New REAL clinical entities.

PURPOSE: In 1994, the International Lymphoma Study Group proposed a "Revised European-American Lymphoma (REAL) Classification of Lymphoid Neoplasms." This classification system was developed because (1) new lymphoid disease entities have been recognized that are not part of the National Cancer Institute Working Formulation (WF) and (2) there was a need to develop a common classification system that could be used internationally. The REAL classification itself had never been tested, however, to determine whether it was reproducible or defined distinct clinicopathologic entities. Therefore, in the past two years, two studies were conducted by the South-west Oncology Group (SWOG) Lymphoma Committee and the Non-Hodgkin's Lymphoma (NHL) Classification Project to validate the REAL classification.

PATIENTS AND

METHODS

The SWOG Lymphoma Committee reviewed the pathology and clinical course of 376 previously untreated patients with stage III or IV disease within WF categories A, B, C, D, or E who received full-dose cyclophosphamide, doxorubicin, vincristine, prednisone (CHOP) in SWOG studies 7204, 7426, and 7713. No patients in this database had localized mucosa-associated lymphoid tissue (MALT) lymphoma. The NHL Classification Project performed a retrospective study of 1,403 consecutive patients with previously untreated NHL seen between 1988 and 1990 at nine sites around the world. Five expert hematopathologists reached a consensus diagnosis on every case by using histologic, clinical, and immunophenotypic data, and 20% of all cases were randomly reviewed.

RESULTS

The most common diagnosis was diffuse large B-cell lymphoma (31%), which combines the large-cell and immunoblastic WF categories (WF G/H). The next most common diagnosis was follicular lymphoma (22%; WF B, C, and D). Marginal zone B-cell (including MALT), peripheral T-cell, small B-lymphocytic, and mantle cell lymphoma each constituted between 5% and 10% of diagnoses. Primary mediastinal large B-cell, anaplastic large T/null cell, high-grade B-cell, Burkitt-like, and precursor T-lymphoblastic lymphoma made up the remaining 10 most frequent diagnoses.

CONCLUSIONS

The analyses conducted by the SWOG Lymphoma Committee and the NHL Classification Project have demonstrated that the REAL classification does define "real" clinical entities that can be diagnosed by expert hematopathologists. The understanding gained from study of "real" entities should permit hematologists/oncologists to better predict the clinical course of their patients and also to develop improved therapy.