Correspondence 2174 penia resolved. Because of clinical evolutivity and further IgG concen- completed in 8 weeks after the first course of Rituximab. Polychemo- tration increase, the anti-CD20 chimeric monoclonal antibody Rituxi- therapy was avoided and the treatment lasted only 4 weeks, with mab (Mabthera, Roche, Basel, Switzerland) was administered at 375 excellent tolerance. We suggest that Rituximab may be introduced as mg/m2/week for 4 weeks. Polyadenopathy and PB plasma cells pro- first-line treatment of EBV-related RA-associated LPDs that do not rap- gressively disappeared. Serum IgG and LDH returned to normal values idly regress after MTX withdrawal. within 2 months. A BM biopsy performed at that time showed disap- pearance of B cell immunoblasts and normal plasma cells. Neutrop- CKelaı¨di11Service d’He´matologie, Hoˆpital Cochin, AP-HP enia did not relapse despite persistence of anti-neutrophils antibodies. M Tulliez2 Paris, France; 2Service d’Anatomopathologie, Nine months after Rituximab therapy, the patient is in complete CLecoq-Lafon 1 Hoˆpital Cochin, AP-HP, Paris, France; 3Service de remission. X-V Pham3 Rhumatologie Paris V, Hoˆpital Cochin, AP-HP, This is the first time that anti-CD20 antibodies have been used in A Kahan3 Paris, France this particular setting, without chemotherapy. Stewart et al3 recently F Dreyfus1 published the case of an EBV-related non-Hodgkin’s lymphoma (NHL) D Bouscary1 occurring in a patient with RA and successfully treated with Rituxi- mab. However, this NHL occurred in the setting of a B cell chronic References lymphocytic leukemia and could not be directly related to RA. More- over, remission was achieved with both polychemotherapy and Ritux- 1 Kamel OW, Van De Rijn M, Weiss LM, Del Zoppo GJ, Kahler imab. The LPD in our case was revealed by an autoimmune neutrop- Hench P, Robbins BA, Montgomery PG, Warnke RA, Dorfman RF. enia. Lymphoproliferation was particular with an expansion of Reversible lymphomas associated with Epstein-Barr virus occur- immunoblasts of B cell phenotype in the LN (Figures 1 and 2) and ring during methotrexate therapy for rheumatoid arthritis and circulating plasma cells that initially suggested a diagnosis of plasma dermatomyositis. N Engl J Med 1993; 328: 1317–1321. cell leukemia. All techniques used for clonality assessment were nega- 2 Kamel OW, Van De Rijn M, LeBrun DP, Weiss LM, Warnke RA, tive. The relation to EBV was affirmed by EBER1, 2 and BHLF-1 Dorfman RF. Lymphoid neoplasms in patients with rheumatoid mRNAs expression in the LN (Figure 3) and positivity for EBV-PCR in arthritis and dermatomyositis: frequency of Epstein–Barr Virus and the blood. BHLF-1 positivity indicated replicative ongoing EBV infec- other features associated with immunosuppression. Hum Pathol tion. We could not detect monoclonal EBV genomic material in the 1994; 25: 638–643. LN of this patient by Southern blot. The presence of anti-EBNA anti- 3 Stewart M, Malkovska V, Krishnan J, Lessin L, Bart W. Lymphoma in bodies at diagnosis excluded a primo-infection to EBV. a patient with rheumatoid arthritis receiving methotrexate treatment: Patients with RA are considered at increased risk of developing lym- successful treatment with rituximab. Ann Rheum Dis 2001; 60: phomas among other neoplasms. In addition to chronic activation of 892–893. the immune system by the autoimmune disorder, EBV-specific T cell 4 Tosato G, Steinberg AD, Blaese RM. Defective EBV-specific sup- function seems to be defective in RA4 and EBV-infected B cells are pressor T-cell function in rheumatoid arthritis. NEnglJMed1981; increased. Immunosuppressive therapy for RA may further increase 305: 1238–1243. the relative risk of developing lymphomas from 2.5 to 10.5 These lym- 5 Kinlen LJ. Incidence of cancer in rheumatoid arthritis and other dis- phomas are often EBV-positive.6 MTX may favor the emergence of orders after immunosuppressive treatment. Am J Med 1985; 78 LPDs and EBV positivity may be predictive of success of MTX with- (Suppl. 1A): 44–55. drawal.7 However, it is still uncertain whether there is a real increase 6 Dawson TM, Starkebaum G, Wood BL, Willkens RF, Gown AM. in the incidence of lymphoma while taking MTX.8 The above data Epstein–Barr virus, methotrexate, and lymphoma in patients with support evidence for the existence of a subgroup of RA-associated rheumatoid arthritis and primary Sjo¨gren’s syndrome: case series. LPDs resembling PT-LPDs in polymorphic histology, EBV positivity J Rheumatol 2001; 28: 47–53. and possible regression after reduction of immunosuppressive ther- 7 Salloum E, Cooper DL, Howe G, Lacy J, Tallini G, Crouch J, Schultz apy. By analogy to PT-LPDs, the polyclonal character of LPD in our M, Murren J. Spontaneous regression of lymphoproliferative disorders case offers further insights into the natural history of LPDs occurring in patients treated with methotrexate for rheumatoid arthritis and in RA. Therapeutic results in PT-LPDs could be extrapolated to RA- other rheumatic diseases. J Clin Oncol 1996; 14: 1943–1949. associated LPDs. In the literature, the delay in response after MTX 8 Starkebaum G. Rheumatoid arthritis, methotrexate, and lym- withdrawal was usually short. In the absence of a rapid response, clin- phoma: risk substitution, or cat and mouse with Epstein–Barr virus? icians almost always proceeded to chemotherapy administration. Pro- J Rheumatol 2001; 28: 2573–2575. 9 Milpied N, Vasseur B, Parquet N, Garnier JL, Antoine C, Quartier gression of the LPD in our case, despite MTX withdrawal, led us to P, Carret S, Bouscary D, Faye A, Bourbigot B, Reguerre Y, Stoppa rapidly introduce anti-CD20 antibody therapy. Rituximab has proven AM, Bourquart P, Hurault de Ligny B, Dubief F, Mathieu-Boue A, efficacy in PT-LPDs.9 Complete response was obtained in our case + Leblond V. Humanized anti-CD20 monoclonal antibody (Rituximab) with a presumed action on the CD20 immunoblastic compartment − in post transplant B-lymphoproliferative disorder: a retrospective and a subsequent decrease in the more mature derived CD20 plas- analysis on 32 patients. Ann Oncol 2000; 11 (Suppl. 1): 113–116. macytic compartment. The response was rapid and progressively

Expression of CD133 on a human cell line lacking CD34

Leukemia (2002) 16, 2174–2175. doi:10.1038/sj.leu.2402652 TO THE EDITOR and the study of Gallacher et al4 demonstrates that only CD133+, CD34−,CD38−, Lin− cells have a clonogenic progenitor capacity We read with interest the review by Bhatia, in which the author high- + lights the potential role and utility of CD133 expression in human equivalent to CD34 stem cells. Unfortunately, these primitive HSCare extremely rare (0.2%) stem cells (HSC). We agree with his proposed working model of the − organization of HSCin which 34 −,38−, Lin− stem cells give rise to among the CD34 cell population and their identification has been CD133+,CD34+ subsets with hematopoietic and endothelial capacity. hampered by the laborious isolation procedure and the small yield of Evidence indicates that the most primitive blood HSCreside among cells obtained by the various methods of isolation used so far. There- cells lacking CD34 and lineage commitment markers (CD34−, Lin−)2,3 fore, a priority is to search for cell lines with HSCcharacteristics in order to facilitate research in this area. In 1994, we established a new fibroblast like cell line, designated GM- 490, from the bone marrow of a patient with acute lymphoblastic leu- Correspondence: C Delfini, Department of Onco-Hematology, S Sal- kaemia.5 GM-490 cells have a morphology typical of adherent fibroblasts. vatore Hospital, 61100 Pesaro, Italy; Fax: 39-0721-364057 GM-490 cells express CD29, CD36, CD71, CD54, CD105 surface Received 26 February 2002; accepted 6 May 2002 antigens but lack CD34, CD38, HLA-DR, CD45, CD14 and all lymph-

Leukemia Correspondence 2175

Figure 1 A GM-490 cell sample was stained with CD133(PE) and CD34(FITC). Cellular debris was gated out on the basis of scatter (a). A total of 10 000 events was acquired. Quadrants were defined by using appropriate isotypic control Mabs (b). Virtually all (98%) GM-490 cells express CD133, but not CD34 (c). oid and myeloid markers. More recently, we have found that a 2 Osawa M, Hanada K, Hamada H, Nakauchi H. Long-term lym- majority of GM-490 cells expresses CD133 (Figure 1). Up to now, phohematopoietic reconstitution by a single CD34− investigative only four human cell lines express CD133: two retinoblastoma (Weri- hematopoietic . Science 1996; 273: 242–245. Rb-1,Y79), one teratocarcinoma (NT2) and one monocytic leukaemia 3 Zanjani ED, Almeida-Porada G, Livingston AG, Flake AW, Ogawa (MUTZ-2) cell line. M. Human bone marrow CD34− cells engraft in vivo and undergo GM-490 is the first known human hematopoietic cell line which multilineage expression that includes giving rise to CD34+ cells. expresses CD133 lacking CD34 antigen. It represents a useful model Exp Hematol 1998; 26: 353–360. to study the hierarchy of HSC. Studies are in progress in our laboratory 4 Gallacher L, Murdoch B, Dongmei M, Karanu FN, Keeney M, in order to induce GM-490 cell differentiation towards hematopoietic Bhatia M. Isolation and characterization of human CD34−Lin− and and endothelial progenitors. CD34+Lin− hematopoietic stem cells using cell surface makers AC133 and CD7. Blood 2000; 95: 2813–2820. CDelfini 11Department of Onco-Hematology, 5 Delfini C, Tabilio A, Falzetti F, Centis F, Annibali M. A human F Centis1 S Salvatore Hospital, Pesaro, Italy; non-hemopoietic cell line expressing thrombospondin receptor F Falzetti22Department of Medicine, (CD36): preliminary characterization. Exp Hematol 1995; 23: A Tabilio2 Hematology and Immunology Section, 101 (Abstr.). University of Perugia, Perugia, Italy

References

1 Bhatia M. AC133 expression in human stem cells. Leukemia 2001; 15: 1685–1688.

Reply to C Delfini et al

Leukemia (2002) 16, 2175. doi:10.1038/sj.leu.2402653 defined in human hematopoietic cells, this cell line provides a unique opportunity to study this novel transmembrane . In addition, The authors demonstrate the existence of a cell line expressing AC133 further characterization of lineage potential, additional cell surface + and devoid of CD34 cell surface expression. Since the role, putative markers, and differentiation ability into CD34 cells may assist in ligand and potential signaling mechanisms of AC133 have yet to be understanding hierarchical organization of primitive subsets of human blood cells. I wish the authors and their collaborators the best of luck.

M Bhatia Stem Cell Biology and Regenerative Medicine, Correspondence: M Bhatia, PO Box 5015, 100 Perth Drive, London, Robarts Research Institute, Ontario, Canada, N6A 5K8; Fax: (519) 663-2982 University of Western Ontario, Received 16 April 2002; accepted 6 May 2002 Faculty of Medicine, Ontario, Canada

Leukemia