Transplantation (2003) 32, 543–548 & 2003 Nature Publishing Group All rights reserved 0268-3369/03 $25.00 www.nature.com/bmt Conditioning Regimens Randomized trial of busulfan vs total body irradiation containing conditioning regimens for children with acute lymphoblastic leukemia: A Pediatric Blood and Marrow Transplant Consortium study

N Bunin1, R Aplenc1, N Kamani2, K Shaw1, A Cnaan1 and S Simms1

1The Children’s Hospital of Philadelphia, Philadelphia, PA, USA; and 2Children’s National Medical Center, Washington, DC, USA

Summary: hypothyroidism, and an increase in secondary malignan- cies.Busulfan (Bu), as an alternative to TBI, is attractive in Conditioning regimens for children with ALL have that radiation sequelae are avoided, and it can be generally included total body irradation (TBI), which administered in small children without the need for may result in significant sequelae. The primary aim of this sedation or anesthesia.Prospective randomized trials study was to evaluate the outcome for children with ALL comparing Bu to TBI have been performed in patients undergoing allogeneic stem cell transplant (SCT) with with AML and CML, and the use of a cyclophosphamide either busulfan (Bu) or TBI regimens. Patients o21 years dose of 120 mg/kg with Bu in adults, and 200 mg/kg in with ALL undergoing allogeneic SCT were eligible. children has resulted in event-free survival (EFS) equivalent Conditioning included either Bu or TBI, with etoposide to that of TBI and 120 mg/kg of cyclophosphamide.1–3 40 mg/kg and cyclophosphamide 120 mg/kg. Randomiza- While Bu has been used in small numbers of patients with tion was stratified based upon duration of remission, ALL,4–7 its efficacy and toxicity as part of the conditioning remission status, and prior cranial irradiation. A total of regimen have never been tested in a randomized fashion. 43 patients were enrolled; 21 received Bu and 22 TBI. To evaluate the efficacy and toxicity of Bu in patients with Median patient age was 8 years (0.5–20 years). Remission ALL, the Pediatric Blood and Marrow Transplant Con- status included 12 patients in CR1, 25 in CR2, and six in sortium developed a randomized trial to compare a Bu CR3. At a median follow-up of 43 months, event-free containing conditioning regimen with a TBI conditioning survival (EFS) is 45% at 3 years, with 29% EFS in the Bu regimen.Etoposide was added to the regimen for potential arm and 58% in the TBI arm (P ¼ 0.03). There was no increase in antileukemic effect while retaining the immuno- significant difference between Bu and TBI for patients suppression of cyclophosphamide to ensure engraftment. who received stem cells from related donors (36 vs 58%, P ¼ 0.3). However, for URD, EFS was 20% for Bu and 57% for TBI (P ¼ 0.04). Relapses were similar in both Patients and methods arms. This randomized prospective study suggests that Bu is inferior to TBI for pediatric patients with ALL This study opened in January 1997 and closed in October undergoing allogeneic SCT. 2000.The Institutional Review Boards of the 11 participa- Bone Marrow Transplantation (2003) 32, 543–548. ting institutions approved the protocol, and parents or doi:10.1038/sj.bmt.1704198 guardians signed informed consent prior to randomization Keywords: ALL; busulfan; TBI of therapy. Conditioning included either oral Bu 1 mg/kg (or 40 mg/m2 for patients p20 kg) every 6 h for 16 doses or TBI Allogeneic stem cell transplant (SCT) may be a curative 1200 cGy/six fractions/3 days, in addition to Etopophos therapy for pediatric patients with ALL resistant to (Etoposide phosphate, Bristol-Myers Squibb, Princeton, .To achieve engraftment and eradicate NJ, USA) 40 mg/kg once daily i.v. day À4 and cyclophos- leukemia, conditioning therapy generally has included total phamide 60 mg/kg once daily i.v. days À3 and À2 (total body irradiation (TBI), in addition to high-dose chemo- dose 120 mg/kg).Bu pharmacokinetics were not obtained therapy.The late effects of TBI in children are well routinely.Patients who received Bu and unrelated donor established and may include growth impairment, cataracts, grafts received antithymocyte globulin, 30 mg/kg/day i.v. days À2, À1, 0 for increased immunosuppression to facilitate engraftment.Additional CNS prophylaxis was given to patients who received Bu, and who had no prior CNS irradiation, as there was concern about the possibility Correspondence: Dr N Bunin, The Children’s Hospital of Philadelphia, Division of Oncology, 34th and Civic Center Blvd., Philadelphia, PA of increased CNS relapse in these high-risk patients. 19104, USA. Email: [email protected] Intrathecal methotrexate, starting on day +100, was Received 7 November 2002; accepted 22 April 2003 administered every 3 months for six doses.Patients with Busulfan vs TBI conditioning for pediatric ALL N Bunin et al 544 prior CNS disease, and no irradiation, received 600 -cGy Table 1 Clinical features cranial irradiation prior to conditioning, if randomized to BuTBI TBI.Patients with prior CNS disease received 1800 cGy before conditioning with Bu.-versus-host disease p6 years 6 8 (GVHD) prophylaxis was cyclosporine 7 methotrexate >6 years 15 14 for matched siblings, cyclosporine and ‘short-course’ Immunophenotype methotrexate for unrelated donor recipients, and cyclos- CD 10+ 16 15 porine/methylprednisolone for cord blood recipients.Sup- CD 10À 23 T cell 1 3 portive care was not uniformly mandated, but all patients NA 2 1 received phenytoin for seizure prophylaxis if they received busulfan. Cytogenetics Normal 4 8 Engraftment was defined as achievement of an absolute Hyperdiploid 6 3 neutrophil count 4500 for 3 consecutive days.Chimerism Hypodiploid 1 0 analysis was not generally available.GVHD was graded by t(9;22) 4 3 commonly accepted criteria.8 t(4,11) 2 4 Other 1 1 Late effects studies included neuropsychologic, pulmon- NA 4 2 ary, cardiac, endocrine, and ophthalmologic evaluations. CR1 6 6 CR2 12 13 CR3 3 3 Statistical analysis Donor MRD 11 10 Patients were randomized based upon stratification for URD 10 12 duration of first remission (less than or greater than 18 months), remission status and prior CNS irradiation. Donor to Patient MM 7 8 Randomization was performed within 1 month of projected MF 4 4 SCT. FM 7 4 Patient status was determined as of 31 January 2003. FF 3 4 Clinical data were collected on case report forms and were NA — 2 then entered into a relational database developed in Access Source 97.Data were exported into STATA 7.0for statistical BM 19 20 analysis.Patient characteristics, toxicity, and GVHD were Cord 1 2 summarized with descriptive statistics.Kaplan–Meier was PSC 1 0 CSA 3 2 used to describe overall survival (OS) and EFS.The CSA/MTX 16 17 Wilcoxon rank-sum test was used to compare survival CSA+methylprednisolone 1 2 between groups.Cox’s proportional-hazards analysis was G or GM-CSF 7 7 used for multivariate analysis.All P-values were two sided. Interim analyses of relapse and nonrelapse mortality were NA ¼ not available; MRD ¼ matched related donor; URD ¼ unrelated performed every 6 months. donor; CSA ¼ cyclosporine; MTX ¼ methotrexate; M ¼ male; F ¼ female.

Table 2 Relapse and nonrelapse mortality Results BuTBI Clinical characteristics and transplant details are listed in Relapses 9 7 Table 1.There were 29 boys and 14 girls enrolled.The Bone marrow 4 4 BM+extramedullary 2 1 median age at SCT was 8.7 years (range 0.7–20 years). For CNS 2 0 patients in first remission (CR1), the median time from Testicular 0 2 diagnosis to BMT was 5.7 months (range 3.6–11.5 months). Other 1 0 All CR1 patients had either t(4;11) or t(9;22).For other Nonrelapse mortality 5 2 MSOF-2, infection-1, Fungus-1, patients, the median time from diagnosis to transplant was GVHD-2 EBV LPD-1 40.2 months (range 9.8–82.4 months). There were 20 related donors, and 19 were matched siblings.One was a parent mismatched at the B locus.Two of four cord bloods were MSOF ¼ multisystem organ failure; EBV LPD ¼ Epstein–Barr virus mismatched at a single locus.All unrelated donors were lymphoproliferative disorder. matched by serology at class I, and matched at DRB1. A total of 42 patients engrafted.One patient, who Relapses occurred in nine patients who received busulfan received Bu, and a related donor mismatched at B, failed to and seven who received TBI at a median of 6.4 months engraft.In total, 21 patients are alive and disease -free at (range 3.2–28.7 months) (Table 2). Five patients had 13–63 months post transplant.The median follow-up for isolated extramedullary relapses, including skin, bone, surviving patients is 43.3 months (range 13–64 months). testes, CNS.Four patients, who had received Bu, under- OS at 3 years is 47% for Bu patients, and 67% for TBI went a second transplant with TBI regimens.Three were re- patients (P ¼ 0.09). EFS is 45% at 3 years, with 29% for Bu transplanted for relapse, and one for graft failure.Three patients and 58% for TBI patients (P ¼ 0.03). are alive and well at 10, 12, and 27 months post second

Bone Marrow Transplantation Busulfan vs TBI conditioning for pediatric ALL N Bunin et al 545 transplant; one patient died of pulmonary insufficiency. Table 3 EFS by patient age and CR status Three other patients remain in remission following addi- BuTBI P-value tional chemotherapy at 3, 22, and 29 months post relapse. Acute GVHD developed in 11 patients, and was grades N EFS (95% CI) N EFS (95% CI) III–IV in two patients.Chronic GVHD occurred in four Age p6 years 6 16.7% (0.01–51.7) 8 60% (19.6–85.2) 0.056 patients, and was limited in two.Seven patients died of Age >6 years 13 36.3% (11.2–62.7) 14 56.3% (27.2–77.6) 0.31 non-relapse causes (Table 2).All but one patient died oday CR 1 5 33.3% (4.6– 67.6) 6 83.3% (27.3–97.5) 0.045 60; the cause of death of the patient who died at day 270 Other CR 14 30.5% (9.8–54.5) 16 46.9% (20.8–69.4) 0.24 was attributable to complications from chronic GVHD. Donor source, patient age, and CR status were determinants of outcome when stratified by conditioning Table 4 Significant Cox’s Proportional-hazards results regimen at 3 years of follow-up.Overall, there was no Variable Hazard ratio P-value 95% CI difference in EFS between patients who received stem cells from related donors vs those who received unrelated donor Conditioning regimen 3.0 0.025 1.2 8.0 cells (46 vs 40%, P ¼ 0.3). There was a trend toward Donor type 2.3 0.083 0.9 5.8 improved EFS for MRD recipients who received TBI, 58%, compared to those who received Bu, 36%, but this was not statistically significant.However, for patients who reach statistical significance and the other evaluated received unrelated donor stem cells, EFS was inferior with covariates were not statistically significant. Bu, 20% in contrast to 57% with TBI (P ¼ 0.04) (Figure 1). Follow-up studies were obtained inconsistently, and age EFS was 42% for patients under 6 years of age and years, precluded pulmonary function studies for young patients. and 44% for those older than 6 years (P ¼ 0.5). Patients less The Lansky activity scores were 90 for one patient with than 6 years had an inferior EFS with Bu conditioning, bronchiolitis obliterans, and 100 for 17 others.Follow-up P ¼ 0.056 (Table 3). There was no significant difference in pulmonary function studies were available for six patients outcome between patients transplanted in first or second who received TBI; one patient has bronchiolitis obliterans remission (TBI EFS of 66% vs Bu EFS 46%) at 3 years. associated with chronic GVHD, and one has mild However, patients transplanted with TBI in CR1 had a obstructive disease.Of the five patients who received Bu significantly improved survival over patients transplanted for whom pulmonary function tests are available, two have in CR1 with Bu, P ¼ 0.045 (Table 3). Only one of seven mild obstructive disease, one has moderate restrictive patients transplanted in third remission remains alive at 18 disease, one has both mild obstructive/restrictive disease, months. and one has normal pulmonary function.Cardiac function Cox’s proportional-hazards analysis was performed showed no change in 13 patients for whom follow-up is using variables defined on univariate analysis as significant available.Neuropsychologic evaluations are available for and variables known to modify EFS.A backward stepwise nine patients.For six patients who received TBI, two selection logic was employed to evaluate six variables: age, showed no change in follow-up evaluations, and four have gender, ethnicity, donor type, CR status, and conditioning. demonstrated decreased performance and/or full scale IQ. In this analysis, conditioning with Bu was significantly For two patients who received Bu, there was no change on associated with an increased risk of relapse with a three- neuropsychologic testing. fold increase in the risk of relapse- or transplant-related mortality (Table 4).This relationship was statistically significant at P ¼ 0.025 despite the sample size requirements Discussion of logistic regression.9 Donor type approached but did not This prospective study was initiated because of concerns regarding the significant sequelae of TBI conditioning Kaplan-Meier survival estimates by conditioning and donor regimens in children.Morbidities of TBI are well estab- 1.00 lished, and include cataracts, hypothyroidism, poor growth, and neurocognitive effects, as well as a significantly 10–14 0.75 higher cancer risk. Other studies have supported the lack of effect upon growth with Bu compared to TBI.15–17 Bu reduced the risk of post-transplant growth impairment, 0.50 thyroid dysfunction, and cataracts.18 While other studies have shown encouraging results with Bu in pediatric ALL 0.25 patients, these were single arm studies limited to single institutions.5–7 Lenarsky reported 14 children who received Bu/cyclophosphamide; 8 received allogeneic marrow.Seven 012345 of these patients were disease free, although follow-up was 5 7 Years short. von Bueltzingsloewen et al used Bu and other agents, including cyclophosphamide or melphalan or Bu/MRD TBI/MRD etoposide, for 21 children less than 4 years with ALL. Bu/URD TBI/URD The overall LFS was 61%, and there were no transplant- Figure 1 Kaplan-Meier survival estimates by conditioning and donor. related deaths.7 Small adult studies have also utilized Bu for

Bone Marrow Transplantation Busulfan vs TBI conditioning for pediatric ALL N Bunin et al 546 ALL.In a study by Copelan et al19 39 adult patients with trial in children and adults with leukemia demonstrated no ALL were treated with a Bu/cyclophosphamide regimen.In statistically significant differences in outcome between Bu the 27 patients in CR1, LFS was 42% at 3 years.For more or TBI regimens, although VOD, hemorrhagic cystitis, and advanced disease, however, the LFS was only 13.5%. severe GVHD were more common in patients receiving Accrual was smaller than anticipated on this study due to Bu.1 This trial included 38 patients with ALL.All patients other institutional commitments within the Pediatric Blood received marrow from matched sibling donors.Bu was and Marrow Transplant Consortium.However, our study associated with more VOD and hemorrhagic cystitis. suggests that the outcome of pediatric patients with ALL Obstructive bronchiolitis was also higher in the Bu group, may be inferior with Bu compared to TBI.Despite the 26% vs 5% with TBI.The 7-year nonrelapse mortality was small number and variability of patients studied, our significantly higher in the Bu group, 21%, in contrast to randomized trial demonstrates a statistically significant 12% for TBI patients.This difference was more marked for difference between Bu and TBI conditioning in 3 year EFS. patients with advanced disease; patients with early disease, This statistically significant difference was seen in both as defined by first remission of acute leukemia or first Kaplan–Meier analysis and Cox’s models of event risk. chronic phase of CML, had no difference in transplant- Thus, despite the small number of patients in this trial, the related mortality between Bu and TBI.28 randomized design and statistically significant results argue Another retrospective matched-pair comparison of the for a difference in efficacy between TBI- and Bu-based two regimens from the European Transplant Group for conditioning regimens for pediatric relapsed ALL.While Blood and Marrow Transplant demonstrated no differ- the number of patients in each conditioning arm was small, ences between regimens.29 There were 246 ALL patients in patients younger than six and those receiving unrelated this study.A recent analysis of four studies of patients with donor BMT fared the worst with Bu conditioning. AML or CML who received either Bu or TBI has However, Bu did not appear to increase the risk of relapse demonstrated no differences in survival or late complica- and may therefore be efficacious for some children.Of tions with at least 7 years of follow-up.30 No differences in note, successful retransplantation may be possible in some late pulmonary complications were noted between the two patients who received Bu, and three patients in this study regimens.These findings are in contrast to those of the were salvaged. Nordic study, which showed an increased risk of chronic Etoposide has been successfully combined with TBI and GVHD, obstructive lung disease, and alopecia.28 None of cyclophosphamide to intensify the preparative regimen and these European studies have shown the increased risk of decrease relapse.20 Etoposide with cyclophosphamide and interstitial pneumonitis found in the recent IBMTR Bu has been used in children with leukemias7,21,22 and this analysis by Davies et al.26 It is possible that preceding combination was the subject of a pilot feasibility study by chemotherapy, including methotrexate, may contribute to the Pediatric Blood and Marrow Transplant Consortium.23 this event in pediatric patients with ALL, but this is In the studies by Emminger, von Bueltzingsloewen, and the unclear. PBMTC, no unexpected toxicities were noted with the Bu pharmacokinetics were not routinely evaluated addition of etoposide, and the decision was made to include during our study.The effects of Bu, both in terms of this drug to decrease relapses potentially.However, the engraftment and toxicity, may be dependent upon phar- addition of etoposide has not been demonstrated to offer macokinetics, particularly in children.31,32 The pharmaco- an advantage when added to Bu/cyclophosphamide in kinetics of Bu are age-dependent; total plasma clearance patients with leukemias in a prospective study.Studies in rate is two to four times higher in children than in adults with leukemias and lymphomas have noted an adults.33,34 This may explain the lower incidence of Bu- increase in hepatotoxicity, including veno-occlusive disease related toxicities, such as mucositis and hepatic toxicity, (VOD), dermatologic toxicity, and mucositis,22–25 without generally seen in young children compared to adults.35 an increase in treatment-related mortality. Larger doses, or dosing based upon body surface area, have A recent retrospective analysis by the IBMTR comparing been proposed for children in an attempt to overcome the Bu to TBI for children with ALL who received a matched problems encountered with lower plasma levels.36 However, sibling donor transplant demonstrated inferior results for these higher doses may increase the incidence and mortality Bu.26 The Bu group included more patients less than 5 years of VOD, or result in neurotoxicity.Higher dosing based of age, and more patients in the TBI group received drugs upon body surface area has failed to overcome the problem in addition to cyclophosphamide, and T-cell-depleted of wide interindividual variation.34 Targeting levels follow- grafts.Bu levels were not obtained routinely.Similar to ing the first Bu dose may alleviate morbidity from under or our prospective study, the incidence of relapse was similar over dosing, but intraindividual variation is also com- in each group, suggesting that Bu may not be inferior to mon.37 Decreased concentration of Bu in young patients TBI in preventing relapse.However, non-relapse mortality, may have contributed to their inferior outcome in our including VOD and interstitial pneumonitis, was higher in study, and most of these young patients also received the Bu group, and LFS was superior in the TBI group (50 transplants from unrelated donors. vs 35%, P ¼ 0.005). In a recent study in children with AML, there was no The small number of patients in our study limits a significant correlation between relapse risk and Bu levels.38 conclusive analysis of the differences in nonrelapse The appropriate range of steady-state concentration of Bu mortality.However, other studies have noted increased may be different in children with ALL, compared to those transplant related toxicity in recipients of Bu, specifically with acute myeloid leukemia or nonmalignant diseases. hemorrhagic cystitis and VOD.27 A randomized Nordic Children with ALL differ from those studied previously, in

Bone Marrow Transplantation Busulfan vs TBI conditioning for pediatric ALL N Bunin et al 547 that most of these patients have had extensive prior 11 Curtis RE, Rowlings PA, Deeg HJ et al.Solid cancers after exposure to hepatotoxic agents, including methotrexate bone marrow transplantation. N Engl J Med 1997; 336: and mercaptopurine.This prior exposure may significantly 897–904. impact upon toxicity with projected Bu plasma levels used 12 Phipps S, Dunavant M, Srivastava DK et al.Cognitive and for patients with other diseases.The prior exposure to academic functioning in survivors of pediatric bone marrow J Clin Oncol hepatotoxins may account for the increased transplant- transplantation. 2000; 18: 1004–11. 13 Sims S, Kazak A, Golomb V et al.Cognitive, behavioral, and related mortality seen in the IBMTR retrospective study. social outcome in survivors of childhood stem cell transplanta- Bu pharmacokinetics and targeting of levels may tion. J Pediatr Hematol/Oncol 2002; 24: 115–119. improve results by decreasing transplant-related mortality, 14 Huma Z, Boulad F, Black P et al.Growth in children after and should be performed in children who receive this agent bone marrow transplantation for acute leukemia. Blood 1995; for conditioning.Prospective, randomized trials are valu- 86: 819–824. able in defining risk factors, and should be encouraged to 15 Giorgiani G, Bozzola M, Locatelli F et al.Role of busulfan evaluate both changes in conditioning regimens and and total body irradiation on growth of prepubertal children supportive care for pediatric patients.Significant concerns receiving bone marrow transplantation and results of treat- regarding late effects, particularly secondary cancers, ment with recombinant human growth hormone. Blood 1995; continue to make conditioning without radiation a 86: 825–831. 16 Afify Z, Shaw PJ, Clavano-Harding A, Cowell CT.Growth potential attractive option, but additional studies are and endocrine function in children with acute myeloid required to develop a safe, effective regimen. leukaemia after bone marrow transplantation using busulfan/ cyclophosphamide. Bone Marrow Transplant 2000; 25: 1087–1092. Acknowledgements 17 Shankar S, Bunin N.Growth in children undergoing bone marrow transplantation using busulfan and cyclophospha- This work was supported by Bristol-Myers Squibb. mide. J Pediatr Hematol/Oncol 1996; 18: 362–366. 18 Michel G, Socie G, Gebhard F et al.Late effects of allogeneic bone marrow transplantation for children with acute myelo- References blastic leukemia in first complete remission: The impact of conditioning regimen without total-body irradiation–a report 1 Ringden O, Ruutu T, Remberger M et al.A randomized trial from the Societe Francaise de Greffe de Moelle. J Clin Oncol comparing busulfan with total body irradiation as condition- 1997; 15: 2238–2246. ing in allogeneic marrow transplantation recipients with 19 Copelan EA, Biggs JC, Avalos BR et al.Radiation-free leukemia: a report from the Nordic Bone Marrow Transplant- preparation for allogeneic bone marrow transplantation in ation Group. Blood 1994; 83: 2723–2730. adults with acute lymphoblastic leukemia. J Clin Oncol 1992; 2 Clift R, Buckner C, Thomas E et al.Marrow transplantation 10: 237–242. for chronic myeloid leukemia: a randomized study comparing 20 Brown RA, Wolf SN, Fay JW et al.High dose etoposide, cyclophosphamide and total body irradiation with busulfan cyclophosphamide, and total body irradiation with allogeneic and cyclophosphamide. Blood 1994; 84: 2036–2043. bone marrow transplantation for patients with acute myeloid 3 Michel G, Gluckman E, Esperou-Bourdeau H et al.Allogeneic leukemia in untreated first relapse: a study by the North bone marrow transplantation for children with acute myelo- American Marrow Transplant Group. Blood 1995; 85: blastic leukemia in first complete remission: impact of 1391–1395. conditioning regimen without total-body irradiation – a report 21 Horstmann M, Kroschke G, Stockschlader M et al.Early from the Societe Francaise de Greffe de Moelle. J Clin Oncol toxicity of intensified conditioning with etoposide combined 1994; 12: 1217–1222. with total body irradiation/cyclophosphamide or busulfan/ 4 Emminger W, Emminger-Schmidmeier W, Haas OA et al. cyclophosphamide in children undergoing autologous or Treatment of infant leukemia with busulfan, cyclophospha- allogeneic bone marrow transplantation. Pediatr Hematol/ mide +/À etoposide and bone marrow transplantation. Bone Oncol 1996; 13: 45–53. Marrow Transplant 1992; 9: 313–318. 22 Zander AR, Berger C, Kroger N et al.High dose chemo- 5 Lenarsky C, Weinberg K, Kohn D et al.Bone marrow therapy with busulfan, cyclophosphamide, and etoposide as transplantation for children with acute lymphoblastic leukemia conditioning regimen for allogeneic bone marrow transplanta- without total body irradiation. Blood 1994; 84 abst: 549a. tion for patients with acute myeloid leukemia in first complete 6 Matsuyama T, Kojima S, Kato K.Allogeneic bone marrow remission. Clin Cancer Res 1997; 3: 2671–2675. transplantation for childhood leukemia following a busulfan 23 Sandler ES, Hagg R, Coppes MJ et al.Hematopoietic Stem and melphalan preparative regimen. Bone Marrow Transplant Cell Transplantation (HSCT) with a conditioning regimen of 1998; 22: 21–26. busulfan, cyclophosphamide, and etoposide for children with 7 von Bueltzingsloewen A, Esperou-Bourdeau H, Souillet G et al. acute myelogenous leukemia AML): a phase I study of the Allogeneic bone marrow transplantation following a busulfan- Pediatric Blood and Marrow Transplant Consortium. Med based conditioning regimen in young children with acute Pediatr Oncol 2000; 35: 403–409. lymphoblastic leukemia: a Cooperative Study of the Societe 24 Crilley P, Topolsky D, Styler MJ et al.Extramedullary toxicity Francaise de Greffe de Moelle. Bone Marrow Transplant 1995; of a conditioning regimen containing busulphan, cyclopho- 16: 521–527. sphamide and etoposide in 84 patients undergoing autologous 8 Deeg H, Storb R.Graft-versus-host disease: pathophysiologi- and allogeneic bone marrow transplantation. Bone Marrow cal and clinical aspects. Ann Rev Med 1984; 3: 11–24. Transplant 1995; 15: 361–365. 9 Katz MH. Multivariable Analysis: A Practical Guide for 25 Copelan EA, Penza SL, Pohlman B et al. Clinicians.University of Cambridge, Cambridge, 1999 to 1992. following busulfan, etoposide and cyclophosphamide in 10 Bhatia S, Louie AD, Bhatia R et al.Solid cancers after bone patients with non-Hodgkin’s lymphoma. Bone Marrow Trans- marrow transplantation. J Clin Oncol 2001; 19: 464–471. plant 2000; 25: 1243–1248.

Bone Marrow Transplantation Busulfan vs TBI conditioning for pediatric ALL N Bunin et al 548 26 Davies SM, Ramsay NKC, Klein JP et al.Comparison of phan and cyclophosphamide conditioning. Bone Marrow preparative regimens in transplants for children with acute Transplant 1994; 13: 583–587. lymphoblastic leukemia. J Clin Oncol 2000; 18: 340–347. 32 Slattery JT, Sanders JE, Buckner CD et al.Graft-rejection and 27 Rozman C, Carreras E, Qian C et al.Risk factors for hepatic toxicity following bone marrow transplantation in relation to veno-occlusive disease following HLA-identical sibling bone busulfan pharmacokinetics. Bone Marrow Transplant 1995; 16: marrow transplantation for leukemia. Bone Marrow Trans- 31–42. plant 1996; 17: 75–80. 33 Grochow L, Krivit W, Blazar B.Busulfan disposition in 28 Ringden O, Remberger M, Ruutu T et al.Increased risk of children. Blood 1990; 75: 1723–1727. chronic graft-versus-host disease, obstructive bronchiolitis, 34 Vassal G, Gouyette A, Hartmann O et al.Pharmacokinetics of and alopecia with busulfan versus total body irradiation: long high-dose busulfan in children. Cancer Chemother Pharmacol term results of a randomized trial in allogeneic marrow 1989; 24: 386–390. recipients with leukemia. Blood 1999; 93: 2196–2201. 35 Hassan M, Ljungman P, Bolme P et al.Busulfan bioavail- 29 Ringden O, Labopin M, Tura S et al.A comparison of ability. Blood 1994; 84: 2144–2150. busulfan versus total body irradiation combined with cyclo- 36 Shaw PJ, Hugh-Jones K, Hobbs JR et al.Busulphan and phosphamide as conditioning for autograft or allograft bone cyclophosphamide cause little early toxicity during displace- marrow transplantation in patients with acute leukemia. Br J ment bone marrow transplantation in fifty children. Bone Haematol 1996; 93: 637–645. Marrow Transplant 1986; 1: 193. 30 Socie G, Clift RA, Blaise D et al.Busulfan plus cyclo- 37 Schuler U, Schroer S, Kuhnle A et al.Busulfan pharmaco- phosphamide compared with total-body irradiation plus kinetics in bone marrow transplant patients: is drug cyclophosphamide before marrow transplantation for myeloid monitoring warranted? Bone Marrow Transplant 1994; 14: leukemia: long-term follow-up of 4 randomized studies. Blood 759–765. 2001; 98: 3569–3574. 38 Baker K, Bostrom B, DeFor T et al.Busulfan pharmaco- 31 Mehta J, Powles RL, Mitchell P et al.Graft failure after bone kinetics do not predict relapse in acute myeloid leukemia. Bone marrow transplantation from unrelated donors using busulf- Marrow Transplant 2000; 26: 607–614.

Bone Marrow Transplantation