The WHO classification of mature T-cell leukemias

We read with interest the comprehensive review of mature T-cell leukemias from Herling et al.¹  If anything, this study, despite the authors' conclusions, supports the robustness of the World Health Organization (WHO) classification. Their clinical and laboratory data, which mimic our own observations, confirm that the disease entities included can be clearly defined on clinico-pathologic criteria. Herling et al¹  contribute further by demonstrating TCL1 expression in T-prolymphocytic leukemia (T-PLL). We do not agree, however, that evidence for heterogeneity of clinical features is a weakness of the WHO classification. In fact, such clinical and hematologic heterogeneity is a fact of life in most leukemias and lymphomas. It is the very heterogeneity of the hematologic disorders that stimulated the adoption of a multiparameter approach to diagnosis by the WHO. Although there may not be unique cytogenetic abnormalities in these diseases, the data on T-PLL are very strong with involvement of breakpoints at 14q11 (TCR α locus) and 14q32 as a consistent feature in 85% of cases, which underscores TCL1 expression, as well as at Xq28, the loci for MTCP1. It is, in fact, our original identification of T-PLL as a disease entity with its associated karyotype²  that led scientists to identify the TCL1 gene. Even the additional cytogenetic changes in T-PLL, such as iso(8q), are rather unique to this disease.³  The 2 leukemias included in the WHO classification—T-PLL and T-large granular lymphocyte (T-LGL) leukemia—are well defined on morphology, immunophenotype, histology, and, in T-PLL, cytogenetics. These disease features continue to be refined with characteristic immunophenotypes^4,5  and bone marrow histology⁶  more recently described in T-LGL leukemia. Use of an extended panel of antibodies including CD57 and CD16 may reduce the overlap described by Herling et al between T-LGL leukemia and other T-cell disorders. Other clinical and laboratory features outlined in that report,¹  such as rising lymphocyte counts in T-PLL and cytopenias in T-LGL leukemia, are useful additions but not defining criteria. Although we have recognized a degree of morphologic heterogeneity in T-PLL,⁷  the consistent chromosome abnormalities and clinical course helped to recognize 2 morphologic variants: with small cells and with cerebriform-like cells. Even in the latter, the question of Sézary syndrome should not arise because of the absence of epidermotropism when skin lesions are present in T-PLL.⁸