Abstract
Diffuse large B-cell lymphoma (DLBCL) is a clinicopathologically heterogeneous group of lymphomas. Gene expression profiling (GEP) has identified two distinct forms of DLBCL: activated B-cell-like (ABC) type and germinal center B-cell-like (GCB) type. Approximately 10% of DLBCL are positive for CD5, and most of CD5-positive (CD5+) DLBCLs are classified as ABC DLBCL. CD5+ DLBCL is also characterized by aggressive clinical course, poor prognosis, and frequent central nervous system relapse (Yamaguchi M, et al. Blood 2002, Haematologica 2008; Miyazaki K, et al. Ann Oncol 2011).
We previously analyzed 90 patients with DLBCL by GEP and identified a signature gene set that enabled categorization of DLBCLs into two groups: CD5+ group and CD5-negative (CD5-) group. Our analysis of samples from 78 patients with ABC DLBCL showed that the CD5+ ABC DLBCL signature gene set comprised many genes which are related to the neuronal system (Miyazaki K, et al. ASH2010, #4164). To clarify the clinical significance of protein expression of these genes not only in CD5+ DLBCL but also in whole DLBCL, we examined the protein expression of the genes in DLBCL cells and analyzed the relationship with clinical features.
The study comprised 167 patients with DLBCL who were diagnosed between 1993 and 2010 at Mie University Hospital. Of the 90 patients who were previously analyzed by GEP, 60 patients were included in the present study. The histologic diagnosis of DLBCL was made according to the 2008 WHO classification. Thirty (18%) of 167 patients were diagnosed as having CD5+ DLBCL. We selected six genes which were included in the CD5+ ABC DLBCL signature gene set: LMO3, SNAP25, SYT5, SV2C, NEUROD1, and SYN2. Protein expression of these genes was analyzed by means of immunohistochemistry using frozen sections.
We identified LMO3 and SNAP25 expressions in DLBCL cells. The other four proteins were not detected in DLBCL cells of the study. SNAP25 was expressed in the cytoplasm of DLBCL cells, whereas LMO3 was expressed in the nucleus of DLBCL cells. LMO3 was positive in 35% (58/164) and SNAP25 was positive in 45% (74/164) of the patients examined. Thirty-three patients were positive for both antigens. For validation of our GEP data, we separately analyzed the 60 patients who were previously analyzed by GEP, and found that CD5 expression in DLBCL showed significant correlation with LMO3 (68%, P = 0.0002) and SNAP25 (73%, P= 0.032) expression.
Analyses of clinical features at diagnosis according to each protein expression revealed that the LMO3+ group was characterized by poor performance status (> 1) (LMO3+ group, 36% vs. LMO3- group, 11%; P= 0.0004). There was no significant difference in clinical features between SNAP25+ and SNAP25- groups.
The most prevalent first-line treatment in the 167 patients was CHOP chemotherapy (142 patients). Forty-four patients received rituximab during first-line chemotherapy. Patients with LMO3+ DLBCL and SNAP25+ DLBCL who attained complete response with the initial therapy accounted for 74% (43/58) and 80% (59/74), respectively, and patients with LMO3- DLBCL and SNAP25- DLBCL for 86% (91/106) and 84% (76/90), respectively. At a median follow-up of 84 months, the LMO3+ group also showed a significantly worse OS than the LMO3- group (P = 0.034, 5-year OS; 46% and 64%, respectively). There was no significant difference in OS between SNAP25+ and SNAP- groups. As the two genes were included originally in the CD5+ ABC DLBCL signature gene set, we divided the patients into two groups by CD5 expression and analyzed the OS of each group. Each protein expression had no significant difference in OS in patients with CD5+ DLBCL; however, LMO3 expression had a significant correlation with a shorter OS (P= 0.048) in the CD5- DLBCL group.
The present study revealed the expressions of LMO3 and SNAP25 in DLBCL cells and an association between LMO3 expression and prognosis in patients with DLBCL. Further studies on LMO3 in DLBCL are warranted. (Supported in part by the National Cancer Center Research and Development Funds: #21-6-3, #23-A-17, and #26-A-4)
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.