Aims: The clinical relevance of the Revised European-American Classification of Lymphoid Neoplasms (REAL) is still debated. To test the clinical validity of the REAL classification in Korea, where the incidence of T-cell lymphoma is higher, we investigated the clinicopathological features of non-Hodgkin's lymphoma (NHL) from Korea Cancer Center Hospital.
Methods and results: Five hundred and one patients with NHL were reclassified according to the REAL classification and clinicopathologically analysed. Immunophenotypically, B-cell lymphoma accounted for 67.9% and T- and NK-cell type for 30.5%. Approximately 48.5% of cases were forms of diffuse large B-cell lymphoma (DLBCL), while only 5.4% were follicular lymphoma. Peripheral T-cell lymphoma unspecified (PTCL-U; 10.8%) and angiocentric lymphomas (11.8%) comprised the majority of T-cell lymphomas. Most of the angiocentric lymphomas presented with localized nasal/nasopharyngeal or tonsillar primaries. All peripheral T-cell lymphomas (PTCL) showed a significantly low overall survival compared to DLBCL (P = 0.02, log rank). Overall survival rates for DLBCL and PTCL-U were also significantly different (P = 0.0043, log rank), though for DLBCL and angiocentric lymphoma there was no significant difference (P = 0.2142, log rank). Angiocentric lymphoma, however, was characterized by a shorter median survival time than DLBCL (54 months vs. 96 months). Among DLBCL patients according to the REAL classification, overall survival was significantly better in nonimmunoblastic type (intermediate-grade, WF-F,G) as compared to large cell immunoblastic type (high-grade, WF-H) (log rank, P < 0.001). The morphological distinction of the immunoblastic and nonimmunoblastic among DLBCL of the REAL classification bears significant prognostic relevance worthy of further consideration.
Conclusion: We conclude that lineage assignment (T vs. B) in the REAL classification is a clinically important distinction, but that it is necessary to subdivide the broad category of DLBCL.