Differential diagnosis of CD3+, CD56+ T-cell leukemias.

Recently, Gentile et a l l reported what they believe to be a new disease, ie, CD3+, CD56' aggressive variant of large granular lymphocyte (LGL) leukemia. At presentation or during the course of the disease, all 3 patients were reported to have an absolute lymphocytosis, a moderate degree of thrombocytopenia, and massive splenomegaly, Splenectomy specimens in all 3 cases showed preferential involvement of the splenic red pulp cords and sinuses. None of the lesions was positive for Epstein-Barr virus (EBV). The differential diagnosis of T-cell lymphoproliferative disorders is complex and, indeed, some diseases may have been reported under more than one name. In their differential diagnosis, the investigators consider several lymphoproliferative disorders of T-cell and natural killer (NK)-cell derivation. However, they omit from their discussion y16 T-cell lymphomaz4 and S-100+, T-cell lymphoproliferative disease: both of which are aggressive T-cell malignancies expressing CD56. y/6 T-cell lymphoma shares many clinicopathologic and immunophenotypic features with the cases presented by Gentile et al.' This disorder presents with marked hepatosplenomegaly, with only modest peripheral blood involvement. The patients are frequently anemic, leukopenic, and thrombocytopenic. The malignant cells show preferential homing to the sinusoids of the splenic red pulp and liver. Prognosis is generally poor, with death reported in 4 of 6 cases in the literature and all 3 of the patients in our study for whom followup was available, despite aggressive Although y/6 T cells are usually double negative for CD4 and CD8, expression of CD8 does occur in y/6 T cells in peripheral lymphoid tissues: and a subset of yl6 T-cell lymphomas are CD8+.3.4 Also significant is the expression of CD56 in all 3 of our cases in which it was studied and in 4 of 6 cases reported by Gaulard et Thus, the phenotype of y/6 T-cell lymphoma is very similar to that of the cases reported by Gentile et a l l : CD2+, CD3'. CD4-, CD5-, CD16+, CD25-, CD56', CD57-, with variable expression of CD8. All cases were aiso negative for EBV sequences? Additional studies would be required to evaluate the relationship of CD3+, CD56' LGL leukemia to yl6 T-cell lymphoma. Notably, the cases reported by Gentile et all were studied only for CD3 and neither PFI nor TCR6 were evaluated. y/6 T-cell lymphomas are consistently PFIand TCR6+; CD3 expression is seen in all instances. At the genotypic level, although all cases of y/6 T-cell lymphoma have shown rearrangement of the T-cell receptor y chain gene (TCRy)? P chain gene rearrangement (TCRP) was seen in 2 of 5 cases ~tudied.~ Thus, TCRP rearrangement does not exclude the diagnosis. S-100+, T-cell lymphoproliferative disease (S-IOO', T-cell LPD) is another aggressive T-cell leukemia to be considered in the differential diagnosis. Marked splenomegaly has been reported in all cases.' In contrast to hepatosplenic y/6 T-cell lymphoma, and similar to CD3+ CD56+ LGL leukemia, most patients exhibit both lymphocytosis and lymphadenopathy. Despite aggressive chemotherapy, the prognosis is also poor, with death reported in all 7 cases reported with adequate follow-up.' Normally, S-100+ T cells are detected exclusively in a small percentage of peripheral blood lymphocytes belonging to the CD4-, CD8' cell compartment.' All 3 of the cases reported by Gentile et al' were CD8+, although in the literature only a proportion of cases of S-100+, T-cell LPD have been positive for CD8.' In S-IO', Tcell LPD, the cells are EBVand express an alp T-cell receptor. Also notable is the expression of several NK-cell-associated markers, including CD16 and CD56, but the absence of CD57. Thus, the phenotype of S-100+, T-cell LPD is also very similar to that of the cases reported by Gentile et al.' Studies of S100 protein would certainly be of interest. The spectrum of diseases characterized as variants of T-cell chronic lymphocytic or prolymphocytic leukemia is broad, with many overlapping clinicopathologic and phenotypic features. As Gentile et al' note, CD56 expression is not specific and is found in many T-cell leukemias and lymphomas. We believe that further studies are required before one concludes that CD3+, CD56' LGL leukemia is a newly described disease. We would advocate that all investigators pursue a multiparameter approach including cytogenetic analyses, if possible, to further delineate this group of T-cell malignancies.