Graft-Versus-Leukemia

A potential graft-versus-leukemia effect after allogeneic hematopoietic stem

cell transplantation for patients with Philadelphia chromosome-positive

acute lymphoblastic leukemia: results from the French Bone Marrow

Transplantation Society

H Esperou1, J-M Boiron2, J-M Cayuela1, O Blanchet3, M Kuentz4, J-P Jouet5, N Milpied6, J-Y Cahn7,C Faucher8, J-H Bourhis9, M Michallet10, M-L Tanguy14, J-P Vernant11, J Gabert12, P Bordigoni13,N Ifrah3, A Baruchel1 and H Dombret1, for the Societe Francaise de Greffe de Moelle – TherapieCellulaire (SFGM-TC)14

1Department of Hematology, Hopital Saint-Louis, Paris, France; 2Department of Hematology, Hopital Haut-Leveque, Pessac,France; 3Department of Hematology, Hopital Angers, France; 4Department of Hematology, Hopital Henri Mondor, Creteil, France;5Department of Hematology, Hopital Calmette-Guerin, Lille, France; 6Department of Hematology, Hotel Dieu, Nantes, France;7Department of Hematology, Hopital Jean Minjoz, Besancon, France; 8Department of Hematology, Institut Paoli-Calmettes,Marseille, France; 9Department of Hematology, Institut Gustave Roussy, Villejuif, France; 10Department of Hematology, HopitalEdouard Herriot, Lyon, France; 11Department of Hematology, Hopital La Pitie Salpetriere, Paris, France; 12Department ofBiochemistry and Molecular Biology, Hopital Nord, Marseille, France; 13Department of Hematology, Hopital d’ Enfants,Vandoeuvre-les-Nancy, France; and 14Societe Franc,aise de Greffe de Moelle – Therapie Cellulaire

Summary:

Philadelphia chromosome (Ph) acute lymphoblastic leu-kemia-positive (ALL) is a subgroup of ALL with a poorprognosis. We sought to evaluate the results of allogeneichematopoietic stem cell transplantation (HSCT) in thissituation. From 1992 to 2000, 121 patients with Ph-positive ALL enrolled in one of the three main Frenchprospective ALL chemotherapy trials and receivingallogeneic HSCT were reported to the registry of theFrench Society of Bone Marrow Transplantation(SFGM-TC). The median age was 35 years (range, 1–53). In all, 76 patients received HSCT in first completeremission and 45 in a more advanced disease stage.Minimal residual disease was evaluated just before thegraft: 35 patients of the 52 patients in first remission hadpersistent BCR-ABL transcript detectable while 17 had nodetectable minimal residual disease. Bone marrow and/orperipheral blood samples from 94 patients were submittedfor reverse transcriptase polymerase chain reactionanalysis at variable points after transplantation. Esti-mated 2-year survival and relapse rate for patients in CR1were 50 and 37%, respectively, significantly better thanfor patients with more advanced disease (P¼ 0.0001 and0.01, respectively). There was no difference in survival orin relapse rates in terms of the donor used. Relapse wasthe most common cause of treatment failure. Hematolo-gical status at the time of transplant and the occurrence ofacute graft-versus-host disease (GvHD) were the only two

prognostic factors identified for relapse (P¼ 0.02 and0.02, respectively). Detection of BCR-ABL transcriptsafter transplantation had a predictive value on relapseoccurrence. Finally, donor lymphocyte infusions weresuccessfully used to treat some relapses. The graft-versus-leukemia effect of HSCT in Ph-positive ALLappears to be supported by (1) the lack of prognosticsignificance of pretransplant BCR-ABL transcript detec-tion, (2) the efficacy of donor lymphocyte infusions incases of relapse, and (3) the role of GvHD as protectingagainst relapse.Bone Marrow Transplantation (2003) 31, 909–918.doi:10.1038/sj.bmt.1703951Keywords: acute lymphoblastic leukemia; Philadelphiachromosome; hematopoietic stem cell transplantation;minimal residual disease; graft-versus-leukemia effect

Introduction

The Philadephia chromosome (Ph) is the hallmark ofchronic myeloid leukemia (CML). It is also found in 5% ofchildren and 20–30% of adults with acute lymphoblasticleukemia (ALL).1 Ph results from the balanced t(9;22)chromosomal translocation, which gives rise to an abnor-mal fusion protein BCR-ABL. This BCR-ABL proto-oncogene is detected in two variant forms according tothe breakpoint cluster region (bcr): p210 for the majorbreakpoint cluster region (M-bcr) or p190 for the minorbreakpoint cluster region (m-bcr). The p210 protein isusually found in CML patients and in 25–50% of adultswith de novo Ph-positive ALL, whereas the p190 protein isfound in 90% of childhood Ph-positive ALL.2,3 Both p210Received 13 August 2002; accepted 19 September 2002

Correspondence: H Esperou, Departement d’ Hematologie, HopitalSaint-Louis, 1 avenue Claude Vellefaux, 75010 Paris, France

Bone Marrow Transplantation (2003) 31, 909–918& 2003 Nature Publishing Group All rights reserved 0268-3369/03 $25.00

www.nature.com/bmt

and p190 transcripts can be detected by specific BCR-ABLreverse transcriptase polymerase chain reaction (RT-PCR)assays.4 The prognostic significance of the bcr subtype hasnot been well established to date. A recent report suggested,however, that the expression of the p190 transcript wasassociated with a significant increase in the risk of relapseafter bone marrow transplantation as compared to theexpression of the p210 variant.5

Ph-positive ALL is a subgroup of ALL with a poorprognosis in both children and adults.6–8 After completeremission (CR) achievement, CR duration is short inpatients treated with conventional intensive chemotherapy,at least in adults. However, it has recently been reportedthat the disease can be controlled by chemotherapy alonein a subset of children with a low leukocyte count atdiagnosis.9 To date, allogeneic hematopoietic stem celltransplantation (HSCT) has been considered to be the besttreatment option in Ph-positive ALL patients in first CR.10

In a smaller series, a 2-year relapse-free survival of about40% is usually observed after allogeneic HSCT either withrelated or unrelated identical donors.11–16 However, relapseremains the main cause of failure with reported post-transplant relapse rates of between 40 and 60%.9,11,12,15

The detection of residual BCR-ABL fusion transcriptsduring the post-transplant period may be useful to diagnoserecurrent disease early, allowing early therapeutic inter-ventions, such as donor lymphocyte infusions or alphainterferon.5,17,18 The detection of residual BCR-ABL fusiontranscripts before transplantation might also provide usefulinformation to stratify patient prognosis.

The aim of the present retrospective study including 121patients, all prospectively included in French multicenteradult and childhood Ph-positive ALL trials between 1992and 2000, was to evaluate the results of allogeneic HSCT inthese patients in terms of outcome and molecular follow-up, and to study the prognostic value of pretransplantcharacteristics, such as BCR-ABL status.

Patients and methods

Patient characteristics

From 1992 to 2000, 150 consecutive allogeneic HSCT forpatients with Ph-positive ALL were reported to the registryof the French Society of Bone Marrow Transplantation(SFGM-TC). In order to minimize biases usually associatedwith registry studies, we chose to analyze only those 121patients previously enrolled in one of the three main Frenchprospective ALL trials used during this period: FRALLE-93 for children,19 GOELAL-1/2,20 and LALA-9421 foradults. Patient characteristics are summarized in Table 1.There were 63 males and 58 females who had receivedHSCT in 27 centers. Two groups were analyzed: 102 adultswith a median age of 38 years (range, 18–53) and 19children with a median age of 8 years (range, 1–18). Medianwhite blood cell count (WBC) was 19� 109/l (range, 1–500)for the entire population, 36� 109/l (range, 1–382) inchildren, and 18� 109/l (range, 1–500) in adults. Allpatients had B-lineage ALL. Among the 102 adults,61 were in first complete remission (CR1) and 41 hadadvanced disease (six in CR2, one in CR3, and 34 in

refractory disease), while 15 children were in CR1, three inCR2, and one had refractory disease. The median timebetween diagnosis and transplantation was 133 days (range,78–303) for the 76 patients grafted in CR1 and 132 days(range, 53–580) for the 35 patients grafted with refractorydisease. All the 10 patients transplanted in CR2 or CR3were early relapsing with a median time between diagnosisand transplantation of 224 days (range, 115–1803). Inpatients transplanted in CR1, the median time to reach CRwas 1.4 months (range, 21 days to 4.5 months).

In all, 74 adults and one child were treated on the LALA-94 trial and received the chemotherapy as scheduled.21 Theinduction course was administered over a 4-week periodand consisted of prednisone, vincristine, cyclophospha-mide, and daunorubicin or idarubicin according to initialrandomization. At day 35, all patients were eligible for asecond course of consolidation (or salvage), whatever theresponse to induction. This course consisted of mitoxan-trone and intermediate-dose cytarabine. Central nervoussystem prophylaxis consisted of five intrathecal methotrex-ate, cytarabine, and steroid injections. Transplantationwith sibling or identified matched unrelated donor stemcells was planned to be performed in CR1 3 months after

Table 1 Patients characteristics

Children(N=19)

Adults(N=102)

All(N=121)

Sex ratioMale/female 10/9 53/49 63/58Median age (range) 8 (1–18) 38 (18–53) 35 (1–53)

Median leukocyte count 36 (1–382) 18 (1–500) 19 (1–500)

Chemotherapy trialFRALLE-93 17 3 20GOELAL-1/2 1 25 26LALA-94 1 74 75

Hematological statusat HSCT and donor type

CR1 15 61 76Sibling 10 48 58Twin 1 1 1Unrelated 3 12 15Mismatched 2 2 2

CR2 3 6 9Sibling 1 4 5Unrelated 2 2 4

CR3 1 1Unrelated 1 1

Refractory disease 1 34 35Sibling 1 23 24Unrelated 11 11 11

BCR subtype at diagnosisM-bcr 2 19 21m-bcr 12 44 56Both split 0 2 2Not available 5 37 42

HSCT: hematopoietic stem cell transplantation; CR1: first completeremission; CR2: second complete remission; M-bcr: major breakpointcluster region; m-bcr: minor breakpoint cluster region.

Allogeneic HSCT in Ph-positve ALLH Esperou et al

910

Bone Marrow Transplantation

enrolment. One or two additional cycles of chemotherapyusing intermediate-dose methotrexate and l-asparaginasewere administered during the pretransplant period. Somepatients (N¼ 22) failed after induction and salvageachieved a CR after various third-line chemotherapyschedules. All patients grafted with advanced disease hadfailed at least three lines of intensive chemotherapy or hadvery early relapse.

In total, 25 adults and one child were treated accordingto the GOELAL-1/2 trial.20 The induction course wasadministered in two phases. The first 4-week phaseconsisted of prednisone, vincristine, cyclophosphamide,and idarubicin, and was followed by a second phase withcyclophosphamide, cytarabine and 6-mercaptopurine.Transplantation was planned after one or two additionalcourses of high-dose methotrexate in CR1 patients andafter one additional course of cyclophosphamide, cytar-abine, and prednisone in patients who had failed this two-phase induction. Central nervous system prophylaxisconsisted of six intrathecal methotrexate and steroidinjections. All patients grafted with advanced disease hadfailed at least three lines of intensive chemotherapy or werein very early relapse.

In total, 17 children and three young adults were treatedaccording to the pediatric FRALLE-93 trial.19 The in-duction course consisted of daunorubicin, vincristine,l-asparaginase, and prednisone administered within a4-week period. Induction was followed by three cycles ofintensive consolidation: (1) two ‘R2’ cycles with intermedi-ate-dose cytarabine, etoposide, and dexamethasone; and (2)one ‘COPADEM’ cycle with vincristine, high-dose metho-trexate, cyclophosphamide, adriamycin, and prednisone.Children received six intrathecal methotrexate, cytarabine,and steroid injections. Transplantation was planned afterthe third cycle of intensive consolidation in children inCR1. One patient who had failed three induction courseswas transplanted with refractory disease.

Diagnostic criteria for Ph-positive ALL

The diagnosis of Ph-positive ALL was based on standardcytogenetic and/or on RT-PCR at diagnosis. Chromosomeanalysis was performed using short unstimulated cultureson bone marrow and/or peripheral blood samples.A minimum of 20 mitoses were fully analyzed, standardcriteria to define a clone were applied, and chromosomalabnormalities were classified according to the InternationalSystem for Human Cytogenetic Nomenclature.22 RT-PCRspecific for BCR-ABL rearrangements was performed onbone marrow and/or peripheral blood samples. The RT-PCR reaction was carried out following a commonprotocol using already published BCR-ABL primers,2,23

as described.24 More recently, the standardized BIOMED-1primers were used.25 Patients for whom the samples had noamplification of control gene expression were excludedfrom the molecular part of analysis but remained in theclinical analysis. Sensitivity of the reaction was assessedduring each experiment using a dilution series of Tom1 (m-BCR) or K562 (M-BCR) RNA. It ranged from 10�4 to10�5 with a median of 10�5. Immunophenotyping of ALLcells was systematically performed on bone marrow and/or

peripheral samples using usual monoclonal antibodies. Asscheduled in the three chemotherapy trials, all cytogenetic,molecular, and immunophenotyping data were prospec-tively and centrally reviewed by three distinct workingcommittees. At diagnosis, 46 patients (two children and 44adults) had M-bcr ALL, 31 (12 children and 19 adults) hadm-bcr ALL, and two adults had mM-bcr ALL. The bcrsubtype was missing in 41 patients and testing was negativein only one case.

Choice of donor

An indication of allogeneic HSCT was scheduled in thethree chemotherapy trials whenever a diagnosis of Ph-positive and/or BCR-ABL-positive ALL was made. If ageno-identical donor was not available, a search for anunrelated donor on National and International bonemarrow donors registries was initiated. The unrelatedallograft was performed as soon as an HLA identicaldonor was identified. The selection criterion was HLAtyping and this varied with time: between 1992 and 1994identity of A and B class I antigens were defined byserology and class II antigens were defined using restrictionfragment length polymorphism. No disparity was allowedin choosing the donor. After 1994, selection criteria weredefined by the identity for class I antigens using PCR-sequence-specific probes (PCR-SSP) and for class IIantigens using restriction fragment length polymorphism.Since this time, only one disparity out of 10 antigens tested(HLA-A, -B, -C, -DR, -DQ) was allowed in selecting thepotential donor. The time of such a search did not exceed 3months after CR and patients without a suitable donorafter this time received alternative treatment of chemo-therapy alone or a scheduled autologous transplant. Themedian time between CR1 and transplantation was 2.6months (range, 0.7–5.5 months) for the whole population,2.3 months (range, 0.7–5 months) for the transplants with asibling donor, and 3.7 months (range, 1.2–5.5 months) forthe grafts with an unrelated donor.

Minimal residual disease analysis before transplantation

Pretransplant RT-PCR was performed in 85/121 (70%)patients, including 52 CR1 and six CR2 in whom RT-PCRanalysis was performed on bone marrow samples. In all,35 of the 52 CR1 patients had persistent BCR-ABLtranscripts, while 17 had no detectable minimal residualdisease. Two of the six CR2 patients had no detectabletranscripts at the time of transplantation.

Transplantation modalities

The conditioning regimen varied between adults andchildren. The majority of children (12 out of 19) receiveda ‘TAMe’, including fractionated total body irradiation(FTBI) (12 grays in six fractions), cytarabine 12 or 18 g/m2,and melphalan 140mg/m2 as previously described.26 Twochildren under 4 years of age received conditioning withoutFTBI. In total, 48 adults received 12 grays in six fractionsFTBI with cyclophosphamide (CY) 120mg/kg given over 2days and etoposide at 50mg/kg in one dose. Nine young

Allogeneic HSCT in Ph-positve ALLH Esperou et al

911

Bone Marrow Transplantation

adults received ‘TAMe’ conditioning, and four receivedconditioning without FTBI (of whom two had a non-myeloablative conditioning regimen). Other adults receiveda combination of the same dose of FTBI and CY with orwithout thiotepa at 10mg/kg in one dose.

In all, 11 children received HSCT from a sibling (eightbone marrow cells, two cord blood cells, and one peripheralblood stem cell), two received family mismatched bonemarrow and peripheral blood stem cells, and six received anunrelated graft (five bone marrow cells and one cord bloodcells). In all, 76 adults received HSC from a sibling (62 withmarrow cells and 14 peripheral blood stem cells), and 26from an unrelated donor (with one cord blood cells). Themedian number of nucleated cells transfused was 2.8� 108/kg (range, 0.1–15.3).

Only two children and six adults were given a T-cell-depleted HSC and the other patients received the usualgraft-versus-host disease (GvHD) prophylaxis with cyclo-sporin A given from day �1 for at least 6 months post-transplant and three or four doses of methotrexate 15mg/m2 on day 1 and 10mg/m2 on days 3, 6711. Engraftmentwas defined as the first of three consecutive days when theabsolute neutrophil count (ANC) exceeded 0.5� 109/l.Granulocyte colony-stimulating factor (G-CSF) was givenonly in cases of delayed engraftment. Acute GvHD wasgraded using Gluckberg’s criteria,27 with modificationsintroduced in 1996.28 Patients were considered at risk forchronic GvHD if they were alive with engraftment beyondday 100 post-transplant and were classified as havinglimited or extensive chronic GvHD.29 For comparison,patients with grade 2 or higher acute GvHD werecompared to those with ograde 2 or no acute GvHD,while patients with limited or extensive chronic GvHD werecompared to those without chronic GvHD. All patientswere transplanted in high-efficiency particle air filtrationrooms. CMV-negative pretransplant patients receivedCMV seronegative blood products. All patients withCMV reactivation had received pre-emptive ganciclovirtreatment. Empiric and curative antibacterial and anti-fungal treatment differed from one center to another butwas based on an intravenous combination of third-generation beta-lactamines with aminoside and/or glyco-peptide and amphotericin B. Cotrimaxazole was given toprevent Pneumocystis pneumonia for the first 6 months,and then lifelong penicillin was started. Chimerism studiesconsisted of in situ hybridization of the Y chromosomeamong patients whose donor was of the opposite sex, ormolecular study of the variable number of tandem repeatpolymorphism.

Minimal residual disease analysis after transplantation

No antileukemia treatment was given systematically aftertransplantation. Minimal residual disease analysis was notuniformly scheduled and left to the discretion of eachcenter. Bone marrow and/or peripheral blood samples from94 patients were submitted to RT-PCR analysis at variouspoints before the graft and between 1 month and five and ahalf years after transplantation, with a mean of three pointsper patient. Nine patients tested during follow-up were nottested before transplantation. Therapeutic approach with

regard to molecular follow-up was not scheduled. In thesame way, treatment of relapse including chemotherapy(combination of vincristine, prednisolone, and variousother drugs), Interferon-a and donor lymphocyte infusion(DLI) was given according to different protocols withoutstandardized recommendation.

Statistical analysis

Follow-up has been updated as of 31 December, 2000. TheFisher exact test was used for binary variable comparison.Duration of CR was calculated from the date of the graftuntil the date of first hematological relapse. Data ontreatment failure were estimated by the Kaplan–Meiermethod30 and compared using the log-rank test.31 Inmultivariate analyses, outcome comparisons were adjustedwith the Cox model32 and tested by the likelihood-ratio test.In univariate as well as in multivariate analysis, acute andchronic GvHD significance were considered as time-depen-dent variables and their impacts on outcome were evaluatedusing multiple-record survival data. A P-value of less than0.05 was considered as reaching statistical significance. Allcalculations were performed using the STATA software,version 7.0 (Stata Corporation, TX, USA).

RESULTS

Overall survival

The median follow-up for the 43 living patients was 29months (range, 1–107 months). At 2 years, the estimatedoverall survival was 37% (95% confidence interval, 28–46%). Prognostic factors for survival are shown in Table 2.Patients receiving HSCT in CR1 had a 2-year survival of50% (95% confidence interval, 37–61%) as compared to17% (95% confidence interval, 8–29%) in patients graftedwith more advanced disease (Po0.0001) (Figure 2a).

CR1 no CR1 n = 38 n = 25

7 1

3 1DLI

1 1

10

2 1

5 4

2

1 7

HSCT

DLI

17

Figure 1 Profiles of minimal residual disease analysis. J, negative PCR;�, positive PCR; ’, diagnosis of relapse; DLI, donor lymphocyte infusion;CR1, first complete remission, CR2, second complete remission. Minimalresidual follow-up was missing for 24 patients on 87 with BCR-ABLtranscript research before the HSCT.

Allogeneic HSCT in Ph-positve ALLH Esperou et al

912

Bone Marrow Transplantation

Estimated survival of patients in CR1 was similar for the 35patients with BCR-ABL positive signals compared to the17 with negative BCR-ABL signal (P¼ 0.69) (Figure 3a).Children had a 2-year survival of 58% (95% confidenceinterval, 33–76%) as compared to 33% (95% confidenceinterval, 23–42%) in adults (P¼ 0.04) (Figure 4a). This wasassociated with a longer survival in patients enrolled in thepediatric FRALLE-93 study as compared to adultsenrolled in both LALA-94 and GOELAL-1/2 studies(P¼ 0.05). Type of donor (sibling or unrelated donor)and type of transcript (M-bcr or m-bcr) had no significantinfluence on survival. The occurrence of acute GvHD(grade 2 or more) or chronic GvHD did not significantlyinfluence survival (P¼ 0.60 and 0.85, respectively, whenboth variables were considered as time-dependent vari-ables).

On multivariable analysis (including hematologicalstatus at the time of HSCT, age considered as a continuousvariable, WBC considered as a continuous variable, andchemotherapy trial), the only independent factor associatedwith longer survival was hematological status (relative riskin patients grafted in conditions other than CR1, 2.6; 95%confidence interval, 1.6–4.2; Po0.0001) (Table 2).

Of note, the estimated 4-year survival for patients whoreceived an allograft in CR1 was 42% (95% confidence

interval, 29–54%) vs 5% (95% confidence interval, 0–18%)for patients with more advanced disease. Only one patientfrom the latter group became a long-term survivor.

Post-transplant morbidity and mortality

Children: All children had myeloid reconstitution exceptone who received an autologous rescue graft after failure ofa mismatched unrelated graft. The median time to recover0.5� 109/l ANC was 21 days (range, 14–43). Chimerismstudies performed in the first 3 months after HSCT showedfull engraftment of donor cells in all the eight patientstested. In all, 10 children experienced grade X2 acuteGvHD (six after sibling HSCT, three after unrelatedHSCT, and one after related mismatched graft). Of 13children alive at 3 months, four had limited chronic GvHDand one extensive chronic GvHD. Six of 19 children (32%)relapsed at a median of 3.5 months (range, 2–40). Eightchildren died: five from relapse, one from acute GvHD, onefrom viral, and one from fungal infection.

Adults: Of 102 patients, 90 had myeloid reconstitutionwith a median time to an ANC of 0.5� 109/l of 20 days(range, 9–59). A total of 11 patients had no hematologicalrecovery and died from infection or grade IV GvHD within

Table 2 Prognostic factors for survival

N patients(N CR1patients)

P-values (RR, 95% CI)

Univariateanalysesa

Multivariateanalysisb

All patients CR1patients

Fourvariable

Hematological statusCR1 vs no CR1 121 o0.0001 — o0.0001

2.6 (1.6–4.2)

AgeAs a continuous variable 121 (76) 0.05 0.12 0.17Less than 18 years 121 (76) 0.04 0.19

Chemotherapy trial 121 (76) 0.05 0.02 0.46

Leukocyte countAs a continuous variable 117 (76) 0.07 0.48 0.13Less than 25� 109/l 117 (76) 0.09 0.43

Conditioning with etoposide 119 (74) 0.62 0.45

Unrelated donor 121 (76) 0.82 0.4

bcr subtype 84 (52) 0.52 0.67

Molecular status 58 (52) — 0.69

Acute GvHDc 119 (74) 0.60 0.78

Chronic GvHDc 106 (69) 0.85 0.48

aUsing the Fisher’s exact test for binary variables and a univariate Coxmodel for continuous variables. bUsing the Cox model; RR: relative risk;CI: confidence interval; GvHD: graft-versus-host disease. cAcute andchronic GvHD were considered as time-dependent variables.

Cu

mu

lati

ve p

erce

nta

ge

0 1 2 30

25

50

75

100

years

At risk: CR1 76 44 27 17 CR2+ 45 15 7 1

Logrank P < 0.001

CR1 76 38 N O

Cu

mu

lati

ve p

erce

nta

ge

0

25

50

75

100

Logrank P = 0.01

CR1 76 23 CR2+ 45 18

N O

0 1 2 3 years

At risk: CR1 76 37 25 17 CR2+ 45 11 5 1

a

b

CR2+ 45 41

Figure 2 Outcome according to hematological status. (a) Kaplan–Meiersurvival estimates. (b) Kaplan–Meier relapse rate estimates.

Allogeneic HSCT in Ph-positve ALLH Esperou et al

913

Bone Marrow Transplantation

the first month after HSCT. One patient without engraftmentafter T-cell-depleted HSCT had myeloid recovery after asecond unmanipulated allograft. In total, 53 (52%) patientsexperienced grade X2 acute GvHD (40 after sibling HSCTand 13 after unrelated HSCT). Of the 76 patients alive at 3months, 33 (43%) had chronic GvHD (20 patients withlimited form and 13 patients with extensive form). In all, 35adults (34%) relapsed within a median of 7 months (range,1–43). In total, 70 patients died: 27 from relapse, 21 fromtoxicity, 18 from infection, and four from GvHD.

Minimal residual disease follow-up

Minimal residual disease was serially assessed before and/or after HSCT in 96 patients, at one to five time points perpatient. Of note, among these, 63 patients had molecularfollow-up before and after HSCT (38 in CR1 and 25 withmore advanced disease) (Figure 1). The results may besummarized as follows.

First, molecular status before HSCT was unknown innine patients in CR1. Six of these patients were continu-ously BCR-ABL negative after HSCT without hematolo-gical relapse. Three patients were continuously BCR-ABLpositive after HSCT and relapsed.

Secondly, 19 patients (17 in CR1 and two in CR2) wereBCR-ABL negative at the time of HSCT. Among the 14

patients with no hematological relapse (74%), eightremained BCR-ABL negative for up to 4 years and sixhad no molecular follow-up (five of them died in the first 3months after HSCT). Five patients (26%) experienced arelapse, all after transient negativity of RT-PCR, and BCR-ABL transcript reappearance at 3, 8,17, 20, and 43 months.Three patients were treated with chemotherapy withoutinterferon-a, but with DLI in two cases. The two othersdied before any treatment.

Thirdly, 68 patients were BCR-ABL positive at the timeof HSCT. In all, 36 were in CR1, 32 with more advanceddisease. Among the 27 patients in CR1 without hematolo-gical relapse (75% of CR1 patients), 19 were then BCR-ABL negative (including two who became BCR-ABLnegative late, at 2 and 6 months postHSCT) and eighthad no molecular follow-up (four of them died early).Among the nine patients who experienced a hematologicalrelapse, one was continuously BCR-ABL positive afterHSCT, seven were transiently RT-PCR negative before thediagnosis of relapse (at 3, 4, 7, 10, 11, 12, and 28 months,respectively), and the remaining one had no molecularfollow-up. Six patients were treated with chemotherapywithout interferon-a, but with DLI in four cases. Amongthe 19 patients with more advanced disease and withouthematological relapse (60% of non-CR1 patients), 11

Cu

mu

lati

ve p

erce

nta

ge

0

25

50

75

100

0 1 2 3 years

At risk:17 13 9 735 22 13 6

BCR-ABL negativeBCR-ABL positive

Cu

mu

lati

ve p

erce

nta

ge

0

25

50

75

100

0 1 2 3 years

At risk:17 12 8 735 19 12 6

BCR-ABL negativeBCR-ABL positive

BCR-ABL negativeBCR-ABL positive

Logrank P = 0.69

17 735 15

N O

17 435 9

N OBCR-ABL negativeBCR-ABL positive Logrank P = 0.50

a

b

Figure 3 Outcome according to BCR-ABL status for patients in CR1. (a)Kaplan–Meier survival estimates. (b) Kaplan–Meier relapse rate estimates.

Cu

mu

lati

ve p

erce

nta

ge

0 1 2 30

25

50

75

100

years At risk: <18 years 20 13 9 4 >18 years 101 46 25 14

Logrank P = 0.04

<18 years 20 9 >18 years 101 70

N O

Cu

mu

lati

ve p

erce

nta

ge

0 1 2 30

25

50

75

100

years

Logrank P = 0.59

<18 years 20 7

>18 years 101 34

N O

At risk: <18 years 20 11 8 >18 years 101 37 22

4 14

a

b

Figure 4 Outcome according to age. (a) Kaplan–Meier survival estimatesaccording to age. (b) Kaplan–Meier relapse rate estimates according to age.

Allogeneic HSCT in Ph-positve ALLH Esperou et al

914

Bone Marrow Transplantation

remained BCR-ABL negative (one became BCR-ABLnegative late, at 2 months), six had no molecular follow-up until transplant-related death, and two died frominfection with a positive BCR-ABL transcript. Among the13 patients with more advanced disease who experienceda relapse, five were continuously BCR-ABL positive, fourwere transiently RT-PCR negative before the diagnosis ofrelapse (at 3, 9, 14, and 16 months, respectively), and fourhad no molecular follow-up (three of them died early).Four patients died before any treatment. Nine patientsreceived chemotherapy with DLI in three cases; all of themdied from relapse.

In five patients (including three BCR-ABL positivepatients at the time of HSCT), BCR-ABL negativity wasobserved at 1, 3, 5, 7, and 10 months only after a post-HSCT BCR-ABL-positive period. In all other patients,BCR-ABL positivity was predictive of relapse.

Hematological relapse

At 2 years, the probability of relapse was 44% (95%confidence interval, 34–56%). Prognostic factors associatedwith relapse are shown in Table 3. The 2-year estimatedrelapse rate was 37% (95% confidence interval 25–51%) inpatients who received HSCT in CR1 as compared to 62%(95% confidence interval, 41–83%) in those with moreadvanced disease (P¼ 0.01) (Figure 2b). For patients in

CR1 the 2 years estimated relapse rate was similar whateverthe molecular status at the time of HSCT (Figure 3b). The2-year estimated relapse rate was similar in children andadults (Figure 4b). On univariate analysis, the occurrenceof acute GvHD was associated with a lower risk of relapse.At 2 years, the risk of relapse was 38% (95% confidenceinterval, 24–56%) in patients with grade II or higher acuteGvHD vs 53% (95% confidence interval, 39–69%) inpatients withograde II acute GvHD (P¼ 0.06, when acuteGvHD was considered as a time-dependent variable).Conversely, the occurrence of chronic GvHD had noinfluence on risk of relapse but the shortness of medianfollow-up does not allow assessment of the role of chronicGvHD, which may not yet have manifested.

A three-variable analysis, including hematological statusat the time of HSCT, WBC considered as a binary variable(more or less than 25� 109/l), and acute GvHD consideredas a time-dependent variable, was performed. The follow-ing two factors were identified as significantly associatedwith a lower risk of relapse: (1) CRl status (relative risk forpatients grafted in conditions other than CR1, 2.4; 95%confidence interval, 1.2–4.7; P¼ 0.02); (2) a grade XIIacute GvHD (relative risk for patients without acuteGvHD, 2.3; 95% confidence interval, 1.2–4.5; P¼ 0.02)(Table 3).

In patients tested for BCR-ABL status between 3 and 6months after HSCT, we found, however, an correlationbetween the occurrence of acute GvHD and molecularstatus at that time. Among the 51 BCR-ABL-negativepatients, 29 had experienced grade X2 acute GvHD;among the 13 BCR-ABL-positive patients, seven hadexperienced acute GvHD (P¼ 1.0, by the w2 test). Amongthe five patients with delayed development of BCR-ABLnegativity, two had experienced a grade X2 acute GvHDand three had not. Of note, there were more patients aliveafter 3 months and without BCR-ABL status assessmentafter HSCT who experienced acute GvHD 19/57 thanpatients with BCR-ABL status assessment after HSCT whoexperienced acute GvHD 13/43.

Donor lymphocyte infusion

Nine patients received one or more donor lymphocyteinfusions (Table 4). For one patient with molecular relapseat 3 years after HSCT this was the only treatment, and twoinfusions of 1� 107 CD3+ cells/kg led to a prolongedcomplete molecular remission. The other patients receivedDLI associated with chemotherapy and always for hema-tological relapse at a median of 10 months. Five patientsachieved a response; for two this was a prolonged completemolecular remission (more than 9 and 12 months), butthree died from relapse or fungal infection after only atransient response. Three patients had no response.

Discussion

Allogeneic HSCT is considered to be the best treatmentoption for patients with Ph-positive ALL in first CR. Thishas been recently confirmed in a French prospective studycomparing this approach with autologous HSCT in 103

Table 3 Factors predicting for relapse

N patients(N CR1patients)

P-values(RR, 95% CI)

Univariateanalysesa

Multivariateanalysisb

All patients CR1patients

Threevariablec

Hematological statusCR1 vs no CR1 121 0.01 — 0.02

2.4 (1.2–4.7)

AgeAs a continuous variable 121 (76) 0.59 0.39Less than 18 years 121 (76) 0.59 0.39

Chemotherapy trial 121 (76) 0.44 0.46

Leukocyte countAs a continuous variable 117 (76) 0.40 0.91Less than 25� 109/l 117 (76) 0.10 0.41 0.25

Conditioning with etoposide 119 (74) 0.88 0.94

Unrelated donor 121 (76) 0.58 0.31

bcr subtype 84 (52) 0.11 0.37

Molecular status 58 (52) — 0.50

Acute GvHDc 119 (74) 0.06 0.20 0.022.3 (1.2–4.5)

Chronic GvHDc 106 (69) 0.82 0.91

aSee Table 2 footnote.

Allogeneic HSCT in Ph-positve ALLH Esperou et al

915

Bone Marrow Transplantation

CR1 patients using the intention-to-treat principle.21 Thereported estimated 2-year survival of 50% in 76 patientswho received allogeneic HSCT in first CR is similar to thereports after allogeneic transplantation in smaller unicenterpopulations of patients with the same disease, as recentlyreviewed by Snyder.10 Survival rates observed in patientstreated with chemotherapy alone are usually much low-er.14,33–35 In a recent worldwide pediatric retrospectivestudy performed on 267 children stratified into threeprognostic subgroups according to age and white cellcount, a significant survival advantage was observed inchildren who received sibling HSCT (N¼ 38) as comparedto those treated with chemotherapy alone (N¼ 147) withineach prognostic subgroup.9

From using retrospective data including variable patientpopulations (children and adults, patients with differenthematological and molecular situations, geno- or pheno-identical donors) we can draw some conclusions.

First, hematological status at the time of HSCT (CR1)appears to be a favorable prognostic factor for overallsurvival and for leukemia-free survival. Transplantationcan thus be recommended at that time: achievement of asecond remission after relapse is rare33 and a worseoutcome after HSCT for patients in second CR or withmore advanced ALL has recently, been emphasized by thestudy of Snyder et al.36 In the present study, only one of the45 patients who received HSCT for advanced ALL wasalive at 4 years. Furthermore, the lack of difference in termsof survival between sibling and unrelated donor transplan-tation observed in our study leads to the recommendationof a search for a matched unrelated donor in CR1 patientswith no sibling donor. Recently, Cornelissen et al37

reported a 37713% DFS for patients with poor-riskALL (who were mainly Ph-positive) who received unrelatedmarrow transplantation in CR1.37 It may be assumed fromthis study that there is a stronger graft-versus-leukemia(GvL) effect after unrelated marrow transplantation, whichoffsets the potential higher transplant-related mortalityrisk. This enhanced GvL effect in Ph-positive ALL can beparalleled to the best known GvL effect after allogeneictransplantation in CML initially reported by Sierra et al.16 Itcan also be linked to the comparative efficacy of DLI in thesepatients with regard to the usual lack of DLI efficacy in ALL.

Secondly, with classical BCR-ABL RT-PCR, the post-allograft monitoring of minimal residual disease gave quiteinteresting results. Analysis of molecular follow-up profilesshows a good correlation between RT-PCR positivity and

relapse. A similar correlation was reported in severalstudies,5,7,18 while not observed in the studies from Snyder10

and Sierra et al,16 who both reported patients withpersistent BCR-ABL transcripts until 2 years after theirgraft who did not relapse. In our study, all patients withreappearance of the BCR-ABL transcript after initialpostgraft negativity relapsed and the only patients whoremained alive after relapse were those who received earlyDLI. Since efficiency is also an argument in favor of asignificant GvL effect of allogeneic transplantation in thissituation. We report a small proportion of patients whowere BCR-ABL positive the first months of follow-up andwho became spontaneously negative a few weeks later. Thisis also reported by Marks et al15 in a pediatric study.

Despite the prognostic influence of acute GvHD onrelapse on multivariate analysis and the high predictivevalue of post-HSCT BCR-ABL status on hematologicalrelapse, we observed no difference in GvHD incidencebetween patients with and without positive minimal resi-dual disease. This lack of correlation may be related to theintroduction of two biases. First, some patients hadhematological relapse before post-HSCT BCR-ABL statusassessment. Secondly, not all eligible patients (ie patients aliveafter 3 months) were tested for post-HSCT BCR-ABL status.The subtype of BCR-ABL transcript does not, in our studyrelate to risk of relapse as reported by Radich et al5 andSnyder et al,38 but not by others inducing Kroger et al.39

All these issues emphasize the place of allogeneic HSCTin the treatment of Philadelphia-positive ALL. The lack ofprognostic significance of the molecular BCR-ABL statusjust before the graft has to be confirmed or refuted byfurther evaluations eventually using quantitative real-timePCR in a larger number of patients. It may be an argumentin favor of a GvL effect after DLIs in patients withmolecular relapse or it may demonstrate the importance ofGvHD as a factor protecting against relapse. The largeproportion of patients who relapse after allogeneictransplantation suggests, however, that further evaluationsshould be undertaken using new antileukemia agents suchas Gleevect either in the pretransplant or posttransplantperiods.

We are indebted to Dr Gerard Socie, Dr Patricia Ribaud,

Dr Philippe Guardiola, Dr Agnes Devergie and Dr Eliane

Gluckman for critical comments and assistance and to Wim

van Putten for providing a Stata package with facilities for

Kaplan–Meier survival curves.

Table 4 Donor lymphocyte infusion characteristics

Name Relapse BCR-ABL Time from HSCT Associated treatment Response Duration of response

DEM hem neg 3m chemo No —MAD hem neg 8m chemo Yes 9m+DRO mol pos 36m — Yes 60m+URV hem pos 11m chemo Yes 3mFOU hem pos 12m chemo Yes 12m+BLO hem pos 7m chemo No —GAW hem pos 11m chemo Yes 3mSAI hem pos 5m chemo Yes 3mMOH hem pos 14m chemo No —

HSCT: hematopoietic stem cell transplantation; BCR-ABL: transcript before HSCT; hem: hematological relapse; mol: molecular relapse; m: month; chemo:chemotherapy.

Allogeneic HSCT in Ph-positve ALLH Esperou et al

916

Bone Marrow Transplantation

References

1 Tuszynski A, Dhut S, Young BD et al. Detection andsignificance of bcr-abl mRNA transcripts and fusion proteinsin Philadelphia-positive adult acute lymphoblastic leukemia.Leukemia 1993; 7: 1504–1508.

2 Faderl S, Kantarjian HM, Talpaz M, Estrov Z. Clinicalsignificance of cytogenetic abnormalities in adult acutelymphoblastic leukemia. Blood 1998; 91: 3995–4019.

3 Suryanarayan K, Hunger SP, Kohler S et al. Consistentinvolvement of the bcr gene by 9;22 breakpoints in pediatricacute leukemias. Blood 1991; 77: 324–330.

4 Maurer J, Janssen JWG, Thiel E et al. Detection of chimericBCR-ABL genes in acute lymphoblastic leukaemia by thepolymerase chain reaction. Lancet 1991; 337: 1055–1058.

5 Radich J, Gehly G, Lee A et al. Detection of bcr-abltranscripts in Philadelphia chromosome-positive acute lym-phoblastic leukemia after marrow transplantation. Blood 1997;89: 2602–2609.

6 Bloomfield CD, Wurster-Hill D, Peng G et al. Prognosticsignificance of the Philadelphia chromosome in adult acutelymphoblastic leukemia. In: Gale RP, Hoelzer D (eds). AcuteLymphoblastic Leukemia. Alan R Liss: New York 1990,pp 101–109.

7 Groupe Francais de Cytogenetique Hematologique. Cytoge-netic abnormalities in adult acute lymphoblastic leukemia:correlations with hematologic findings and outcome.A collaborative study of the Groupe Francais de Cytogene-tique Hematologique. Blood 1996; 87: 3135–3142.

8 Beyermann B, Adams HP, Henze G. Philadelphia chromo-some in relapsed childhood acute lymphoblastic leukemia:a matched-pair analysis. Berlin–Frankfurt–Munster StudyGroup. J Clin Oncol 1997; 15: 2231–2237.

9 Arico M, Valsecchi MG, Camitta B et al. Outcome oftreatment in children with Philadelphia chromosome-positiveacute lymphoblastic leukemia. N Engl J Med 2000; 342:998–1006.

10 Snyder DS. Allogeneic stem cell transplantation for Philadel-phia chromosome -positive acute lymphoblastic leukemia. BiolBlood Marrow Transplantation 2000; 6: 597–603.

11 Barrett AJ, Horowitz MM, Ash RC et al. Bone marrowtransplantation for Philadelphia chromosome-positive acutelymphoblastic leukemia. Blood 1992; 79: 3067–3070.

12 Dunlop LC, Powles R, Singhal S et al. Bone marrowtransplantation for Philadelphia chromosome-positive acutelymphoblastic leukemia. Bone Marrow Transplant 1996; 3:365–369.

13 Chao NJ, Blume KG, Forman SJ et al. Long-term follow-up of allogeneic bone marrow recipients for Philadelphiachromosome-positive acute lymphoblastic leukemia. Blood1995; 85: 3353–3354.

14 Thomas X, Thiebaut A, Olteanu N et al. Philadelphiachromosome positive adult acute lymphoblastic leukemia:characteristics, prognostic factors and treatment outcome.Hematol Cell Ther 1998; 40: 119–128.

15 Marks DI, Bird JM, Cornish JM et al. Unrelated donor bonemarrow transplantation for children and adolescents withPhiladelphia-positive acute lymphoblastic leukemia. J ClinOncol 1998; 16: 931–936.

16 Sierra J, Radich J, Hansen JA et al. Marrow transplants fromunrelated donors for treatment of Philadelphia chromosome-positive acute lymphoblastic leukemia. Blood 1997; 90: 1410–1414.

17 Miyamura K, Tanimoto M, Morishima Y et al. Detection ofPhiladelphia chromosome positive acute lymphoblastic leuke-mia by polymerase chain reaction: possible eradication of

minimal residual disease by marow transplantation. Blood1992; 79: 1366–1370.

18 Preudhomme CN, Henic B, Cazin JL et al. Good correlationbetween RT-PCR analysis and relapse in Philadelphia-positiveacute lymphoblastic leukemia. Leukemia 1997; 11: 294–298.

19 Michel G, Landman-Parker J, Auclerc MF et al. Use ofrecombinant human granulocyte colony-stimulating factor toincrease chemaotherapy dose-intensity: a randomized trial invery high risk childhood acute lymphoblastic leukemia. J ClinOncol 2000; 18: 1517–1524.

20 Ifrah N, Witz F, Jouet JP et al. Intensive short term therapywith granulocyte-macrophage colony-stimulating factor sup-port, similar to therapy for acute myeloblastic leukemia, doesnot improve overall results for adults with acute lymphoblasticleukemia. GOELAMS Group. Cancer 1999; 86: 496–505.

21 Dombret H, Gabert J, Boiron JM et al. for the Groupe d’etudeet de traitement de la leucemie aigue lymphoblastique de l’adulte (GET-LAL group). Outcome of treatment in adultswith Philadelphia chromosome-positive acute lymphoblasticleukemia: results of the prospective multi-center LALA-94trial. Blood 2002; 100: 2357–2366.

22 ISCN (International System for Human Cytogenetic Nomen-clature). Guidelines for cancer cytogenetics. In: Mitelman F(ed.). Supplement to an International System for HumanCytogenetic Nomenclature. Karger: Basel, 1991, pp 1–53.

23 Kawasaki ES, Clark SS, Coyne MY et al. Diagnosis of chronicmyeloid and acute lymphocytic leukemias by detection ofleukemia-specific mRNA sequences amplified in vitro. ProcNatl Acad Sci USA 1988; 85: 5698–5702.

24 Raynaud S, Mauvieux L, Cayuela JM et al. TEL-AML1 fusiongene is a rare event in adult acute lymphoblastic leukemia.Leukemia 1996; 10: 1529–1530.

25 van Dongen JJ, Macintyre EA, Gabert JA et al. StandardizedRT-PCR analysis of fusion gene transcripts from chromosomeaberrations in acute leukemia for detection of minimal residualdisease. Report of the BIOMED-1 concerted action: investiga-tion of minimal residual disease in acute leukemia. Leukemia1999; 13: 1901–1928 (Review).

26 Deconinck E, Cahn JY, Milpied N et al. Allogeneic bonemarrow transplantation for high-risk acute lymphoblasticleukemia in first remission: long-term results for 42 patientsconditioned with an intensified regimen (TBI, high-dose Ara-Cand melphalan). Bone Marrow Transplant 1997; 9: 731–735.

27 Glucksberg H, Storb R, Fefer A et al. Clinical manifestationsof graft-versus-host disease in human recipients of marrowfrom HL-A-matched sibling donors. Transplantation 1974; 4:295–304.

28 Przepiorka D, Weisdorf D, Martin P et al. 1994 Consensusconference on acute GVHD grading. Bone Marrow Transplant1995; 6: 825–828.

29 Flowers MED, Kansu E, Sullivan KM. Pathophysiology andtreatment of graft-versus-host disease. Hematol Oncol ClinNorth Am 1999; 13: 1091.

30 Kaplan E, Meier P. Nonparametric estimation from incom-plete observations. J Am Stat Assoc 1958; 53: 457–481.

31 Peto R, Peto J. Asymptotically efficient rank invariant testprocedures. J R Stat Soc A 1972; 135: 185–206.

32 Cox DR. Regression models and life-tables. J R Stat Soc B1972; 34: 187–220.

33 Annino L, Ferrari A, Cedrone M et al. Adult Philadelphia-chromosome-positive acute lymphoblastic leukemia: experi-ence of treatments during a ten-year period. Leukemia 1994; 8:664–667.

34 Faderl S, Kantarjian HM, Thomas DA et al. Outcome ofPhiladelphia chromosome-positive adult acute lymphoblasticleukemia. Leukemia Lymphoma 2000; 36: 263–273.

Allogeneic HSCT in Ph-positve ALLH Esperou et al

917

Bone Marrow Transplantation

35 Uckun FM, Nachman JB, Sather HN et al. Clinicalsignificance of Philadelphia chromosome positive pediatricacute lymphoblastic leukemia in the context of contemporaryintensive therapies: a report from the Children’s CancerGroup. Cancer 1998; 83: 2030–2039.

36 Snyder DS, Nademanee AP, O’Donnell MR et al. Long-termfollow-up of 23 patients with Philadelphia chromosome-positive acute lymphoblastic leukemia treated with allogeneicbone marrow transplant in first complete remission. Leukemia1999; 13: 2053–2058.

37 Cornelissen JJ, Carston M, Kollman C et al. Unrelated marrowtransplantation for adult patients with poor-risk acute lympho-blastic leukemia: strong graft-versus-leukemia effect and riskfactors determining outcome. Blood 2001; 97:1572–1577.

38 Snyder DS, Negrin RS, Blume KG et al. Allogeneic bonemarrow transplantation for BCR-ABL positive acute lympho-blastic leukemia [abstract]. Ann Hematol 1999; 78 (Suppl II): 59.

39 Kroger N, Kruger W, Wacker-Backhaus G et al. Intensifiedconditioning regimen in bone marrow transplantation forPhiladelphia chromosome-positive acute lymphoblastic leuke-mia. Bone Marrow Transplant 1998; 22: 1029–1033.

Appendix

The members of the participating teams from the SFGM-TC are: A Baruchel, A Devergie, H Dombret, H Esperou,

E Gluckman, Ph Guardiola, T Leblanc, P Ribaud, PhRousselot, G Socie (Hopital Saint-Louis, Paris); JMBoiron, J Reiffers (Hopital du Haut Leveque, Pessac); MKuentz, C Cordonnier, J Beaune (Hopital Henri Mondor,Creteil); I Yacoub-Agha, JP Jouet (Hopital Huriez, Lille);E Deconinck, JY Cahn (Hopital Jean Minjoz, Besancon);Ph Moreau, N Milpied, B Gueglio (Hotel Dieu, Nantes);JH Bourhis, L Jolibois (Institut Gustave Roussy, Villejuif);M Michalet, V Lheritier (Hopital Edouard Herriot, Lyon);C Faucher, D Blaise, D Reportier (Institut Paoli Calmette,Marseille); P Bordigoni (Hopital d’enfants Vandoeuvre lesNancy); N Ifrah, M Truchan-Graczyk (Centre Hospitalier,Angers); N Dhedin, JP Vernant, L Sutton, M Cornillot(Hopital de la Pitie-Salpetriere, Paris); M Attal (HopitalPurpan, Toulouse); B Lioure (Hopital de Haute Pierre,Strasbourg) N Fegeux, Ph Quittet, JF Rossi (Centrehospitalier, Montpellier); A Buzyn, A Fisher, I Hirsh(Hopital Necker, Paris); A Stamatoulas-Bastard, H Tilly(Centre Henri Becquerel, Rouen); D Guyotat (HopitalNord Saint-Etienne); F Demeocq, JL Stephan, JO Bay(Clermond Ferrand); M Duval, E Vilmer, R Larche(Hopital Robert Debre, Paris); F Garban, JJ Sotto (HopitalMichallon, Grenoble); A Sadoun (Hopital Jean Bernard,Poitiers); N Gratecos (Hopital de L’Archet, Nice); T deRevel, G Nedellec (Hopital Inter Armees Percy, Clamart).

Allogeneic HSCT in Ph-positve ALLH Esperou et al

918

Bone Marrow Transplantation