Observations regarding hairy cell leukemia and chronic lymphocytic leukemia within the proposed new classification of lymphoid neoplasms.

In their proposal for a new classification of lymphoid neoplasms Harris et all indeed simplify lymphoma classification to a list of pathologically defined disease entities. From a clinician's viewpoint the purpose of a lymphoma classification is to support treatment decisions. Thus, the proposal from the International Lymphoma Study Group can be used only by comparing it regularly to the Kiel classification (in our region) to decide whether aggressive combined chemotherapy is justified or if the disease is a low-grade non-Hodgkin's lymphoma (NHL), regarded to be incurable in the majority of cases. Within the classification by Harris et al, the admixture of chronic lymphocytic leukemia (CLL), prolymphocytic leukemia (PLL) and lymphoplasmocytoid immunocytoma is provocative and in view of the heterogeneity among these diseases it would seem better to leave them separated. Furthermore, recent results in the treatment of CLL (with fludarabin) for the first time suggest the possibility of a cure.' Thus the "available therapy" mentioned by the authors probably refers to prednisone and chlorambucile, and in the more advanced cases, COP (cyclosphosphamide, oncovine, prednison) or CHOP (cyclophosphamide, doxorubicin, oncovine, prednison). Also, while fludarabin has been found to be very useful in PLL, there are not enough data on the treatment of lymphoplasmocytoid immunocytoma with the drug, making the distinction clinically necessary.? In the case of hairy cell leukemia (HCL), it should be emphasized especially from the pathologic point of view, that, although rare cases without bone marrow involvement have been published, i t is the spleen that is always involved.' This is important, as HCL is presumed to originate from the spleen. Though lymph node involvement is clinically indeed uncommon, recent data show that retroperitoneal lympadenopathy is not unusual in HCL.5 When describing the immunophenotype of HCL, the negativity of the B-cell marker CD21 as compared with its positivity on BCLL cells seems worth mentioning. Monoclonal antibodies such as anti-Tac (CD25), HC-2, RAB-I, and B-ly-7 (now designated CD103) are directed against activation molecules present on hairy Thus, the postulated normal counterpart of the hairy cell is an activated B cell, that is relatively well differentiated.'.'" Though CD103 is indeed an important marker in differentiating HCL from other B-cell leukemias such as prolymphocytic leukemia (PLL) or CLL, it is less useful than HC2 in the differentiation of HCL from the HCL variant and from the splenic lymphoma with villous lymphocytes." Also, in rare cases of typical HCL, the neoplastic cells were found to be T cells by cell surface marker" and even by gene rearrangement studies.I3 This implies the existence of HCL of Tcell origin. Finally, if mentioning therapeutic options in HCL, the difference in the capability of the listed drugs to induce remission is important: it is generally recognized by now that interferon a does not induce complete remission and is not curative, whereas with 2-deoxycofomycin and 2-chloro-deoxyadenosine, a considerable proportion of patients reach complete remission and, potentially, cure. With the presumed pathogenetic role of the spleen in the pathogenesis of HCL in mind, residual splenic disease after successful treatment of HCL with interferon-a2bI4 and splenic relapse in a HCL patient during bone marrow remission after 2-chloro-deoxyadenosine therapy are of utmost interest." In conclusion, although advances in the treatment of certain lowgrade NHLs (like CLL and HCL) are considerable, it undoubtedly remains necessary to wait for a modified lymphoma classification even after the publication of the proposal by Harris et al.'.'6