Syndromes Myélodysplasiques (SMD) et Allogreffes de...

Syndromes Myélodysplasiques (SMD) et

Allogreffes de CSH

Pr Ibrahim Yakoub-Agha

Introduction •  Maladie hétérogène

•  Histoire naturelle: •  > 50 ans mais peut toucher les jeunes (très jeunes..) •  Evolution variable (mois – décennies) •  Transformation en LA (état préleucémique) •  Evolution fatale

•  Absence de traitement spécifique curatif (transfusions, FCH, immunosuppresseurs, chimio, déméthylation , Imides..)

•  Allogreffe de CSH seul traitement potentiellement curatif

•  Donneur •  Lourdeur et toxicité

Sutton et al, Blood. 1996

1) Plateau à 30%

3) Transplant-Related Mortality (TRM), élevée & précoce!

2) Survie ~ EFS

L’allogreffe ne peut être le traitement standard

de tous les SMD

Quel source de

CSH?

Indication, quelle

maladie?

Quel traitement préalable?

Quel donneur?

Quel conditionnement?

entretien?

G

Prise en charge de la rechute post-greffe

Quel source de

CSH?

Indication, quelle

maladie? Quand?


Quel donneur?


entretien?

G


Quelle maladie?

•  International Prognostic Scoring System (IPSS)

•  Faible (= 0) (6 ans)

•  Intermédiaire-1 (0.5-1.0) (4 ans)

•  Intermédiaire-2 (1.5-2.0) (2 ans)

•  Élevé (> 2.0) (< 1an)

Greenberg, P. et al. Blood 1997;89:2079-2088

0 0.5 1.0 1.5 2.0

Blastes MO (%) <5 5-10 11-20 21-30

Caryotype* favorable intermédiaire défavorable

Cytopénies 0/1 2/3 * Favorable, normal, -Y, del(5q), del(20q); défavorable, complexe (3 anomalies) ou anomalie du Ch7; Intermédiaire, autres anomalies.

Greenberg, P. et al. Blood 1997;89:2079-2088

Int2 & élevé Indication ++

Faible & Int1 Indication discutable

•  Trop tard risque de transformation •  Trop tôt TRM inacceptable

•  Modèle de Markov •  Pertinence de l’IPSS •  Absence de greffe versus allo-SCT apparentée

Cutler, Blood 2004

Quand?

Cutler, C. S. et al. Blood 2004;104:579-585

Markov decision model

All patients began alive in the Alive-MDS state and were able to transition after each 6-month cycle to other health states. Patients could have remained in an alive state for any number of cycles without transitioning to another health state. The BMT state was transitory and all subjects entering the BMT state transitioned to another health state by the end of the cycle.

Cutler, C. S. et al. Blood 2004;104:579-585

Net benefit or loss of overall discounted life expectancy for the 4 IPSS risk groups are shown above and below the x-axis

Greffe retardée Absence de

risque

Greffe retardée délétère

Quand? •  IPSS

•  Intermédiaire-2 •  Élevé •  Faible •  Intermédiaire-1

Plus vite

Plus tard

Moment optimal pour allo-SCT •  Apparition de cytopénies importantes, •  Développement de nouvelles anomalies cytogénétiques, •  Changement de groupe IPSS, •  Echec du traitement (AZA, Rev…) •  Anomalies génétiques acquises? Protocole national en cours (Marie Robin)

Quel source de

CSH?

Indication, quelle

maladie? Quand?


Quel donneur?


entretien?

G


Traitement préalable à l’allogreffe?

•  Pourquoi? –  Pour gagner du temps « bridging » –  Maladie en RC > maladie progressive *

* Castro-Malaspina et al, Biol Blood Marrow Transplant. 2008

* Yakoub-Agha, J Clin Oncol. 2000 (SFGM-TC)

Allo-SCT for MDS and Blasts %

Sierra, Blood 2002

Trois stratégies

-‐  Chimiothérapie d’induc6on (CT)

-‐  Agents hypométhylants (AHM)

-‐  Greffe d’emblée « up-‐front »

Pa#ent condi#on Disease characteris#cs ICT HMA Up-‐front allo-‐SCT

Cytogene6cs* Marrow blasts %

Fit pa#ents (without comorbidi#es)

< high risk < 5 5-‐10 > 10

High risk < 5 5-‐10 > 10

ICT: induc6on-‐type chemotherapy; HMA: hypomethyla6ng agents; allo-‐SCT: allogeneic stem cell transplanta6on: *as assessed by IPSS: NI: not indicated; BO: best op6on;

I. Yakoub-‐Agha and J. Deeg, BBMT, 2014

decision-‐making algorithm based on disease characteris#cs and pa#ent age and comorbidi#es. In the absence of prospec/ve trial.

Chimiothérapie d’induc/on type LAM

Epidemiology

•  Incidence: 3.4 par 100,000 inhabitants

Rollison DE, et al. Blood. 2008;112:45-52.

Inci

denc

e R

ate

(per

100

,000

)

Age at Diagnosis (Yrs) *P for trend < .05

0

10

20

30

40

50

< 40 40-‐49 50-‐59 60-‐69 70-‐79 ≥ 80

2.6 3,5 7,4

16

33,8

36.4*

High-risk MDS and AML, > 50 years old

Estey E et al. Blood 2007;109:1395-1400

©2007 by American Society of Hematology

Estey E et al. Blood 2007;109:1395-1400


Total: 14/259 (5.4%!!!!!) CR: 14/99

(15%)

High-risk MDS and AML, > 50 years old

Prognostic factors in adult de novo MDS treated by intensive chemotherapy. P. Fenaux, Br J Haematol. 1991

CR rate according to FAB RAEB-‐T (> 19% of marrow blasts) at diagnosis : 69% other FAB subtypes : 19% (P = 0.008)

. DFS according to karyotype normal : median 16.5 months abnormal : median 4 months (P = 0.018).

15 RAEB-‐T at diagnosis and normal karyotype -‐ CR rate of 80% -‐ median actuarial DFS of 18 months,.




< high risk < 5 No 5-‐10 possible > 10 BO

High risk < 5 No 5-‐10 No > 10 possible

ICT: induc6on-‐type chemotherapy; HMA: hypomethyla6ng agents; allo-‐SCT: allogeneic stem cell transplanta6on: *as assessed by IPSS: NI: not indicated; BO: best op6on;



Agents hypométhylants (AHM)

Agents hypométhylants (AHM)

Fenaux et al, lancet oncol, 2012

HMA and Ch 5 and 7 abnormalities

JCO.2012.44.3499 29

30

OUTCOME

Damaj et al, JCO 2012

HMA and response rates




< high risk < 5 No BO 5-‐10 possible BO > 10 BO possible

High risk < 5 No possible 5-‐10 No possible > 10 possible possible

ICT: induc6on-‐type chemotherapy; HMA: hypomethyla6ng agents; allo-‐SCT: allogeneic stem cell transplanta6on: *as assessed by IPSS: NI: not indicated; BO: best op6on; **: if pa6ent can undergo allo-‐SCT rapidly within less than 3 months.



Allogreffe d’emblée (Up-‐front)

Avantages: •  Une bonne option pour les patients ne pouvant pas répondre à une CT ou AHM •  Pas de toxicité avant greffe •  La greffe peut être réalisable dès qu’un donneur est identifié.

Allogreffe d’emblée Up-‐front

transplantation in the no donor group) and gave similar resultscon!rming that transplantation signi!cantly improved survival(Supplementary data).

DISCUSSIONThe present study shows in a prospective way, when consideringall other treatments available, that HSCT can signi!cantly improvethe survival in patients with a higher risk MDS. Long-termsurvivors were only seen in the transplanted group. This is toour knowledge the !rst time this assumption is con!rmedprospectively. The 'donor versus non donor analysis' avoids usualbiases because of the comparison of transplanted and non-transplanted patient cohorts, for whom follow-up starts at thetime of transplant. The constantly made remark about the fact thattransplanted patient cohorts represent the selected cohorts couldbe con!rmed here. One bias is the exclusion of patients whoacquire a contraindication to transplant or who die beforetransplant. These patients are usually included in the non-transplant cohort, worsening the outcome of that cohort (andimproving that of the transplant-cohort). The advantage of our'donor versus no donor' study was that the probability to acquire

these potential contraindications might have been be similar inboth groups, particularly because baseline clinical, biological andnon-transplant treatment characteristics did not signi!cant differbetween the two groups. Furthermore, to better compare thesurvival of the donor and no donor groups, when !nding a donormay take a few weeks, we used a landmark analysis. Othermethods, such as the Mantel–Byar method, could have been usedif the exact date at which the donor was found had beenrecorded. As only two patients died within the !rst 3 months offollow-up, such an alternative method is unlikely to have alteredthe results. To account for other sources of time-dependentconfounding bias, we used inverse probability of censoringweighting to account for mismatched transplant in the donorgroup, as well as marginal structural models and inverseprobability of treatment weighting to estimate a causal effect ofHLA-matched transplant on survival. Given all these precautions,the OS was signi!cantly better in patients with an HLA matcheddonor (donor group): 37 versus 15% in the no donor group. This4-year OS can appear modest in patients transplanted withreduced intensity conditioning regimen , but it takes into accountpatients who died before transplantation or those who hadacquired a contraindication to the transplant.Survival curves between the donor and no donor groups

separated only after the second year post transplantation for atleast two reasons: (1) patients could not receive transplantation atinclusion due to a necessary logistical delay for transplantationand (2) transplanted patients had a maximal mortality risk duringthe !rst year post transplant. This observation had already beenreported by Platzbecker et al.5 in a large multicentric analysis of atransplanted and non-transplanted cohort. The transplantedcohort showed a better survival but only after the second yearof follow-up. This variability in the mortality risk over time explainsin part the dif!culty of comparing patients retrospectively.6 Theuse of reduced intensity conditioning regimen -transplantationhas decreased non-relapse mortality, allowing transplantation inolder and comorbid patients. However, causes of death aftertransplant are still shared between graft-versus-host disease andrelapse with a balance between both complications according toconditioning intensity.11

The fact that the majority of patients with a donor weretransplanted in the study made the interpretation of the datapossible. Indeed, 72% of patients with a donor underwent atransplantation which is consistent with studies based on donorselection.25,26 Nevertheless, some patients did not receivetransplantation because of disease progression or comorbidityacquisition, raising the issue of delaying transplantation by usingpre-transplant treatment. This study was not designed to answerthis question. Concerning the type of pre-transplant treatment, a

Table 2. Outcome according to the donor group

Outcome No donor HLA matched donor P

Treatment after inclusionIntensive chemotherapy, n (%) 12 (24) 33 (29) 0.57Demethylating agents, n (%) 44 (88) 79 (71) 0.017Any, n (%) 46 (92) 98 (88) 0.59Probability of achieving o10% blasts at 6 months (95% CI) 68% (53–79) 57% (47–66) 0.27a

Probability of remission at 6 months (95% CI) 22% (10–33) 21% (14–28) 0.78a

Probability of AML at 6 months (95% CI) 4% (0–9) 8% (4–14) 0.28a

Probability of death with disease at 6 months (95% CI) 0% (0–6) 4% (1–8) 0.048a

Overall survival at 48 months (95% CI) 15% (6–39) 37% (28–48) 0.020b

Probability of AML at 48 months (95% CI) 30% (12–45) 24% (16–33) 0.84b

Probability of death with the disease at 48 months (95% CI) 73% (47–87) 37% (27–46) 0.001b

Abbreviations: AML, acute myeloblastic leukemia; CI, con!dence interval; HLA, human leukocyte antigen. aRobust score test in Fine–Gray model over the 6!rst months. bRobust score test in the Cox model over whole follow-up, with time-varying effects for overall survival and death with the disease.

Months

Pro

babi

litie

s

Alive w/o SCT

Alive after SCT

Death after SCT

Death w/o SCT

3 12 24 36 48 600.0

0.2

0.4

0.6

0.8

1.0

1

0.8

0.6

0.4

0.2

0

Figure 2. Predicted probabilities of being alive or dead with orwithout transplantation in patients with an HLA matched donor.

Survival in MDS patients according to donor availabilityM Robin et al

4

Leukemia (2015) 1 – 6 © 2015 Macmillan Publishers Limited

Robin et al, Leukemia 2015

ORIGINAL ARTICLE

HLA-matched allogeneic stem cell transplantation improvesoutcome of higher risk myelodysplastic syndrome Aprospective study on behalf of SFGM-TC and GFMM Robin1,2,3, R Porcher4,5, L Adès6, E Raffoux7, M Michallet8, S François9, J-Y Cahn10, A Delmer11, E Wattel8, S Vigouroux12, J-O Bay13,J Cornillon14, A Huynh15, S Nguyen16, M-T Rubio17, L Vincent18, N Maillard19, A Charbonnier20, RP de Latour1,2,3, O Reman21,H Dombret2,6, P Fenaux2,6 and G Socié1,2,3

Allogeneic hematopoietic stem cell transplantation (HSCT) is considered the only a curative treatment in patients with higher riskmyelodysplastic syndrome (MDS), although demethylating agents (DMA) have been reported to improve survival. The advantage ofHSCT over other treatment comes from retrospective studies and the aim of the current study was to prospectively test thishypothesis, analyzing in particular patients from the pre-transplant period to avoid the selection bias of performing transplantation.This study was conducted to compare overall survival in MDS patients candidates to transplantation according to donor availability.The majority of patients (76%) received a treatment with DMA after registration, 69% had a human leukocyte antigen (HLA)-identicaldonor, 70% of whom were transplanted. Baseline patient and disease characteristics were similar according to donor availability. Four-year overall survival was signi!cantly better in patients with an HLA matched donor (37%) compared to patients without donor (15%).There was also evidence that this overall survival advantage was because of transplantation. Mortality risk was decreased aftertransplantation but it became signi!cant only after the second year post transplant, because of early transplant-related mortality. Ourresults appear to justify, in higher risk MDS, a transplantation approach in all potential candidates who have an HLA identical donor.

Leukemia advance online publication, 10 March 2015; doi:10.1038/leu.2015.37

INTRODUCTIONMyelodysplastic syndromes (MDS) are bone marrow stem celldisorders predominating in the elderly, characterized by ineffectivehematopoiesis with marrow dysplasia, blood cytopenias and a highrisk of transformation to acute myeloblastic leukemia (AML).Prognosis of MDS is largely determined by an InternationalPrognostic Scoring System (IPSS),1 recently revised,2 where higherrisk patients (that is, patients with intermediate 2 (int-2) or high IPSS)have a poor survival (median of less than 18 months). Althoughintensive AML-like chemotherapy (IC) and the demethylating agents(DMA), especially azacitidine, can induce remission and/or improvesurvival in higher risk MDS, they are generally considered asnon-curative. By contrast, allogeneic hematopoietic stem celltransplantation (HSCT) can cure a proportion of MDS patients,including elderly patients increasingly referred to HSCT because ofthe development of a reduced intensity conditioning regimen andthe improvement of overall transplantation management diminish-ing post-transplant morbidity and mortality.Most retrospective studies have concluded that patients with

higher risk MDS have a survival advantage if they can bene!t fromearly HSCT.3–6 However, treatment-related mortality remains high

in HSCT, between 10 and 40%, particularly after myeloablativeconditioning regimen or in the elderly7–12 and long-term survivalranges from 30 to 60% depending on patient characteristics,disease risk, type of donor, source of stem cells and complicationsoccurring after HSCT. Furthermore, HSCT is offered after a relativelystringent selection based on age (younger patients), performancestatus (good) and often disease evolution (not too rapidly evolving),which preclude any adequate retrospective comparison betweentransplant- and non-transplant-cohorts and explain why a minorityof MDS patients are !nally transplanted. Prospective studies thatwould include patients before HSCT are therefore justi!ed.13

The objective of the present study was to prospectivelydetermine in a large cohort of patients with higher risk MDS withno obvious contraindication to HSCT at inclusion, the clinicalbene!t of HSCT over conventional treatments, comparing out-come of patients with or without a donor.

PATIENTS AND METHODSEligibility criteriaEligibility criteria were (1) age 50–70 years, (2) de novo or therapy-relatedMDS,14 with int-2 or high IPSS, or int-1 IPSS with poor risk cytogenetics or

1Hematologie - Transplantation; Assistance Publique Hôpitaux de paris (AP-HP), Hôpital Saint Louis, Paris, France; 2University Paris VII, Denis Diderot, Paris, France; 3Inserm U 1160,Paris, France; 4Team METHODS, Epidemiology and Statistics Sorbonne Paris Cité Research Centre UMR 1153, Inserm, Paris Descartes University, Paris, France; 5AP-HP, HôpitalHôtel-Dieu, Centre d’Épidémiologie Clinique, Paris, France; 6Hématologie sénior, Assistance Publique Hôpitaux de paris (AP-HP), Hôpital Saint Louis and Hôpital Avicenne, Paris,France; 7Hématologie adulte, Assistance Publique Hôpitaux de paris (AP-HP), Hôpital Saint Louis, Paris, France; 8Hématologie, CHU, Lyon, France; 9Hématologie, CHU, Angers,France; 10Hématologie, CHU,UMR 5525 CNRS-UJF, Grenoble, France; 11Hématologie, CHU, Reims, France; 12Hématologie, CHU, Bordeaux, France; 13Hématologie, CHU, Clermont-Ferrand, France; 14Hématologie, Institut de Cancérologie de la Loire, Saint-Etienne, France; 15Hématologie, CHU, Toulouse, France; 16Hématologie, hôpital de La Pitié SalpétrièreAssistance Publique Hôpitaux de paris (AP-HP), Hôpital Saint Louis, Paris, France; 17Hématologie, hôpital Saint-Antoine, Assistance Publique Hôpitaux de paris (AP-HP), HôpitalSaint Louis, Paris, France; 18Hématologie, CHU, Montpellier, France; 19Hématologie, CHU, Poitiers, France; 20Hématologie CHU, Amiens, France and 21Hématologie CHU, Caen,France. Correspondence: Dr M Robin, Service d’hématologie – greffe, Hôpital St Louis, 1 Avenue Claude Vellefaux, Paris Cedex 10 75475, France.E-mail: [email protected] 26 September 2014; revised 26 November 2014; accepted 1 December 2014; accepted article preview online 13 February 2015

Leukemia (2015), 1–6© 2015 Macmillan Publishers Limited All rights reserved 0887-6924/15

www.nature.com/leu

Greffe d’emblée (Up-‐front)

Contre: •  Risque de transformation en LA avant greffe

•  Augmentation du risque de rechute post-greffe

Up-‐front allo-‐SCT

Against: •  Risk of transformation to AML before allo-SCT

•  Increased risk of post-transplant relapse in patients transplanted with progressive disease.

Greenberg, blood, 1997
















Months

Pro

babi

litie

s

Alive w/o SCT

Alive after SCT

Death after SCT

Death w/o SCT

3 12 24 36 48 600.0

0.2

0.4

0.6

0.8

1.0

1

0.8

0.6

0.4

0.2

0



4

Leukemia (2015) 1 – 6 © 2015 Macmillan Publishers Limited


ORIGINAL ARTICLE











www.nature.com/leu


recent French retrospective study suggested that patients treatedwith DMA or IT had similar outcomes.27 We similarly found nodifference between those pretransplant treatments (data noshown). The question of cytoreductive treatment before trans-plantation is still pending.28

This study had some limits. First, this was not a randomizedprospective controlled study with strictly similar treatment plans,but more an observation study based on practicability. Theoptimal way to perform this study would be the randomization totransplantation in patients with a donor. The strong evidence thatHSCT is the only curative option and further endorsement byconsensual recommendations4 possibly preclude performing suchprospective randomized trials in patients with a donor.Patients with a mismatched donor (9/10) were also transplanted

in this study but the probability to identify a 9/10 donor in theabsence of 10/10 donor remains relatively low and this grouprepresented only 16 patients which precluded any reliablestatistical analysis for this subset. We thus decided to censorthese patients at time of their transplantation and analyzed themin the no donor group after weighting them with potentialconfounders.Another potential issue was the high probability of identifying

an HLA matched donor in the present study (69%), but it wasconcordant with previous studies.29 Furthermore, completeness ofpatient inclusion was veri!ed in the two largest centers(representing 50% of the patients) in order to check that patientswithout a donor were all registered.In conclusion, after adjusting for all known prognostic factors,

including cytogenetics and response to non-transplant treatment,patients with a donor had a survival improvement, whichmay be attributed to the transplantation. The majority ofpatients with higher risk MDS have a donor and our !ndingsstrongly support that they should be transplanted before theprogression or acquisition of a non-hematological contraindica-tion to transplantation.

CONFLICT OF INTERESTThe authors declare no con"ict of interest.

ACKNOWLEDGEMENTSWe thank members of GFM and SFGM-TC for their participation and support. Wethank Alix O’Meara for the English review. We thank Pierre Fabre for theircontribution to this study.

AUTHOR CONTRIBUTIONSMR, GS, HD, ER, LA and PF designed research, RP and MR analyzed the data; allco-authors performed the research and recruited patients, MR, GS, LA, PF andRP wrote the paper and all authors gave their comments.

REFERENCES1 Greenberg P, Cox C, LeBeau MM, Fenaux P, Morel P, Sanz G et al. Scoring system

for evaluating prognosis in myelodysplastic syndromes. Blood 1997; 89:2079–2088.

2 Greenberg, PL, Tuechler H, Schanz J, Sanz G, Garcia-Manero G, Sole f et al.Revised international prognostic scoring system for myelodysplastic syndromes.Blood 2012; 120: 2454–2465.

3 Cutler CS, Lee SJ, Greenberg P, Deeg HJ, Perez WS, Anasetti C et al. A decisionanalysis of allogeneic bone marrow transplantation for the myelodysplasticsyndromes: delayed transplantation for low-risk myelodysplasia is associated withimproved outcome. Blood 2004; 104: 579–585.

4 Koreth J, Pidala J, Perez WS, Deeg HJ, Garcia-Manero G, Malcovati L et al. Role ofreduced-intensity conditioning allogeneic hematopoietic stem-cell transplanta-tion in older patients with de novo myelodysplastic syndromes: an internationalcollaborative decision analysis. J Clin Oncol 2013; 31: 2662–2670.

5 Platzbecker U, Schetelig J, Finke J, Trenschel R, Scott BL, Kobbe G et al. Allogeneichematopoietic cell transplantation in patients age 60-70 years with de novo high-risk myelodysplastic syndrome or secondary acute myelogenous leukemia:comparison with patients lacking donors who received azacitidine. Biol BloodMarrow Transplant 2012; 18: 1415–1421.

6 Brand R, Putter H, van Biezen A, Niederwieser D, Martino R, Mufti G et al.Comparison of allogeneic stem cell transplantation and non-transplant approa-ches in elderly patients with advanced myelodysplastic syndrome: optimal sta-tistical approaches and a critical appraisal of clinical results using non-randomized data. PLoS One 2013; 8: e74368.

7 Sierra J, Perez WS, Rozman C, Carreras E, Klein JP, Rizzo GD et al. Bone marrowtransplantation from HLA-identical siblings as treatment for myelodysplasia. Blood2002; 100: 1997–2004.

8 Guardiola P, Runde V, Bacigalupo A, Ruutu T, Locatelli F, Boogaerts MA et al.Retrospective comparison of bone marrow and granulocyte colony-stimulatingfactor-mobilized peripheral blood progenitor cells for allogeneic stem celltransplantation using HLA identical sibling donors in myelodysplastic syndromes.Blood 2002; 99: 4370–4378.

9 de Lima M, Anagnostopoulos A, Munsell M, Shahjahan M, Ueno N, Ippoliti C et al.Nonablative versus reduced-intensity conditioning regimens in the treatment ofacute myeloid leukemia and high-risk myelodysplastic syndrome: dose is relevant

Table 3. Factors associated with death in all patients and factors associated with performing transplantation in patients with the donor group(multivariable models)

Covariates Death Transplantation

HR (95% CI) P HR (95% CI) P

Age (per year) 1.01 (0.97–1.06) 0.56 0.98 (0.93–1.03) 0.40Marrow blasts at inclusion (per %) 1.02 (1–1.04) 0.014 0.94 (0.89–0.99) 0.021Intermediate cytogenetics 0.67 (0.36–1.28) 0.23 1.06 (0.57–1.95) 0.86Poor cytogenetics 1.90 (1.18–3.05) 0.008 0.74 (0.41–1.35) 0.33Response to treatmentComplete remission 0.55 (0.28–1.09) 0.088 5.63 (3.09–10.2) o0.0001Partial remission 0.75 (0.33–1.72) 0.50 6.76 (3.37–13.6) o0.0001Progressive disease 1.62 (0.73–3.58) 0.23 4.92 (1.67–14.5) 0.004Persistent AML 2.36 (1.25–4.44) 0.008 0.50 (0.11–2.21) 0.036Marrow blasts 410% during follow-up 1.56 (0.89–2.74) 0.12 1.19 (0.46–3.09) 0.72

Abbreviations: AML, acute myeloblastic leukemia; CI, con!dence interval; HR, hazards ratio.

Table 4. Causal effect of stem cell transplantation estimated by usinga marginal structural model

Variable HR (95% CI) P

o1 year after HSCT 1.07 (0.60–1.91) 0.821–2 year after HSCT 0.69 (0.28–1.69) 0.3142 years after HSCT 0.024 (0.003–0.22) 0.0009

Abbreviations: CI, con!dence interval; HR, hazards ratio; HSCT, hemato-poietic stem cell transplantation.


5

© 2015 Macmillan Publishers Limited Leukemia (2015) 1 – 6

ORIGINAL ARTICLE











www.nature.com/leu
















Months

Pro

babi

litie

s

Alive w/o SCT

Alive after SCT

Death after SCT

Death w/o SCT

3 12 24 36 48 600.0

0.2

0.4

0.6

0.8

1.0

1

0.8

0.6

0.4

0.2

0



4

Leukemia (2015) 1 – 6 © 2015 Macmillan Publishers Limitedcord blood and 1 from a haploidentical related donor. Thirty-onepatients in the donor group did not receive HSCT, 5 of whomwere still alive at last follow-up. Reasons not to perform thetransplantation were: progressive disease with marrow blasts410% despite treatment (n= 16), acquisition of a comorbiditycontraindicating the transplant (n= 9), death during IC or DMA inresponders or before evaluation (n= 4), patient refusal (n= 1) andsocial reason (n= 1).

Outcome at 48 monthsWith a 43-month (range 9–60) median follow-up, the probabilityof death with the disease was 73% in the no donor group and 37%in the donor group (P= 0.001). Sixteen percent of patients in thedonor group versus 6% in the no donor with a disease in completeremission. The incidence of AML post MDS at 48 months was 30and 24% in the no donor and the donor group. Four-year survivalwas 37% in the donor group compared with 15% in the no donorgroup (P= 0.02) (Figure 1). As shown in the Figure 1, survivalbetween the two groups appeared similar during the !rst

24 months of follow-up, the difference becoming apparent afterthat date. The survival gain in the donor group was likely becauseof transplantation. The early transplant-related mortality and thedelay in transplantation likely explained the absence of signi!cantdifference between the donor and no donor groups during the!rst 2 years.Figure 2, representing only patients with a donor, shows that

survival progressively declined in non-transplanted patients too5% at 48 months, while it remained relatively stable 430% intransplanted patient after 18 months. In the donor group, becausepatients transplanted differed from those not transplanted, wetested the effect of transplantation as a time-dependent variablein a marginal structural Cox model, controlling for time-dependentconfounding by an inverse probability of treatment weighting (seemethod). Brie"y, the weights are based on the inverse of eachpatient’s probability of undergoing the transplant. This aims atcreating a statistical population in which performing HSCT at anytime would be unrelated to the observed prognostic factors andthus affected by indication bias. Prognostic covariates (timedependent or not) associated with transplant were determined byusing Cox models with time-dependent variables, and then usedto derive the time-dependent weights. The following parameterswere used in the weighting model: age, marrow blast % atinclusion, poor cytogenetics, response to the treatment (completeresponse, PR, stable, progressive, refractory AML and bone marrowblasts 410% at any time point) during follow-up (Table 3). Ofnote, poor cytogenetics were associated with poor survival butnot with performing or not transplantation (Table 3). Response topre-transplant treatment was associated with better survival butnot with performing or not transplantation except if the patientremained with refractory AML, which dramatically decreased theprobability to be transplanted (Table 3). The marginal structuralmodel shows that the mortality risk related to the transplantationwas not constant over time and three periods were individualized.During the 12 !rst months post transplant, the mortality risk wasslightly higher; it decreased between 12 and 23 months andbecame signi!cantly lower after 24 months (Table 4) !nallycon!rming Figure 1 results. This model was applied in the donorgroup and in the whole cohort (mimicking the effect of

Table 1. Patient characteristics at inclusion

Covariates No donor HLA matcheddonor

P

Number of patients 50 112Median age in years (range) 61 (50–70) 60 (50–70) 0.47

Gender 0.86Female, n (%) 16 (32) 39 (35)Male, n (%) 34 (68) 73 (65)

Inclusion at diagnosis, n (%) 27 (55) 81 (72) 0.081

Time since diagnosis in monthso12, n (%) 11 (22) 11 (10)12–36, n (%) 6 (12) 14 (12)436, n (%) 5 (10) 6 (5)

Therapy related MDS, n (%) 10 (20) 24 (21) 40.99

WHO classi!cation, n (%) 0.62AML 5 (10) 7 (6)CMML 1, 2 8 (16) 12 (11)RAEB1 12 (24) 17 (15)RAEB2 21 (42) 59 (53)RCMD 4 (8) 11 (10)RA/MDS-U 0 6 (6)

Percent marrow blasts (IQRa) 12 (8–17) 12 (7–14) 0.56! 5%, number, n (%) 45 (90) 93 (84) 0.34

Cytogenetics (IPSS), n (%) 0.22Good 10 (20) 37 (33)Intermediate 14 (29) 23 (21)Poor 25 (51) 51 (46)

Cytogenetics (IPSS-R), n (%) 0.90Very good 0 1 (1)Good 14 (29) 38 (35)Intermediate 10 (21) 19 (17)Poor 14 (29) 32 (29)Very poor 10 (21) 19 (17)

IPSS score, n (%) 0.53Intermediate 1 5 (10) 8 (7)Intermediate 2 28 (56) 75 (67)High 12 (24) 22 (20)

Abbreviations: AML, acute myeloblastic leukemia; CMML, chronic myelo-monocytic leukemia; HLA, human leukocyte antigen; IPSS, InternationalPrognostic Scoring System; IQR, interquartile range; MDS, myelodysplasticsyndromes; MDS-U, unclassi!ed MDS; RA, refractory anemia; RAEB,refractory anemia with excess blast; RCMD, refractory cytopenia withmultilineage dysplasia; WHO, World Health Organization. aIQR.

Ove

rall

surv

ival

3 12 24 36 48 60Months from inclusion

0.0

0.2

0.4

0.6

0.8

1.0

No. at risk:HLA!matched donor 112 85 63 50 39 33 23 14 8 5

No matched donor 50 29 21 15 8 6 3 2 0 0

HLA!matched donorNo HLA!matched donor

Figure 1. Overall survival according to donor availability: no donorgroup and donor group. Donor group include all patients witheither an HLA matched sibling or an unrelated donor.


3

© 2015 Macmillan Publishers Limited Leukemia (2015) 1 – 6


Damaj et al, BBMT, 2014

Upfront Allogeneic Stem Cell Transplantation afterReduced-Intensity/Nonmyeloablative Conditioning for Patientswith Myelodysplastic Syndrome: A Study by the SociétéFrançaise de Greffe de Moelle et de Thérapie Cellulaire

Gandhi Damaj 1, Mohammad Mohty 2, Marie Robin 3, Mauricette Michallet 4, Patrice Chevallier 5,Yves Beguin 6, Stephanie Nguyen 7, Pierre Bories 8, Didier Blaise 9, Natacha Maillard 10,Marie Therese Rubio 2, Nathalie Fegueux 11, Jerome Cornillon 12, Aline Clavert 13, Anne Huynh 14,Lionel Adès 15, Anne Thiébaut-Bertrand 16, Olivier Hermine 17, Stephane Vigouroux 18,Pierre Fenaux 15, Alain Duhamel 19, Ibrahim Yakoub-Agha 20,*

1Departement d’Hématologie, Centre Hospitalier Universitaire, Université de Caen, Faculté de Médecine, Caen, France2Departement d’Hématologie, Assistance Publique-Hôpitaux de Paris, Hôpital Saint-Antoine, Université Paris 6, Paris, France3Departement d’Hématologie-Transplantation, Assistance Publique-Hôpitaux de Paris, Hôpital St. Louis, Université Paris 7, Paris, France4Departement d’Hématologie, Centre Hospitalier Universitaire Lyon Sud, Lyon, France5Departement d’Hématologie, Centre Hospitalier Universitaire, Nantes, France6Departement d’Hematologie, Centre Hospitalier Universitaire and University of Liège, Belgium7Departement d’Hématologie, Assistance Publique-Hôpitaux de Paris, Hôpital de la Pitié-Salpêtrière, Université Paris 6, Paris, France8Departement d’Hématologie, Hôpitaux Universitaires de Strasbourg, France9Departement d’Hématologie, Institut Paoli-Calmettes, Marseille, France10Departement d’Hématologie, Centre Hospitalier Universitaire, Poitiers, France11Departement d’Hématologie, Centre Hospitalier Universitaire, Montpellier, France12Departement d’Hématologie, Centre Hospitalier Universitaire, Rennes, France13Departement d’Hématologie, Centre Hospitalier Universitaire, Angers, France14Departement d’Hématologie, Centre Hospitalier Universitaire Purpan, Toulouse, France15Departement d’Hématologie, Assistance Publique-Hôpitaux de Paris, Hôpital Avicenne, Université Paris 13, Bobigny, France16Departement d’Hématologie, Centre Hospitalier Universitaire et Unité Mixte de Recherche 5525 Centre National de Recherche Scienti!que-Université JosephFournier, Grenoble, France17Departement d’Hématologie, Assistance Publique-Hôpitaux de Paris, Hôpital Necker Enfants-Malades, Université Paris 5, Paris, France18Departement d’Hématologie, Centre Hospitalier Universitaire, Bordeaux, France19Departement de Biostatistics, Centre d’Etudes et de Recherche en Informatique Médicale, Centre Hospitalier Universitaire, Lille, France20Departement d’Hématologie, Centre Hospitalier Régionale Universitaire, Lille, France

Article history:Received 2 February 2014Accepted 4 May 2014

Key Words:TransplantationAzacitidineBest supportive careMDS

a b s t r a c tCytoreduction before allogeneic stem cell transplantation (allo-SCT) for patients with myelodysplastic syn-dromes remains a debatable issue. After excluding patients who had received preconditioning inductionchemotherapy, we analyzed 128 consecutive patients with myelodysplastic syndrome who received reduced-intensity or nonmyeloablative conditioning (RIC/NMA) allo-SCT. Among them, 40 received azacitidine (AZA)before transplant (AZA group) and 88 were transplanted up front (best supportive care [BSC] group). Atdiagnosis, 55 patients had intermediate 2 or high-risk scores per the International Prognostic Scoring Systemand 33 had a high cytogenetic risk score. Progression to a more advanced disease before allo-SCT wasrecorded in 22 patients. Source of stem cells were blood (n ! 112) or marrow (n ! 16) from sibling (n ! 78) orHLA-matched unrelated (n ! 50) donors. With a median follow-up of 60 months, 3-year overall survival,relapse-free survival, cumulative incidence of relapse, and nonrelapse mortality were, respectively, 53%versus 53% (P ! .69), 37% versus 42% (P ! .78), 35% versus 36% (P ! .99), and 20% versus 23% (P ! .74), for theAZA group and BSC group, respectively. Multivariate analysis con!rmed the absence of statistical differencesin outcome between the AZA and BSC groups, after adjusting for potential confounders using the propensity

Financial disclosure: See Acknowledgments on page 1355.* Correspondence and reprint requests: Ibrahim Yakoub-Agha, MD, PhD,

UAM allogreffes de CSH, CHRU, F-59037 Lille CEDEX, France.

E-mail address: [email protected] (I. Yakoub-Agha).

1083-8791/$ e see front matter ! 2014 American Society for Blood and Marrow Transplantation.http://dx.doi.org/10.1016/j.bbmt.2014.05.010

Biol Blood Marrow Transplant 20 (2014) 1349e1355

Biology of Blood andMarrow Transplantationjournal homepage: www.bbmt.org

P=0,69

HR= 1.27, 95%CI (0.78-2.34)

P=0,99

HR= 1.15, 95%CI (0.62-2.13)

P=0,78

HR= 1.04, 95%CI (0.61-1.75)

P=0,74

HR= 1.56, 95%CI (0.64-3.85)

AZA versus BSC

Damaj et al, BBMT 2014

BSC AZA




< high risk < 5 No BO possible 5-‐10 possible BO possible > 10 BO possible possible

High risk < 5 No possible BO** 5-‐10 No possible BO** > 10 possible possible BO**

**: if pa6ent can undergo allo-‐SCT rapidly within less than 4 months.



Quel source de

CSH?

Indication, quelle

maladie? Quand?


Quel donneur?


entretien?

G


62.9% ! 6.3%, 55.6% ! 6.5%, 19.4% ! 5.3%, and 25% ! 5.7%,respectively. As listed in Table 3, none of the following variablesseemed to influence the outcome of transplantation: recipient age, sexof donor, sex of recipient, sex mismatch, donor cytomegalovirus(CMV) serology, or disease status. No factor among initial patient ordonor characteristics was found to influence relapse. Overall survivalwas adversely influenced by recipient CMV-positive serology and bydonors older than 37 years. TRM was affected by the same variablesand by recipient/donor CMV serology mismatch (negative donor andpositive recipient). EFS rates were found to be lower for patients withpositive CMV serology (Fig 1).

Acute GVHD. One hundred sixteen patients (49%) developedacute grades II to IV GVHD, including 83 patients (46%) from thesibling group and 33 (60%) from the unrelated group. Fifty patients(21%) developed acute grades III to IV GVHD; 34 patients (19%)from the sibling group and 16 (29%) from the unrelated group. Therisk of developing acute grades III to IV GVHD was lower for patientswith acute leukemia and for patients whose time from diagnosis toalloSCT was less than 5.5 months. The risk of acute grades II to IVGVHD was higher for patients with CML or for patients who hadundergone alloSCT from an unrelated donor (Fig 1). The risk waslower for patients with acute leukemia or positive CMV serology.

Multivariable AnalysisTable 4 lists hazard ratios (HR) and confidence intervals associ-

ated with donor type and with significant other risk factors.The effect of donor type was nonsignificant for overall survival,

EFS, relapse, TRM, or acute GVHD. Overall survival and TRM wereadversely influenced by a recipient CMV-positive serology, donor ageolder than 37 years, and the occurrence of acute grades II to IV GVHD.EFS rates were higher for patients with recipient CMV-positive serol-ogy. Acute grades II to IV GVHD rates were higher for patients withCML, but unexpectedly lower for patients with recipient CMV-positive serology. No factor was found to influence either relapse oracute grades III to IV GVHD.

DISCUSSION

To our knowledge, this is the first study which prospectively comparesresults of sibling transplantation to HLA-allellically matched (10/10)unrelated transplantation.

We observed 2-year overall survival, EFS, relapse, and TRM ratesof 62.9%, 55.6%, 19.4%, and 25%, respectively. Our results are similar

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

400 800 1,200 1,600 2,000Time (days)

400 800 1,200 1,600 2,000Time (days)

Time (days) 400 800 1,200 1,600 2,000

Time (days)

400 800 1,200 1,600 2,000Time (days)

N DeathsSibling 181 67Unrelated 55 212

Cox P = .551

20 40 60 80 100

Time (days)20 40 60 80 100


Cox P = .533

N GVHDSibling 181 83Unrelated 55 33

Cox P = .049

N GVHDSibling 181 34Unrelated 55 16

Cox P = .108


Cox P = .962


Cox P = .522

Ove

rall S

urviv

al (%

)

Tran

spla

ntat

ion-

Rela

ted

Mor

tality

(%)

Even

t-Fre

e Su

rviva

l (%

)

Acut

e G

rade

II-IV

GVH

D (%

)

Rela

pse

(%)

Acut

e G

rade

III-I

V G

VHD

(%)

A B

C D

E F

Fig 1. Cumulative incidence estimates of (A) overall survival, (B) event-free survival, (C) transplantation-related mortality, (D) acute grades II to IV graft-versus-hostdisease (GVHD), (E) relapse, and (F) grades III to IV GVHD in the 236 patients, according to the donor type.

Sibling v HLA-Matched Unrelated Allo-SCT

www.jco.org 5699Downloaded from jco.ascopubs.org on November 25, 2014. For personal use only. No other uses without permission.

Copyright © 2006 American Society of Clinical Oncology. All rights reserved.

Yakoub-Agha et al, JCO, 2006

Allogeneic Marrow Stem-Cell Transplantation From HumanLeukocyte Antigen–Identical Siblings Versus HumanLeukocyte Antigen–Allelic–Matched Unrelated Donors (10/10)in Patients With Standard-Risk Hematologic Malignancy:A Prospective Study From the French Society of Bone MarrowTransplantation and Cell TherapyIbrahim Yakoub-Agha, Florence Mesnil, Mathieu Kuentz, Jean Michel Boiron, Norbert Ifrah, Noel Milpied,Sami Chehata, Helene Esperou, Jean-Paul Vernant, Mauricette Michallet, Agnes Buzyn, Nicole Gratecos,Jean Yves Cahn, Jean Henri Bourhis, Zina Chir, Colette Raffoux, Gerard Socié, Jean Louis Golmard,and Jean-Pierre Jouet

A B S T R A C T

PurposeTo investigate the influence of donor type (human leukocyte antigen [HLA] -identical sibling donorversus HLA-A–, HLA-B–, HLA-Cw–, HLA-DRB1–, and HLA-DQB1–identical unrelated donors, orso-called 10/10) on the outcome of patients who underwent allogeneic stem-cell transplantation(alloSCT), adjusting for other prognostic factors, in patients with standard-risk hematologic malignancy.Patients and MethodsBetween March 2000 and January 2003, we prospectively investigated the outcome of 236consecutive patients with standard-risk malignancy from 12 French centers. Fifty-five patientsunderwent alloSCT from an unrelated HLA-identical donor at the allelic level, whereas 181 patientsreceived an alloSCT from an HLA-identical sibling. Diagnoses included acute leukemia (n ! 175),chronic myeloid leukemia (n ! 43), and myelodysplastic syndrome (MDS; n ! 18). All patientsreceived unmodified marrow graft following myeloablative conditioning with cyclophosphamideand total-body irradiation. Graft-versus-host disease (GVHD) prophylaxis consisted of cyclosporineand short-course methotrexate in all patients.ResultsIn multivariable analysis, overall survival and transplantation-related mortality were adverselyinfluenced by recipient cytomegalovirus (CMV) -positive serology, age of donor older than 37years, and the occurrence of acute grade ! II GVHD. Event-free survival rates were lower forpatients with recipient CMV-positive serology. Acute grades II to IV GVHD rates were higher forpatients with chronic myeloid leukemia (CML). No factor was found to influence either relapse oracute grades III to IV GVHD. The effect of donor type was nonsignificant for all criteria.ConclusionIn patients with standard-risk malignancy, transplantation from unrelated HLA-allellically matcheddonors led to outcomes similar to those from HLA-identical sibling donors.

J Clin Oncol 24:5695-5702. © 2006 by American Society of Clinical Oncology

INTRODUCTION

Allogeneic stem-cell transplantation (alloSCT) of-fers potential curative treatment for a wide range ofotherwise fatal hematologic diseases.1-6 However,only one third of patients have a human leukocyteantigen (HLA) -identical sibling donor. Indeed, withthe increase in the number of single-child families,stem-cell grafts from unrelated donors are beingincreasingly used.

The success of unrelated alloSCT is influencedby the degree of HLA compatibility between donorand patient.7 In contrast, the presence of donor-recipient mismatching is associated with increasedrisk of post-transplantation complications, includ-ing graft rejection, acute and chronic graft-versus-host disease (GVHD), and mortality; these risks areincreased by multiple HLA mismatches.8-12 Thismight explain why transplantation from unrelatedor related mismatched donors has been restricted topatients with high-risk malignancy.13-18

From Lille; Agence de la biomedecine,St Denis; Creteil; Bordeaux; Angers;Nantes; Institute Gustave Roussy, Ville-juif; Hospital St Louis, Paris; HopitalPitie-Salpetrieere, Paris; HopitalEdouard Herriot, Lyon; Hopital Necker,Paris; Nice; Besancon; Societe Fran-caise de Greffe De Moelle et TherapieCellulaire, St Denis; register FranceGreffe de Moelle, St Denis; Unite debiostatistique, Universite de Jussieu,Paris, France.

Submitted June 29, 2006; acceptedOctober 2, 2006; published online aheadof print at www.jco.org on November 20,2006.

Authors’ disclosures of potential con-flicts of interest and author contribu-tions are found at the end of thisarticle.

Address reprint requests to Jean-PierreJouet, MD, UAM allogreffes de CSH,Maladies du Sang, CHRU de Lille,F-59037, France; e-mail: [email protected].

© 2006 by American Society of ClinicalOncology

0732-183X/06/2436-5695/$20.00

DOI: 10.1200/JCO.2006.08.0952

JOURNAL OF CLINICAL ONCOLOGY O R I G I N A L R E P O R T

VOLUME 24 ! NUMBER 36 ! DECEMBER 20 2006

5695Downloaded from jco.ascopubs.org on November 25, 2014. For personal use only. No other uses without permission.

Copyright © 2006 American Society of Clinical Oncology. All rights reserved.

95% CI, 0.90-1.26; P ! .46) or DFS (RR, 1.02; 95% CI,0.86-1.20; P ! .86) in the multivariate analysis. There was anassociation with increased acute GVHD (RR, 1.43; 95% CI,1.16-1.76; P ! .001). There was a suggestion of an increasedrisk of TRM (RR, 1.22; 95% CI, 0.99-1.52; P ! .06) anddecreased risk of relapse (RR, 0.74; 95% CI, 0.57-0.96;P ! .02), although these were not significant.

Number of relevant mismatches

Increasing the number of HLA mismatches was associated withclinically important and statistically significantly worse outcomes(Table 5). In particular, comparison of 8/8, 7/8, and 6/8 HLA-matched cohorts showed a 1-year survival of 52%, 43%, and 33%,respectively, suggesting that each additional HLA mismatch isassociated with absolute unadjusted survival differences of 9% to10% (Table 3). This difference was highly statistically significantfor 7/8 versus 8/8 (RR, 1.25; 95% CI, 1.13-1.37; P " .001) and for6/8 versus 7/8 (RR, 1.31; 95% CI, 1.17-1.47; P " .001).

Figure 1 shows the association of the number of HLA mis-matches with survival for patients with early, intermediate, oradvanced disease. More advanced disease before HCT is associatedwith a greater absolute impact on survival compared with increas-ing HLA mismatching. For example, 1-year survival for a patientwith early-stage disease is 63% for an 8/8 matched donor and 52%for a 7/8 matched donor. However, for a patient with intermediate-risk disease, the survival with an 8/8 matched donor is 48%compared with 40% with a 7/8 donor.

Younger patients (" 40 years) with early-stage disease and 8/8donors had the best survival. For this group, 1-year survival was67% (95% CI, 62%-71%) and 5-year survival was 54% (95% CI,49%-58%).

Patient and donor characteristics versus HLA matching

Results of a multivariate model including 8/8, 7/8, and 6/8mismatched pairs are shown in Table 6. The relative risk for asingle HLA mismatch was 1.25 (25% higher risk for death witha single mismatch), similar to the relative risk associated withunmodifiable patient characteristics such as patient age, patient

Table 4. Unadjusted clinical outcomes by degree of HLA-A, -B, -C,-DRB1 match

8/8 7/8 6/8

No. patients 1840 985 633

Overall survival at 5 y 37 (35-40) 29 (26-32) 22 (19-26)

Overall survival at 1 y 52 (50-54) 43 (40-46) 33 (30-37)

Disease-free survival at 1 y 47 (44-49) 38 (35-42) 29 (26-33)

Treatment-related mortality at 1 y 36 (34-38) 45 (42-49) 55 (51-59)

Relapse at 1 y 18 (16-19) 16 (14-18) 15 (13-18)

Chronic GVHD at 1 y 44 (41-46) 36 (33-39) 32 (29-36)

Acute GVHD grade III-IV at 100

days

28 (26-30) 37 (34-40) 44 (40-48)

Failure to engraft at 28 days 10 (9-11) 13 (10-15) 17 (14-20)

Data are given as percent (95% CI).

Table 5. Effect of HLA mismatching on survival

No. RR (95% CI) P

Fully matched (8/8)* 1840 1.00 —

Single mismatch (7/8) 985 1.26 (1.15-1.39) ".001

A 274 1.36 (1.17-1.58) ".001

B 116 1.16 (0.92-1.47) .20

C 478 1.19 (1.05-1.35) .006

DRB1 117 1.48 (1.19-1.85) .001

Double mismatch (6/8) 633 1.66 (1.48-1.85) ".001

A#B 41 1.13 (0.77-1.65) .53

A#C 130 1.68 (1.37-2.07) ".001

A#DRB1 20 1.96 (1.19-3.23) .008

B#DRB1 29 1.51 (1.00-2.27) .05

B#C 284 1.87 (1.62-2.16) ".001

C#C 36 1.73 (1.18-2.54) .005

C#DRB1 72 1.27 (0.96-1.67) .09

Others 241 — —

Triple mismatch (5/8) 275 1.64 (1.42-1.91) ".001

A#B#C 97 1.77 (1.40-2.24) ".001

A#C#DRB1 20 1.82 (1.11-2.99) .02

B#C#C 41 1.96 (1.40-2.75) ".001

B#C#DRB1 48 1.64 (1.19-2.25) .002

B#B#C 25 0.93 (0.56-1.56) .79

Others 44 — —

Quadruple mismatch (4/8) 91 2.05 (1.61-2.60) ".001

A#B#C#DRB1 23 2.39 (1.51-3.77) ".001

Others 68 — —

— indicates not calculated.*Matched at HLA-A, -B, -C, -DRB1.

0.

1.

2.

3.

4.

5.

6.

7.

8.

9.

0.1

06846342210

Su

rv

iv

al

Su

rv

iv

al

Su

rv

iv

al

100.<=Pknar-goL

dehctaMALH8/8

dehctaMALH8/7

tnalpsnartretfashtnoM

dehctaMALH8/6

0.

1.

2.

3.

4.

5.

6.

7.

8.

9.

0.1

06846342210

dehctaMALH8/8

dehctaMALH8/7


dehctaMALH8/6

0.

1.

2.

3.

4.

5.

6.

7.

8.

9.

0.1

06846342210

20.=Pknar-goL

dehctaMALH8/8

dehctaMALH8/7


dehctaMALH8/6

100.<=Pknar-goL

A

B

C

(n = 241)

(n = 835)

(n = 378)

(n = 674)

(n = 410)

(n = 268)

(n = 327)

(n = 195)

(n = 123)

Figure 1. Survival of patients with early, intermediate, and advanced diseasedepending on degree of HLA matching (8/8, 7/8, and 6/8) for HLA-A, -B, -C, and-DRB1. (A) Early-stage disease for 8/8, 7/8, and 6/8, respectively: 1-year survival63%, 52%, and 39%; 5-year survival 50%, 39%, and 28%. (B) Intermediate-stagedisease for 8/8, 7/8, and 6/8, respectively: 1-year survival 48%, 40%, and 32%;5-year survival 32%, 27%, and 22%. (C) Advanced-stage disease for 8/8, 7/8, and6/8, respectively: 1-year survival 31%, 29%, and 24%; 5-year survival 17%, 15%, and10%.

4580 LEE et al BLOOD, 15 DECEMBER 2007 ! VOLUME 110, NUMBER 13

For personal use only.on November 28, 2014. by guest www.bloodjournal.orgFrom

Lee et al; Blood 2007.

Low-risk group

High-risk group

intermediate-risk group

Role of HLA-matching in patient outcome following an unrelated-donor allo-SCT

Quel source de

CSH?

Indication, quelle

maladie? Quand?


Quel donneur?


entretien?

G


CONDITIONING REGIMEN IS A MAJOR STEP OF ALLOGENEIC TRANSPLANT

•  The ideal preparative regimen for marrow transplantation of patients with malignant diseases should

1.  Have sufficient immunosuppressive effect … to avoid graft rejection

2.  Be capable of eradicating malignancy 3.  Have tolerable morbidity without mortality

1) Have sufficient immunosuppressive effect … to avoid graft rejection

Cornelissen et al, blood 2017

2) Be capable of eradicating malignancy GVL effect

2) Be capable of eradicating malignancy GVL effect

2) Be capable of eradicating malignancy Intensity of conditioning

BMT, 2013

CIR PFS

2) Be capable of eradicating malignancy Intensity of conditioning

3) Have tolerable morbidity without Intensity and NRM

BMT, 2013

NRM

Yakoub-Agha et al, JCO 2000

3) Have tolerable morbidity without Intensity and NRM

Outcome according to the intensity of conditioning

NMA RIC MA Hyper-MA Engraftment Disease control by GVL Disease control by conditioning TRM


NMA RIC MA Hyper-MA Engraftment yes yes yes yes Disease control by GVL Disease control by conditioning TRM


NMA RIC MA Hyper-MA Engraftment yes yes yes yes Disease control by GVL Disease control by conditioning No TRM

NRM and REL cumulative incidence estimates (36-month) from a competing risk model, estimated separately for both conditioning regimens.

Rodrigo Martino et al. Blood 2006;108:836-846


MAC versus RIC

Unacceptable!!

Unacceptable

Disease control: is chemotherapy always useful?

.

Fenaux, Br J Haematol. 1991

La réponse à la chimiothérapie des maladies présentant:

- peu de blastes médullaires - et ceux présentant des anomalie

cytogénétiques, Est très mauvaise!


Transplantation, 2015

Sorror et al, JCO 2014

NRM: impact of the age?

Comparisons of outcome stratifications by the hematopoietic cell transplantation–comorbidity index (HCT-CI) and the composite comorbidity/age index (HCT-CI/age).

Mohamed L. Sorror et al. JCO 2014;32:3249-3256

©2014 by American Society of Clinical Oncology

•  In patients with MDS candidate for allo-SCT:

–  Disease characteristics •  Cytogenetic (High risk) •  Marrow blasts ( <10%)

–  Composite HCT-CI/age •  0 •  1-2 •  3-4 •  ≥ 5

Conditioning: How to make the decision?

0

5

HC

T-C

I/age

Disease characteristics

Likely to respond Unlikely to respond

MAC

RIC

NMA

< high risk cytogenetic High risk cytogenetic

M blasts ≥ 10% Or in CR

M blasts < 10% M blasts ≥ 10% M blasts < 10%

HCT-CI/age 0 MAC RIC Sequential RIC/NMA

HCT-CI/age 1-2 MAC RIC RIC/NMA NMA

HCT-CI/age 3-4 RIC NMA NMA

HCT-CI/age ≥ 5 NMA NMA

Decision making table








Likely to respond to chemotherapy and in good or acceptable conditions









Unlikely to respond to chemotherapy and/or in bad conditions












Unlikely to respond to chemotherapy and/or in bad conditions

Unlikely to respond to chemotherapy and in bad conditions

Quel source de

CSH?

Indication, quelle

maladie? Quand?


Quel donneur?


entretien?

G


n engl j med 367;16 nejm.org october 18, 2012 1487

The new england journal of medicineestablished in 1812 october 18, 2012 vol. 367 no. 16

Peripheral-Blood Stem Cells versus Bone Marrow from Unrelated Donors

Claudio Anasetti, M.D., Brent R. Logan, Ph.D., Stephanie J. Lee, M.D., M.P.H., Edmund K. Waller, M.D., Ph.D., Daniel J. Weisdorf, M.D., John R. Wingard, M.D., Corey S. Cutler, M.D., M.P.H., Peter Westervelt, M.D., Ph.D.,

Ann Woolfrey, M.D., Stephen Couban, M.D., Gerhard Ehninger, M.D., Laura Johnston, M.D., Richard T. Maziarz, M.D., Michael A. Pulsipher, M.D., David L. Porter, M.D., Shin Mineishi, M.D., John M. McCarty, M.D., Shakila P. Khan, M.D.,

Paolo Anderlini, M.D., William I. Bensinger, M.D., Susan F. Leitman, M.D., Scott D. Rowley, M.D., Christopher Bredeson, M.D., Shelly L. Carter, Sc.D., Mary M. Horowitz, M.D., and Dennis L. Confer, M.D.,

for the Blood and Marrow Transplant Clinical Trials Network*

A BS TR AC T

From H. Lee Moffitt Cancer Center and Re-search Institute, Tampa (C.A.), and Shands Cancer Center, University of Florida, Gainesville ( J.R.W.) — both in Florida; Medical College of Wisconsin, Milwaukee (B.R.L., M.M.H.); Fred Hutchinson Cancer Research Center, Seattle (S.J.L., A.W., W.I.B.); Winship Cancer Institute, Emory Univer-sity, Atlanta (E.K.W.); University of Minne-sota Medical Center, Minneapolis (D.J.W.), Mayo Clinic, Rochester (S.P.K.), and the National Marrow Donor Program, Minne-apolis (D.L.C.) — all in Minnesota; Dana–Farber Cancer Institute, Boston (C.S.C.); Siteman Cancer Center, Washington Uni-versity, St. Louis (P.W.); Dalhousie Univer-sity, Halifax, NS (S.C.), and Ottawa Hos-pital, Ottawa (C.B.) — both in Canada; University Hospital Dresden, Dresden, Ger-many (G.E.); Stanford University Medical Center, Stanford, CA (L.J.); Oregon Health and Science University, Portland (R.T.M.); Primary Children’s Medical Center, Univer-sity of Utah, Salt Lake City (M.A.P.); Univer-sity of Pennsylvania, Philadelphia (D.L.P.); University of Michigan, Ann Arbor (S.M.); Massey Cancer Center, Virginia Common-wealth University, Richmond (J.M.M.); University of Texas M.D. Anderson Cancer Center, Houston (P.A.); National Institutes of Health, Bethesda (S.F.L.), and EMMES Corporation, Rockville (S.L.C.) — both in Maryland; and John Theurer Cancer Cen-ter, Hackensack University, Hackensack, NJ (S.D.R.). Address reprint requests to Dr. Anasetti at the Moffitt Cancer Center, 12902 Magnolia Dr., Tampa, FL 33612, or at [email protected].* Additional members of the Blood and

Marrow Transplant Clinical Trials Network who contributed to this study are listed in the Supplementary Appendix, available at NEJM.org.

N!Engl!J!Med!2012;367:1487-96.DOI:!10.1056/NEJMoa1203517Copyright © 2012 Massachusetts Medical Society.

BACKGROUNDRandomized trials have shown that the transplantation of filgrastim-mobilized peripheral-blood stem cells from HLA-identical siblings accelerates engraftment but increases the risks of acute and chronic graft-versus-host disease (GVHD), as com-pared with the transplantation of bone marrow. Some studies have also shown that peripheral-blood stem cells are associated with a decreased rate of relapse and im-proved survival among recipients with high-risk leukemia.METHODSWe conducted a phase 3, multicenter, randomized trial of transplantation of periph-eral-blood stem cells versus bone marrow from unrelated donors to compare 2-year survival probabilities with the use of an intention-to-treat analysis. Between March 2004 and September 2009, we enrolled 551 patients at 48 centers. Patients were randomly assigned in a 1:1 ratio to peripheral-blood stem-cell or bone marrow trans-plantation, stratified according to transplantation center and disease risk. The me-dian follow-up of surviving patients was 36 months (interquartile range, 30 to 37).RESULTSThe overall survival rate at 2 years in the peripheral-blood group was 51% (95% confidence interval [CI], 45 to 57), as compared with 46% (95% CI, 40 to 52) in the bone marrow group (P = 0.29), with an absolute difference of 5 percentage points (95% CI, !3 to 14). The overall incidence of graft failure in the peripheral-blood group was 3% (95% CI, 1 to 5), versus 9% (95% CI, 6 to 13) in the bone marrow group (P = 0.002). The incidence of chronic GVHD at 2 years in the peripheral-blood group was 53% (95% CI, 45 to 61), as compared with 41% (95% CI, 34 to 48) in the bone marrow group (P = 0.01). There were no significant between-group differences in the incidence of acute GVHD or relapse.CONCLUSIONSWe did not detect significant survival differences between peripheral-blood stem-cell and bone marrow transplantation from unrelated donors. Exploratory analyses of secondary end points indicated that peripheral-blood stem cells may reduce the risk of graft failure, whereas bone marrow may reduce the risk of chronic GVHD. (Funded by the National Heart, Lung, and Blood Institute–National Cancer Institute and others; ClinicalTrials.gov number, NCT00075816.)

The New England Journal of Medicine Downloaded from nejm.org at INSERM DISC DOC on November 28, 2014. For personal use only. No other uses without permission.

Copyright © 2012 Massachusetts Medical Society. All rights reserved.

Peripher al-Blood Stem Cells vs. Bone Marrow


Surv

ival

!(%)

100

8090

7060

40

10

50

20

00 1 2 3

Years!since!Transplantation

A

P=0.33

Peripheral blood

Bone marrow

Dis

ease

-free

!Sur

viva

l!(%

)

100

8090

7060

4030 30

10

50

20

00 1 2 3


B

P=0.38

Peripheral blood

Bone marrow

Cum

ulat

ive!

Inci

denc

e!of

!Dea

thU

nrel

ated

!to!R

elap

se!(%

)

100

8090

7060

4030

10

50

20

00 1 2 3


C

P=0.99

Peripheral blood

Bone marrow

Cum

ulat

ive!

Inci

denc

e!of

!Rel

apse

(%)

100

8090

7060

4030

10

50

20

00 1 2 3


D

P=0.74

Peripheral blood

Bone marrow

Cum

ulat

ive!

Inci

denc

e!of

!Neu

trop

hil

Engr

aftm

ent!(

%)

100

8090

7060

4030

10

50

20

00 14 4228 56

Days!since!Transplantation

E

P<0.001Peripheral blood

Bone marrow

Cum

ulat

ive!

Inci

denc

e!of

!Pla

tele

tEn

graf

tmen

t!(%

)C

umul

ativ

e!In

cide

nce!

of!C

hron

icG

VHD

!(%)

100

8090

7060

4030

10

50

20

00 60 120 180


F

P<0.001 Peripheral blood

Bone marrow

Cum

ulat

ive!

Inci

denc

e!of

!Acu

teG

VHD

!(%)

100

8090

7060

4030

10

50

20

00 60 120 180


G

P=0.77

Peripheral blood

Bone marrow

100

8090

7060

4030

10

50

20

00 12 24

Months!since!Transplantation

H

P=0.01

Peripheral blood

Bone marrow



Anasseti et al, NEJM, 2012

Peripher al-Blood Stem Cells vs. Bone Marrow


risk were conducted with the use of an interac-tion test in the Cox proportional-hazards model. The statistical analyses were performed with the use of SAS software, version 9.2 (SAS Institute).

R ESULT S

CHARACTERISTICS OF THE PATIENTS AND DONORSThe characteristics of the patients and donors are shown in Table 1, along with details of the treat-ments. The bone marrow and peripheral-blood

groups were well balanced with respect to age, sex, Karnofsky performance-status score, diag-nosis, disease risk, positive result on serologic testing for CMV, and race. Patients were treated in 48 transplantation centers in the United States and Canada. Donors were from 54 National Marrow Donor Program–affiliated donor centers in the United States, Canada, and Germany. The propor-tion of donors who were fully matched for HLA-A, B, C, and DRB1 and other donor characteris-tics were similar between the two study groups.

Table!1.!(Continued.)

CharacteristicBone!Marrow

(N!=!278)

Peripheral-Blood!!Stem!Cells!!(N!=!273)

GVHD prophylaxis — no. (%)

Cyclosporine and methotrexate 67 (24) 59 (22)

Tacrolimus and methotrexate 183 (66) 196 (72)

Other 28 (10) 18 (7)

Did not undergo transplantation — no. (%) 14 (5) 11 (4)

Antithymocyte globulin treatment — no./total no. (%)§§ 65/258 (25) 72/255 (28)

No. of donor mismatches at HLA-A, B, C, and DRB1 — no./total no. (%)¶¶

0 200/264 (76) 209/262 (80)

1 55/264 (21) 50/262 (19)

2 7/264 (3) 3/262 (1)

3 2/264 (1) 0/262

CD34+ cell dose per kilogram (!10!6)""

Median 2.75 7.70

Interquartile range 1.94–4.53 5.43–11.28

* The CD34+ cell counts differed significantly between the two groups (P<0.001). There were no other significant between-group differences. GVHD denotes graft-versus-host disease, RAEB refractory anemia with excess of blasts, RARS refractory anemia with ring sideroblasts, and RCMD refractory cytopenia with multilineage dysplasia.

† RAEB-1 is characterized by 5 to 9% blasts in bone marrow, and RAEB-2 by 10 to 19% blasts in bone marrow.‡ High-risk disease includes acute myeloid leukemia in third or subsequent remission or not in remission, acute lympho-

blastic leukemia not in remission, the myelodysplastic syndrome with excess blasts in transformation, chronic myeloid leukemia in blast phase, and chronic myelomonocytic leukemia in any stage.

§ Race was determined by the investigators.¶ Scores on the Karnofsky performance-status scale range from 0 to 100, with higher scores indicating better performance

status." The required minimum doses for this conditioning regimen were cyclophosphamide at a dose of 120 mg per kilogram

of body weight and a fractionated total-body irradiation of 12 Gy.** The required minimum doses for this conditioning regimen were cyclophosphamide at a dose of 120 mg per kilogram

given intravenously and busulfan at a dose of 14 mg per kilogram given orally or 11.2 mg per kilogram given intrave-nously or an average targeted serum concentration greater than 600 ng per milliliter.

†† The required minimum doses for this conditioning regimen were fludarabine at a dose of 120 mg per square meter of body-surface area and busulfan at a dose of 8 mg per kilogram or 250 mg per square meter. The protocol did not specify a minimum dose for antithymocyte globulin.

‡‡ The required minimum doses for this conditioning regimen were fludarabine at a dose of 120 mg per square meter and melphalan at a dose of 140 mg per square meter.

§§ Data on treatment with antithymocyte globulin were not collected for patients who did not undergo transplantation. Data were also missing for several patients who did undergo transplantation.

¶¶ Data on donor HLA mismatch were not collected for patients who did not undergo transplantation."" Data on CD34+ cell dose were missing for 121 patients (44%) in the bone marrow group and 25 (9%) in the peripheral-

blood group.



Anasseti et al, NEJM, 2012

Articles

www.thelancet.com/oncology Vol 10 September 2009 855

Standard graft-versus-host disease prophylaxis with or without anti-T-cell globulin in haematopoietic cell transplantation from matched unrelated donors: a randomised, open-label, multicentre phase 3 trialJürgen Finke, Wolfgang A Bethge, Claudia Schmoor, Hellmut D Ottinger, Matthias Stelljes, Axel R Zander, Liisa Volin, Tapani Ruutu, Dominik A Heim, Rainer Schwerdtfeger, Karin Kolbe, Jiri Mayer, Johan A Maertens, Werner Linkesch, Ernst Holler, Vladimir Koza, Martin Bornhäuser, Hermann Einsele, Hans-Jochem Kolb, Hartmut Bertz, Matthias Egger, Olga Grishina, Gérard Socié, for the ATG-Fresenius Trial Group*

SummaryBackground Graft-versus-host disease (GVHD) is a major cause of morbidity and mortality after allogeneic haematopoietic cell transplantation from unrelated donors. Anti-T-cell globulins (ATGs) might lower the incidence of GVHD. We did a prospective, randomised, multicentre, open-label, phase 3 trial to compare standard GVHD prophylaxis with ciclosporin and methotrexate with or without anti-Jurkat ATG-Fresenius (ATG-F).

Methods Between May 26, 2003, and Feb 8, 2007, 202 patients with haematological malignancies were centrally randomly assigned using computer-generated centre-stratifi ed block randomisation between treatment groups receiving ciclosporin and methotrexate with or without additional ATG-F. One patient in the ATG-F group did not undergo transplantation, thus 201 patients who underwent transplantation with peripheral blood (n=164; 82%) or bone marrow (n=37; 18%) grafts from unrelated donors after myeloablative conditioning were included in the full analysis set, and were analysed according to their randomly assigned treatment (ATG-F n=103, control n=98). The primary endpoint was severe acute GVHD (aGVHD) grade III–IV or death within 100 days of transplantation. The trial is registered with the numbers DRKS00000002 and NCT00655343.

Findings!The number of patients in the ATG-F group who had severe aGVHD grade III–IV or who died within 100 days of transplantation was 12 and 10 (21·4%, 95% CI 13·4–29·3), respectively, compared with 24 and nine (33·7%, 24·3–43·0) patients, respectively, in the control group (adjusted odds ratio 0·59, 95% CI 0·30–1·17; p=0·13). The cumulative incidence of aGVHD grade III–IV was 11·7% (95% CI 6·8–19·8) in the ATG-F group versus 24·5% (17·3–34·7) in the control group (adjusted hazard ratio [HR] 0·50, 95% CI 0·25–1·01; p=0·054), and cumulative incidence of aGVHD grade II–IV was 33·0% (n=34; 95% CI 25·1–43·5) in the ATG-F group versus 51·0% (n=50; 95% CI 42·0–61·9) in the control group (adjusted HR 0·56, 0·36–0·87; p=0·011). The 2-year cumulative incidence of extensive chronic GVHD was 12·2% (n=11; 95% CI 7·0–21·3) versus 42·6% (n=34; 95% CI 33·0–55·0; adjusted HR 0·22, 0·11–0·43; p<0·0001). There were no di! erences between treatment groups with regard to relapse, non-relapse mortality, overall survival, and mortality from infectious causes.

Interpretation!The addition of ATG-F to GVHD prophylaxis with ciclosporin and methotrexate resulted in decreased incidence of acute and chronic GVHD without an increase in relapse or non-relapse mortality, and without compromising overall survival. The use of ATG-F is safe for patients who are going to receive a haematopoietic cell transplantation from matched unrelated donors.

Funding!Fresenius Biotech GmbH.

IntroductionGraft-versus-host disease (GVHD) is a major cause of morbidity and mortality after allogeneic haematopoietic cell transplantation. Haematopoietic cell transplantation from unrelated donors results in an increased risk of severe GVHD compared with matched sibling donor transplants using the same regimen for GVHD prophylaxis, which usually contains ciclosporin and short-course methotrexate.1–3 Strategies using intensifi ed GVHD prophy laxis including T-cell depletion did not result in better outcome due to increased risks of infection and relapse.4,5 Di! erent types of anti-T-cell globulin (ATG)

preparations have been tested as part of conditioning regimens to achieve in-vivo T-cell depletion to prevent GVHD. ATG preparations can di! er substantially in potency and dose due to di! erent production methods, namely the cells used for immunisation of the source animal. In a sequential randomised trial testing two doses of ATG (Thymoglobulin; Genzyme, Boston, MA, USA) in 109 patients, the lower dose of ATG did not lead to a di! erence in incidence and severity of acute GVHD compared with standard prophylaxis, whereas the higher dose reduced severe GVHD but was accompanied by a higher risk of lethal infections.6

Lancet Oncol 2009; 10: 855–64

Published Online August 19, 2009DOI:10.1016/S1470-2045(09)70225-6

See Refl ection and Reactionpage 839

*For full list of investigators see webappendix

Department of Hematology and Oncology, Universitätsklinikum Freiburg, Germany (Prof J Finke MD, Prof H Bertz MD, M Egger MD); Department of Hematology and Oncology, Universitätsklinikum Tübingen, Germany (W A Bethge MD); Clinical Trials Center Universitätsklinikum Freiburg, Germany (C Schmoor PhD, O Grishina MD); Klinik und Poliklinik für KMT, Universitätsklinikum Essen, Germany (H D Ottinger MD); Department of Hematology and Oncology, Universitätsklinik Münster, Germany (M Stelljes MD); Department of Internal Medicine, Universitätsklinikum Hamburg-Eppendorf, Hamburg, Germany (Prof A R Zander MD); Department of Medicine, Helsinki University Central Hospital, Finland (L Volin MD, T Ruutu MD); Department of Hematology, Universitätsspital Basel, Switzerland (D A Heim MD); Centre of Blood and Bone Marrow Transplantation, Stiftung Deutsche Klinik für Diagnostik, Wiesbaden, Germany (Prof R Schwerdtfeger MD); Department of Internal Medicine, Universitätsmedizin der Johannes-Gutenberg-Universität Mainz, Germany (K Kolbe MD); Department of Internal Medicine, Hematooncology, University

Articles


Standard graft-versus-host disease prophylaxis with or without anti-T-cell globulin in haematopoietic cell transplantation from matched unrelated donors: a randomised, open-label, multicentre phase 3 trialJürgen Finke, Wolfgang A Bethge, Claudia Schmoor, Hellmut D Ottinger, Matthias Stelljes, Axel R Zander, Liisa Volin, Tapani Ruutu, Dominik A Heim, Rainer Schwerdtfeger, Karin Kolbe, Jiri Mayer, Johan A Maertens, Werner Linkesch, Ernst Holler, Vladimir Koza, Martin Bornhäuser, Hermann Einsele, Hans-Jochem Kolb, Hartmut Bertz, Matthias Egger, Olga Grishina, Gérard Socié, for the ATG-Fresenius Trial Group*

SummaryBackground Graft-versus-host disease (GVHD) is a major cause of morbidity and mortality after allogeneic haematopoietic cell transplantation from unrelated donors. Anti-T-cell globulins (ATGs) might lower the incidence of GVHD. We did a prospective, randomised, multicentre, open-label, phase 3 trial to compare standard GVHD prophylaxis with ciclosporin and methotrexate with or without anti-Jurkat ATG-Fresenius (ATG-F).

Methods Between May 26, 2003, and Feb 8, 2007, 202 patients with haematological malignancies were centrally randomly assigned using computer-generated centre-stratifi ed block randomisation between treatment groups receiving ciclosporin and methotrexate with or without additional ATG-F. One patient in the ATG-F group did not undergo transplantation, thus 201 patients who underwent transplantation with peripheral blood (n=164; 82%) or bone marrow (n=37; 18%) grafts from unrelated donors after myeloablative conditioning were included in the full analysis set, and were analysed according to their randomly assigned treatment (ATG-F n=103, control n=98). The primary endpoint was severe acute GVHD (aGVHD) grade III–IV or death within 100 days of transplantation. The trial is registered with the numbers DRKS00000002 and NCT00655343.

Findings!The number of patients in the ATG-F group who had severe aGVHD grade III–IV or who died within 100 days of transplantation was 12 and 10 (21·4%, 95% CI 13·4–29·3), respectively, compared with 24 and nine (33·7%, 24·3–43·0) patients, respectively, in the control group (adjusted odds ratio 0·59, 95% CI 0·30–1·17; p=0·13). The cumulative incidence of aGVHD grade III–IV was 11·7% (95% CI 6·8–19·8) in the ATG-F group versus 24·5% (17·3–34·7) in the control group (adjusted hazard ratio [HR] 0·50, 95% CI 0·25–1·01; p=0·054), and cumulative incidence of aGVHD grade II–IV was 33·0% (n=34; 95% CI 25·1–43·5) in the ATG-F group versus 51·0% (n=50; 95% CI 42·0–61·9) in the control group (adjusted HR 0·56, 0·36–0·87; p=0·011). The 2-year cumulative incidence of extensive chronic GVHD was 12·2% (n=11; 95% CI 7·0–21·3) versus 42·6% (n=34; 95% CI 33·0–55·0; adjusted HR 0·22, 0·11–0·43; p<0·0001). There were no di! erences between treatment groups with regard to relapse, non-relapse mortality, overall survival, and mortality from infectious causes.

Interpretation!The addition of ATG-F to GVHD prophylaxis with ciclosporin and methotrexate resulted in decreased incidence of acute and chronic GVHD without an increase in relapse or non-relapse mortality, and without compromising overall survival. The use of ATG-F is safe for patients who are going to receive a haematopoietic cell transplantation from matched unrelated donors.

Funding!Fresenius Biotech GmbH.

IntroductionGraft-versus-host disease (GVHD) is a major cause of morbidity and mortality after allogeneic haematopoietic cell transplantation. Haematopoietic cell transplantation from unrelated donors results in an increased risk of severe GVHD compared with matched sibling donor transplants using the same regimen for GVHD prophylaxis, which usually contains ciclosporin and short-course methotrexate.1–3 Strategies using intensifi ed GVHD prophy laxis including T-cell depletion did not result in better outcome due to increased risks of infection and relapse.4,5 Di! erent types of anti-T-cell globulin (ATG)

preparations have been tested as part of conditioning regimens to achieve in-vivo T-cell depletion to prevent GVHD. ATG preparations can di! er substantially in potency and dose due to di! erent production methods, namely the cells used for immunisation of the source animal. In a sequential randomised trial testing two doses of ATG (Thymoglobulin; Genzyme, Boston, MA, USA) in 109 patients, the lower dose of ATG did not lead to a di! erence in incidence and severity of acute GVHD compared with standard prophylaxis, whereas the higher dose reduced severe GVHD but was accompanied by a higher risk of lethal infections.6

Lancet Oncol 2009; 10: 855–64

Published Online August 19, 2009DOI:10.1016/S1470-2045(09)70225-6

See Refl ection and Reactionpage 839

*For full list of investigators see webappendix

Department of Hematology and Oncology, Universitätsklinikum Freiburg, Germany (Prof J Finke MD, Prof H Bertz MD, M Egger MD); Department of Hematology and Oncology, Universitätsklinikum Tübingen, Germany (W A Bethge MD); Clinical Trials Center Universitätsklinikum Freiburg, Germany (C Schmoor PhD, O Grishina MD); Klinik und Poliklinik für KMT, Universitätsklinikum Essen, Germany (H D Ottinger MD); Department of Hematology and Oncology, Universitätsklinik Münster, Germany (M Stelljes MD); Department of Internal Medicine, Universitätsklinikum Hamburg-Eppendorf, Hamburg, Germany (Prof A R Zander MD); Department of Medicine, Helsinki University Central Hospital, Finland (L Volin MD, T Ruutu MD); Department of Hematology, Universitätsspital Basel, Switzerland (D A Heim MD); Centre of Blood and Bone Marrow Transplantation, Stiftung Deutsche Klinik für Diagnostik, Wiesbaden, Germany (Prof R Schwerdtfeger MD); Department of Internal Medicine, Universitätsmedizin der Johannes-Gutenberg-Universität Mainz, Germany (K Kolbe MD); Department of Internal Medicine, Hematooncology, University

Patients: 201 including 164 with PBSC

Articles

860 www.thelancet.com/oncology Vol 10 September 2009

(range 12–38; p<0·0001; see webappendix). Engraftment with platelets of 50!10"/L or greater at day 100 was achieved in 68·0% of patients in the ATG-F group after a median of 37 days (range 1–756), and in 84·7% of patients in the

control group after a median of 20 days (range 10–376; p<0·0001; see webappendix).

At day 100, the proportion of patients who had reached the primary e# cacy endpoint (severe aGVHD grade III–IV or death) was 21·4% (95% CI 13·4–29·3) in the ATG-F group (aGVHD III–IV n=12, death before aGVHD III–IV n=10), versus 33·7% (24·3–43·0) in the control group (aGVHD III–IV n=24, death before aGVHD III–IV n=9). The adjusted odds ratio for risk of aGVHD III–IV or death in the ATG-F group compared with the control group was 0·59 (95% CI 0·30–1·17; p=0·13).

The cumulative incidence of aGVHD grade III–IV in the ATG-F group was 11·7% (95% CI 6·8–19·8), compared with 24·5% in the control group (17·3–34·7); adjusted hazard ratio (HR) 0·50 (95% CI 0·25–1·01; p=0·054; fi gure 2A). The cumulative incidence of aGVHD grade II–IV in the ATG-F group was 33·0% (n=34; 95% CI 25·1–43·5), versus 51·0% (n=50; 42·0–61·9) in the control group (adjusted HR 0·56, 95% CI 0·36–0·87; p=0·011; fi gure 2B), and the cumulative incidence of grade I–IV aGVHD was 56·3% in the ATG-F group (n=58; 95% CI 47·5–66·8) versus 74·5% (n=73; 66·3–83·6; adjusted HR 0·61, 95% CI 0·43–0·88; p=0·007).

Of the 170 patients alive without second transplantation at day 100 (ATG-F n=90, control n=80), cGVHD (limited or extensive) was seen in 74 patients (ATG-F n=27, control n=47). The 2-year cumulative incidence of cGVHD in the ATG-F group was 30·8% (95% CI 22·4–42·3), versus 58·8% (48·9–70·6) in the control group (fi gure 2C; adjusted HR 0·34, 95% CI 0·21–0·55; p<0·0001).

Extensive cGVHD was seen in 45 patients (ATG-F n=11, control n=34). The 2-year cumulative incidence of extensive cGVHD in the ATG-F group was 12·2% (95% CI 7·0–21·3), versus 42·6% (33·0–55·0) in the control group (fi gure 2D; adjusted HR 0·22, 0·11–0·43; p<0·0001). Analyses of acute type cGVHD are shown in the webappendix, but did not alter the overall results.

There was no di$ erence between the two groups regarding the incidence of non-relapse mortality in the fi rst 100 days after transplantation (ATG-F n=11, 10·7%; control n=13, 13·3%). During the total follow-up period, non-relapse mortality was noted in 49 patients (ATG-F n=20; control n=29). The 2-year cumulative incidence of non-relapse mortality in the ATG-F group was 19·6% (95% CI 13·2–29·0), versus 28·9% (20·8–40·1) in the control group (fi gure 3A; adjusted HR 0·68, 0·38–1·22; p=0·20). Relapse occurred in 56 patients (ATG-F n=31; control n=25). The 2-year cumulative incidence of relapse in the ATG-F group was 28·9% (21·2–39·3), versus 23·6% (16·5–33·7) in the control group (fi gure 3B; adjusted HR 1·18, 0·69–2·03; p=0·55). The 2-year disease-free survival rate in the ATG-F group was 51·6% (95% CI 41·8–61·4) versus 47·5% (37·1–57·9) in the control group (fi gure 3C; adjusted HR 0·91, 0·62–1·36, p=0·65).

12 patients (11·7%) in the ATG-F group died within 100 days of transplantation (of relapse n=1, non-relapse mortality n=11). The death rate in the control group was

0

10398

7565

6155

4939

3420

1811

156

62

20

10

00

Number at riskATG-F

Control

10398

7565

6155

4939

3420

1811

156

62

20

10

00

Number at riskATG-F

Control

10398

7565

6155

4939

3420

1811

156

62

20

10

00

Number at riskATG-F

Control

10398

8371

6959

5640

3920

2113

187

73

21

10

00

Number at riskATG-F

Control

0·25

0·50

0·75

1·00

A

0

0·25

0·50

0·75

1·00

B

Incid

ence

of n

on-re

laps

e m

orta

lity

Incid

ence

of r

elap

seDi

seas

e-fre

e sur

viva

lOv

eral

l sur

viva

l

0

0·25

0·50

0·75

1·00

C

0

0·25

0·50

0·75

1·00

D

0 6 12 18 24 30Months post transplantation

36 42 48 54 60

ATG-F/crudeControl/crudeATG-F/adjustedControl/adjusted

Figure !: Rates of non-relapse mortality, relapse, disease-free survival, and overall survival, by treatment groupAdjusted rates are adjusted for disease status and stem-cell source.

Finke et al, Lancet Oncol, 2009

Articles


sensitivity analysis using the package cmprsk in R, version 2.8.1, for fi tting the Fine and Gray model.13 As the results were similar, only the comparisons of the event-specifi c hazard functions are presented.

In the Cox regression and the Fine and Gray models, the disease status (early vs advanced) and stem-cell source (bone marrow vs peripheral blood) were included for adjustment of the treatment e! ect. As an estimate of e! ect size, the hazard ratio of ATG-F versus control was calculated with 95% CI. The e! ect on the risk of the di! erent events was modelled. As a consequence, a hazard ratio of ATG-F versus control below one refers to a reduction in the risk of the occurrence of the event in question corresponding with a prolongation of the time to the event in question’s occurrence. In addition to the crude rates (Kaplan–Meier and cumulative incidence rates) as estimators of probability of event over time, Cox model-based being adjusted for covariates were calculated.14,15

The incidence of infections and adverse events was calculated as the number of patients who experienced at least one infection or adverse event of a certain category as a percentage of the total number of patients in the safety population. The 95% CI for the di! erence between rates of infections and adverse events was calculated using the normal approximation of the binomial distribution. Treatment groups were judged to be similar for a specifi c adverse event if the corresponding 95% CI for the di! erence between incidences included zero. Otherwise they were judged to be di! erent, although this could not be considered statistically di! erent because no confi rmatory tests were done, and 95% CI are to be regarded as descriptive. All p values of statistical tests performed with regard to secondary endpoints are to be regarded as descriptive.

The study is registered with WHO primary registers, numbers DRKS00000002 and NCT00655343.

Role of the funding sourceThis study was designed, conducted, analysed and interpreted by the corresponding author and the clinical trials centre. The sponsor approved the study protocol, the statistical analysis plan, and the clinical study report. The corresponding author, the senior author, and the authors from the clinical trials centre had full access to all clinical data. All authors had fi nal responsibility for the decision to submit the manuscript for publication.

Results202 patients were recruited from 31 centres. One centre recruited 30 patients, six centres recruited 11–20 patients, six centres recruited six to ten patients, and 18 centres recruited less than six patients. Figure 1 shows the trial profi le. The full analysis set consists of 201 patients (ATG-F n=103, control n=98). Median follow-up for secondary endpoints was 2 years (range 0·42–4·53). Patient, disease, and transplant characteristics are summarised in table 1.

Engraftment with neutrophils at 1·0"10#/L or greater at day 100 was achieved in 97·1% of patients in the ATG-F group after a median of 26 days (range 3–63), and in 94·9% of patients in the control group after a median of 19 days

0

10398

10295

9687

9278

8973

8971

8669

8568

8367

8166

8165

Number at riskATG-F

Control

0·25

0·50

0·75

1·00

A

0

10398

10192

8974

7960

7052

6950

6548

6446

6244

6044

5942

Number at riskATG-F

Control

0·25

0·50

0·75

1·00

B

0

9080

6338

4214

349

194

101

91

10

00

Number at riskATG-F

Control

0·25

0·50

0·75

1·00

C

0

9080

0 6 12 18 24 30Months post transplantation

36 42 48

0 10 20

Incid

ence

of a

GvHD

III–I

VIn

ciden

ce o

f aGv

HD II

–IVIn

ciden

ce o

f cGv

HD (l

im/e

xt)

Incid

ence

of c

GvHD

(ext

ensiv

e)

30 40 50Days post transplantation

60 70 80 90 100

7043

5426

4317

288

144

131

40

20

Number at riskATG-F

Control

0·25

0·50

0·75

1·00

D

ATG-F/crudeControl/crudeATG-F/adjustedControl/adjusted

Figure !: Rates of acute and chronic graft-versus-host disease (aGvHD and cGvHD) by treatment grouplim/ext=limited/extensive. Adjusted rates are adjusted for disease status and stem-cell source.

ORIGINAL ARTICLE

Antithymocyte globulins and chronic graft-vs-host disease after myeloablative allogeneicstem cell transplantation from HLA-matched unrelated donors: a report from the SocieteFrancaise de Greffe de Moelle et de Therapie Cellulaire

M Mohty1, M Labopin2, ML Balere3, G Socie4, N Milpied1, R Tabrizi5, N Ifrah6, Y Hicheri7, N Dhedin8, M Michallet9, A Buzyn10,J-Y Cahn11, J-H Bourhis12, D Blaise13, C Raffoux3, H Esperou3,4 and I Yakoub-Agha14

1Hematology Department, Centre Hospitalier Universitaire Hotel-Dieu, Universite de Nantes, CRCNA INSERM U892 and Centred’investigation Clinique (CI2C), Nantes, France; 2Hopital Saint-Antoine, Assistance Publique des Hopitaux de Paris and UPMCUniversite Paris 06, UMR-S 938, Paris, France; 3Agence de Biomedecine, France Greffe de Moelle, Paris, France; 4Servicede greffe de moelle osseuse, Hopital Saint-Louis, Paris, France; 5Hematology Department, CHU Haut-Leveque, Pessac, France;6Hematology Department, CHU d’Angers, Angers, France; 7Hematology Department, Hopital Henri Mondor, Creteil, France;8Hematology Department, Hopital de la Pitie –Salpetriere, Paris, France; 9Hematology Department, CHU de Lyon, Lyon, France;10Hematology Department, Hopital Necker-Enfants maladies, Paris, France; 11Hematology Department, CHU de Grenoble andTIMC-TheREx (UMR 5525 CNRS-UJF), Grenoble, France; 12Hematology Department, Institut Gustave-Roussy, Villejuif, France;13Unite de Transplantation et de Therapie cellulaire (U2T), Universite de la Mediterranee, INSERM UMR 891, InstitutPaoli-Calmettes, Marseille, France and 14Hematology Department, CHU de Lille, Lille, France

This retrospective report assessed the impact of rabbitantithymocyte globulins (ATG), incorporated within a standardmyeloablative conditioning regimen prior to allogeneic stemcell transplantation (allo-SCT) using human leukocyte antigen-matched unrelated donors (HLA-MUD), on the incidence ofacute and chronic graft-vs-host disease (GVHD). In this seriesof leukemia patients, 120 patients (70%) did not receive ATG(‘no-ATG’ group), whereas 51 patients received ATG (‘ATG’group). With a median follow-up of 30.3 months, the cumulativeincidence of grade 3–4 acute GVHD was 36% in the no-ATGgroup and 20% in the ATG group (P!0.11). The cumulativeincidence of extensive chronic GVHD was significantly lower inthe ATG group as compared to the no-ATG group (4 vs 32%,respectively; P!0.0017). In multivariate analysis, the absenceof use of ATG was the strongest parameter associated with anincreased risk of extensive chronic GVHD (relative risk)!7.14,95% CI: 1.7–33.3, P! 0.008). At 2 years, the probability ofnonrelapse mortality, relapse, overall and leukemia-free survi-vals was not significantly different between the no-ATG andATG groups. We conclude that the addition of ATG to GVHDprophylaxis resulted in decreased incidence of extensivechronic GVHD without an increase in relapse or nonrelapsemortality, and without compromising survival after myeloabla-tive allo-SCT from HLA-MUD.Leukemia (2010) 24, 1867–1874; doi:10.1038/leu.2010.200;published online 30 September 2010Keywords: ATG; GVHD; unrelated donors; allogeneic stem celltransplantation

Introduction

Allogeneic hematopoietic stem cell transplantation (allo-SCT) isan effective treatment modality for a number of hematologicalmalignancies that are resistant to standard chemotherapy.1

Allo-SCT is presently considered the only treatment modality

with curative potential for different disease categories. Unfortu-nately, more than 65% of the patients who could benefit fromallo-SCT do not have a human leukocyte antigen (HLA)-matched sibling. The lack of suitable HLA-matched relateddonors has led to the use of alternative donors, such as HLA-matched unrelated donors (HLA-MUD).2–4 However, despitesignificant improvements in HLA matching techniques,5,6

allo-SCT from MUD is still limited by the immunologicalrecognition and destruction of host tissues, termed graft-vs-hostdisease (GVHD). Both in its acute and chronic forms, GVHDcontinues to be the major source of morbidity and mortalityfollowing allo-SCT, and the risk of severe GVHD is welladmitted to increase with the level of HLA mismatches betweenrecipient and donor.7,8 In one study, the rate of clinicallysignificant acute GVHD increased from 29% for recipients ofHLA genotypically identical sibling grafts to 63% for recipientsof one antigen incompatible unrelated donor grafts.9 In thestandard myeloablative conditioning (MAC) allo-SCT settingusing HLA-matched related or unrelated donors, the classicalprophylaxis of GVHD consists of administration of a calcineurininhibitor (cyclosporine A or tacrolimus) combined withlow-dose methotrexate. This regimen results in approximately35 and 45% incidence of grade 2–4 acute and chronic GVHDafter allo-SCT using matched-related donors,10,11 whereas theseincidences are 60 and 70% in the MUD transplant setting.12

Attempts are therefore needed to improve GVHD prophylaxis inthe MUD allo-SCT setting.13 The aim of this multicenterretrospective report was to assess the impact of rabbitantithymocyte globulins (ATG), incorporated within a standardMAC regimen prior to allo-SCT using HLA-MUD, on theincidence and severity of both acute and chronic GVHD.

Materials and methods

Study design, inclusion criteria, clinical evaluation anddata collectionThis was a multicenter, retrospective, registry-based study thatincluded patients with acute leukemia and myelodysplasticsyndromes (MDS) who received allo-SCT from HLA-MUD andreported to the registry of the French Society of Bone Marrow

Received 13 May 2010; revised 3 July 2010; accepted 30 July 2010;published online 30 September 2010

Correspondence: Professor M Mohty, Hematologie Clinique, CHU deNantes, Universite de Nantes, INSERM UMR U892, Place ARicordeau, F-44093 Nantes Cedex, France.E-mail: [email protected]

Leukemia (2010) 24, 1867–1874& 2010 Macmillan Publishers Limited All rights reserved 0887-6924/10

www.nature.com/leu

ORIGINAL ARTICLE

Antithymocyte globulins and chronic graft-vs-host disease after myeloablative allogeneicstem cell transplantation from HLA-matched unrelated donors: a report from the SocieteFrancaise de Greffe de Moelle et de Therapie Cellulaire

M Mohty1, M Labopin2, ML Balere3, G Socie4, N Milpied1, R Tabrizi5, N Ifrah6, Y Hicheri7, N Dhedin8, M Michallet9, A Buzyn10,J-Y Cahn11, J-H Bourhis12, D Blaise13, C Raffoux3, H Esperou3,4 and I Yakoub-Agha14

1Hematology Department, Centre Hospitalier Universitaire Hotel-Dieu, Universite de Nantes, CRCNA INSERM U892 and Centred’investigation Clinique (CI2C), Nantes, France; 2Hopital Saint-Antoine, Assistance Publique des Hopitaux de Paris and UPMCUniversite Paris 06, UMR-S 938, Paris, France; 3Agence de Biomedecine, France Greffe de Moelle, Paris, France; 4Servicede greffe de moelle osseuse, Hopital Saint-Louis, Paris, France; 5Hematology Department, CHU Haut-Leveque, Pessac, France;6Hematology Department, CHU d’Angers, Angers, France; 7Hematology Department, Hopital Henri Mondor, Creteil, France;8Hematology Department, Hopital de la Pitie –Salpetriere, Paris, France; 9Hematology Department, CHU de Lyon, Lyon, France;10Hematology Department, Hopital Necker-Enfants maladies, Paris, France; 11Hematology Department, CHU de Grenoble andTIMC-TheREx (UMR 5525 CNRS-UJF), Grenoble, France; 12Hematology Department, Institut Gustave-Roussy, Villejuif, France;13Unite de Transplantation et de Therapie cellulaire (U2T), Universite de la Mediterranee, INSERM UMR 891, InstitutPaoli-Calmettes, Marseille, France and 14Hematology Department, CHU de Lille, Lille, France

This retrospective report assessed the impact of rabbitantithymocyte globulins (ATG), incorporated within a standardmyeloablative conditioning regimen prior to allogeneic stemcell transplantation (allo-SCT) using human leukocyte antigen-matched unrelated donors (HLA-MUD), on the incidence ofacute and chronic graft-vs-host disease (GVHD). In this seriesof leukemia patients, 120 patients (70%) did not receive ATG(‘no-ATG’ group), whereas 51 patients received ATG (‘ATG’group). With a median follow-up of 30.3 months, the cumulativeincidence of grade 3–4 acute GVHD was 36% in the no-ATGgroup and 20% in the ATG group (P!0.11). The cumulativeincidence of extensive chronic GVHD was significantly lower inthe ATG group as compared to the no-ATG group (4 vs 32%,respectively; P!0.0017). In multivariate analysis, the absenceof use of ATG was the strongest parameter associated with anincreased risk of extensive chronic GVHD (relative risk)!7.14,95% CI: 1.7–33.3, P! 0.008). At 2 years, the probability ofnonrelapse mortality, relapse, overall and leukemia-free survi-vals was not significantly different between the no-ATG andATG groups. We conclude that the addition of ATG to GVHDprophylaxis resulted in decreased incidence of extensivechronic GVHD without an increase in relapse or nonrelapsemortality, and without compromising survival after myeloabla-tive allo-SCT from HLA-MUD.Leukemia (2010) 24, 1867–1874; doi:10.1038/leu.2010.200;published online 30 September 2010Keywords: ATG; GVHD; unrelated donors; allogeneic stem celltransplantation

Introduction

Allogeneic hematopoietic stem cell transplantation (allo-SCT) isan effective treatment modality for a number of hematologicalmalignancies that are resistant to standard chemotherapy.1

Allo-SCT is presently considered the only treatment modality

with curative potential for different disease categories. Unfortu-nately, more than 65% of the patients who could benefit fromallo-SCT do not have a human leukocyte antigen (HLA)-matched sibling. The lack of suitable HLA-matched relateddonors has led to the use of alternative donors, such as HLA-matched unrelated donors (HLA-MUD).2–4 However, despitesignificant improvements in HLA matching techniques,5,6

allo-SCT from MUD is still limited by the immunologicalrecognition and destruction of host tissues, termed graft-vs-hostdisease (GVHD). Both in its acute and chronic forms, GVHDcontinues to be the major source of morbidity and mortalityfollowing allo-SCT, and the risk of severe GVHD is welladmitted to increase with the level of HLA mismatches betweenrecipient and donor.7,8 In one study, the rate of clinicallysignificant acute GVHD increased from 29% for recipients ofHLA genotypically identical sibling grafts to 63% for recipientsof one antigen incompatible unrelated donor grafts.9 In thestandard myeloablative conditioning (MAC) allo-SCT settingusing HLA-matched related or unrelated donors, the classicalprophylaxis of GVHD consists of administration of a calcineurininhibitor (cyclosporine A or tacrolimus) combined withlow-dose methotrexate. This regimen results in approximately35 and 45% incidence of grade 2–4 acute and chronic GVHDafter allo-SCT using matched-related donors,10,11 whereas theseincidences are 60 and 70% in the MUD transplant setting.12

Attempts are therefore needed to improve GVHD prophylaxis inthe MUD allo-SCT setting.13 The aim of this multicenterretrospective report was to assess the impact of rabbitantithymocyte globulins (ATG), incorporated within a standardMAC regimen prior to allo-SCT using HLA-MUD, on theincidence and severity of both acute and chronic GVHD.

Materials and methods

Study design, inclusion criteria, clinical evaluation anddata collectionThis was a multicenter, retrospective, registry-based study thatincluded patients with acute leukemia and myelodysplasticsyndromes (MDS) who received allo-SCT from HLA-MUD andreported to the registry of the French Society of Bone Marrow

Received 13 May 2010; revised 3 July 2010; accepted 30 July 2010;published online 30 September 2010

Correspondence: Professor M Mohty, Hematologie Clinique, CHU deNantes, Universite de Nantes, INSERM UMR U892, Place ARicordeau, F-44093 Nantes Cedex, France.E-mail: [email protected]

Leukemia (2010) 24, 1867–1874& 2010 Macmillan Publishers Limited All rights reserved 0887-6924/10

www.nature.com/leu

and Gray20 for other outcomes. All P-values are two sided withtype I error rate fixed at 0.05. Statistical analyses were performedwith SPSS (SPSS Inc., Chicago, IL, USA) and R 2.9.0 softwarepackages (R Development Core Team, Vienna, Austria).

Results

Patients and donors characteristicsA total of 171 patients were included in this retrospectiveanalysis. The median age was 33 (range, 15–62) years. Thecohort included 57% male recipients, 35% female donors, 44%AML, 43% ALL, 11% MDS and 2% unclassified leukemias. Thestem cell source was bone marrow in 72.5% of patients,whereas G-CSF-mobilized peripheral blood stem cells wereused in 27.5% of cases. In total, 81% of patients weretransplanted from 10/10 allelic HLA-MUD and 19% from anMUD with at least one allelic difference. In this series, 120patients (70%) did not receive ATG (no-ATG group), whereas 51patients received ATG (ATG group) as part of the MAC regimen.

Days

Cum

ulat

ive

inci

denc

e of

acu

teG

VH

D g

rade

3!4

0 20 40 60 80 1000.0

0.2

0.4

0.6

0.8

1.0a

b

0.0

0.2

0.4

0.6

0.8

1.0

Months

Cum

ulat

ive

inci

denc

e of

ext

ensi

vech

roni

c G

VH

D

0 2 4 6 8 10 12 14 16 18 20 22 24

Figure 1 (a) Cumulative incidence of grade 3–4 acute GVHD (solidline: no-ATG group; dotted line: ATG group; P!0.11); (b) cumulativeincidence of extensive chronic GVHD (solid line: no-ATG group;dotted line: ATG group; P!0.0017).

Table 3 Univariate analysis of risk factors for grade 3–4 acuteGVHD

Patient ageo33 years (median) 37±6%433 years 26±5% 0.10

Donor ageo37 years (median) 23±6%437 years 32±6% 0.17

Patient sexMale 38±5%Female 21±5% 0.03

Donor sexMale 31±5%Female 28±7% 0.98

CMV serologic statusSeronegative donor–recipient pair 36±5%Other 24±6% 0.17

ABOMismatch 32±5%Other 30±7% 0.96

DiagnosisAcute myeloid leukemia 28±6%Acute lymphoblastic leukemia 34±6%Myelodysplastic syndrome 25±11% 0.61

Disease statusStandard risk 30±4%Advanced disease 33±7% 0.93

Stem cell sourceBone marrow 31±7%PBSC 31±4% 0.96

Donor typeMatched unrelated donor (10/10allelic match)

26±4%

Matched unrelated donor (other) 55±10% 0.004

Conditioning regimen categoryCyclophosphamide+TBI 34±4%Chemotherapy without TBI 17±8% 0.08

GVHD prophylaxisCsA+short-course methotrexate 34±4%CsA+other immunosuppressive drugs 17±8% 0.19

Time (months) between diagnosis and allo-SCTo Median 31±5%4 Median 31±6% 0.96

Nucleated cellso Median 30±6%4 Median 30±6% 0.87

Prior stem cell transplantationNo 30±4%Yes 41±14% 0.46

ATG as part of the conditioning regimenNo 36±5%Yes 20±6% 0.11

Abbreviations: allo-SCT, allogeneic stem cell transplantation; ATG,antithymocyte globulins; CMV, cytomegalovirus; CsA, cyclosporine A;GVHD, graft-vs-host disease; PBSC, peripheral blood stem cells; TBI,total body irradiation.

ATG and GVHD after HLA-MUD allo-SCTM Mohty et al

1870

Leukemia

Clinical Research

Antithymocyte Globulin before AllogeneicStem Cell Transplantation for ProgressiveMyelodysplastic Syndrome: A Study fromthe French Society of Bone MarrowTransplantation and Cellular Therapy

Rémy Duléry 1, Mohamad Mohty 2,3,4, Alain Duhamel 5,Marie Robin 6, Yves Beguin 7, Mauricette Michallet 8,Stéphane Vigouroux 9, Bruno Lioure 10, Alice Garnier 11,Jean El Cheikh 12, Claude-Eric Bulabois 13, Anne Huynh 14,Jacques-Olivier Bay 15, Etienne Daguindau 16,Patrice Ceballos 17, Laurence Clément 18, Charles Dauriac 19,Natacha Maillard 20, Faezeh Legrand 21, Jérôme Cornillon 22,Gaëlle Guillerm 23, Sylvie François 24, Simona Lapusan 2,Patrice Chevallier 25, Gandhi Damaj 26,Ibrahim Yakoub-Agha 1,*

1Department of Hematology and Bone Marrow Transplantation, CHRU Lille, Lille, France2Department of Hematology and Bone Marrow Transplantation, Saint Antoine Hospital, Paris, France3Division of Hematology, University Pierre et Marie Curie, Paris, France4 INSERM, UMRs 938, Paris, France5Department of Biostatistics, CRHU Lille, Lille, France6Department of Hematology-Transplantation, Saint-Louis Hospital, Paris, France7Department of Hematology, CHU of Liège and University of Liège, Liège, Belgium8Department of Hematology, University Hospital, Lyon, France9Department of Hematology, University Hospital, Bordeaux, France10Department of Hematology, University Hospital, Strasbourg, France11Department of Hematology, Pitié-Salpêtrière Hospital, Paris, France12Department of Hematology, Institut Paoli-Calmettes, Marseille, France13Department of Hematology, University Hospital, Grenoble, France14Department of Hematology, University Hospital Purpan, Toulouse, France15Department of Hematology, University Hospital, Clermont Ferrand, France16Department of Hematology, University Hospital, Besançon, France17Department of Hematology, University Hospital Montpellier, Montpellier, France18Department of Hematology, University Hospital, Nancy, France19Department of Hematology, University Hospital, Rennes, France20Department of Hematology, University Hospital, Poitiers, France21Department of Hematology, University Hospital, Nice, France22Department of Hematology, Institut de Cancérologie de la Loire, Saint Priest en Jarez, France23Department of Hematology, University Hospital, Brest, France24Department of Hematology, University Hospital, Angers, France25Department of Hematology, University Hospital, Nantes, France26Department of Hematology, University Hospital-CHU Sud, Amiens, France

Article history:Received 2 December 2013Accepted 17 January 2014

Key Words:Conditioning regimenGraft-versus-host diseaseMyelodysplastic syndromeAllogeneic stem celltransplantationAntithymocyte globulin

a b s t r a c tWe investigated the impact of rabbit antithymocyte globulins (ATG) on patient outcomes after allogeneic stemcell transplantation (allo-SCT) for progressive myelodysplastic syndrome (MDS). Of the 242 consecutive pa-tients who underwent allo-SCT for progressive MDS between October 1999 and December 2009, 93 receivedATG (ATG group) at the median dose of 5 mg/kg, whereas 149 patients did not (no-ATG group). Donors weresibling (n ! 153) or HLA-matched unrelated (n ! 89). Patients received blood (n ! 90) or marrow (n ! 152)grafts after either myeloablative (n ! 109) or reduced-intensity (n ! 133) conditioning. Three-year overall andevent-free survival, nonrelapse mortality, relapse, and chronic graft-versus-host disease (GVHD) developmentwere not signi!cantly different between the 2 groups. In contrast, acute grade II to IV GVHD occurred moreoften in the no-ATG group (55% of the patients) than in the ATG group (27%, P < .0001). Similar results wereobserved with acute grade III to IV GVHD (28% and 14% in the no-ATG group and ATG group, respectively;P ! .009). In multivariate analysis, after adjustment with propensity score, the absence of ATG was the

Financial disclosure: See Acknowledgments on page 653.* Correspondence and reprint requests: Ibrahim Yakoub-Agha, MD, PhD,

UAM Allogreffes de CSH, CHRU de Lille, F-59037 Lille CEDEX, France.E-mail address: [email protected] (I. Yakoub-Agha).

1083-8791/$ e see front matter ! 2014 American Society for Blood and Marrow Transplantation.http://dx.doi.org/10.1016/j.bbmt.2014.01.016

Biol Blood Marrow Transplant 20 (2014) 646e654

American Society for BloodASBMTand Marrow Transplantation

strongest parameter associated with an increased risk of acute grade II to IV GVHD (hazard ratio, 2.13; 95%con!dence interval, 1.35 to 3.37; P ! .001]. ATG had no impact on overall and event-free survival or cumulativeincidence of the relapse. In conclusion, the addition of ATG to allo-SCT conditioning did not increase theincidence of relapse of patients with progressive MDS. The incidence of acute GVHD was decreased withoutcompromising outcomes.

! 2014 American Society for Blood and Marrow Transplantation.

INTRODUCTIONAllogeneic hematopoietic stem cell transplantation (allo-

SCT) remains the only potentially curative therapeuticapproach in patients with myelodysplastic syndrome (MDS).Despite the bene!cial effects of allo-SCT, these patients are at asubstantial risk of relapse after transplantation [1-4]. Diseasestatus is amajor factor that in"uences patientoutcomewith anincreased risk of relapse, especially in patients with progres-sive disease [5]. Indeed, disease status at transplantation canbe broken down into 2 categories according to the Interna-tional Working Group 2006 response criteria: (1) responding,for patients with complete, marrow, and partial remission orstable disease with hematological improvement; and (2) pro-gressive disease including refractory, relapsing, progressiveand stable disease without hematological improvement [6].

Although signi!cant improvements in HLA matchingtechniques have been accomplished [7], allo-SCT is stilllimited by the immunological recognition and destruction ofhost tissues, termed graft-versus-host disease (GVHD). Both inits acute and chronic forms, GVHD continues to be the majorsource of morbidity and mortality after allo-SCT [8]. Severeacute GVHD has a poor prognosis, with 5-year overall sur-vival of 25% for grade III and 5% for grade IV disease [9].

One of the strategies developed to reduce the risk ofGVHD is ex vivo T cell depletion of the graft. Although thishas been proven to be very effective to prevent GVHD, this isalso associated with a signi!cant increase in graft failure andrisk of relapse [10,11]. An alternative strategy is to providein vivo T cell depletion in blood and lymphoid tissues usingantithymocyte globulin (ATG), a set of polyclonal antibodiesdirected against a wide range of immune cell epitopes [12].However, the use of ATG, incorporated within the condi-tioning regimen before allo-SCT, is still controversial, espe-cially for patients with progressive disease. Indeed, the riskof GVHD seems to be reduced in various proportions withATG [13-20] but a signi!cant increase of disease relapse hasalso been observed [21,22]. In addition, the impact of ATG onthe incidence of relapse is still unknown in the subgroup ofpatients with progressive disease.

In an attempt to assess the impact on outcomes of rabbitATG incorporated within the conditioning regimen, wereport a multicenter retrospective study of 242 consecutivepatients who underwent an allo-SCT for progressive MDS.

PATIENTS AND METHODSThe study was approved by the French Society of Bone Marrow Trans-

plantation and Cell Therapy board and conducted according to the decla-ration of Helsinki.

Patient SelectionTransplantationmodalities were made as homogenous as possible using

the following inclusion criteria: patients older than 18 years with MDS whowere referred for !rst allo-SCT. The source of stem cells was the bonemarrow or blood from either a sibling or an unrelated donor that was HLA-A,-B, -Cw, -DR, and -DQ identical at allelic level. Patients with chronic myelo-monocytic leukemia and those who received allo-SCT from an HLA-mismatched donor or cord blood or a T celledepleted graft were excluded.Thymoglobuline (Genzyme Corporation, Cambridge, MA) was the only ATGadministered to the patients, as this is the only brand approved in France foruse in allo-SCT.

Participating centers veri!ed the data recorded for each patient in theFrench Bone Marrow Transplantation Registry and provided additional in-formation. Quality of the data and HLA matching were controlled by using acomputerized search for discrepancy errors. Consequently, 461 consecutivepatients who underwent allo-SCT between October 1999 and December2009 in 24 French and Belgian centers were identi!ed. Thirty-six patientswere excluded because their !les lacked data. Because the objective of thisstudy was to investigate the impact of ATG on patients with progressiveMDS, we excluded the 183 patients who responded to pretransplantationtreatment according to InternationalWorking Group 2006 criteria [6]. Of the242 remaining patients, 93 received ATG during conditioning (ATG group)and 149 did not (no-ATG group, n ! 149) (Figure 1).

Patient and Donor Characteristics and Transplantation ModalitiesDisease morphology was classi!ed according to the French-American-

British and World Health Organization (WHO) classi!cations [23,24] andthe International Prognostic Scoring System (IPSS) score was calculated atdiagnosis according to Greenberg et al. [25]. Progression to more advanceddisease between diagnosis and transplantation, responses, and disease sta-tus at transplantation were evaluated according to standard criteria [6,26].

Patient and donor initial characteristics at diagnosis and transplantationare shown in Table 1. The use of ATG was center-based in line with nationalguidelines [27]. The ATG and no-ATG groups were unbalanced in terms ofrecipient age, stem cell source, number of CD34" cells in the graft, condi-tioning, use of total body irradiation, and GVHD prophylaxis. However, therewas no difference between the 2 groups regarding the other patient char-acteristics, including French-American-British/WHO classi!cation, IPSSscore, and cytogenetic risk category at diagnosis. In the ATG group, the drugwas delivered at a dose of < 5 mg/kg (n ! 6), 5 mg/kg (n ! 53, 57%), 7.5 mg/kg (n! 11), or 10 mg/kg (n ! 10). ATG was infused over 1 day (n! 6), 2 days(n ! 45, 48%), 3 days (n ! 19), 4 days (n ! 7), or 5 (n ! 4) days.

STATISTICAL ANALYSESThe analysis was performed on the reference date of April

1, 2011. Overall survival (OS) was de!ned as the time elapsedfrom allo-SCT to death, regardless of the cause of death.Event-free survival (EFS) was de!ned as survival with noevidence of relapse. Relapse was de!ned as the presence ofmore than 5% marrow blasts and/or reappearance of majormyelodysplastic features associated with evidence of

Figure 1. Flow chart for patient selection strategy. *Patients whose !les weremissing data for at least 1 of the following were excluded: initial French-American-British (FAB)/World Health Organization (WHO) category, Interna-tional Prognostic Scoring System (IPSS) score and cytogenetic features atdiagnosis, disease status before transplantation, and use of antithymocyteglobulin (ATG) as part of the conditioning.

R. Duléry et al. / Biol Blood Marrow Transplant 20 (2014) 646e654 647

strongest parameter associated with an increased risk of acute grade II to IV GVHD (hazard ratio, 2.13; 95%con!dence interval, 1.35 to 3.37; P ! .001]. ATG had no impact on overall and event-free survival or cumulativeincidence of the relapse. In conclusion, the addition of ATG to allo-SCT conditioning did not increase theincidence of relapse of patients with progressive MDS. The incidence of acute GVHD was decreased withoutcompromising outcomes.

! 2014 American Society for Blood and Marrow Transplantation.

INTRODUCTIONAllogeneic hematopoietic stem cell transplantation (allo-

SCT) remains the only potentially curative therapeuticapproach in patients with myelodysplastic syndrome (MDS).Despite the bene!cial effects of allo-SCT, these patients are at asubstantial risk of relapse after transplantation [1-4]. Diseasestatus is amajor factor that in"uences patientoutcomewith anincreased risk of relapse, especially in patients with progres-sive disease [5]. Indeed, disease status at transplantation canbe broken down into 2 categories according to the Interna-tional Working Group 2006 response criteria: (1) responding,for patients with complete, marrow, and partial remission orstable disease with hematological improvement; and (2) pro-gressive disease including refractory, relapsing, progressiveand stable disease without hematological improvement [6].

Although signi!cant improvements in HLA matchingtechniques have been accomplished [7], allo-SCT is stilllimited by the immunological recognition and destruction ofhost tissues, termed graft-versus-host disease (GVHD). Both inits acute and chronic forms, GVHD continues to be the majorsource of morbidity and mortality after allo-SCT [8]. Severeacute GVHD has a poor prognosis, with 5-year overall sur-vival of 25% for grade III and 5% for grade IV disease [9].

One of the strategies developed to reduce the risk ofGVHD is ex vivo T cell depletion of the graft. Although thishas been proven to be very effective to prevent GVHD, this isalso associated with a signi!cant increase in graft failure andrisk of relapse [10,11]. An alternative strategy is to providein vivo T cell depletion in blood and lymphoid tissues usingantithymocyte globulin (ATG), a set of polyclonal antibodiesdirected against a wide range of immune cell epitopes [12].However, the use of ATG, incorporated within the condi-tioning regimen before allo-SCT, is still controversial, espe-cially for patients with progressive disease. Indeed, the riskof GVHD seems to be reduced in various proportions withATG [13-20] but a signi!cant increase of disease relapse hasalso been observed [21,22]. In addition, the impact of ATG onthe incidence of relapse is still unknown in the subgroup ofpatients with progressive disease.

In an attempt to assess the impact on outcomes of rabbitATG incorporated within the conditioning regimen, wereport a multicenter retrospective study of 242 consecutivepatients who underwent an allo-SCT for progressive MDS.

PATIENTS AND METHODSThe study was approved by the French Society of Bone Marrow Trans-

plantation and Cell Therapy board and conducted according to the decla-ration of Helsinki.

Patient SelectionTransplantationmodalities were made as homogenous as possible using

the following inclusion criteria: patients older than 18 years with MDS whowere referred for !rst allo-SCT. The source of stem cells was the bonemarrow or blood from either a sibling or an unrelated donor that was HLA-A,-B, -Cw, -DR, and -DQ identical at allelic level. Patients with chronic myelo-monocytic leukemia and those who received allo-SCT from an HLA-mismatched donor or cord blood or a T celledepleted graft were excluded.Thymoglobuline (Genzyme Corporation, Cambridge, MA) was the only ATGadministered to the patients, as this is the only brand approved in France foruse in allo-SCT.

Participating centers veri!ed the data recorded for each patient in theFrench Bone Marrow Transplantation Registry and provided additional in-formation. Quality of the data and HLA matching were controlled by using acomputerized search for discrepancy errors. Consequently, 461 consecutivepatients who underwent allo-SCT between October 1999 and December2009 in 24 French and Belgian centers were identi!ed. Thirty-six patientswere excluded because their !les lacked data. Because the objective of thisstudy was to investigate the impact of ATG on patients with progressiveMDS, we excluded the 183 patients who responded to pretransplantationtreatment according to InternationalWorking Group 2006 criteria [6]. Of the242 remaining patients, 93 received ATG during conditioning (ATG group)and 149 did not (no-ATG group, n ! 149) (Figure 1).

Patient and Donor Characteristics and Transplantation ModalitiesDisease morphology was classi!ed according to the French-American-

British and World Health Organization (WHO) classi!cations [23,24] andthe International Prognostic Scoring System (IPSS) score was calculated atdiagnosis according to Greenberg et al. [25]. Progression to more advanceddisease between diagnosis and transplantation, responses, and disease sta-tus at transplantation were evaluated according to standard criteria [6,26].

Patient and donor initial characteristics at diagnosis and transplantationare shown in Table 1. The use of ATG was center-based in line with nationalguidelines [27]. The ATG and no-ATG groups were unbalanced in terms ofrecipient age, stem cell source, number of CD34" cells in the graft, condi-tioning, use of total body irradiation, and GVHD prophylaxis. However, therewas no difference between the 2 groups regarding the other patient char-acteristics, including French-American-British/WHO classi!cation, IPSSscore, and cytogenetic risk category at diagnosis. In the ATG group, the drugwas delivered at a dose of < 5 mg/kg (n ! 6), 5 mg/kg (n ! 53, 57%), 7.5 mg/kg (n! 11), or 10 mg/kg (n ! 10). ATG was infused over 1 day (n! 6), 2 days(n ! 45, 48%), 3 days (n ! 19), 4 days (n ! 7), or 5 (n ! 4) days.

STATISTICAL ANALYSESThe analysis was performed on the reference date of April

1, 2011. Overall survival (OS) was de!ned as the time elapsedfrom allo-SCT to death, regardless of the cause of death.Event-free survival (EFS) was de!ned as survival with noevidence of relapse. Relapse was de!ned as the presence ofmore than 5% marrow blasts and/or reappearance of majormyelodysplastic features associated with evidence of

Figure 1. Flow chart for patient selection strategy. *Patients whose !les weremissing data for at least 1 of the following were excluded: initial French-American-British (FAB)/World Health Organization (WHO) category, Interna-tional Prognostic Scoring System (IPSS) score and cytogenetic features atdiagnosis, disease status before transplantation, and use of antithymocyteglobulin (ATG) as part of the conditioning.


increasing relapse and NRM, and without compromising OS.Data from Mohty et al. [13] con!rmed these results with ATGThymoglobuline in a retrospective report on 120 patients withacute leukemia and MDS undergoing allo-SCT from matchedunrelated donors after MAC.

The impact of rabbit ATG was also analyzed with RIC in aretrospective multicenter study on 1676 adults undergoingtransplantation for hematologic malignancies. Soiffer et al.[21] reported a higher risk of relapse with ATG compared toTcellereplete regimens (51% and 38%, respectively; P< .001).Although these data are supported by a large series, theirinterpretation may be confounded by the lack of homoge-neity of the underlying diseases (acute and chronic leukemia,MDS, osteomyelo!brosis, and lymphoma) and the condi-tioning regimens. Conversely, in a recent phase 2 prospectivestudy [16], patients with hematologic malignancies under-went allo-SCT from HLA-identical sibling and were ran-domized between 2 different strategies of conditioning:"udarabine was associated with either busulfan and ATG

(n ! 69) or total body irradiation alone (n ! 70). After 5-yearfollow-up, the busulfan-ATG regimen was associated withgreater disease control. However, because of higher NRMrate, this did not translate into better OS. Recently, deMassonet al. [22] assessed the outcomes of 37 cases of advanced-stage primary cutaneous T cell lymphomas treated withallo-SCT in a multicenter retrospective analysis. In multi-variate analysis, the use of rabbit ATG was the only factorassociated with an increased incidence of relapse (HR, 4.8;95% CI, 1.8 to 12.9; P ! .002) and a reduced progression-freesurvival (HR, 2.9; 95% CI, 1.3 to 6.2; P ! .04). This demon-strates that ATG may have a different impact on relapsedepending on the disease treated with allo-SCT.

We, therefore, chose to restrict our study to only 1 diseaseand focused on progressive patients, hypothesizing that ATGbefore transplantation could have a signi!cant impact onoutcome in this setting. Because the outcome of allo-SCTdepends on the degree of donor-recipient HLA matching[31], we only included patients who received allo-SCT from

Figure 2. Kaplan-Meier estimates of (A) 3-year overall survival, (B) 3-year event-free survival, (C) cumulative incidence of 3-year relapse, and (D) grade II to IV acutegraft-versus-host disease (GVHD) in 242 patients, according to the administration of antithymocyte globulin (ATG) in the conditioning. HR indicates hazard ratio;CI, con!dence interval; ns, not signi!cant.

Table 4Bivariate Analysis by Key Subsets in the Sub-Group of Patients Who Progressed to a More Advanced Disease before Transplantation: Three-Year Overall Survival,Event-Free Survival, Relapse, and Nonrelapse Mortality Rates

No. of Patients Overall Survival EFS Relapse NRM

% P* % P* % Py % Py

Antithymocyte globulinNo 43 33 .42 26 .68 37 .65 37 .22Yes 25 46 33 46 21

EFS indicates event-free survival; NRM, nonrelapse mortality.* Log-rank.y Gray (cumulative incidence).


Quel source de

CSH?

Indication, quelle

maladie? Quand?


Quel donneur?


entretien?

G


AZA - DLI

Protocole national de phase II

Quel source de

CSH?

Indication, quelle

maladie? Quand?


Quel donneur?


entretien?

G


Guièze et al, BBMT 2015.

Conclusion • Quelle maladie? IPSS int-2 & élevé

• Quand? • Quel traitement préalable?

Selon caractéristiques de la maladies et possibilité de faire une greffe rapidement

• Quels donneurs? géno = phéno 10/10 > mismatch

•  Conditionnement? Selon l’âge et co-morbidités

•  Source de CSH? MO ou CSP avec ATG

•  IPSS •  Tôt : Int-2 & élevé •  Tard: Faible & Int-1

Syndromes Myélodysplasiques (SMD) et Allogreffes de...

Documents

Transcript of Syndromes Myélodysplasiques (SMD) et Allogreffes de...