier comprising 87 genes that resulted in the successful classification of these tumors. We next validated the approach using an independent cohort of FFPE tissue biopsies (12 BL, 7 genetic “double hit” lymphomas, and 38 DLBCL lacking MYC and BCL2 rearrangements). Targeted profiling and molecular classification correctly diagnosed 100% of tumors as either BL or DLBCL. For DLBCLs, the molecular classifier correctly predicted the MYC expression in 87% of cases when compared to a well-validated IHC assay. We conclude that a targeted gene expression profile, using the Nanostring nCounter platform, coupled with a validated molecular classifier can effectively distinguish DLBCL from BL and quantify MYC activity and BCL2 expression in DLBCL. This protocol will be useful for the routine diagnostic and prognostic stratification of aggressive B-cell lymphomas in clinical practice. Citation Format: Christopher D. Carey, Daniel Gusenleitner, Bjoern Chapuy, Heather Sun, Azra Ligon, Alexandra E. Kovach, Long P. Le, Aliyah R. Sohani, Margaret Shipp, Stefano Monti, Scott J. Rodig. A targeted molecular classifier of MYC activity and BCL-2 expression in aggressive B-cell lymphomas, designed for clinical practice. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 5582. doi:10.1158/1538-7445.AM2014-5582