Introduction

Diffuse large B-cell lymphoma (DLBCL) is a group of heterogeneous disease with distinct molecular subtypes. The most established subtyping algorithm, the Cell-of-Origin (COO) model, categorizes DLBCL into activated B-cell (ABC) and germinal center B-cell (GCB)-like subgroups through gene expression profiling. COO subtyping is mandatory for every newly diagnosed DLBCL patients, as it is critical for determining the therapeutic and surveillance strategies. We evaluated a newly developed assay using 32-gene expression profiling to determine the COO of DLBCL with formalin-fixed paraffin-embedded (FFPE) tissue.

Methods

The DLBCL-COO Test is a qPCR-based 32-gene expression assay for COO determination in FFPE samples. Biopsy of DLBCL patients with paired FFPE and fresh tissue were identified to assign COO, based on the immunohistochemistry (IHC) algorithm (Han's algorithm), DLBCL-COO qPCR assay and global gene expression profiling with RNA-seq, respectively. The global gene expression profiling with RNA-seq was taken as the "gold standard" for reference. Clinical information including the survival data were collected.

Results

160 cases of DLBCL with evaluable COO assignments with IHC, DLBCL-COO 32-gene assay and global gene expression profiling with RNA-seq were identified. Comparing with the 77.5% concordance between IHC algorithm and gold standard, there is 91.9% concordance between DLBCL-COO 32-gene assay and gold standard (P =0.005). 72 patients assigned as ABC subtype and 14 patients assigned as Type-3 subtype demonstrated a significantly inferior overall survival than 42 patients assigned as GCB subtype using DLBCL-COO assay (P =0.023). However, COO based the IHC algorithm failed to provide the predictive value regarding overall survival (P =0.09).

Conclusions

DLBCL-COO assay provides flexibility and accuracy in DLBCL subtype characterization. These subtype distinctions should help guide disease prognosis and treatment options within DLBCL clinical practice.

Disclosures

Sun:Canhelp Genomics: Employment. Xu:Canhelp Genomics: Employment.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution