|
|||||||||
|
|
|
|
|
|
|
|
|
|
|
Blood, 1 November 2004, Vol. 104, No. 9, pp. 2989-2990.
CORRESPONDENCE To the editor:The WHO classification of mature T-cell leukemiasWe 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 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 immunophenotypes4,5 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
Correspondence: Daniel Catovsky, Haematology/Oncology, Royal Marsden Hospital/ICR, Fulham Road, London SW3 6JJ, United Kingdom; e-mail: daniel.catovsky{at}icr.ac.uk. References
Response:Tumor biology should drive classification effortsWe readily acknowledge the many contributions of Professor Daniel Catovsky's group to the current understanding of T-cell leukemias. However, we fundamentally disagree with the philosophy expressed by them in their present letter and elsewhere.1 In our opinion, tumor categorization follows understanding and does not precede it. Such classifications are always provisional and subject to improvement based on new knowledge. The current World Health Organization (WHO) scheme is no exception.2Our paper3 was intended to provide an overview of the types of mature T-cell leukemias encountered at a large referral center and how they relate to the WHO entities. The relative rarity of these tumors means that the unambiguous classification of T-cell tumors remains a difficult task for many hospitals, including our own. We identified helpful parameters for assigning a given T-cell leukemia to a particular WHO category, but also discuss how difficult-to-classify cases and overlap cases point to areas for further study. A short list of these unresolved issues would include (1) the nature of the 20% to 30% of cases currently classified as T-cell prolymphocytic leukemia (T-PLL) that lack TCL1 expression and their relationship to classical TCL1-expressing T-PLL; (2) the histogenesis and appropriate classification of TdT-negative, high-grade T-cell leukemias with blastoid morphology; (3) heterogeneity in the behavior of large granular lymphocyte proliferations; (4) the pathogenetic relationship of mycosis fungoides to primary Sezary syndrome, which, despite the assertion of the authors above, often lacks epidermotropism on skin biopsy; and (5) the mechanism of action of TCL1, which is the only well-established oncoprotein specifically associated with mature T-cell leukemias, remains a puzzle. Finally, we believe that future tumor classifications will require more than just diagnostic categorization. Classifications will also incorporate data on risk stratification, pattern of disease progression, and underlying pathogenesis to help guide the optimal selection and timing of therapy. Although the current WHO scheme touches on these areas, we fully expect that future versions will continue the progress in that direction.
Correspondence: Dan Jones, Dept of Hematology, Box 72, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030; e-mail: dajones{at}mdanderson.org.
References
  CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati What's this?
Related Article in Blood Online:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|||||||
 |
 |
 |
 |
 |
 |
 |
 |
||
| Copyright © 2004 by American Society of Hematology Online ISSN: 1528-0020 | |||||||||
