Decreased Nonrelapse Mortality after Unrelated Cord Blood Transplantation for Acute Myeloid Leukemia Using Reduced-Intensity Conditioning: A Prospective Phase II Multicenter Trial Academic research paper on "Clinical medicine"

AS BMI

American Society for Blood and Marrow Transplantation

Biology of Blood and Marrow Transplantation

journal homepage: www.bbmt.org

Decreased Nonrelapse Mortality after Unrelated Cord Blood Transplantation for Acute Myeloid Leukemia Using Reduced-Intensity Conditioning: A Prospective Phase II Multicenter Trial

Bernard Rio1, Sylvie Chevret2, Stéphane Vigouroux3, Patrice Chevallier4, Sabine Fürst5, Anne Sirvent6, Jacques-Olivier Bay 7, Gérard Socié 8, Patrice Ceballos 9, Anne Huynh10, Jérôme Cornillon11, Sylvie Françoise12, Faezeh Legrand13, Ibrahim Yakoub-Agha14, Gérard Michel15, Natacha Maillard16, Geneviève Margueritte17, Sébastien Maury18, Madalina Uzunov19, Claude Eric Bulabois20, Mauricette Michallet21, Laurence Clement22, Charles Dauriac23, Karin Bilger24, Eliane Gluckman25, Annalisa Ruggeri25,26, Agnès Buzyn27, Stéphanie Nguyen28, Tabassome Simon29, Nöel Milpied3, Vanderson Rocha25,30,*, on behalf of Société Française de Greffe de Moelle et de Thérapie Cellulaire and Eurocord

1 Service d'Hématologie, Hôtel-Dieu Assistance Publique-Hôpitaux de Paris, Paris, France

2 Department de Bioinformatique et Statistique Médicale, Hopital Saint-Louis, Paris, France

3 Service d'Hématologie, Service d'Hématologie clinique et de thérapie cellulaire, CHU de Bordeaux Hopital du Haut-Lévèque, Pessac, France 4Service d'Hématologie, Hematology Department, CHU de Nantes, Nantes, France

5 Service d'Hématologie, Service de Greffe de Moelle, Institut Paoli Calmettes, Marseille, France 6Service d'Hématologie, Hematologie Clinique, Hopital de l'Archet I, Nice, France 7 Service d'Hématologie, Service d'Hématologie Clinique, CHU Estaing, Clermont-Ferrand, France 8Service d'Hématologie, Hematology-Bone Marrow Transplantation, Hopital Saint-Louis, Paris, France

9 Service d'Hématologie, Hôpital Lapeyronie, Montpellier, France

10 Service d'Hématologie, Hématologie Clinique, CHU, Toulouse, France

11 Service d'Hématologie, Hematology, Institut de Cancérologie de la Loire, Loire, France

12 Service d'Hématologie, CHU Angers, Angers, France

13 Service d'Hématologie, Hematology, CHU Besançon, Besancon, France

14 Service d'Hématologie, Hematology, CHU de LILLE, Lille, France

15 Department of Hematology, Hôpital La Timone, Marseille, France

16 Service d'Hématologie, CHU Poitiers, Poitiers, France

17 Service d'Hématologie, CHU Montpellier Pédiatrie, Montellier, France

18 Service d'Hématologie, Service d'Hematologie, Hôpital Henri Mondor, Creteil, France

19 Service d'Hématologie, Pitié-Salpêtrière (AP-HP), Paris, France 20Service d'Hématologie, CHU Grenoble, Grenoble, France

21 Service d'Hématologie, Hôpital Edouard Herriot, Lyon, France 22Service d'Hématologie, CHU Nancy, Nancy, France

23 Service d'Hématologie, CHU Rennes, Rennes, France

24 Service d'Hématologie, CHRU Strasbourg, Strasbourg, France

25Service d'Hématologie, Eurocord Office, Hôpital Saint-Louis, Paris, France

26Service d'Hématologie, Hématologie Clinique et Thérapie Cellulaire, Hôpital Saint-Antoine, Paris, France 27 Service d'Hématologie, Comité Scientifique, SFGM-TC, Hôpital Necker (AP-HP), Paris, France 28Service d'Hématologie, Hôpital Pitié Salpêtrière (AP-HP), Paris, France 29Service d'Hématologie, URC Est Saint Antoine (AP-HP), Paris, France 30 Department of Haematology, Churchill Hospital, Oxford, United Kingdom

^^^ CrossMark

Financial disclosure: See Acknowledgments on page 452.

* Correspondence and reprint requests: Professor Vanderson Rocha, MD, PhD, Department of Clinical Haematology, Level 2, Cancer and Haematology Centre, Churchill Hospital, Old Road, Headington, Oxford OX3 7LE UK. E-mail address: vanderson.rocha@ouh.nhs.uk (V. Rocha).

http://dx.doi.org/10.1016/j.bbmt.2014.11.009

1083-8791/® 2015 American Society for Blood and Marrow Transplantation.

Article history: Received 5 August 2014 Accepted 7 November 2014

Key Words: Nonrelapse mortality Umbilical cord blood transplantation

Reduced-intensity conditioning Acute myeloid leukemia

ABSTRACT

A prospective phase II multicenter trial was performed with the aim to obtain less than 25% nonrelapse mortality (NRM) after unrelated cord blood transplantation (UCBT) for adults with acute myeloid leukemia (AML) using a reduced-intensity conditioning regimen (RIC) consisting of total body irradiation (2 Gy), cyclophosphamide (50 mg/kg), and fludarabine (200 mg/m2). From 2007 to 2009, 79 UCBT recipients were enrolled. Patients who underwent transplantation in first complete remission (CR1) (n = 48) had a higher frequency of unfavorable cytogenetics and secondary AML and required more induction courses of chemotherapy to achieve CR1 compared with the others. The median infused total nucleated cells (TNC) was 3.4 x 107/kg, 60% received double UCBT, 77% were HLA mismatched (4/6), and 40% had major ABO incompatibility. Cumulative incidence of neutrophil recovery at day 60 was 87% and the cumulative incidence of 100-day acute graft-versus-host disease (II to IV) was 50%. At 2 years, the cumulative incidence of NRM and relapse was 20% and 46%, respectively. In multivariate analysis, major ABO incompatibility (P = .001 ) and TNC (<3.4 x 107/kg; P = .001) were associated with increased NRM, and use of 2 or more induction courses to obtain CR1 was associated with increased relapse incidence (P = .04). Leukemia-free survival (LFS) at 2 years was 35%, and the only factor associated with decreased LFS was secondary AML (P = .04). In conclusion, despite the decreased NRM observed, other RIC regimens with higher myelosuppression should be evaluated to decrease relapse in high-risk AML. (EUDRACT 2006-005901-67).

© 2015 American Society for Blood and Marrow Transplantation.

INTRODUCTION

Allogeneic hematopoietic stem cell transplantation (HSCT) using a high-dose or myeloablative conditioning (MAC) regimen is a curative option for the treatment of acute myeloid leukemia (AML), especially in young adult patients. However, this procedure is not often used for older patients and/or those patients with comorbidities because of the higher mortality rates related to transplantation in these populations. Non-myeloablative and reduced-intensity conditioning (RIC) regimens have been developed to minimize the effects of toxicity and death of allogeneic HSCT, mainly decreasing nonrelapse mortality (NRM) [1]. Nevertheless, preparative regimens with greater antileukemic intensity and with minimal extramedullary toxicity should be better evaluated in patients with AML [2,3].

Many studies have shown that allogeneic RIC, compared with MAC, HSCT in patients with AML decreases NRM, but it is associated with an increased relapse rate and, because of this balance, it does not impact leukemia-free survival (LFS) [4,5].

A MAC regimen before single or double unrelated cord blood transplantation (UCBT) has been frequently used in adult patients with leukemia, with high morbidity and mortality rates (34% to 57%), probably due to delayed hematopoietic recovery [6-10]. Therefore, strategies to decrease mortality and improve engraftment, such as use of RIC, double UCBT, cord blood intrabone infusion, and in vitro expansion [8,11-15] have been sought.

These strategies are under investigation and may minimize the high incidence of NRM related to the delayed neutrophil recovery and toxicity seen after MAC UCBT. The use of single or double UCBT associated with low-dose total body irradiation (TBI) (2 Gy), cyclophosphamide, and fludarabine (TCF) showed the benefits of obtaining a prompt and complete neutrophil recovery and decreased mortality related to transplantation [11,16-18]. However, outcomes after RIC UCBT using TCF have been analyzed for patients with various hematological malignancies who underwent transplantation in remission or advanced phase of the disease [18]; nonetheless, the use of such conditioning regimen for patients with AML needs further investigation.

With the aim of obtaining a cumulative incidence of NRM lower than 25% after RIC UCBT using TCF in AML patients, the Societé Française de Greffe de Moelle Osseuse et Therapie Cellulaire and Eurocord conducted a phase II multicenter prospective trial.

METHODS

Study Design and Patients

The primary objective of this prospective, multicenter phase II trial was to obtain a cumulative incidence of NRM at 2 years lower than 25% using an established conditioning regimen, namely TCF, in AML patients. Criteria of inclusion were the following: patients between 15 and 65 years of age with de novo or secondary AML for whom an HLA-identical related or unrelated hematopoietic stem cell donor, defined as 10/10 or 9/ 10 HLA match (considering an allelic HLA typing of class I and class II), were not available. Patients or their legal guardians gave their informed consent to participate in the study, according to the Declaration of Helsinki. Exclusion criteria included creatinine clearance below 50 mL/ minute, left ventricular ejection fraction below 40%, severe abnormalities of functional respiratory tests, bilirubin level higher than twice the limits of normal range, alanine aminotransferase level more than 4 times the upper limits of normal range, poorly controlled hypertension, and a Karnofsky score <70.

Patients were recruited in 23 centers of the Societé Française de Greffe de Moelle Osseuse et Therapie Cellulaire between October 2007 and August 2009 and were followed for 2 years after inclusion. This study was approved by the internal review board of Hôtel-Dieu Assistance Publique-Hôpitaux de Paris.

Procedures

Conditioning regimen and graft-versus-host disease prophylaxis

The conditioning regimen, consisted of cyclophosphamide 50 mg/kg/ day on day -6, fludarabine 40 mg/m2/day from day-6 to day-2, and TBI at 2 Gy at day-1.

Graft-versus-host disease (GVHD) prophylaxis consisted of cyclosporine with mycophenolate mofetil. Cyclosporine was started at day-3 and maintained until day +100, then it was slowly tapered until 6 months and stopped in absence of GVHD. Mycophenolate mofetil was started at day-3 at 30 mg/kg/day and stopped at day +30.

Supportive care

Lenograstim was started at day +1 at 150 mg/m2/day until achieving an absolute neutrophil count (ANC) >.5 x 109/L for 3 consecutive days. Then, it was given every 2 days until recovery was confirmed. All patients were treated in laminar air flow rooms and received oral antibiotics against gram-positive bacteria (oracillin, amoxicillin, or josamycine), fluconazole 400 mg per day, and valacyclovir or acyclovir for herpes simplex and zoster and cytomegalovirus (CMV) prophylaxis. Fluconazole was stopped when patients had an ANC above 1 x 109/L and were off corticosteroids. Oral antibiotics were continued for 1 year after transplantation in patients without any sign of chronic GVHD or until discontinuation of immuno-suppressive treatment. Prophylaxis against Pneumocystis jirovecii and/or toxoplasmosis (trimethoprim/sulfamethoxazole or sulfadoxine/pyrimeth-amine, respectively) was started after neutrophil recovery and continued for a minimum of 6 months or until discontinuation of immunosuppressive treatment. Patients were tested once each week during the first 100 days after UCBT by real-time PCR for CMV and Epstein-Barr virus, toxoplasmosis, human herpesvirus 6, and adenovirus in case of any clinical sign of viral infection. Assessment for Aspergillus infection using serum galactomanan test was performed once each week in the first 100 days.

Graft Characteristics

The minimum total nucleated cell (TNC) dose required for a single unit UCBT was 3 x 107/kg before freezing and the maximum HLA disparity permitted was 4/6 between patient and units. In case of a double-unit UCBT, there was no requirement for a minimum cell dose of each unit. The total amount of nucleated cells of both units should be more than 3 x 107/kg before freezing. Also, an HLA 4/6 matching was recommended between the units in case of double cord blood transplantation. HLA was defined by low resolution for HLA-A and-B and high resolution for HLA-DRB1; the highest HLA disparity between cord blood and patients was taken into consideration in double cord blood transplantation. ABO was classified as the highest ABO incompatibility between cord blood and patients in case of double UCBT.

Definitions of Endpoints

Overall survival (OS) was calculated from the date of UCBT until death or last observation alive. LFS was calculated from the date of UCBT until relapse or last disease free follow-up. Relapse and death from any cause were considered events. NRM was defined as death without prior relapse. Neutrophil recovery was defined as achieving ANC >.5 x 109/L for 3 consecutive days. Full donor chimerism was defined as >95% leukocytes of donor origin in peripheral blood or marrow samples, measured by different techniques, according to transplantation centers. Autologous reconstitution was defined as >95% leukocytes of recipient origin. Mixed chimerism was defined by the presence of >5% but <95% of leukocytes of donor origin. The diagnosis and grading of acute and chronic GVHD was assigned by the transplantation centers using standard criteria [19,20].

Sample Size and Justification

The primary objective of this multicenter phase II trial was to obtain a cumulative incidence of NRM lower than 25% based on information gathered from previous studies; 1 of them showing an NRM of 45% after UCBT using a MAC regimen for adults with hematological malignancies [8] and another showing an NRM of 25% [16,17] in UCBT recipients given RIC before transplantation. Therefore, with a type I and type II error rate of 5% and the hypothesis of decreasing the NRM from 45% to 25%, from a 2-sided test based on 1 sample, at least 76 patients enrolled in the study was required.

Statistical Analysis

Median values and ranges were used for continuous variables and percentages were used for categorical variables. Patient-, disease-, and transplantation-related variables of the groups (CR1 and others) were compared using chi-square or exact Fisher's test for categorical variables and Mann-Whitney test for continuous variables. The distributions of OS and LFS were calculated using the Kaplan-Meier [21] method and then compared by the log-rank test. The cumulative incidence of neutrophil engraftment, grade II to IV acute and chronic GVHD, relapse, and NRM were calculated within a competing risks setting, where deaths before the event of interest were considered competing risks. The following variables were tested in the univariate analyses: age and weight (introduced as continuous variables) at transplantation, gender, CMV serology before transplantation, comorbidity index, cytogenetic risk, French-American-British classification (FAB), de novo or secondary AML, number of courses of induction to achieve CR, extramedullary leukemia, time intervals between diagnosis and CR1 and between CR1 and UCBT, graft source (single or double), number of nucleated cells infused per kilogram of body weight, number of HLA disparities, and ABO compatibility.

Multivariate analyses were performed using Cox proportional hazards regression model [22] for LFS and OS, and the cause-specific Cox proportional hazards model was used for neutrophil engraftment, grade II to IV acute and chronic GVHD, relapse, and NRM. Variables that reached a P value of .15 in the univariate analyses were then included in the multivariable models, as well as the propensity to have reached CR1. Indeed, given the low sample size, we summarized the predictive information of the CR1 outcome using the linear score derived from the multivariable logistic model, including median number of TNC infused (3.4 x 107/kg), FAB classification, cytogenetics, time from diagnosis to CR1, and comorbidity score. Variable selection was based on a stepwise procedure at the .05 level.

Statistical analyses were performed with SAS 9.3 (SAS Inc, Cary, NC) and R (http://www.R-project.org/) software.

RESULTS

Patient's Characteristics

Characteristics of 79 patients are listed in Table 1. The median age at UCBT was 51 years (range, 15 to 65), median weight was 65 kg (range, 49 to 105), 53% of patients were female, and 49% had positive CMV serology before UCBT. The

comorbidity score [5] was 1 in 16%, 2 in 12%, and 3 or higher in 24%; 48% of the patients had no comorbidities. Cytogenetics was normal in 33 patients (52%); of those 1 of 33 were Fms-like tyrosine kinase 3 internal tandem duplication (FLT3/1TD) positive and abnormal in 48%, including 36% with a complex karyotype and/or abnormality of chromosomes 5, 7,11 and inversion of chromosome 3. Nine (14%) patients had previously undergone transplantation with autologous peripheral blood stem cells.

Forty-eight patients underwent transplantation in CR1, 30 in CR2, and 1 in CR4. Table 1 also lists patients-, disease-, and transplantation-related factors by disease status at transplantation and the statistical differences between patients who underwent transplantation in CR1 and other disease status (CR2 or more).

Patients in CR1 with favorable cytogenetics were not offered transplantation, with the exception of 1 patient who underwent transplantation in CR1 with favorable cytogenetics because CR was achieved only after 3 courses of chemotherapy. Compared with patients who underwent transplantation in CR2 or more, CR1 patients had higher frequency of unfavorable cytogenetics (60%), and among those, 19% had secondary leukemia. Almost 50% of patients who underwent transplantation in CR1 achieved remission after 2 courses of induction treatment. The time between diagnosis and transplantation was 6.6 (range, 5.6 to 8) and 19 months (range, 13.8 to 23) for the first and second remission groups, respectively.

The majority of patients received 2 cord blood units. After thawing the cord blood units, patients who underwent transplantation with a single CBU received a median number of nucleated cells of 2.92 x 107/kg and a median of .92 CD34 x 105/kg. Patients who underwent transplantation with 2 CBU received 3.5 107/kg and 1.1 x 105/kg, respectively. Regarding HLA compatibility, 2% of the units were HLA matched, 21% were 5/6, and 77% were 4/6. ABO major incompatibility was observed in 44% of the patients (in double cord blood, the highest incompatibility was considered).

Outcomes

Neutrophil and platelet recovery

Seventy patients achieved neutrophil counts >500/mL. The cumulative incidence of neutrophil recovery at day +60 was 89%. Median time for recovery was 15 days (range, 4 to 53). Nine patients did not engraft: 2 died in aplasia at day +23 (central nervous system hemorrhage) and day +116 (after a second transplantation), 4 died after autologous reconstitution and relapse, 2 were alive with autologous reconstitution at last follow-up (more than 2 years after transplantation), and 1 was alive after a second transplantation with a mismatched related donor.

1n univariable analyses for neutrophil recovery, younger patients (continuous variable) (hazard ratio [HR], 1.02; 95% confidence interval [95% CI], 1 to 1.03; P=.04), higher cell dose (continuous variable) (HR, 1.28; 95% CI, 1.02 to 1.61 ; P = .03), positive CMV serology before transplantation (HR, 1.77; 95% CI, 1.09 to 2.86; P = .02), and undergoing transplantation with a major incompatibility ABO graft (HR, 1.75; 95% CI, 1.07 to 2.85; P = .026) were associated with greater neutrophil recovery. All factors above were independent risk factors for neutrophil recovery in a multivariable model. There was no statistical difference for patients who underwent transplantation with a single or double UCBT. Forty-nine patients achieved platelet counts >20,000/mL at a median time of 36 days (range, 6 to 77). At day 180, the cumulative incidence of platelet recovery was 62%.

Table 1

Patient, Disease, Transplantation, and Donor-related Factors: Total Number by Disease Status at Transplantation

Characteristics CR1 Patients (n = 48) >CR1 Patients (n = 31) P Value Total (n = 79)

n % or median (95% CI) CR1 n % or median (95% CI) >CR1 n % or median (95% CI)

Age at UCBT 48 51.4 (36.7-58.4) 31 48.2 (35.1-57.0) .35 79 51.0 (35.1-58.0)

Gender

Male 19 39.5% 18 58.1% .17 37 46.8%

Female 29 60.4% 13 41.9% 42 53.2%

CMV serology

Negative 21 43.8% 19 61.3% .17 40 50.6%

Positive 27 56.3% 12 38.7% 39 49.4%

Comorbidity index (Sorror)

0 20 43.5% 17 54.8% .57 37 48.1%

1 7 15.2% 5 16.1% 12 15.6%

>2 19 41.3% 9 29.0% 28 36.4%

NA 2 0 2

Weight

Kg 47 64.0 (56.5-76.0) 31 66.0 (58.5-73.5) .65 78 65 (57.3-74.8)

Previous fungal Infection

Yes 11 22.9% 6 19.4% 17 21.5%

FAB classification

M0,M1,M2 17 35.4% 18 58.1% .003 35 44.3%

M4 9 18.8% 10 32.3% 19 24.1%

Others 22 45.8% 3 9.7% 25 31.7%

Secondary AML*

No 39 81.3% 26 83.9% 1.00 65 82.3%

Yes 9 18.8% 5 16.1% 14 17.7%

Cytogenetic risk

Favorable 1 2.1% 5 16.1% .01 6 7.6%

Intermediate 18 37.5% 17 54.8% 35 44.3%

Unfavorable 29 60.4% 9 29.0% 38 48.1%

Interval from diag. to UCBT

Months 47 6.6 (5.6-8.0) 30 18.9 (13.4-23.2) <.0001 77 8.0 (5.8-16.3)

Interval from last CR to UCBT 48 4.6 (3.5-6,0) 28 3.05 (2.3-4.2) .01 76 4.0 (2.5-5.3)

Interval from diag. to CR1

Days 47 57.0 (41.5-81.5) 27 41.0 (32.0-51.5) .02 74 48.0 (38.0-67.8)

No. of inductions

1 20 51.3% 19 86.4% .02 39 63.9%

2 14 35.9% 3 13.6% 17 27.9%

3 5 12.8% 0 0% 5 8.2%

NA 9 9 18

Extramedullary involvement

No 41 85.4% 30 96.8% .14 71 89.9%

Yes 7 14.6% 1 3.2% 8 10.1%

Year of transplantation

2007 1 2.1% 3 9.7% .05 4 5.1%

2008 19 39.9% 18 58.1% 37 46.8%

2009 28 58.3% 10 32.3% 38 48.1%

Type of graft

Single 15 31.3% 11 35.5% .81 26 32.9%

Double 33 68.8% 20 64.5% 53 67.1%

TNC x 107/kg 48 3.5 (2.8-4.2) 31 3.5 (3.1-3.8) .64 79 3.5 (3.0-4.1)

CD34/kg x 105/kg infused 48 .1 (.07-.14) 31 .1 (.08-.178) .21 79 .1 (.08-.15)

CD3/kg x 106/kg infused 29 5.6 (4.2-7.0) 20 5.3 (3.8-6.3) .89 49 5.6 (3.8-7.0)

CFU-GM x 104/kg infused 31 1.0 (.3-1.6) 20 1.4 (.2-3.8) .45 51 1.0 (.23-2.16)

HLA match

5/6 or 6/6 12 25.0% 6 19.4% .60 18 22.8%

4/6 or 3/6 36 75.0% 25 80.7% 61 77.2%

Matched 15 31.3% 11 35.5% .53 26 32.9%

Minor 13 27.1% 5 16.1% 18 22.8%

Major 20 41.7% 15 48.4% 35 44.3%

NA indicates not available; Diag indicates diagnosis; CFU-GM, colony forming unit-granulocyte and monocyte.

* Secondary AML to myeloproliferative disorder (n = 1); myelodysplastic syndrome (n = 6), and secondary myelodysplastic syndrome (n = 7; breast cancer n = 4, previous autologous transplantation, for lymphoma = 3). Statistically significant P values are shown in bold.

Chimerism data were available for 70 patients who engrafted at days +30 and +60 and months 3 and 6 after transplantation. Full donor chimerism was observed in 47% for single UCBT and 77% for double UCBT recipients at day +30, and in 75% and 92%, at day 60+, respectively.

Almost the same figures were observed at 3 and 6 months after UCBT. Figure 1 shows the chimerism data by type of graft (single, n = 22; double, n = 48) during different time periods after UCBT. In the group of patients who underwent transplantation with a single CB unit who had engrafted by

Figure 1. Chimerism data for patients who engrafted by single (left) and double (right) UCBT, and the time period in which they engrafted.

day +60, almost all patients had full donor chimerism. In the group of patients who underwent transplantation with a double CB unit who had engrafted, full chimerism for most patients was also observed at day +60; however, dual chimerism was observed during 3 months in few patients.

Acute and chronic GVHD

Acute GVHD was observed in 39 patients; 14 had grade I, 16 patients had grade II, 18 patients had grade III, and 1 patient had grade IV. The cumulative incidence of acute GVHD (grades II to IV) at day +100 was 50% ± 6% and 24% ± 3% for grade III and IV. In univariate analyses, increased incidence of acute GVHD was associated with ABO major incompatibility

(HR, 2.21; 95% CI, 1.04 to 4.7; P = .04). In multivariate analysis, we did not find any risk factor associated with acute GVHD. Thirty-one patients had skin involvement (grade 1, n = 4; grade 2, n = 15; and grade 3, n = 12), 8 patients had liver involvement (grade 2, n = 5; grade 3, n = 3), and 23 had gastrointestinal involvement (grade 1, n = 4; grade 2, n = 8; grade 3, n = 7; and grade 4, n = 4). GVHD treatment was mainly steroids and it was associated with antithymocyte globulin in 4 patients (grade 3 or 4) and with monoclonal antibody in another patient (grade 4). Of the 19 patients having grade 3 or 4 GVHD, 13 died, mostly of infections.

Chronic GVHD was diagnosed in 16 of 66 patients at risk: 12 had limited GVHD and 4 extensive GVHD.

NRM and causes of death

The median follow-up for survivors was 25 months. At 2 years, the cumulative incidence of NRM was 20.25% (Figure 2A). Table 2 shows the risk factors for NRM in uni-variate analyses. In multivariable analysis, 2 factors were associated with increased risk of NRM: lower TNC(<3.4 x 107/ kg) (HR, 3.3; 95% CI, 1.1 to 8.3; P = .003) (Figure 2B) and major ABO incompatibility (HR, 12.6; 95% CI, 2.77 to 57; P = .001) (Figure 2C). Sixteen patients died of transplantation-related causes: 6 from GVHD, 4 of whom had associated infections (sepsis, n = 1; adenovirus and HHV6, n = 1; CMV disease, n = 2) and 2 other causes. Causes of death for the remaining 10 patients were sepsis (n = 1), central nervous system hemorrhage (n = 1), toxicities (n = 4, multiorgan failure, renal, and or cardiac failure), and 4 unknown or other causes.

Because ABO compatibility was an important factor associated with NRM, we looked at causes of death by ABO

Figure 2. (A) Cumulative incidence of nonrelapse mortality overall. (B) Incidence of transplantation-related mortality according to TNC infused ( x ;107/kg) and (C) by ABO compatibility.

Table 2

Univariable Analysis for Nonrelapse Mortality

Characteristics Values n HR 95% CI P Value

Age (continuous), 79 1.01 (.97-1.04) .78

Gender Male 37 1.00

Female 42 1.14 (.42-3.06) .80

CMV status Neg 40 1.00

Pos 39 .59 (.21-1.62) .30

Comorbidity 0 37 1.00

index (Sorror)

1 12 2.03 (.48-8.50) .33

>=2 28 1.92 (.61-6.04) .27

Weight 78 1.03 (1-1.07) .073

(continuous), kg

Disease status >CR1 31 1.00

CR1 48 1.31 (.48-3.62) .60

FAB classification M0, M1, or M2 35 1.00

M4 19 .52 (.14-1.97) .34

Others 25 .78 (.26-2.40) .67

Secondary AML No 65 1.00

Yes 14 1.58 (.45-5.58) .48

Cytogenetic risk Favorable 6 1.00

Intermediate 35 1.43 (.18-11.65) .74

Unfavorable 38 1.40 (.18-11.24) .75

Interval from 77 .96 (.9-1.03) .26

diag. to UCBT

Interval from 76 .98 (.79-1.23) .89

CR1 to UCBT

Days from 74 1.00 (1.00-1.01) .22

diag. to CR1

No. of inductions 1 39 1.00

2 17 .23 (.03-1.75) .15

3 5 3.35 (.68-16.46) .14

Extramedullary No 71 1.00

involvement

Yes 8 0 (0-lnf) 1.00

Type of graft Single 26 1.00

Double 53 .99 (.34-2.86) .99

TNC infused 79 .84 (.47-1.48) .55

CD34/kg infused 79 .01 (0-56.60) .30

CD3/kg infused 49 1.04 (.81-1.33) .75

CFU-GM infused 51 .76 (.45-1.29) .31

HLA match 5/6 or 6/6 18 1.00

4/6 or 3/6 61 .82 (.27-2.56) .74

ABO compatibility Matched 26 1.00

Minor 18 1.90 (.12-30.33) .65

Major 35 12.08 (1.59-91.92) .02

ABO major No 44 1.00

incompatibility

Yes 35 9.23 (2.10-40.62) .003

compatibility. 1n fact, 1 patient died of infection associated with graft failure, out of 26 patients given an ABO compatible graft; 1 patient with graft failure died of hemorrhage, out of 18 patients given a minor ABO mismatched graft, and 14 patients died of GVHD or complications (9 GVHD, 2 infections, and 3 toxicities) out of 35 given a major ABO-incompatible graft.

Relapse

Thirty-seven patients relapsed after cord blood transplantation. At 2 years, the cumulative incidence of relapse was 46.35% (Figure 3A); it was 56% for patients who underwent transplantation in CR1 and 32% for patients who underwent transplantation in CR2 or more. 1n univariate analyses, increased relapse risk was associated with (1) first complete remission (HR, 2.06; 95% CI, 1.02 to 4.19; P = .045), (2) secondary AML (HR, 2.2; 95% CI, 1.0 to 4.86; P = .05); and (3) 2 or more inductions course of chemotherapy to

obtain first complete remission (HR, 2.16; 95% CI, 1.07 to 4.35; P = .036). There was no statistical association of number of cord blood units transplanted with relapse incidence. Other patient-, disease-, and transplantation-related factors were not statistically associated with relapse. At time of analysis, 30 patients died after relapse.

In a multivariate analysis, only patients having more than

1 induction course of chemotherapy to obtain remission had a higher risk of relapsing after RIC UCBT (Figure 3B).

LFS and OS

Estimated LFS at 2 years was 35% (Figure 4A). It was 26% for patients who underwent transplantation in CR1 and 48% for patients who underwent transplantation in other remission (HR, 1.79; 95% CI, 1.0 to 3.2; P = .045) (Figure 4B). Other factors associated with decreased LFS rate were secondary AML (HR, 1.99; 95% CI, 1.02 to 3.89; P = .043) and longer time from diagnosis to CR1 (HR, 1.05; 95% CI, 1.01 to 1.09; P = .027). In a multivariate model, only patients with secondary leukemia had decreased LFS (P = .043) (Figure 4C).

Estimated OS at 2 years was 44%. At last follow-up, 46 patients died; 30 of relapse and 16 of the transplantation-related causes described above.

DISCUSSION

UCBT is considered an alternative option of allogeneic HSCT for patients lacking an HLA-matched donor [23]. Even if outcomes after single or double UCBT were compared with those of HLA-identical siblings or unrelated donor recipients after MAC regimen, delayed engraftment and toxicity leads to a mortality rate of 34% to 57% [6-10].

In the last 6 years, UCBT has been frequently used in adults with hematological malignancies, mainly after RIC. Data from Eurocord registry show that two thirds of all adult UCBT patients receive RIC and, of those, roughly two thirds use double cord blood units (V. Rocha, Eurocord registry, unpublished data).

In 2003, Barker et al. [24], on behalf of the University of Minnesota, described the feasibility of UCBT using a RIC regimen. Two types of regimen were used: 200 cGy TBI, busulfan, fludarabine, or 200 cGy TBI, cyclophosphamide, and fludarabine (TCF). The latter conditioning regimen was associated with better engraftment rates. Later on, the same group showed an NRM of 26% after UCBT with this conditioning regimen [11,17,18]. In a registry-based analysis, the Eurocord group has shown that outcomes of UCBT recipients using TCF were associated with better outcomes when compared with those using other regimens, with an NRM at

2 years of 12% for TCF and 51% for other regimens [25]. This was confirmed in a recent retrospective study that showed 2-year NRM of 19% in patients conditioned with TCF compared with 52% for those receiving other regimens [18]. Recently, a multicenter prospective study reported the efficacy of double UCBT using TCF in 50 adult patients with different hematological malignancies and disease status [26]. Cumulative incidence of engraftment was 94% and 1-year NRM was 24% [26].

All the above studies show that RIC UCBT using single or double cord blood units and TCF is feasible, with encouraging results, and that NRM varies between 12% to 26% for adults with heterogeneous diagnoses of malignant disorders and disease status. However, it was not yet clear if outcomes after RIC UCBT using a homogeneous conditioning (TCF) and GVHD prophylaxis in patients with AML could be reproducible in a prospective multicenter trial.

With our prospective, multicenter trial in France, we were able to demonstrate an NRM of 20% with the use of TCF and, at the same time, evaluate other outcomes of RIC UCBT using this regimen. Only patients with high-risk AML lacking a 10/ 10 or 9/10 HLA unrelated donor and with comorbidities or advanced age were included. We showed that NRM at 2 years was 20% after TCF UCBT in 79 AML patients in remission, which is very similar to what has been observed after TCF UCBT for other diseases.

We found that a TNC dose >3.4 x 107/kg was an independent risk factor for decreased NRM. This finding agrees with what has been previously seen in the literature: cell dose is an important risk factor associated with greater engraftment and decreased NRM. The minimum cell dose for UCBT in this protocol was pre-established at 3 x 107/kg before freezing. We were able to find a single unit for one third of the patients, and the other two thirds of the patients underwent transplantation with a double CB unit. In this

Figure 4. (A) Cumulative incidence of LFS overall, (B) according to status (CR1 or not at transplantation), and (C) according to diagnosis.

prospective trial, the number of CB units used (single or double CB) was not associated with NRM. Interestingly, another risk factor found to be associated with increased NRM was major ABO incompatibility. In univariate analysis, ABO major incompatibility was also associated with the incidence of acute GVHD; however, the association was not observed in the multivariate analysis. One could argue about the possible rationale for this finding. In fact, it has been demonstrated that the ABO glycosyltransferase-derived peptides have the potential to function as minor histocom-patibility antigens and elicit an in vitro T cell response [27]. ABO mismatch is well known to increase the risk of graft failure in solid organ transplantation, although allogeneic hematopoietic cell transplantation is successfully performed across the ABO barrier [28]. Although reports suggested poor disease-free survival with ABO-mismatch in UCBT recipients, there are limited data on the impact of ABO mismatch on GVHD after UCBT [25,29]. Despite the finding in our study of the association of ABO major incompatibility with acute GVHD in univariate analysis, we were not able to find any association in the multivariate model. More recent studies analyzing UCBT recipients did not find any association with acute or chronic GVHD or NRM [30,31]. Therefore, further investigations are needed to explain the reasons of this association using a more homogeneous series of UCBT after RIC. Other factors, such as number of HLA mismatches, age, and comorbidity index did not affect NRM.

As expected, patients in our study who underwent transplantation in CR1 had very high unfavorable risk of AML, such as higher frequency of unfavorable cytogenetics, secondary AML, FAB classification (M5/M6/M7), and greater number of induction courses of chemotherapy to achieve CR1, compared with patients who underwent transplantation in CR2 or more. These features can explain a 2-year relapse incidence of 56% for patients who underwent transplantation in CR1. In a recent registry-based analysis, relapse incidence was 44% after double UCBT using the TCF regimen for patients with acute leukemias (AML and acute lymphoblastic leukemia), including all disease statuses [18]. Patients who underwent transplantation in CR2 or in nonremission had higher risk of relapse than patients who underwent transplantation in CR1 [18].

In our analysis, incidence of relapse was not statistically different in multivariate analysis according to disease status, despite a higher incidence of relapse for CR1 patients. In fact, the only factor increasing relapse incidence was the previous number of induction chemotherapy courses to obtain first CR. LFS was also decreased in patients who underwent transplantation in CR1; however, in multivariate analysis, the only factor associated with decreased LFS was the presence of secondary AML. Those AML features of patients who underwent transplantation in CR1 may explain the higher incidence of relapse and decrease of LFS using this RIC regimen.

In conclusion, UCBT using TCF is a feasible procedure for patients with high-risk AML without an unrelated donor and for whom MAC is contraindicated. Despite the decreased NRM, the relapse incidence is quite high, and the LFS rate decreased mainly for patients with high-risk AML. Strategies including adding other drugs such as thiotepa, to this TCF regimen or increasing the TBI dose, are under investigation with very promising results [32]. Other strategies, such as targeting leukemic cells with monoclonal antibodies, should be further investigated to decrease the relapse incidence and improve LFS without increasing toxicity after UCBT.

ACKNOWLEDGMENTS

The authors thank all transplantation centers in France participating in this study, the Societé Française de Greffe de Moelle Osseuse et Therapie Cellulaire (SFGM-TC), and staff in Eurocord office.

Financial disclosure: The protocol was granted by the Institut National du Cancer (A0M06206) with a promotion of Assistance Publique-Hôpitaux de Paris (P060206). V.R. is funded by the National Institute Health Research-Biomedical Research Centre funding scheme. Eurodact 2006-005901-67 is registered in www.clinical trials: NCT00797758.

Conflict of Interest Statement: There are no conflicts of interest to report.

Authorship contributions: B.R. is the first investigator for his center and coordinated the study, designed the study, analyzed and interpreted data, and wrote the manuscript; S.C. designed research, performed data validation, analyzed and interpreted data, and performed statistical analysis; S.V., P.C., S.F., A.S., J.O.B., G.S., P.C., A.H., J.C., S.F., F.L., I.Y-A., G.M., N.M., G.M., S.M., M.U., C.E.B., M.M., L.C., C.D., and K.B. are the first investigators for their center; they collected data and validated clinical information; E.G. revised the manuscript, A.R validated clinical data (Eurocord); A.B. is the director of scientific program of SFGM-TC and participated in designing the study; S.N. participated in the biological program and design of the study; T.S. was involved in clinical monitoring and participated in revising the manuscript; N.M. is the director of the SFGM-TC and participated in revising the manuscript; and V.R designed research, performed research, analyzed and interpreted data, and wrote the manuscript. All authors edited and approved the manuscript.

REFERENCES

1. Bacigalupo A, Ballen K, Rizzo D, et al. Defining the intensity of conditioning regimens: working definitions. Biol Blood Marrow Transplant. 2009;15:1628-1633.

2. Appelbaum FR. Dose intensity of preparative regimens for acute myeloid leukemia - one-size-fits-all or tailor-made? Best Pract Res Clin Haematol. 2010;23:509-517.

3. Martino R, de Wreede L, Fiocco M, et al. Comparison of conditioning regimens of various intensities for allogeneic hematopoietic SCT using HLA-identical sibling donors in AML and MDS with <10% BM blasts: a report from EBMT. Bone Marrow Transplant. 2013;48:761-770.

4. Aoudjhane M, Labopin M, Gorin NC, et al. Comparative outcome of reduced intensity and myeloablative conditioning regimen in HLA identical sibling allogeneic haematopoietic stem cell transplantation for patients older than 50 years of age with acute myeloblastic leukaemia: a retrospective survey from the Acute Leukemia Working Party (ALWP) of the European group for Blood and Marrow Transplantation (EBMT). Leukemia. 2005;19:2304-2312.

5. Sorror ML, Sandmaier BM, Storer BE, et al. Comorbidity and disease status based risk stratification of outcomes among patients with acute myeloid leukemia or myelodysplasia receiving allogeneic hematopoi-etic cell transplantation. J Clin Oncol. 2007;25:4246-4254.

6. Oran B, Wagner JE, DeFor TE, et al. Effect of conditioning regimen intensity on acute myeloid leukemia outcomes after umbilical cord blood transplantation. Biol Blood Marrow Transplant. 2011;17:1327-1334.

7. Eapen M, Rocha V, Sanz G, et al. Effect of graft source on unrelated donor haemopoietic stem-cell transplantation in adults with acute leukaemia: a retrospective analysis. Lancet Oncol. 2010;11:653-660.

8. Rocha V, Labopin M, Ruggeri A, et al. Unrelated cord blood transplantation: outcomes after single-unit intrabone injection compared with double-unit intravenous injection in patients with hematological malignancies. Transplantation. 2013;95:1284-1291.

9. Laughlin MJ, Eapen M, Rubinstein P, et al. Outcomes after transplantation of cord blood or bone marrow from unrelated donors in adults with leukemia. N Engl J Med. 2004;351:2265-2275.

10. Brunstein CG, Gutman JA, Weisdorf DJ, et al. Allogeneic hematopoietic cell transplantation for hematologic malignancy: relative risks and benefits of double umbilical cord blood. Blood. 2010;116:4693-4699.

11. Brunstein CG, Barker JN, Weisdorf DJ, et al. Umbilical cord blood transplantation after nonmyeloablative conditioning: impact on transplantation outcomes in 110 adults with hematologic disease. Blood. 2007;110:3064-3070.

12. Frassoni F, Gualandi F, Podesta M, et al. Direct intrabone transplant of unrelated cord-blood cells in acute leukaemia: a phase I/II study. Lancet Oncol. 2008;9:831-839.

13. Delaney C, Heimfeld S, Brashem-Stein C, et al. Notch-mediated expansion of human cord blood progenitor cells capable of rapid myeloid reconstitution. Nat Med. 2010;16:232-236.

14. de Lima M, McNiece I, Robinson SN, et al. Cord-blood engraftment with ex vivo mesenchymal-cell coculture. N Engl J Med. 2012;367: 2305-2315.

15. Rocha V, Broxmeyer HE. New approaches for improving engraftment after cord blood transplantation. Biol Blood Marrow Transplant. 2010; 16:S126-S132.

16. Barker JN, Weisdorf DJ, Wagner JE. Creation of a double chimera after the transplantation of umbilical-cord blood from two partially matched unrelated donors. N Engl J Med. 2001;344:1870-1871.

17. Barker JN, Weisdorf DJ, DeFor TE, et al. Transplantation of 2 partially HLA-matched umbilical cord blood units to enhance engraftment in adults with hematologic malignancy. Blood. 2005;105:1343-1347.

18. Brunstein CG, Eapen M, Ahn KW, et al. Reduced-intensity conditioning transplantation in acute leukemia: the effect of source of unrelated donor stem cells on outcomes. Blood. 2012;119:5591-5598.

19. Glucksberg H, Storb R, Fefer A, et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HL-A-matched sibling donors. Transplantation. 1974;18:295-304.

20. Terwey TH, Vega-Ruiz A, Hemmati PG, et al. NIH-defined graft-versus-host disease after reduced intensity or myeloablative conditioning in patients with acute myeloid leukemia. Leukemia. 2012;26:536-542.

21. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457-481.

22. Cox DR. Regression models and life tables. J R Stat Soc. 1972;34:187.

23. Rocha V, Gluckman E. Improving outcomes of cord blood transplantation: HLA matching, cell dose and other graft- and transplantation-related factors. Br J Haematol. 2009;147:262-274.

24. Barker JN, Weisdorf DJ, DeFor TE, et al. Rapid and complete donor chimerism in adult recipients of unrelated donor umbilical cord blood transplantation after reduced-intensity conditioning. Blood. 2003;102: 1915-1919.

25. Rocha V, Mohty M, Gluckman E, Rio B. Reduced-intensity conditioning regimens before unrelated cord blood transplantation in adults with acute leukaemia and other haematological malignancies. Curr Opin Oncol. 2009;21(Suppl 1):S31-S34.

26. Brunstein CG, Fuchs EJ, Carter SL, et al. Alternative donor transplantation after reduced intensity conditioning: results of parallel phase 2 trials using partially HLA-mismatched related bone marrow or unrelated double umbilical cord blood grafts. Blood. 2011;118: 282-288.

27. Eiz-Vesper B, Seltsam A, Blasczyk R. ABO glycosyltransferases as potential source of minor histocompatibility antigens in allogeneic peripheral blood progenitor cell transplantation. Transfusion. 2005;45: 960-968.

28. Klumpp TR. Immunohematologic complications of bone marrow transplantation. Bone Marrow Transplant. 1991;8:159-170.

29. Berglund S, Le Blanc K, Remberger M, et al. Factors with an impact on chimerism development and long-term survival after umbilical cord blood transplantation. Transplantation. 2012;94:1066-1074.

30. Romee R, Weisdorf DJ, Brunstein C, et al. Impact of ABO-mismatch on risk of GVHD after umbilical cord blood transplantation. Bone Marrow Transplant. 2013;48:1046-1049.

31. Konuma T, Kato S, Ooi J, et al. Effect of ABO blood group incompatibility on the outcome of single-unit cord blood transplantation after mye-loablative conditioning. Biol Blood Marrow Transplant. 2014;20: 577-581.

32. Ponce DM, Sauter C, Devlin S, et al. A novel reduced-intensity conditioning regimen induces a high incidence of sustained donor-derived neutrophil and platelet engraftment after double-unit cord blood transplantation. Biol Blood Marrow Transplant. 2013;19:799-803.