Bone Marrow Transplantation, (1999) 23, 1161–1165  1999 Stockton Press All rights reserved 0268–3369/99 $12.00 http://www.stockton-press.co.uk/bmt Myelodysplasia and acute leukemia following high-dose and autologous bone marrow or peripheral blood stem cell transplantation

RM Sobecks1, MM Le Beau1, J Anastasi2 and SF Williams1

1Section of /, Department of and 2Section of Hematopathology, Department of , The University of Chicago, Chicago, IL, USA

Summary: (t-MDS)/acute myeloid leuke- mia (t-AML) has been described after ABMT/PBSCT and -related myelodysplastic syndrome (t-MDS)/ most experiences have involved leukemia arising after acute myeloid leukemia (t-AML) has been reported treatment for Hodgkin’s disease (HD) or non-Hodgkin’s after autologous bone marrow or peripheral blood stem lymphoma (NHL).1–4 More recently, t-MDS/t-AML after cell transplantation (ABMT/PBSCT) for various malig- ABMT/PBSCT has been reported in patients treated for nancies. We retrospectively reviewed all adult breast carcinoma,5,6 multiple myeloma7 and metastatic sem- ABMT/PBSCT cases performed at the University of inoma.8 Chicago Medical Center from 1985 to 1997 in order to These reports suggest that t-MDS/t-AML results from determine the incidence of therapy-related leukemia. chemotherapy and/or given for the pri- Among 649 patients, seven (1.1%) developed therapy- mary malignancy prior to transplant rather than from the related acute lymphoblastic leukemia (one patient) or t- actual high-dose chemotherapy administered for the trans- MDS/t-AML (six patients). Of these seven, primary plant preparative regimen. This hypothesis is supported by malignancies included one case of breast carcinoma, five the time interval required for the disease to develop in cases of Hodgkin’s disease (HD) and one case of non- ABMT/PBSCT patients (approximately 4 to 6 years after Hodgkin’s lymphoma (NHL). Disease-specific inci- initial diagnosis).1–4 This duration is similar to that dences for therapy-related leukemia occurring after observed for t-MDS/t-AML following conventional-dose ABMT/PBSCT were one in 354 (0.3%) for breast carci- chemotherapy.9,10 In contrast, the time from ABMT/PBSCT noma, five in 79 (6.3%) for HD and one in 103 (1%) to the diagnosis of t-MDS/t-AML is substantially less for NHL. The median latency periods for the develop- (median, 2.7 years).1–4 ment of therapy-related leukemia from the time of Most cases of therapy-related leukemia are myeloid. initial diagnosis and of ABMT/PBSCT were 5.5 and 1.5 However, therapy-related acute lymphoblastic leukemia (t- years, respectively, for the combined HD and NHL ALL) occurs in some cases. The term therapy-related leuke- group of patients and 4.4 and 2.8 years, respectively, for mia will be used herein to refer to t-MDS/t-AML as well as the one breast carcinoma patient. All seven patients had t-ALL. We report our institution’s experience of secondary clonal cytogenetic abnormalities, and five had recurring MDS and leukemia after high-dose chemotherapy and abnormalities typical of myeloid disorders. Given the ABMT/PBSCT for various malignancies. Our analysis similar latency period observed in patients treated with includes detailed cytogenetic evaluation. conventional chemotherapy alone, our findings support the hypothesis that therapy-related leukemia after ABMT/PBSCT likely results from pre-transplant ther- Patients and methods apy. Early detection of therapy-related leukemia is therefore critical to exclude these patients from Patients undergoing ABMT/PBSCT. Keywords: therapy-related leukemia; autologous bone From 1985 to 1997, a total of 649 adult patients underwent marrow/peripheral blood stem cell transplantation; cyto- ABMT/PBSCT at the University of Chicago Center. Based on a review of our Bone Marrow Transplant Database (Chicago, IL, USA), the following clinical characteristics were collected for patients who developed therapy-related leukemia: age at time of transplant, sex, dis- Autologous bone marrow or peripheral blood stem cell ease histology and stage, pre-transplant treatment regimens transplantation (ABMT/PBSCT) is widely used for the and radiation therapy (including total dose and fields treatment of many different malignancies. Therapy-related involved), type of transplant (ABMT or PBSCT), prepara- tive regimen, post-transplant consolidative radiation ther- Correspondence: Dr RM Sobecks, The University of Chicago Medical apy, response after transplant, time from initial diagnosis Center, 5841 S Maryland Ave, MC2115, Chicago IL, 60637–1470, USA to transplant and to the development of therapy-related leu- Received 8 October 1998; accepted 6 January 1999 kemia, and post-transplant clinical outcome. All patients Therapy-related leukemia after ABMT/PBSCT RM Sobecks et al 1162 continued to be followed after ABMT/PBSCT and none board and all patients signed informed consent prior to were lost to follow-up. Bone marrow morphology was ABMT/PBSCT. based on both marrow aspirate and trephine morphology. Pre-transplant bone marrow examinations for all patients were usually performed 2–3 months prior to transplant. Therapy-related leukemia Morphology review for all patients was without evidence Patients were closely followed after ABMT/PBSCT and of MDS or acute leukemia. Bone marrow cytogenetic bone marrow examinations were performed routinely on analysis was not performed routinely prior to transplant. patients suspected of having developed therapy-related leu- kemia (ie patients whose blood counts did not recover post- Treatment regimens transplant or patients who developed cytopenias). Cases were classified according to the FAB classification system The transplant preparative regimen for patients was deter- when possible or as refractory cytopenia with multilineage mined by their diagnosis, response to prior therapy and the dysplasia (RCMD)11 when appropriate. Cytogenetic analy- disease-specific treatment protocol available at the time of sis using a trypsin-Giemsa banding technique was perfor- transplantation. Table 1 demonstrates each patients trans- med on cells from aspirated bone marrow, bone marrow plant preparative regimen. All treatment protocols were biopsy specimens and/or peripheral blood samples obtained approved by the University of Chicago institutional review at the time of secondary bone marrow dysfunction. Meta-

Table 1 Clinical characteristics and prior treatment for patients with therapy-related leukemia post-ABMT/PBSCT

Patient Age at Disease Initial stage Pre-ABMT therapy ABMT therapy Source of Response Time from Dx transplant (No. cycles) blood stem after ABMT to ABMT/MDS (years)/Sex cells or Leuk (months)

1 25/F Breast IIa CAF and Hexadrol (8) Cy-7500 mg/m2 BM CR 19/53 Ca LOMAC (3) Thi-675 mg/m2 RTb 2 32/F HD-NS IVA MOPP (6) E-900 mg/m2 PBSC CR 22/67 ABVD (3) Thi-14.4 mg/kg Cy-6 g/m2 BCNU-300 mg/m2 RTc 3 36/M HD-NS IVB MOPP (6) E-900 mg/m2 PBSC CR 45/64 RT to L-spine Thi-14.4 mg/kg MOPP (3) Cy-6 mg/m2 ABVD (6) BCNU-300 mg/m2 4 32/M HD IA RTd E-900 mg/m2 PBSC PR 116/134 ABVD (6)/MOPP (5) Thi-14.4 mg/kg MOPP (8) Cy-6 g/m2 BCNU-300 mg/m2 5 33/F HD-NS IV ABVD/PAV (6) E-900 mg/m2 BM CR 72/92 MOPP (7) Thi-14.4 mg/kg Cy-6 g/m2 BCNU-300 mg/m2 RTe 6 32/M HD-NS IVA MOPP/ABVD (6) E-900 mg/m2 BM CR 20/37 RTf Thi-14.4 mg/kg Cy-6 g/m2 BCNU-300 mg/m2 7 58/M NHL-MC IV COMLA/ABP (4) Bu-16 mg/kg PBSC CR 50/52 Mit/E/Dex (2) Cy-120 mg/kg Mit/E/Dex/VCR/IFN (3) AraC-1500 mg/m2 DHAP (3)

Cy = cyclophosphamide; Thi = thiotepa; E = etoposide; Mit = mitoxantrone; Dex = dexamethasone; VCR = vincristine; Cis = cisplatin; IFN = interferon- alpha; AraC = cytarabine; RT = radiation therapy; CAF = cyclophosphamide, doxorubicin, fluorouracil; MOPP = mechlorethamine, vincristine, procarbaz- ine and prednisone; ABVD = doxorubicin, bleomycin, vinblastine and dacarbazine; EVA = etoposide, vinblastine and doxorubicin; COMLA = cyclophosphamide, vincristine, methotrexate, leucovorin and cytarabine; PAV = procarbazine, adriamycin and vinblastine; LOMAC = leucovorin, vincristine, methotrexate, doxorubicin and cyclophosphamide; DHAP = dexamethasone, cytarabine and cisplatin; MC = mantle cell NHL; HD = Hodgkin’s disease; BM = bone marrow; PBSC = peripheral blood stem cells; CR = complete response; PR = partial response. aMetastatic disease at time of ABMT. b5000 cGy adjuvant RT to chest and axilla. cModified mantle RT. d4900 cGy extended mantle initially, then after ABVD/MOPP 4010 cGy inverted Y, 1050 cGy left inguinal boost, and 3000 cGy to mediastinum. e4000 cGy inverted mantle and pelvis post-ABMT. f4500 cGy full mantle RT. Therapy-related leukemia after ABMT/PBSCT RM Sobecks et al 1163 phase cells from short-term (24 and 48 h) unstimulated cul- chemotherapy regimen for ALL, she remained in complete tures were examined. Chromosomal abnormalities are remission for 24 months. She then developed relapsed ALL described according to the International System for Human and subsequently underwent an allogeneic bone marrow Cytogenetic Nomenclature (ISCN, 1995).12 transplant using total body irradiation and high-dose etopo- side as the preparative regimen. This patient died 1 month post-allogeneic transplant of respiratory and hepatic failure. Results Hodgkin’s disease Of 649 patients who received ABMT or PBSCT, seven (1.1%) subsequently developed MDS or leukemia. This The median age at the time of transplant for the five Hodg- included one patient with breast carcinoma, five patients kin’s disease patients who developed therapy-related leuke- with Hodgkin’s disease and one patient with NHL. The mia was 32 years (range, 32–36). These patients received overall incidence of MDS or leukemia post-transplant was a median of two chemotherapy regimens prior to transplant one in 354 (0.3%) breast carcinoma patients, five in 79 (range, 2–3) and three patients received radiation therapy (6.3%) Hodgkin’s disease patients, and one in 103 (1%) as well (see Table 1). Four of the patients (2, 3, 5 and 6) NHL patients. No cases of therapy-related leukemia were achieved a complete remission post-ABMT/PBSCT. Three observed among the remaining 113 ABMT/PBSCT patients of these four later developed recurrent disease at a median who had diagnoses of multiple myeloma, ovarian carci- of 19 months post-transplant (range, 19–28 months). noma, or various other solid tumors. Patient characteristics, Patient 2 received salvage chemotherapy for relapsed treatment and outcome are described in Tables 1 and 2. Hodgkin’s disease post-PBSCT with etoposide, vinblastine and doxorubicin (EVA) followed by three cycles of mesna, Breast carcinoma ifosfamide, mitoxantrone and etoposide (MINE). She died 5 months later with congestive heart failure, t-MDS and Patient 1 had metastatic breast carcinoma and had achieved progressive HD. Patient 3 had recurrent HD in his bone a complete remission post-ABMT with no evidence of dis- marrow at the time that t-MDS was diagnosed. He sub- ease recurrence. However, 34 months post-transplant she sequently received cytarabine and daunorubicin followed developed CD10-positive acute lymphoblastic leukemia by a clinical trial with vincristine in which stellazine and with a hyperdiploid karyotype (Ͼ50 chromosomes). After verapamil were employed as potential modulators of multi- a standard induction, intensification and maintenance drug resistance. He died of an infectious pneumonitic pro-

Table 2 Post-ABMT/PBSCT presentation of therapy-related leukemia

Patient Bone marrow morphology Bone marrow Cytogenetic abnormalities Status of primary Survival status in diagnosis malignancy after months after ABMT/PBSCT ABMT/PBSCT (t-MDS/leukemia)

1 BM cellularity: 35% ALL (common 54, XX, +X, +4, +6, +14, +17, +18, NED D 67 (33) 55% lymphoblasts precursor B) +21, +21[3]/46, XX [23] 2 BM cellularity: 75% t-MDS c/w RAEB 45, X, t(X;1)(p22;q32), der (5) recurrent HD 28 D 47 (2) 8% blasts in BM t(5;?;7)(q11;?;p14), −7, months post-PBSCT trilineage dysplasia t(9;13)(q22;q11), −18, +mar1 [19]/46, XX [1] 3 mixed BM cellularity (20–90%) t-MDS c/w RAEB 46, XY, t(1;2)(p36;p21), recurrent HD 19 D 23 (4) blasts: 15% BM and 8% PB t(2;12)(p21;p11), del(5)(q23q35 or months post-PBSCT trilineage dysplasia recurrent HDa q22q34), t(14;21)(q22;q22) [20]/45, X, −Y, t(1;2), del(5),t(14;21)[11] 4 BM cellularity: 45–50% FAB-unclassified c/w 46, XY, t(4;14)(q21;q13), progressive HD D 27 (9) 2% blasts in BM RCMD t(10;?16)(q24;p13) [2]/46, XY [17] granulocytic/ 5 BM cellularity progressively t-AML 46, XX, −7, +mar [19]/46, idem, NED D 36 (4) increased from 30 to 90%b del(15)(q2?2q2?5) [2] trilineage dysplasia 33% blasts in PB 6 BM cellularity: 15% FAB-unclassified c/w 45, XY, −13 [4]/46, XY [27] recurrent HD 19 D 23 (7) 2% blasts in BM RCMD months post-ABMT* granulocytic/erythroid dysplasia 7 BM cellularity progressively FAB-unclassified c/w 46, XY, t(15;21)(q22;q22) [3]/46, recurrent NHL 14 D 14 (12) increased from 15 to 90%c RCMD XY [19] months post-PBSCT trilineage dysplasia

PB = peripheral blood; BM = bone marrow; c/w = consistent with; D = dead; A = alive; RAEB = refractory with excess blasts; RCMD = refractory cytopenia with multilineage dysplasia; * = recurrent HD only detected in right cervical lymph node. aHD located in hypocellular areas. bOver a 3-month interval. cOver a 9-month interval. Therapy-related leukemia after ABMT/PBSCT RM Sobecks et al 1164 cess with possible progressive HD 4 months post relapse probability of developing t-MDS/t-AML was 1.6%. Median (no autopsy was performed). Patient 4 died of progressive latency periods from the time of initial therapy and from HD 27 months post-PBSCT. Patient 5 developed a dissemi- ABMT to the diagnosis of t-MDS/t-AML were 47 months nated varicella zoster infection 3 years post-ABMT and (range, 26–71 months) and 28 months (range, 2–41 died of probable sepsis. Patient 6 developed recurrent HD months), respectively. Median survival after the diagnosis limited to a right cervical lymph node. However, he died of t-MDS/t-AML was 9 months. Roman-Unfer et al6 ident- 4 months later after having developed nocardia, cytomega- ified two breast carcinoma patients who developed t- lovirus and Enterococcus faecium pneumonia with sub- MDS/t-AML after ABMT, both of whom had recurring sequent multiorgan system failure. cytogenetic abnormalities (one patient with an 11q23 trans- Three of the five HD patients (2, 3 and 5) had character- location and one patient with a del(5q)). The median lat- istic cytogenetic abnormalities with loss or deletions of ency period for the development of t-MDS/t-AML post- chromosomes 5 and/or 7 (Table 2). The other two HD ABMT was 15.5 months (range, 8–23 months). patients (4 and 6) had clonal abnormalities detected: t(4;14) The single patient treated for breast carcinoma in our and t(10;?16) for patient 4, and monosomy 13 for patient series developed ALL. This patient was included in our 6. No patient had a translocation involving 11q23 which report since she developed a leukemia post-ABMT. In occur in t-AML after treatment with topoisomerase II addition, therapy-related ALL has been described pre- inhibitors. viously in breast carcinoma patients after treatment with The median times from initial diagnosis to prior doxorubicin-containing regimens.13 However, ther- ABMT/PBSCT and to therapy-related leukemia were 45 apy-related ALL has thus far only been clearly established months (range, 20–116 months) and 64 months (range, 37– in the setting of 11q23 translocations (MLL rearrange- 134 months), respectively. The median overall survival for ment). Thus our patient’s leukemia may well have been the HD patients was 27 months post transplant (range, 23– a second primary malignancy rather than therapy-related 47 months) and 4 months after developing therapy-related disease. Though she did not have an 11q23 translocation, leukemia (range, 2–9 months). the latency periods from the time of initial diagnosis and from ABMT to the development of her leukemia (53 and 34 months, respectively) were comparable to those reported Non-Hodgkin’s lymphoma by Laughlin et al5 (see above). Of note, the time period One patient had mantle cell NHL (patient 7, Table 1) and from initial diagnosis is shorter for breast carcinoma developed t-MDS 2 months post-transplant. He died 12 patients than for HD and NHL patients (47–53 months vs months later after developing a recurrence of his primary 66 months, respectively), whereas the latency period from lymphoma which transformed to a blastic variant with ABMT to therapy-related leukemia is similar (15.5–34 leukemic manifestations concurrent with t-MDS. This months vs 18–32.4 months, respectively). This earlier patient had a translocation involving 21q22, which has occurrence of therapy-related leukemia in breast carcinoma been observed in a small subset of t-MDS/t-AML patients may be due to the small patient numbers in the (approximately 3%). reported series. Investigators at the University of Arkansas reported 188 multiple myeloma patients, 71 of whom were enrolled in a Discussion total therapy program which allowed no more than one cycle of prior standard chemotherapy (median duration of The risk of t-MDS/t-AML has increased with further fol- pre-transplant therapy was 7.6 months).7 The remaining low-up time after ABMT/PBSCT. Reported crude inci- 117 patients had more prolonged pretransplant chemo- dence rates range from 2.3–7.6% for HD and NHL therapy (median, 24 months). Therapy-related MDS patients1–4 and 0.6–5.7% for breast carcinoma patients.5,6 developed in seven patients from the latter group, whereas Our series demonstrates a relatively low overall incidence no cases were observed among the ‘total therapy’ group of (1.1%) of therapy-related leukemia after ABMT/PBSCT; patients with limited prior standard chemotherapy. These however, the incidence for our HD patients (6.3%) was investigators concluded that pretransplant therapy was the comparable to prior experiences.1–4 main risk factor for t-MDS post-autologous transplantation In our HD and NHL patients, the median latency periods in multiple myeloma. Among 28 multiple myeloma patients for the development of t-MDS/t-AML from the time of at our institution no cases of therapy-related leukemia were initial diagnosis and from the time of ABMT/PBSCT were identified after PBSCT. Similar to the ‘total therapy’ group similar to those in earlier reports.1–4 From the time of initial from the University of Arkansas series, most of our patients diagnosis we observed a 5.5 year median latency period had received three cycles of vincristine, doxorubicin and (range, 4.3–11.2 years) as compared with 5.5 years (range, dexamethasone, followed by high-dose cyclophosphamide 3.9–5.8 years) in other series. After ABMT/PBSCT, our for stem cell mobilization and one cycle of EDAP patients had a median latency period of 1.5 years (range, (etoposide, dexamethasone, cytarabine and cisplatin) prior 0.2–3.8 years) while that of other institutions was 2.7 years to transplant.14 However, several of our patients had also (range, 1.4–3.7 years). received additional prior therapy including melphalan. The development of t-MDS/t-AML has recently been All seven patients in this report had clonal cytogenetic recognized post-ABMT/PBSCT for breast carcinoma.5,6 abnormalities. Patient 1 with therapy-related ALL had Laughlin et al5 reported five patients, one of whom had a hyperdiploid karyotype with Ͼ50 chromosomes, a a poor-risk cytogenetic abnormality [del(7q)]. The 4-year recurring abnormality in ALL. Of the six patients with Therapy-related leukemia after ABMT/PBSCT RM Sobecks et al 1165 t-MDS/t-AML, five had a recurring abnormality typical of acute myelogenous leukemia following high-dose chemoradi- myeloid disorders. In particular, three patients (2, 3 and 5) otherapy and autologous stem-cell transplantation for lymph- had loss or deletion of chromosome 5 and/or 7, character- oid malignancies. J Clin Oncol 1994; 12: 2527–2534. istic of t-AML arising after alkylating agent therapy, and 2 Traweek ST, Slovak ML, Nademanee AP et al. Clonal kary- one (patient 7) had a translocation involving 21q22, charac- otypic hematopoietic cell abnormalities occurring after auto- teristic of t-AML arising after therapy with topoisomerase logous bone marrow transplantation for Hodgkin’s disease and II inhibitors. Patient 6 had loss of chromosome 13. Overall, non-Hodgkin’s lymphoma. Blood 1994; 84: 957–963. 3 Stone RM, Neuberg D, Soiffer R et al. Myelodysplastic syn- the cytogenetic pattern of these t-MDS/AML cases arising drome as a late complication following autologous bone mar- after ABMT/PBSCT is similar to that observed in our series row transplantation for non-Hodgkin’s lymphoma. J Clin of patients with t-MDS/AML occurring after cytotoxic Oncol 1994; 12: 2535–2542. therapy alone. 4 Miller JS, Arthur DC, Litz CE et al. Myelodysplastic syn- In our recently updated series of 270 consecutive patients drome after autologous bone marrow transplantation: an with t-MDS/t-AML, 252 (93%) had a clonal chromosomal additional late complication of curative cancer therapy. Blood abnormality (Ref. 15 and Le Beau et al, unpublished), and 1994; 83: 3780–3786. 191 patients (71%) had a clonal abnormality leading to loss 5 Laughlin MJ, McGaughey DS, Crews JR et al. Secondary or deletion of chromosome 5 and/or 7. Overall, 115 patients myelodysplasia and acute leukemia in breast cancer patients (42%) had abnormalities of chromosome 5, and 138 (51%) after autologous bone marrow transplant. J Clin Oncol 1998; had abnormalities of chromosome 7. Nine patients (3.3%) 16: 1008–1012. had a translocation of 11q23, and eight (2.9%) patients had 6 Roman-Unfer S, Bitran JD, Hanauer S et al. Acute myeloid a translocation of 21q22. Three patients (1%) had −13 or leukemia and myelodysplasia following intensive chemo- del(13q). therapy for breast cancer. Bone Marrow Transplant 1995; 16: 163–168. Our current report supports the hypothesis that therapy- 7 Govindarajan R, Jagannath S, Flick JT et al. Preceding stan- related leukemia after ABMT/PBSCT likely results from dard therapy is the likely cause of MDS after autotransplants pre-transplant therapy. However, the high-dose chemo- for multiple myeloma. Br J Haematol 1996; 95: 349–353. therapy may have had a contributory role since no patient 8 Chan T, Juneja S, Wolf M et al. Secondary myelodysplastic had evidence of therapy-related leukemia immediately prior syndrome following bone marrow transplantation: report of to ABMT/PBSCT. Since the use of ABMT/PBSCT will two cases. Bone Marrow Transplant 1994; 13: 145–148. probably continue to increase in the future, therapy-related 9 Coleman CN, Williams CJ, Flint A et al. Hematologic neo- leukemia will remain a significant post-transplant risk. plasia in patients treated for Hodgkin’s disease. New Engl J Recognizing those patients with evidence of therapy-related Med 1977; 297: 1249–1252. leukemia prior to transplant is important in order to exclude 10 Pedersen-Bjergaard J, Ersbol J, Sorensen HM et al. Risk of them from undergoing ABMT/PBSCT. As in the case of acute nonlymphocytic leukemia and preleukemia in patients HD,16 routine bone marrow cytogenetic analysis prior to treated with cyclophosphamide for non-Hodgkin’s lymphoma. ABMT/PBSCT may allow early detection. This should be Ann Intern Med 1985; 103: 195–200. considered particularly for those patients extensively 11 Rosati S, Mick R, Xu F et al. Refractory cytopenia with multi- treated prior to their transplant. In addition, patients should lineage dysplasia: further characterization of an ‘unclassifi- able’ myelodysplastic syndrome. Leukemia 1996; 10: 20–26. have ABMT/PBSCT performed earlier in their disease 12 ISCN (1995). An International System for Human Cytogenetic course, or at least have their blood stem cells harvested Nomenclature. Mitelman F (ed). Karger: Basel, 1995. before being heavily pretreated. Finally, further efforts in 13 Archimbaud E, Charrin C, Guyotat D et al. Acute leukaemia the treatment of therapy-related leukemia are clearly war- with t(4;11) in patients previously exposed to carcinogens. Br ranted given the uniformly poor outcome of these patients. J Haematol 1988; 69: 467–470. 14 Sobecks RM, Zimmerman TM, Grinblatt DL et al. A phase II trial of high-dose late intensification therapy in patients with chemotherapy sensitive multiple myeloma. Blood 1997; 90 Acknowledgements (Suppl. 1, part 2): 408b (Abstr. 4582). 15 Le Beau MM, Albain KS, Larson RA et al. Clinical and cyto- genetic correlations in 63 patients with therapy-related myelo- We very much appreciate data management support from Mrs dysplastic syndrome and acute nonlymphocytic leukemia: Sheila Huffman/Dertz, Ms Karen Welborne and Ms Marjorie further evidence for characteristic abnormalities of chromo- Isaacson. We also are grateful to Dr Richard Larson for providing somes 5 and 7. J Clin Oncol 1986; 4: 325–345. information from a t-MDS/t-AML database and for the assistance 16 Chao NJ, Nademanee AP, Long GD et al. Importance of bone of the technologists in the Hematology/Oncology Cytogenetics marrow cytogenetic evaluation before autologous bone mar- Laboratory at the University of Chicago. This work was supported row transplantation for Hodgkin’s disease. J Clin Oncol 1991; in part by PHS grant CA 40046 (MM LeBeau). 9: 1575–1579.

References

1 Darrington DL, Vose JM, Anderson JR et al. Incidence and characterization of secondary myelodysplastic syndrome and