Leukemia (1997) 11, 1983–1996 1997 Stockton Press All rights reserved 0887-6924/97 $12.00

CORRESPONDENCE

CASE REPORT

CD32CD41 cells with a Th2-like pattern of cytokine production in the peripheralblood of a patient with cutaneous T cell lymphomaD Brugnoni1, P Airo1, C Tosoni1, M Taglietti1, F Lodi-Rizzini1, P Calzavara-Pinton2, C Leali2 and R Cattaneo1

Servizio di 1Immunologia Clinica and 2Clinica Dermatologica, Spedali Civili, Brescia, Italy

A case of cutaneous T cell lymphoma associated with mild autoimmune disorders were also ruled out. A diagnosis ofeosinophilia and rise of IgE levels is reported. A population of CTCL was made and treatment with IFN-a was started (9 MUCD32CD41 cells was observed in the peripheral blood. After × 3/week), but was stopped 2 months later due to lack of com-activation, these purified CD32CD41 cells showed a Th2 pattern

pliance without clinical benefit. Psoralene plus UV-A (P-UVA)of cytokine production, secreting higher levels of IL-5 and IL-therapy induced complete regression of the mycosis fungoides4 and lower levels of IFN-g compared to the patient’s and

controls’ CD31CD41 cells. Moreover, high levels of IL-5 and lesions, whereas follicular mucinosis was not modified by thesoluble CD30, a marker of Th2 cell activation, were detected in treatment. However, lesions slowly relapsed in the followingthe patient’s serum. years.Keywords: Th2 cytokines, cutaneous T cell lymphoma; In September 1995, in spite of a normal blood lymphocyteCD32CD41 cells

count (1620/ml), two-colour immunophenotypical studiesrevealed a small CD3−CD4+ T cell population (Table 1). Rarelymphocytes with convoluted nuclei were observed in the

Introduction peripheral blood smears. Eosinophil count was 842/ml andtotal IgE level 2888 KU/l. A new cutaneous biopsy demon-

Over the last few years, it has become clear that human T cells strated a dense dermal infiltration by CD4+CD7− cells in partcan be divided into distinct subsets according to the pattern of lacking TCRb expression, whereas a bone marrow biopsycytokine production: Th1 cells secrete mainly IL-2 and inter- showed only moderate eosinophilia.feron-g, whereas Th2-cells produce IL-4 and IL-5.1 The enriched CD3−CD4+ population, obtained by depletion

Recently, it has been suggested that cutaneous T cell lym- of CD3+ cells from PBMC using magnetic beads, as describedphoma (CTCL) represents clonal proliferation of Th2-cells.2 elsewhere,4 showed defective response to mitogens activatingMoreover, two cases of clonal expansion of CD3−CD4+ T the cells through crosslinking of membrane receptors (PHAcells, without clinical evidence of lymphoma, secreting Th2- and CD3), but proliferative response, and expression of acti-type cytokines have been recently described.3,4 We report vation markers such as CD69 and CD25, was partiallyhere a new case of CTCL associated with a population of cir-culating CD3−CD4+ cells with a Th2-pattern of cytokineproduction.

Table 1 Immunophenotypical analysis

Marker SV HealthyCase report controls

SV, a 49-year-old female, was admitted to our hospital inPBL (%)

1992 for erythematous eczema-like plaques involving face, CD3 58 61–85axillas, breast, abdomen and feet, which appeared in 1985 CD4 58 34–54and slowly enlarged afterwards. No other abnormalities were CD8 15 15–33

CD19 7 2–14observed at physical examination; laboratory parameters wereHLA-DR 23 2–20normal except for mild eosinophilia (606 ml) and raised IgECD3+CD4+ 37 34–54levels (1595 kilo units/l). Histologic examination of a skinCD3−CD4+ 21 ,1

biopsy demonstrated a dense dermal infiltrate of small lym-Healthy controlsphocytes with convoluted nuclei and CD3+CD4+CD7+/−

(CD3+CD4− cells)CD8−CD30+/− phenotype, and follicular mucinosis. A bonemarrow biopsy showed mild eosinophilia, without

Purified CD3−CD4+ cells (%)lymphomatous infiltration. In peripheral blood smears, SezaryCD3 cytoplasmic 30 (faint expression) NA

cells were not detected. Total peripheral blood lymphocyte TCR ab ,1 .99count was 1090/ml (CD3, 68%, CD4+, 55%, CD8+, 13%, CD2 99 .99CD19, 8%). No signs of lymphoma were detected by periodic CD5 98 .90

CD28 93 95–100echographic and X-ray examination. Allergic, parasitic andCD45R0 93 30–60CD62L 1 40–80CD40La 87 80–90Correspondence: D Brugnoni, Servizio di Immunologia Clinica, Sped-

ali Civili, 25100 Brescia, Italy; Fax: +30 382796 +30 3995085Received 24 January 1997; accepted 24 July 1997 aAfter stimulation with PMA + ionomycin for 16 h.

Correspondence

1984restored using the transmembrane activators PMA + iono- Recently, three patients have been reported in whom pro-mycin (data not shown). liferation of a CD3−CD4+ T cell population was associated

Unstimulated PBMC and CD3−CD4+ cells from the patient with a Th2-like pattern of cytokine secretion and with hyper-secreted very low levels of IL-4, IL-5, IL-2 and IFN-g, as did eosinophilia with3 or without4,5 hyper IgE. Although in somePBMC from controls (data not shown). After activation, puri- of these cases there was no clinical evidence of lymphoma,fied CD3−CD4+ cells produced higher levels of IL-5 and IL-4 the clonal nature of the proliferation was demonstrated byand lower levels of IFN-g compared to the patient’s and con- molecular biology studies which showed monoclonaltrols’ CD3+CD4+ cells (Figure 1). Moreover, high levels of IL- rearrangement3 or deletion of TCRB gene.4 In three other5 (94 pg/ml; controls 33 ± 30 pg/ml) and soluble CD30 patients, a lymphoma with CD3−CD4+ phenotype and(26 U/ml; controls 10 ± 8), believed to be a marker of Th2 cell cutaneous involvement developing after several years of hyp-activation, were detected in the patient’s serum. ereosinophilia and raised IgE levels was described.6–8 In one

Interestingly, activation of CD3−CD4+ cells with PMA + case,6 the presence of CD3−CD4+ lymphocytes in the periph-ionomycin did not induce CD30 membrane expression, sug- eral blood also preceded the clinical manifestations of malig-gesting therefore that other mechanisms of activation were nant lymphoma. The case presented here provides a furtherneeded to induce the CD30 expression observed in the link between both CTCL and CD3−CD4+ phenotype, andcutaneous biopsy. On the contrary, it induced a high Th2-cytokine hyperproduction.expression of CD70 (another molecule of the TNFR family) by For the lack of the TCR/CD3 complex on surface mem-CD3−CD4+ cells (92 vs 12 ± 6% by purified CD4+ cells from

brane, T cells with this phenotype cannot receive signals bynormal controls).the ‘classical’ pathway of T cell activation. On the other hand,T cell proliferation and Th2-cytokine secretion can be inducedin vitro, bypassing the TCR/CD3 complex through directDiscussionstimulation of protein kinase C.4 As raised levels of Th2-cyto-kine can be detected in vivo in the sera of these patients,3–5

Recent data have led to the hypothesis that CTCL represent thein vivo cytokine secretion in these cases is either spontaneousproliferation of malignant helper T cells with a Th2 productionor induced by activation through ‘alternative’ pathways.pattern; in fact, PBMC from patients with CTCL have a defec-

These observations provide new insights in our comprehen-tive capacity to produce IL-2 and IFN-g, but can producesion of the pathogenesis of CTCL, and might have therapeuticincreased quantity of IL-4. Moreover, mRNA for Th2-typeimplications leading to treatment with Th2 antagonistcytokines (IL-4, IL-5, IL-10) has been demonstrated in skincytokines such as IFN-a, IFN-g and IL-12.2lesions from patients with CTCL.2

Figure 1 Cytokine production after stimulation with PMA (5 ng/ml) + ionomycin (500 ng/ml) for 72 h, as measured by sandwich ELISAcommercial kits. Horizontal bars show 1 s.d. Ten healthy individuals with no history of atopic diseases were used as controls.

Correspondence

19855 Schandene L, Del Prete GF, Cogan E, Stordeur P, Crusiaux A,ReferencesKennes B, Romagnani S, Goldman M. Recombinant interferon-alpha selectively inhibits the production of Interleukin-5 by humanCD4+ T cells. J Clin Invest 1996; 97: 309–315.1 Romagnani S. Human Th1 and Th2 subsets: doubt no more.

Immunol Today 1991; 12: 256–259. 6 O’Shea JJ, Jaffe ES, Lane HC, MacDermott RP, Fauci AS. PeripheralT-cell lymphoma presenting as hypereosinophilia with vasculitis.2 Rook AH, Heald P. The immunopathogenesis of cutaneous T-cell

lymphoma. Hematol Oncol Clin N Am 1995; 9: 997–1010. Am J Med 1987; 82: 539–545.7 Bagot M, Bodemer C, Wechsler J, Divine M, Haioun C, Capesius3 Cogan E, Schandene L, Crusiaux A, Cochaux P, Velu T, Goldman

M. Clonal proliferation of type-2 helper T cells in a man with the C, Saal F, Cabotin P, Roubertie E, De Prost Y, Lorette G, Revuz J.Lymphome T non epidermotrope precede pendant plusieurshypereosinophilic syndrome. New Engl J Med 1994; 330: 535–

538. annees par un syndrome hypereosinophilique. Ann DermatolVenereol 1990; 117: 883–885.4 Brugnoni D, Airo’ P, Rossi G, Bettinardi A, Simon HU, Garza L,

Tosoni C, Cattaneo R, Blaser K, Tucci A. A case of hypereosino- 8 Moraillon I, Bagot M, Bournerias I, Prin L, Wechsler J, Revuz J.Syndrome hypereosinophilique avec pachydermie precedant unphilic syndrome is associated with the expansion of a CD3− CD4+

T-cell population able to secrete large amounts of interleukin 5. lymphome. Traitement par l’interferon alpha. Ann DermatolVenereol 1991; 118: 883–885.Blood 1996; 87: 1416–1422.

Correspondence

1986

LETTERS TO THE EDITOR

Burkitt’s type acute lymphoblastic transformation associated with t(8;14) in a case ofB cell chronic lymphocytic leukemia

Table 1 Laboratory data in a patient with B-CLLTO THE EDITOR

July May FebruaryB cell chronic lymphocytic leukemia (B-CLL) has the propen-1990 1991 1993sity to transform into high-grade diseases; diffuse large cell

lymphoma known as Richter’s syndrome, or prolymphocyticHb (g/dl) 11.0 10.1 11.4transformation.1,2 In rare cases, patients with CLL developPlatelet (×109/l) 191.0 165.0 48.0acute lymphoblastic leukemia (ALL) and multiple myeloma.1 Leukocyte (×109/l) 19.3 44.0 132.2

Among B cell neoplasms, transformation of follicular lym- Lymphocyte (×109/l) 15.1 39.2 80.6phoma to a Burkitt’s lymphoma (BL) has been well docu- Neutrophil (×109/l) 4.1 4.8 10.6

Lymphocyte in BM (%) 50.4 NT 42.4mented. This transformation has frequently been associatedBlast in BM (%) 1.2 NT 53.6with clonal acquisition of the translocation between chromo-LDH (U/L) 364 408 8118somes 8 and 14, 2, or 22 generally associated with de novoIgG (mg/dl) 1089 977 917BL, in which activation of c-myc gene expression may play a CD19 (%) 88* NT 96* (98**)

role in transformation.3 A case of BL transformation of CLL m (%) 36 NT 99 (99)was also reported in association with such Burkitt’s translo- k (%) 2 NT 5 (3)

l (%) 79 NT 91 (98)cation t(8;22).4 Recently, mutations of the p53 gene wereCD5 (%) 92 NT 91 (99)found in BL, Burkitt’s type ALL and B-CLL, particularly inRai stage 0 III IVRichter’s syndrome.5 Several steps in the clinical evolution of

these diseases are associated with a variety of recurrent chro-BM, bone marrow; LDH, lactate dehydrogenase; NT, not tested;mosome changes and genetic alterations that could proveImmunophenotypes of the mononuclear cells in the peripheralessential to the understanding of leukemogenesis. We would blood (*) and bone marrow (**).

like to report a case of B-CLL that developed Burkitt’s typeALL in the bone marrow (BM), which corresponded to clonalevolution of CLL cells with the occurrence of a secondary oculomotor and facial nerve palsy. He was treated with both

whole brain irradiation and intrathecal methotrexate, cytara-t(8;14), and prolymphocytic transformation in the peripheralblood (PB). bine, and prednisolone which was followed by improvement

of central nervous system leukemia. In December 1994, heA 76-year-old Japanese man was diagnosed in July 1990 asB-CLL Rai stage 0, manifested by a leukocyte count of 19.3 × died of pneumonia, 22 months after transformation although

his hematological findings showed chronic phase of CLL. At109/l with 63% lymphocytes (Table 1). Lymphocytes in the PBwere positive for HLA-DR, CD19, CD20, CD5, CD25 anti- diagnosis and transformation, all PB cells showed a normal

karyotype. At acute transformation, 13 of 19 BM cells hadgens, m and l chains of immunoglobulin (Ig), but not CD10or CD2 antigens. Treatment had not been prescribed until karyotypic changes consisting of t(8;14)(q24;q32) together

with trisomy 17 and 3q+.May 1991 when the leukocyte counts had increased to 44 ×109/l together with mild anemia and splenomegaly, and then Whether transformation of B-CLL corresponds to clonal

evolution of the initial lymphocyte population or to an inde-the patient received cyclophosphamide at 50 mg daily. In Feb-ruary 1993, the patient developed low-grade fever. The leuko- pendent malignancy with the emergence of a second unre-

lated clone arising in a susceptible host has been contro-cyte counts increased to 132.2 × 109/l with 61% lymphocytesand 26.5% prolymphocytes (Figure 1a).The BM aspirate was versial.1 In our case, almost all MNC in the BM and PB at

transformation expressed the same Ig isotypes (ml), as well ashypercellular with 42.4% lymphocytes and 53.6% blasts.Morphologically, blasts were consistent with those of Burkitt’s PBMNC at other times. Furthermore, blasts in the BM also

expressed CD5 antigen. Immature B cells from ALL includingtype leukemia, L3 in FAB classification (Figure 1b). Only afew same blasts were found in the PB. At this time, more than Burkitt’s type are generally negative for CD5 antigen.6 These

phenotypic findings therefore indicate that not only prolym-90% of mononuclear cells (MNC) from PB and BM includingprolymphocytes and blasts, respectively, had the same immu- phocytes but also blasts are derived from an identical clone

of CLL. We found the identical rearrangement of the Ig heavynophenotypes as the initial CLL cells (Table 1). Two coursesof chemotherapy with aclarubicin, cyclophosphamide, vincri- chain joining region (JH) gene in PB and BM MNC at all time-

points, confirming the clonal evolution of the initial CLL cellsstine and prednisolone were followed by the chronic phase ofCLL in April 1993. The patient received five courses of further (Figure 2a).

It is well documented that abnormal karyotypes are signifi-chemotherapy with vincristine and prednisolone weekly. InAugust 1993, he developed meningeal leukemia with right cantly correlated with advanced clinical stage and shortened

survival in CLL.7 In our case, blasts at transformation morpho-logically resembled those of Burkitt’s type ALL and showedt(8;14), suggesting that evolution of CLL cells occurred inCorrespondence: N Asou, Second Department of Internal Medicine,association with c-myc gene translocation. There are someKumamoto University School of Medicine, 1-1-1 Honjo, Kumamotoreports that CLL cells at chronic phase also have Burkitt’s type860, Japan

Received 22 November 1996; accepted 30 June 1997 chromosome translocation.8 It is possible that we could not

Correspondence

1987

Figure 2 Southern blot analysis of the Ig heavy chain (a) and c-myc (b) genes. BglII-digested DNA samples were hybridized with theJH probe (C76R51A) and then the filter was reprobed with the c-mycexon II fragment (pMyc6514-R2). Lane 1, K562 cell DNA served asa control for the germline configuration; lane 2, the patient’s DNAfrom PBMNC at initial diagnosis of B-CLL; lanes 3 and 4, the patient’sDNA from PBMNC and BMMNC at acute transformation; lanes 5 and6, PBMNC and BMMNC at the chronic phase after transformation. Asthe intensity of the germline band of the JH gene was weaker thanthose of the rearranged band at each phase, one allele of the JH genein leukemia cells may have been deleted with the germline bandderived from contaminating normal cells. The rearranged c-myc banddid not comigrate with the rearranged JH band.

Figure 1 Prolymphocytes with mature nuclei and nucleoli in theperipheral blood (a), and blasts with deep basophilic cytoplasm andvacuoles in the bone marrow (b) at acute transformation (May–Giemsa staining).

detect abnormal karyotypes in B-CLL cells and prolympho- Acknowledgementscytes because of lacking mitosis. However, we demonstratedthat only blasts had t(8;14) because only BM cells at acutetransformation showed c-myc gene rearrangement (Figure 2b). We thank Dr TH Rabbitts at the Laboratory of Molecular

Biology at Cambridge, UK and the Japanese Cancer ResearchOur case, therefore, indicated that activation of the c-mycgene is closely associated with clonal evolution even in CLL. Resources Bank (JCRB) for providing DNA probes of human

IgJH and c-myc genes, respectively. This work was partly sup-Recently, a similar case of Burkitt’s type ALL supervening onB-CLL was reported in association with t(8;22).9 In addition, ported by a Grant-in-Aid for Scientific Research from the Min-

istry of Education, Science and Culture of Japan, a Grant-in-another case of ALL transformation in CLL carrying t(8;12)which involved c-myc and BTG1 genes was presented.10 Aid for Cancer Research from the Ministry of Health and Wel-

fare of Japan and a grant to NA from the Okukubo MemorialMoreover, blasts from at least two cases of ALL superveningon CLL resembled those of Burkitt’s type ALL.11,12 These Fund for Medical Research in Kumamoto University School

of Medicine.observations suggest the potential importance of c-myc geneactivation in acute transformation of B-CLL. In addition, we

N Asou Second Department ofdetected no mutation of p53 gene exons 5–8 by SSCP-PCRanalysis in PB or BM MNC (data not shown). M Osato Internal Medicine,

K Horikawa Kumamoto University SchoolThe transformations detected in CLL are generallyaccompanied by drug resistance and poor prognosis.1 In K Nishikawa of Medicine, Kumamoto,

O Sakitani Japanaddition, adult Burkitt’s type ALL is an aggressive type of leu-kemia which shows rapid dissemination throughout the lymph L Li

H Yamasakinodes. It is interesting that Burkitt’s type lymphoblastic trans-formation in the present case was almost limited to the BM S Nishimura

T Okuboand responded well to relatively mild chemotherapy. It ispossible that secondary t(8;14) does not have the same patho- H Suzushima

K Takatsukigenetic impact as primary t(8;14).

Correspondence

1988 cal importance of myeloid antigen expression in adult acute lym-Referencesphoblastic leukemia. New Engl J Med 1987; 316: 1111–1117.

7 Juliusson G, Oscier DG, Fitchett M, Ross FM, Stockdill G, MackieMJ, Parker A, Castoldi GL, Cuneo A, Knuutila S, Elonen E, Gahrton1 Foon KA, Thiruvengadam R, Saven A, Bernstein ZP, Gale RP. Gen-

etic relatedness of lymphoid malignancies. Transformation of C. Prognostic subgroups in B-cell chronic lymphocytic leukemiadefined by specific chromosome abnormalities. New Engl J Medchronic lymphocytic leukemia as a model. Ann Intern Med 1993;

119: 63–73. 1990; 323; 720–724.8 Geisler C, Philip P, Plesner T, Andersson P, Zeuthen J, Guld-2 Melo JV, Catovsky D, Galton DAG. The relationship between

chronic lymphocytic leukemia and prolymphocytic leukemia. II. hammer B, Hansen MM. Simultaneous presence of translocationt(14;18) and t(2;8) in a case of chronic lymphocytic leukemia.Patterns of evolution of ‘prolymphocytoid’ transformation. Br J

Haematol 1986; 64: 77–86. Cancer Genet Cytogenet 1986; 22: 35–44.9 Mohamed AN, Compean R, Dan ME, Smith MR, Al-Katib A.3 de Jong D, Voetdijk BMH, Beverstock GC, van Ommen GJB, Wil-

lemze R, Kluin PM. Activation of the c-myc oncogene in a precur- Clonal evolution of chronic lymphocytic leukemia to acute lym-phoblastic leukemia. Cancer Genet Cytogenet 1996; 86: 143–146.sor-B cell blast crisis of follicular lymphoma, presenting as com-

posite lymphoma. New Engl J Med 1988; 318: 1373–1378. 10 Rimokh R, Rouault J-P, Wahbi K, Gadoux M, Lafage M, Archim-baud E, Charrin C, Gentilhomme O, Germain D, Samarut J,4 Litz CE, Arthur DC, Gajl-Peczalska KJ, Rausch D, Copenhaver C,

Coad JE, Brunning RD. Transformation of chronic lymphocytic Magaud J-P. A chromosome 12 coding region is juxtaposed to theMYC proto-oncogene locus in a t(8;12)(q24;q22) translocation inleukemia to small non-cleaved cell lymphoma: a cytogenetic,

immunological, and molecular study. Leukemia 1991; 5: 972– a case of B cell chronic lymphocytic leukemia. Genes ChromosCancer 1991; 3: 24–36.978.

5 Gaidano G, Ballerini P, Gong JZ, Inghirami G, Neri A, Newcomb 11 McPhedran P, Heath CW Jr. Acute leukemia occurring duringchronic lymphocytic leukemia. Blood 1970; 35: 7–11.EW, Magrath IT, Knowles DM, Dalla-Favera R. p53 mutations in

human lymphoid malignancies: association with Burkitt lym- 12 Torelli UL, Torelli GM, Emilia G, Selleri L, Venturelli D, Artusi T,Donelli A, Colo A, Fornieri C. Simultaneously increasedphoma and chronic lymphocytic leukemia. Proc Natl Acad Sci

USA 1991; 88: 5413–5417. expression of the c-myc and m chain genes in the acute blastictransformation of a chronic lymphocytic leukaemia. Br J Haematol6 Sobol RE, Mick R, Royston I, Davey FR, Ellison RR, Newman R,

Cuttner J, Griffin JD, Collins H, Nelson DA, Bloomfield CD. Clini- 1987; 65: 165–170.

Pulmonary embolism in a patient with acute promyelocytic leukemia treated withall-trans retinoic acid

TO THE EDITOR firmed by flebography and Doppler study. Pulmonary embol-ism was confirmed by perfusion scintigraphy. Fibrinolytic

Acute promyelocytic leukemia (APL) accounts for about 10– treatment was administered with r-TPA followed by heparin. Aweek later, new perfusion scintigraphy showed an important15% of acute myeloid leukemias.1 APL has been historically

characterized by a high rate of early death, particularly from improvement of the perfusion of both lungs.Thromboembolic complications may also occur in patientsintracranial bleeding. The hemorrhagic complications have

been attributed to a combination of intravascular thrombin treated with ATRA alone or in combination with chemo-therapy, as seen in our case, but are very uncommon. It is ofgeneration, excessive fibrinolysis and/or proteolytic activities

released from blast cells.2 A recent approach for induction interest that several cases of thromboembolic events, rarelyclinically evident in patients treated with conventionaltreatment in APL is differentiation therapy with all-trans-reti-

noic acid (ATRA). Several trials reported complete remission chemotherapy, have been reported in association with ATRA,especially with ATRA-related hyperleukocytosis.4–7 The rapidwith progressive improvement in the bone marrow over 1 to

3 months and disappearance of the coagulopathy within 1 or correction of the fibrinogenolysis/proteolysis syndrome in APLpatients treated with ATRA may generate a hypercoagulable2 weeks.3 Thromboembolic complications in patients treated

with ATRA are very uncommon and most of them are associa- state during the early phase of treatment, predisposing somepatients to display thromboembolic complications.2 It couldted with ATRA-related hyperleukocytosis.

We report on a 32-year-old patient who presented an APL be necessary to establish a preventive heparinotherapy inthese patients.without clinical and laboratory evidence of coagulopathy. At

diagnosis he did not show hyperleukocytosis, with total leuko-V Jimenez-Yuste Department of Hematologycytes of 1.17 × 109/l. Treatment with ATRA (45 mg/m2) was

then started. After 8 days, chemotherapy treatment was MP Martın Hospital La PazM Canales Madrid, Spainadministered (daunorubicin). On day 27, when he was in

complete remission with a leukocyte count of 9.3 × 109/l, the E OjedaF Hernandez-Navarropatient developed acute dyspnea and chest pain with clinical

signs of acute thrombosis of both legs. Thrombosis was con-

ReferencesCorrespondence: V Jimenez-Yuste, Servicio de Hematologıa y Hemo-terapia, Residencia General, 6a planta Diagona, Paseo de la Castel-lana 261, 28046 Madrid, Spain; Fax: 34 1 7293898 1 Warrell RP, de The H, Wang ZY, Degos L. Acute promyelocytic

leukemia. New Engl J Med 1993; 329: 177–189.Received 26 May 1997; accepted 19 June 1997

Correspondence

19892 Dombret H, Scrobohaci ML, Ghorra P, Zini JM, Daniel MT, Cas- 5 De Lacerda JF, Do Carmo JA, Guerra ML, Geraldes J, De Lacerdataigne S, Degos L. Coagulation disorders associated with acute JMF. Multiple thrombosis in acute promyelocytic leukemia afterpromyelocytic leukemia: corrective effect of all-trans retinoic acid tretinoin. Lancet 1993; 342: 114–115.treatment. Leukemia 1993; 7: 2–9. 6 Hashimoto S, Koike T, Tatewaki W, Seki Y, Sato N, Ueno M, Arak-

3 Castaigne S, Chomienne C, Daniel MT, Berger R, Fenaux P, Degos awa M, Shibata A. Fatal thromboembolism in acute promyelocyticL. All-trans-retinoic acid as a differentiation therapy for acute pro- leukemia during all-trans retinoic acid therapy combined withmyelocytic leukemia. I Clinical results. Blood 1990; 76: 1704– antifibrinolytic therapy for prophylaxis of hemorrhage. Leukemia1709. 1994; 8: 1113–1115.

4 Runde V, Aul C, Heyll A, Schneider W. All-trans retinoic acid: not 7 Pogliani EM, Rossini F, Casaroli I, Maffe P, Porneo G. Thromboticonly a differentiating agent, but also an inducer of thromboem- complications in acute promyelocytic leukemia during all-transbolic events in patients with M3 leukemia. Blood 1992; 79: retinoic acid therapy. Acta Haematol 1997; 97: 228–230.534–535.

Assessment of chimerism in sex-mismatched allogeneic bone marrow transplants(allo-BMT) by in situ hybridization and cytogenetics. Is host cell percentagepredictive of relapse?

TO THE EDITOR an apparently normal chromosome pattern. X-Y FISH was car-ried out on cytogenetic preparations; a total of 1000

Several techniques have been applied to detect and monitor interphase cells per specimen was scored for every patient. Inchimerism after allogeneic bone marrow transplantation (allo- order to determine the confidence limits, the sensitivity andBMT), but none is quite satisfactory.1 Recently, in sex-mis- the specificity of the technique, we followed the approach ofmatched allo-BMT recipients, it has been demonstrated that Dewald et al.2 In our study the normal range for XX cells inFISH with sex chromosome probes (X-Y FISH) is a simple, males was fixed up to 0.791 ± 0.013% and that for XY cellsquick and very sensitive technique.2 In fact, X-Y FISH exam- in females was fixed up to 0.477 ± 0.03%. The mean percent-ines both proliferating and G0 cells and provides more accur- age of host and recipient cells, calculated for every patient atate information about the presence and the number of residual three different periods (14–100, 101–365 and ,365 days),host cells in the recipient marrow. Therefore, X-Y FISH and was always corrected for the cut-offs reported above.conventional cytogenetic studies were carried out in 29 acute After the transplant, 16 patients were in CR and 13 relapsed,leukemia patients (19 acute non-lymphocytic leukemia nine in the marrow and four in extramedullary sites. In the(ANLL) and 10 acute lymphoblastic leukemia (ALL), who statistical analysis, these last four patients were considered inreceived a non-T cell-depleted allo-BMT from an opposite sex CR, as FISH results were obtained only from bone marrowdonor in the period September 1994 to October 1996. Data cells. The median follow-up time determined from the lastwere analyzed as of October 1996. Mean recipient age was FISH and/or cytogenetic study was 262 days (range 42–171530 years (range 14–48); 12 patients were females and 17 days) for the 20 cases in CR and 158 days (range 38–332 days)males. At the time of the transplant seven ANLL (one M2, four for the nine bone marrow relapses. Within the first 20 daysM3 and two M4) were in complete remission (CR) and 12 post-transplant X-Y FISH was the technique of choice to study(one M0, four M2 and seven M4) were in relapse; four ALL chimerism, being successful in all nine cases examined; chro-were in CR and six, including three Ph1+ ALL, were in relapse. mosome analysis, on the contrary, provided results in 4/9The conditioning regimen consisted of busulfan 16 mg/kg over cases. Host cell persistence was observed in 24/24 and in the4 days and CY 60 mg/kg on each of 2 days (BU-CY) in 11 19/23 patients (mean recipient cell percentages were 6.7 andcases, etoposide (VP) 30 mg/kg on day −4, between busulfan 7.3%) studied by X-Y FISH in the first and in the second per-(16 mg/kg) and CY (60 mg/kg on each of 2 days) in eight, CY iod, and in 9/24 and in 5/20 cases examined by cytogeneticsand total body irradiation (TBI) 12 GY (total dose) in 10, all in the same periods. Too few cases were evaluable in the thirdALL. GVHD prophylaxis was carried out with methotrexate,

time-interval. Therefore our results confirm the persistence ofcyclosporine or a combination of the two.recipient cells post-transplant independently of disease statusChromosome studies, performed on bone marrow cells, dis-as already reported by various authors.1–3 The host cell per-played no irregularity in the sex chromosome complementcentage detected is in line with that obtained in other series,and identified clonal karyotype rearrangements in 27/29being comprised between 5 and 33%.4,5 This variability inpatients. In ANLL a t(8;21), accompanied by trisomy 8, wasrecipient cell numbers among the different laboratories isdetected in two cases, a t(15;17) in four, a t(9;11) in four, anpartly due to stromal cell contamination, higher in direct boneinv(16) with additional rearrangements in two, and a t(3;21)marrow smears than in conventional cytogenetic preparationsin three. With regard to ALL, three cases showed the Ph1 chro-(3% vs less than 1%).1 The presence of stromal cells is prob-mosome, two hyperdiploidy, two other abnormalities and twoably the reason why X-Y FISH does not foretell relapse butmerely defines it. In our study we applied linear regression tocorrelate variation in host cell percentage with time and withCorrespondence: P Bernasconi, Divisione di Ematologia, Policlinicorelapse and to compare recipient cell numbers in cases in CRSan Matteo IRCCS, Piazzale Golgi N5, 27100 Pavia, Italy

Received 24 December 1996; accepted 24 June 1997 and in relapse. We observed that FISH host cell percentages

Correspondence

1990were significantly associated wtih relapse (P = 0.001), also in developed an abdominal lymphoadenopathy, the other a leu-

kemic bone marrow infiltration, despite chronic GVHD.patients examined in the first and second period (P = 0.01,Relapse also occurred in a patient with hyperdiploidy. There-P = 0.0008). An identical correlation was seen between cyto-fore in ALL cases, the Ph1 always defined a bad prognosis,genetic results obtained in the second period in cases in CReven if chronic GVHD occurs.and those in relapse (P = 0.007). No correlation with time was

observed. Subsequently logistic regression and Cox test dem-P Bernasconi1 1Istituto di Ematologia,onstrated that FISH mean host cell percentage at 30 days andPM Cavigliano1 Universita di Pavia,at 100 days post-transplant could not predict clinical outcomeE Genini1 Divisione di Ematologia,at 100 or at 160/200 days, respectively. Therefore our dataEP Alessandrino1 Policlinico San Matteo IRCCS;are in contrast with those of Bourhis et al,4 who showed thatA Colombo1 and 2Unita di Biometria,patients with .5% host cells had a higher probability ofC Klersy2 Direzione Scientifica,relapse. In the present study, as in others,5 a progressiveL Malcovati1 Policlinico San Matteo IRCCS,increase of recipient cell is the only factor really predictive ofG Biaggi1 Pavia, Italydisease recurrence. This last is better foretold by the cyto-G Martinelli1genetic abnormality present at diagnosis. In our series, bothS Calatroni1the two ANLL-M2 with the t(8;21) and additional defects (oneM Caresana1

in CR and one with active disease at the time of allo-BMT),M Boni1who did not develop chronic graft-versus-host diseaseC Astori1(GVHD), finally relapsed, one in an extramedullary siteC Bernasconi1(inguinal granulocytic sarcoma) and the other in the marrow.

All four ANLL-M3 cases with the t(15;17), transplanted in CR,Referencesare disease-free 107–1127 days post-transplant and only three

of the four ANLL-M4 cases (three active disease and one CR 1 Wessman M, Popp S, Ruutu T, Volin L, Cremer T, Knuutila S.pretransplant) with the t(9;11) and mild chronic GVHD are in Detection of residual host cells after bone marrow transplantation

using non-isotopic in situ hybridization and karyotype analysis.CR 186–254 days post-transplant. One of the two cases withBone Marrow Transplant 1993; 11: 279–284.inv(16) and additional rearrangements achieved CR after a

2 Dewald GW, Schad CR, Christensen ER, Law ME, Zinsmeister AR,chronic GVHD, the other relapsed in the marrow. All threeStalboerger PG, Jalal SM, Ash RC, Jenkins RB. Fluorescence in situ

cases with t(3;21), transplanted with active disease died, two hybridization with X and Y chromosome probes for cytogeneticof disease recurrence and one of acute GVHD. Therefore in studies on bone marrow cells after opposite sex transplantation.

Bone Marrow Transplant 1993; 12: 149–154.our study t(15;17) patients had the best outcome, followed by3 Durnam DM, Anders KR, Fisher L, O’Quigley J, Bryant EM,those with t(9;11) and chronic GVHD, suggesting that this

Thomas ED. Analysis of the origin of marrow cells in bone marrowallo-BMT complication might overcome the unfavorable sig-transplantation recipients using a Y-chromosome-specific in situ

nificance of the karyotype. The impact of chronic GVHD on assay. Blood 1989 74: 2220–2226.disease-free survival was further stressed by the fact that we 4 Bourhis J-H, Jagiello CA, Black P, O’Reilly RJ, Gerritsen WR.

Mixed chimerism is unfavourably determining the outcome of annoticed a progressive disappearance of the initial chromo-allogeneic bone marrow transplantation. Blood 1991; 78: 86some defect still detected post-transplant in 9/11, 5/7 and 1/3(Abstr.).cases studied at each period and a significantly lower FISH

5 Palka G, Stuppia L, Di Bartolomeo P, Morizio E, Peila R, Guancialihost cell percentage (P = 0.004) in cases suffering from allo- Franchi P, Calabrese G. FISH detection of mixed chimerism in 33BMT complication. With regard to ALL, two of the three Ph1+ patients submitted to bone marrow transplantation. Bone Marrow

Transplant 1996; 17: 231–236.cases, all transplanted with active disease, relapsed. One

Secondary leukemia after autologous peripheral blood progenitor cell transplantationfor breast cancer

TO THE EDITOR dence of secondary acute myeloid leukemia (sAML) or myelo-dysplastic syndrome (MDS) after treatment with doxorubicinand cyclophosphamide.3 Moreover, there are few reports ofAdjuvant chemotherapy for breast cancer has been widely

used as routine practice and benefits derived from this acute leukemia and MDS after autologous transplantationprimarily for breast cancer4 or other solid tumours.5,6 Weapproach have been clearly recognized.1 In the past few years

high-dose chemotherapy followed by hemopoietic stem cell recently observed a 38-year-old white female who developedinfiltrating lobular carcinoma of the right breast in 1993. Atinfusion for breast cancer has also been employed and the

number of patients receiving this approach is rapidly increas- that time right quadrantectomy was performed. Eleven out of19 lymph nodes contained metastases and estrogen and pro-ing.2 Recently some concerns have been raised about the inci-gesterone receptors were positive. One month later chemo-therapy with fluorouracil, epi-doxorubicin and cyclophos-phamide (FEC) was initiated. Drug dosages in the first cycle

Correspondence: S Sica were as follows: cyclophosphamide 1500 mg/m2, 4-epi-Received 14 May 1997; accepted 3 July 1997doxorubicin 120 mg/m2 and fluorouracil 1000 mg/m2. G-CSF

Correspondence

19915 mg/kg was administered after the end of chemotherapy in and treatment options include combination chemotherapy but

for some of these patients the therapeutic window is extremelyorder to mobilize and collect peripheral blood progenitor cells(PBPC). The patient then received three standard cycles of narrow since they have already reached the maximum cumu-

lative dose of anthracyclines and have already been exposedFEC. High-dose chemotherapy with carboplatin 1200 mg/m2,etoposide 900 mg/m2 and melphalan 100 mg/m2 was perfor- to high-dose cyclophosphamide or melphalan. Therefore,

FLAG chemotherapy could be particularly indicated in sAMLmed in October 1993 followed by PBPC reinfusion. InOctober 1996 a mild leukopenia and thrombocytopenia was and excellent results in this setting have already been

reported.16 The question whether high-dose chemotherapyobserved. Three months later the patient was referred to ourDivision of Hematology. Pancytopenia and the presence of per se or cumulative exposure to leukemogenic drugs is

responsible for sAML/MDS is as yet unanswered.blast cells in peripheral blood was evident at this time. Bonemarrow aspiration was diagnostic for acute myelogeneous

S Sica Department of Hematologyleukemia M2 according to the FAB classification. CytogeneticP Salutari Chair of Hematologyanalysis showed a normal female karyotype. No evidence ofM Ciarli Universita’ Cattolica Sacro Cuorebreast cancer recurrence was observed. The patient was sub-N Piccirillo Rome, Italymitted to induction chemotherapy with FLAG regimenL Laurenti(fludarabine 30 mg/m2 days 1–5, cytosine arabinoside 2 g/m2

G Menichelladays 1–5 and G-CSF 400 mg/m2 from day 0 to neutrophilG Leonerecovery) and achieved complete remission. Consolidation

chemotherapy with FLAG was performed in March 1993. Thepatient is in complete remission with short follow-up. She has

Referencesno HLA-identical sibling and an autologous bone marrow orperipheral blood progenitor cell transplantation is planned. 1 Consensus Conference. Adjuvant chemotherapy for breast cancer.

JAMA 1985; 254: 3461–3463.High-dose chemotherapy followed by autologous PBPC rein-2 Greenberg PAC, Hortobagyi GN, Smith TL, Ziegler LD, Frye DK,fusion is considered new and promising for the treatment of

Buzdar AU. Long term follow up of patients with completehematological and non-hematological malignancies. Theremission following combination chemotherapy for metastaticreduction of transplantation-related mortality determined by breast cancer. J Clin Oncol 1996; 14: 2197–2205.

the introduction of PBPC and growth factors, the amelioration 3 Diamandidou E, Buzdar AU, Smith TL, Frye D, Witjaksono M,of supportive treatment and the reduction of procedure-related Hortobagyi GN. Treatment related leukemia in breast cancer

patients treated with fluorouracil-doxorubicin-cyclophosphamidecosts contribute to the growing popularity of such ancombination adjuvant chemotherapy: the University of Texas MDapproach. Nevertheless, there are some concerns about theAnderson Cancer Center experience. J Clin Oncol 1996; 14:occurrence of sAML or MDS as late complication of curative2722–2730.high-dose chemotherapy. These concerns originally derived 4 Roman-Unfer S, Bitran JD, Hanauer S, Johnson L, Rita D, Booth

from retrospective studies of lymphoma patients submitted to C, Chen K. Acute myeloid leukemia and myelodysplasia followinghigh-dose chemotherapy.7–9 The overall incidence in lym- intensive chemotherapy for breast cancer. Bone Marrow Trans-

plant 1995; 16: 163–168.phoma patients is about 5% in a recent survey from the5 Guyotat D, Coiffer B, Campos L, Archimbaud E, Treille D, EhrsamEBMTG with an actual risk of developing sAML/MDS of 2.6

A, Fiere D. Acute leukemia following high dose chemoradio-at 5 years.10 Independent risk factors identified in this largetherapy with bone marrow rescue for ovarian teratoma. Actastudy were the diagnosis of Hodgkin’s disease and the amount Haematol 1988; 80: 52–53.

of previous treatment. Cytogenetic alterations when observed 6 Chan T, Juneja S, Wolf M, Januszewicz E, Cooper I. Secondarywere highly consistent with the diagnosis of therapy-related myelodysplastic syndrome following bone marrow transplan-

tation: report of two cases. Bone Marrow Transplant 1994; 13:sAML/MDS involving chromosome 5 and 7 or complex145–148.changes.10 We believe that the lesson learnt by the hematol-

7 Chao NJ, Nademanee AP, Long GD, Schmidt GM, Donlon TA,ogy community should be shared with oncologists. In factParker P, Slovak ML, Nadasawa LS, Blume KG, Forman SJ. Impor-from 1988 there are reports on the occurrence of sAML/MDS tance of bone marrow cytogenetic evaluation before autologous

after high-dose chemotherapy for solid tumors. The incidence bone marrow transplantation for Hodgkin’s disease. J Clin Oncolis currently unknown but it should be stressed that the number 1991; 9: 1575–1579.

8 Miller JS, Arthur DC, Litz CE, Neglia JP, Miller WJ, Weisdorf DJ.of patients annually submitted to high-dose chemotherapy forMyelodysplastic syndrome following autologous bone marrowsolid tumors is still increasing. The role of alkylating agents,transplantation: an additional late complication of curative cancertopoisomerase II inhibitors and epipodophyllotoxins has beentherapy. Blood 1993; 83: 3780–3785.elucidated on the basis of cytogenetic and molecular stud- 9 Stone RM. Myelodysplastic syndrome after autologous transplan-

ies.11–13 The leukemogenic potential is dose dependent for tation for lymphoma: the price of progress? Blood 1994; 83:alkylating agents and epipodophyllotoxins and their combi- 3437–3438.

10 Milligan DW, Ruiz de Elvira C, Taghipour G, Goldstone AH, Kolbnation with different drugs such as cisplatin and their scheduleHJ. MDS and secondary leukemia after autografting for lym-are probably relevant factors. The role of hemopoietic growthphoma. A final report from the EBMT registry. Bone Marrow Trans-factors should also be considered in the development ofplant 1997; 19: (Suppl. 1) S209a (Abstr. 0834a).sAML/MDS as recently suggested by Brodsky et al.14 Growth 11 Pedersen-Bjergaard J, Sigsgaard T, Nielsen D, Gjedde SB, Philip

factors are widely used after chemotherapy in a variety of P, Hansen M, Larsen SO, Rorth M, Mouridsen H, Dombernowskytumors and they are also currently employed for PBPC mobil- P. Acute monocytic or myelomonocytic leukemia with balanced

chromosome translocations to band 11q23 after therapy with 4-ization and after transplantation. In conclusion, the occur-epi-doxorubicin and cisplatin or cyclophosphamide for breastrence of sAML/MDS, although expected to be rare, should becancer. J Clin Oncol 1992; 10: 1444–1451.discussed with patients before entering experimental trial; fol-

12 Persen-Bjergaard J, Philip P. Balanced translocations involvinglow-up of transplanted patients should also include periodical chromosome bands 11q23 and 21q22 are highly characteristicassessment of hemopoietic function and, as recommended by of myelodysplasia and leukemia following therapy with cytostaticRohatiner, cytogenetic analysis should be performed prior to agents targeting at DNA-topoisomerase II. Blood 1991; 78:

1147–1148.high-dose chemotherapy.15 Prognosis of sAML/MDS is poor

Correspondence

1992 13 Le Beau MM, Albain KS, Larson RA, Vardiman JW, Davis EM, 15 Rohatiner A. Myelodysplasia and acute myelogenous leukemiaafter myeloablative therapy with autologous stem cell transplan-Blough RR, Golomb HM, Rowley JD. Clinical and cytogenetic cor-

relations in 63 patients with therapy-related myelodysplastic syn- tation. J Clin Oncol 1994; 12: 2521–2523.16 Visani G, Tosi P, Zinzani PG, Manfrai E, Testoni N, Clavio M,dromes and acute nonlymphocytic leukemia: further evidence for

characteristic abnormalities of chromosomes nos 5 and 7. J Clin Cenacchi A, Gamberi B, Canara P, Gobbi M, Tura S. FLAG(Fludarabine + high dose cytarabine + G-CSF): an effective andOncol 1986; 4: 325–345.

14 Brodsky RA, Bedi A, RJ Jones. Are growth factors leukemogenic? tolerable protocol for the treatment of ‘poor risk’ acute myeloidleukemias. Leukemia 1994; 8: 1842–1846.Leukemia 1996; 10: 175–177.

Internal tandem duplication in the juxtatransmembrane domain of the flt3 is notinvolved in blastic crisis of chronic myeloid leukemia

TO THE EDITOR

Fms-like tyrosine kinase 3 (flt3), a member of class III receptortyrosine kinase, is expressed in the majority of patients withB-lineage acute lymphoblastic leukemia (ALL) and acutemyeloid leukemia (AML).1,2 In a study on the incidence anddistribution of the expression of the flt3 gene by amplifyingits juxtatransmembrane domain, we found the internal tandemduplication of this region in approximately 17% (5/30) of AMLpatients.3 This length mutation was not found in ALL (0/50)and was preferentially seen in AML with monocytic prolifer-ation. We examined 38 samples from 24 patients with chronicmyeloid leukemia (CML) for the flt3 expression and theinternal tandem duplication. Nineteen samples from 17patients and 19 samples from 11 patients were obtained atchronic phase (CP) and at blastic crisis (BC), respectively.Among 11 patients with BC, seven showed myelocytic pro-liferation, two monocytic, one lymphocytic, and one mega-karyocytic. PCR with two sets of primers, one for the flankingjuxtatransmembrane domain of flt3 and the other for b-actin,was performed in one tube after the synthesis of cDNA. Inchronic phase, flt3 mRNA was not expressed in 13 out of 19samples (68%). Hyperexpression of flt3 mRNA was detectedin 14 out of 19 samples (74%) from eight patients in BC. Fivesequential samples from a patient (case 23) with megakary-ocytic BC showed no expression of flt3 mRNA. All fourpatients (cases 14, 16, 18 and 20) on whom sequential analy-sis at both CP and BC was performed, showed increasedexpression of flt3 mRNA in BC. All 20 amplified productsexhibiting visible bands were in the same size and no lengthmutation was detected even in two BC patients who showedmonocytic proliferation (Figure 1). Although the number ofcases in the present study is rather limited, our findingsindicated that the internal tandem duplication of the flt3 mightnot be involved in the transformation into BC in CML patients.

Figure 1 Thirty-eight cDNAs from 24 patients with CML and onefrom a healthy individual were co-amplified in one tube by two setsof PCR primers, one for flt3 and the other for b-actin. The primersequences were flt3-S; 59-TGTCGAGCAGTACTCTAAACA-39, flt3-AS;59-ATCCTAGTACCTTCCCAAACTC-39, b-actin-S; 59-CTTCCTGGG-CATGGAGTC-39, and b-actin-AS; 59-CGCTCAGGAGGAGCAAT-GAT-39. The product size for flt3 is 366 bp and that for b-actin isCorrespondence: S Yokota, Department of Internal Medicine III, Kyoto

Prefectural University of Medicine, Kawaramachi-Hirokoji, Kamigyo- 210 bp. The numbers above the panels are case numbers. SM, sizemarkers: pBR322/MspI digests; C, chronic phase; B, blastic crisis; N,ku, Kyoto 602, Japan

Received 13 November 1996; accepted 22 January 1997 normal control.

Correspondence

19932 Drexler HG. Expression of FLT3 receptor and response to FLT3T Iwai Department of Internal Medicine III,ligand by leukemic cells. Leukemia 1996; 10: 588–599.S Yokota Kyoto Prefectural

3 Nakao M, Yokota S, Iwai T, Kaneko H, Horiike S, Kashima K,M Nakao University of Medicine,Sonoda Y, Fujimoto T, Misawa S. Internal tandem duplication ofH Kaneko Kyoto, Japan the flt3 gene found in acute myeloid leukemia. Leukemia 1996;

H Nakai 10: 1911–1918.K KashimaS Misawa

References

1 Brasel K, Escobar S, Anderberg R, de Vries P, Gruss H-J, LymanSD. Expression of the flt3 receptor and its ligand on hematopoieticcells. Leukemia 1995; 9: 1212–1218.

Two Burkitt-type lymphoma/leukemia-derived cell lines presenting 3q27translocations and immunoglobulin/BCL6 chimeric transcripts

TO THE EDITOR the J3 sequence, and with 14q32 chromosomal breakpointwithin the area between JH and Sm (Figure 1).

3q27 translocations, affecting the BCL6 gene, are recognized Next, we constructed cDNA libraries from MD901 andMD903 cells and isolated clones, in which the BCL6 exon 1as common specific chromosomal abnormalities in B cell

non-Hodgkin’s lymphomas.1–6 We established two Burkitt cell was substituted by sequences from Jl2–Cl2 intron within theIgl gene (MD901) or JH sequences within the IgH genelines carrying 3q27 translocations besides the Burkitt-specific

8q24 translocations, and analyzed the structural alterations of (MD903). We analyzed the 5′ end structure of these chimerictranscripts using the rapid amplification of cDNA end (RACE)5′ portion of the BCL6 gene.

The cell lines, designated as MD901 and MD903, were technique, employing primers from the BCL6 exons 3 and 2as gene-specific primers. The structure of each RACE productderived from a 60-year-old male patient with Burkitt-type lym-

phoma and a 38-year-old male with L3-type acute lympho- is described in Figure 2.For MD901 cells, two mRNA species (designated as P1 andblastic leukemia, respectively. Both lines were negative for

Epstein–Barr virus. MD901 cells presented the immunopheno- P2) were identified. The overlapping 400 bp Igl parts weretranscribed from Jl2–Cl2 intron in the antisense direction withtype of a mature B cell, expressing surface IgM and k. The

immunophenotype of MD903 cells was that of a pre-B cell, 78 bp additional sequence at the 3′ portion in P2 type, andspliced to the BCL6 exon 2 at the cryptic donor–acceptor sites.expressing cytoplasmic m chains, but not surface immuno-

globulins (Igs). For MD903 cells, three types of RACE product were obtainedThe karyotypes of MD901 and MD903 cells were deter-

mined as 48,XY, t(3;22)(q27;q11), del(6)(q21q25),t(8;22)(q24;q11), add (14)(q32), +2mar and 47,XY,t(3;14)(q27;q32),+7, t(8;22)(q24;q11), dir ins (13;13)(q22;q14q22),der(19)t(1;19)(q23;q13), respectively. In both lines, Southern blotanalysis did not detect rearrangements within the 50 kb areaaround the c-MYC exons, but Northern blot analysis presentedhigh expresssion of the c-MYC mRNA (data not shown).

3q27 breakpoints were located within the first intron of theBCL6 gene in both lines. In MD901 cells, we have alreadyrevealed the juxtaposition of the Igl and the BCL6 genes inhead to head manner, with 22q11 breakpoint at the 5′ end ofthe Jl2 exon.7 We constructed a MD903 genomic library andisolated a recombinant clone, containing the reciprocal der(3)junctions with breakpoints in the IgH and the BCL6 genes.The der(3) chromosome carried the IgH locus with rearrangedD-J region, in which the DA4 sequence was recombinated to

Figure 1 Chromosomal breakpoint analysis in MD903 cells.(a) Restriction maps of the breakpoint region on chromosomes 3q27and 14q32. Relevant functional domains are indicated for chromo-some 14 (D, diversity region; J1–6, joint region; Em, IgH transcriptionalenhancer; Sm, switch m region; Cm, Cm exons). The vertical arrowCorrespondence: Y Nakamura, Third Department of Internal Medi-

cine, Dokkyo University School of Medicine, 880 Kitakobayashi, indicates the chromosomal breakpoint. Restriction enzyme abbrevi-ations are; B, BamHI; E, EcoRI; H, HindIII; G, BglII. (b) NucleotideMibu-machi, Shimotsuga-gun, Tochigi, 321-02, Japan

Received 10 October 1996; accepted 23 May 1997 sequence around the breakpoint. Asterisks show nucleotide identity.

Correspondence

1994Acknowledgements

We thank Miss Furuta, Okada, and Mr Akimoto for theirexcellent technical support. We also thank Dr Hiroyuki Ham-aguchi (Musashino Red Cross Hospital, Tokyo) for providingus with cells from a patient. This work is supported by Grants-in-Aid for Scientific Research from the Ministry of Education,Science, Sports and Culture of Japan.

Y Nakamura1 1Third Department of Internal Medicine,T Miki2 Research Center for Cellular andN Kawamata2 Molecular Biology,K Ohashi2 Dokkyo University School of Medicine,S Hirosawa2 Tochigi;H Kobayashi3 2First Department of Internal Medicine,N Maseki3 Tokyo Medical and Dental University,Y Kaneko3 Tokyo;K Saito1 and 3Department of Hematology Clinic,H Enokihara1 Saitama Cancer Center,S Furusawa1 Saitama, Japan

References

1 Bastard C, Tilly H, Lenormand B, Bigorne C, Boulet D, Kunlin A,Monconduit M, Piguet H. Translocations involving band 3q27 andIg gene regions in non-Hodgkin’s lymphoma. Blood 1992; 79:2527–2531.

2 Ye BH, Lista F, Lo Coco F, Knowles DM, Offit K, Chaganti RSK,Dalla-Favera R. Alterations of a zinc finger-encoding gene, BCL-6, in diffuse large-cell lymphoma. Science 1993; 262: 747–750.

3 Kerchaert J-P, Deweindt C, Tilly H, Quief S, Lecocq G, Bastard C.LAZ3, a novel zinc-finger encoding gene, is disrupted by recurring

Figure 2 Characterization of the chimeric transcripts in MD901 chromosome 3q27 translocations in human lymphomas. Natand MD903 cells. (a) Altered BCL6 genomic locus and structures of Genet 1993; 5: 66–70.the chimeric mRNAs. (b) Nucleotide sequences of fusion junctions of 4 Miki T, Kawamata N, Hirosawa S, Aoki N. Gene involved in thethe Ig/BCL6 chimeric transcripts in MD901 and MD903 cells. 3q27 translocation associated with B cell lymphoma, BCL5,

encodes a Kruppel-like zinc-finger protein. Blood 1994; 83: 26–32.

5 Lo Coco F, Ye BH, Lista F, Corradini P, Offit K, Knowles DM,Chaganti RSK, Dalla-Favera R. Rearrangements of the BCL6 genein diffuse large cell non-Hodgkin’s lymphoma. Blood 1994; 83:1757–1759.

6 Bastard C, Deweindt C, Kerckaert JP, Lenormand B, Rossi A, Pez-and cloned. The shortest one, designated as P1, presented onlyzella F, Fruchart C, Duval C, Monconduit M, Tilly H. LAZ315 bp JH consensus sequence spliced to the BCL6 exon 2. Therearrangements in non-Hodgkin’s lymphoma: correlation with his-other two mRNA species, designated as P2 and P3, contained tology, immunophenotype, karyotype and clinical outcome in 217

0.4 kb (P2 type) or 1.2 kb (P3 type) from intronic and exonic patients. Blood 1994; 83: 2423–2427.JH sequences spliced at the 3′ end of the J5 (P2) or J6 (P3). 7 Miki T, Kawamata N, Arai A, Ohashi K, Nakamura Y, Kato A,

Hirosawa S, Aoki N. Molecular cloning of the breakpoint for 3q27The formation of these Ig/BCL6 chimeric transcripts wastranslocation in B cell lymphomas and leukemias. Blood 1994;confirmed by Northern blot and RT-PCR analysis (data not83: 217–222.shown). In both lines, the RT-PCR using the BCL6 exon 1 and

8 Ye BH, Chaganti S, Chang C-C, Niu H, Corradini P, Chaganti RSK,exon 2 primers did not amplify the band, indicating that the Dalla-Favera R. Chromosomal translocations cause deregulatednormal BCL6 allele was silent (data not shown). BCL6 expression by promotor substitution in B cell lymphoma.

We also performed the Western blot analysis using a rabbit EMBO J 1995; 14: 6209–6217.9 Dallery E, Galiegue-Zouitina S, Collyn-d’Hooghe M, Quief S,polyclonal antiserum against the bacterially produced BCL6

Denis C, Hildebrand M-P, Lantoine D, Deweindt C, Tilly H, Bas-protein. Both MD901 and MD903 cells expressed the BCL6tard C, Kerckaert J-P. TTF, a gene encoding a novel small G pro-protein of unaltered molecular weight, though at lower levelstein, fuses to the lymphoma-associated LAZ3 gene by t(3;4)compared to Burkitt cell line Raji (data not shown). chromosomal translocation. Oncogene 1995; 10: 2171–2178.

The formation of chimeric transcripts from the BCL6 and the 10 Galiegue-Zoutina S, Quief S, Hildebrand M-P, Denis C, Lecocqjuxtaposed genes (IgH, TTF, BOB1/OBF1) has been reported G, Collyn-d’Hooghe M, Bastard C, Yuille M, Dyer MJS, Kerckaert

JP. The B cell transcriptional coactivator BOB1/OBF1 gene fusesrecently.8–10 The structures of the chimeric portion are variousto the LAZ3/BCL6 gene by t(3;11)(q27;q23.1) chromosomal trans-in cases with 3q27 translocation, though the normal BCL6location in a B cell leukemia cell line (Karpas 231). Leukemiaprotein was produced from each species of mRNA.1996; 10: 579–587.Our two cell lines, together with other 3q27-translocated 11 Nacheva E, Dyer MJS, Fischer P, Stranks G, Heward JM, Marcus

lines, Ly8,8 VAL,9 and Karpas 231,11 are considered to be RE, Grace C, Karpas A. C-MYC translocations in de novo B cellrepresentative of the subset of B cell lymphoma with dual lineage acute leukemias with t(14;18) (cell lines Karpas 231 and

353). Blood 1993; 82: 231–240.translocations involving Ig genes.

Correspondence

1995

Isolated chloroma of breast preceding acute nonlymphatic leukemia

TO THE EDITOR showed a phenotype consistent with AML (CD34+/CD33+/CD38+/HLA-DR+/CD71+). Chromosome studies of

Chloroma is an extramedullary tumor consisting of immature bone marrow showed a karyotype of 47,XX,+8,inv(16)(p13;q22) in 100% cells examined. The revised his-myeloid precursors. It usually occurs as a complication in the

course of acute myeloid leukemia (AML), and occasionally in tology of the breast tumor showed that the cells infiltrating thebreast tissue were identical to those from bone marrow biopsychronic myeloid leukemia (CML) or other myeloproliferative

disorders (MPD). Isolated chloromas had been described rela- (Figure 1a) and were positive for lysozyme and myeloperoxi-dase (Figure 1b). These findings indicated that the originaltively rarely. They may occur in various sites, however their

presentation in breast is very uncommon and may be misdiag- tumor of the breast was an isolated chloroma, which precededthe acute myelogenous leukemia. The breast tissue examinednosed, mainly as lymphoma.1 As most of the cases invariably

progress to AML, standard induction chemotherapy for AML is by FISH technique for chromosome 8 did not demonstrate tri-somy 8. The patient was treated with fludarabine phosphaterecommended at the time of diagnosis.2 Several chromosomal

abnormalities were reported in patients with chloroma. They and high-dose cytarabine3 and achieved complete remission(CR). After an additional course of consolidation treatmentwere generally associated with bone marrow involvement,

with t(8;21)(q22;) and inv(16)(p13;q22) being found most with high-dose cytarabine, she successfully underwent anfrequently.

We describe a patient with a tumor of the breast which wasdiagnosed originally as a large cell immunoblastic lymphomaand was treated with intensive chemotherapy. Nineteenmonths later she developed AML. Revised histological exam-ination of the original breast tumor confirmed a diagnosis ofchloroma. The cytogenetic studies of her bone marrow at thediagnosis of AML showed a combination of two chromosomalabnormalities: trisomy of chromosome 8 and inversion(16)(p13;q22), in all the metaphases examined.

A 23-year-old female patient was referred to our hospital inJanuary 1994 with a painless mass in her right breast. Thesurgically removed mass revealed diffuse infiltration by malig-nant cells, which stained negative for keratine and commonleukocytic antigen (CLA), but some of the large cells showedpositive staining for B cells. Morphologic features of sclerosis,starry sky and presence of reticulin in the tumor lead to adiagnosis of diffuse large cell immunoblastic lymphoma withsclerosis. Physical examination was normal except for a scaron her right breast. The clinical staging which included com-plete blood count, kidney and liver tests, CT scan of neck,chest and abdomen, gallium scan, bone marrow aspirationand bone marrow biopsy was normal. The patient was treatedwith an intensive chemotherapy protocol for high-risk lym-phoma, which included doxorubicin. Her follow-up wasuneventful.

Nineteen months later she was admitted to the hospital withfever and chills, her hemoglobin was 9 g/dl, WBC count1.3 × 109/l with absolute neutrophil count 0.176 × 109/l andplatelet count 89 × 109/l. Blood smear revealed 56% blastcells with deeply basophilic cytoplasm and nucleoli. Bonemarrow aspiration showed 90% immature cells with positivestaining for myeloperoxidase. Bone marrow biopsy demon-strated diffuse infiltration by immature cells which stainedpositive for lysozyme. Flow cytometry of bone marrow cells

Figure 1 (a) Breast tissue showing diffuse infiltration by immatureCorrespondence: M Quitt, Institute of Hematology and Blood Bank,Lady Davis Carmel Hospital, 7 Michal St 34362, Haifa, Israel myeloid cells (H&E ×40). (b) The myeloid malignant cells shown in

(a) stained positive for myeloperoxidase (paraffin sections ×40).Received 20 February 1997; accepted 24 June 1997

Correspondence

1996allogeneic bone marrow transplantation and remained in therapy regimen for AML, which includes cytosine arabino-

side significantly improves survival rate in this disease.3 AnCR 22 months after the diagnosis of AML.Isolated chloroma is defined as an extramedullary focus of early diagnosis in our patient would have affected the initial

management of her disease. As a majority of the extramedul-nonlymphatic leukemia in the absence of a previous historyof AML, MPD or myelodysplastic syndrome and the presence lary myeloid cell tumors can be successfully detected by fix-

ation-resistant monoclonal antibodies such as antimyelo-of normal bone marrow. Our patient presented with an iso-lated chloroma of the breast, which is a very rare form of peroxidase and antilysosyme associated antigen (CD68), these

stains should be used in all cases of atypical lymphoidextramedullary leukemia (EML). In a retrospective study of 16patients with isolated granulocytic sarcoma only one infiltrates, which do not stain with lymphoid monoclonal

antibodies.presented with a tumor of the breast4 and a recent review of154 cases with extramedullary myeloid tumors described only10 patients with breast involvement.1 The typical pathological

Acknowledgementspicture of chloroma shows myeloid cells in various stages ofmaturation, but in many cases the morphology resembles

We thank Ms Hanna Bar-El and Mira Ziv for the chromosomelarge cell lymphoma and may lead to misdiagnosis, especiallyanalysis, Dr Aliza Amiel for FISH study and Ms Rebecca Sorekin the absence of bone marrow involvement. As many as 75%for the editorial help.of these tumours were initially diagnosed as lymphoma.4 Our

patient’s breast tumor was diagnosed initially as a large cellM Quitt1 1Institute of Hematology,immunoblastic lymphoma with sclerosis, mainly on the basisI Elmalach2 2Institute of Pathology,of morphological features. In the absence of the bone marrowH Dar3 Lady Davis Carmel Medicalinvolvement at that time, the suspicion for EML was low.E Aghai1 Center andLater, during bone marrow involvement by AML, the lyso-

3The Simon Winter Institute ofzyme and myeloperoxidase staining of the breast tumor con-Human Genetics, Bnai Zionfirmed the diagnosis of EML.

Medical Center,In addition, she had trisomy 8 and inv chromosome 16 atTechnion, Faculty ofthe diagnosis of acute leukemia. Trisomy 8 as a single

Medicine, Haifa, Israelchromosomal abnormality or with additional aberrations hadbeen reported previously in various hematological malig-nancies. A patient with MPD who developed a granulocytic

Referencessarcoma of the sternum had trisomy 8 and t(3;4).5

Inv(16)(p13;q22) is a cytogenetic abnormality associated with1 Byrd JC, Edenfield WJ, Shields DJ, Dawson NA. Extramedullaryacute myelomonocytic leukemia with an excess of eosino-

myeloid cell tumours in acute nonlymphatic leukemia: a clinicalphils.6 It had been demonstrated in 32 patients with EML. review. J Clin Oncol 1995; 13: 1800–1816.Only three of them had this chromosomal aberration at the 2 Imrie KR, Kovacs MJ, Selby D, Lipton J, Patterson BJ, Pantalonydiagnosis of the primary EML; in the rest of the patients it was D, Poldre P, Ngan BY, Keating A. Isolated chloroma: the effect of

early antileukemic therapy. Ann Intern Med 1995; 123: 351–353.diagnosed with EML concomitant with acute leukemia.13 Visani G, Tosi P, Zinzani PL, Manfroi S, Ottaviani E, Testoni N,Though our patient had inv(16)(p13;q22) at the presentation

Clavio M et al. FLAG (fludarabine + high-dose cytarabine + G-of acute leukemia, there was no evidence of monoblasts orCSF): an effective and tolerable protocol for the treatment of poor

eosinophilia by morphological examination, cytochemical risk acute myeloid leukemias. Leukemia 1994; 8: 1842–1846.staining or immunophenotyping. 4 Meis JM, Butler JJ, Osborne BM, Manning JT. Granulocytic sar-

Our patient received intensive chemotherapy, including coma in nonleukemic patients. Cancer 1986; 58: 2697–2709.5 Myint H, Chacko J, Mould S, Ross F, Oscier DG. Karyotypic evol-anthracycline and yet developed AML within a short period

ution in a granulocytic sarcoma developing in a myeloproliferativeof time. This is not surprising in view of the results of otherdisorder with a novel (3;4) translocation. Br J Haematol 1995; 90:studies showing that 90% of patients treated as lymphoma462–464.

with anthracycline regimens developed acute nonlymphatic 6 Le Beau MM, Larson RA, Bitter MA et al. Association of inversionleukemia at the median of 12 months. As mentioned pre- of chromosome 16 with abnormal marrow eosinophils in acute

myelomonocytic leukemia. New Engl J Med 1983; 309: 630–636.viously, initial treatment of EML with an intensive chemo-