Scholarly article on topic 'Impact of the Direction of HLA Mismatch on Transplantation Outcomes in Single Unrelated Cord Blood Transplantation'

Impact of the Direction of HLA Mismatch on Transplantation Outcomes in Single Unrelated Cord Blood Transplantation Academic research paper on "Clinical medicine"

CC BY-NC-ND
0
0
Share paper
OECD Field of science
Keywords
HLA incompatibility / Graft-versus-host direction / Host-versus-graft direction / Overall survival / Nonrelapse mortality

Abstract of research paper on Clinical medicine, author of scientific article — Junya Kanda, Yoshiko Atsuta, Atsushi Wake, Tatsuo Ichinohe, Minoko Takanashi, et al.

Abstract The impact of the direction of HLA mismatch (MM) on outcome in unrelated cord blood (UCB) transplantation has not yet been clarified. We conducted a retrospective study using national registry data on 2977 patients who underwent transplantation using a single UCB for leukemia or myelodysplastic syndrome. HLA matching was assessed by serologic data for HLA-A, -B, and -DR loci. The median age of the recipients at transplantation was 41 years (range, 0-82 years), and 2300 recipients (77%) were age ≥16 years. The 2-year overall survival rate was 0.46. The presence of MM only in the graft-versus-host direction or only in the host-versus-graft direction was not associated with overall mortality (hazard ratio [HR], 0.88; P = .317 and HR, 0.95; P = .670, respectively) compared with 1 bidirectional MM. This finding was consistent in both the child and adult cohorts. The presence of MM only in the graft-versus-host direction was associated with a lower incidence of nonrelapse mortality (HR, 0.65; P = .040), significant only in the child cohort. No MM category was associated with relapse. Our findings suggest that the direction of HLA MM does not have a significant impact on overall survival after UCB transplantation.

Academic research paper on topic "Impact of the Direction of HLA Mismatch on Transplantation Outcomes in Single Unrelated Cord Blood Transplantation"

Impact of the Direction of HLA Mismatch on Transplantation Outcomes in Single Unrelated Cord Blood Transplantation

Junya Kanda1, Yoshiko Atsuta 2, Atsushi Wake 3, Tatsuo Ichinohe 4,

Minoko Takanashi5, Yasuo Morishima 6, Shuichi Taniguchi3,

Satoshi Takahashi7, Hiroyasu Ogawa8, Kazuteru Ohashi9, Yuju Ohno10,

Nobuyuki Aotsuka11, Yasushi Onishi12, Koji Kato13,

Tokiko Nagamura-Inoue14, Yoshinobu Kanda1,*, on behalf of the HLA

Working Group of the Japan Society for Hematopoietic Cell Transplantation

1 Division of Hematology, Saitama Medical Center, Jichi Medical University, Saitama, Japan

2 Department of Hematopoietic Stem Cell Transplantation, Data Management/Biostatistics, Nagoya University School of Medicine, Nagoya, Japan

3 Department of Hematology, Toranomon Hospital, Tokyo, Japan

4 Division of Hematology, Respiratory Medicine and Oncology, Department of Internal Medicine, Faculty of Medicine, Saga University, Saga, Japan

5 Japanese Red Cross Kanto-Koshinetsu Block Blood Center, Tokyo, Japan

6 Division of Epidemiology and Prevention, Aichi Cancer Center Research Institute, Nagoya, Japan

7 Department of Molecular Therapy, Advanced Clinical Research Center, Institute of Medical Science, University of Tokyo, Tokyo, Japan

8 Division of Hematology, Department of Internal Medicine, Hyogo College of Medicine, Hyogo, Japan

9 Hematology Division, Tokyo Metropolitan Cancer and Infectious Diseases Center, Komagome Hospital, Tokyo, Japan

10 Department of Internal Medicine, Kitakyushu Municipal Medical Center, Kitakyushu, Japan

11 Department of Hematology and Oncology, Japanese Red Cross Narita Hospital, Narita, Japan

12 Department of Hematology and Rheumatology, Tohoku University Hospital, Sendai, Japan

13 Division of Hematology Oncology, Children's Medical Center, Japanese Red Cross Nagoya First Hospital, Nagoya, Japan

14 Department of Cell Processing and Transfusion, Research Hospital, Institute of Medical Science, University of Tokyo, Tokyo, Japan

American Society for Blood and Marrow Transplantation

Article history:

Received 22 August 2012

Accepted 22 September 2012

Key Words: HLA incompatibility Graft-versus-host direction Host-versus-graft direction Overall survival Nonrelapse mortality

ABSTRACT

The impact of the direction of HLA mismatch (MM) on outcome in unrelated cord blood (UCB) transplantation has not yet been clarified. We conducted a retrospective study using national registry data on 2977 patients who underwent transplantation using a single UCB for leukemia or myelodysplastic syndrome. HLA matching was assessed by serologic data for HLA-A, -B, and -DR loci. The median age of the recipients at transplantation was 41 years (range, 0-82 years), and 2300 recipients (77%) were age >16 years. The 2-year overall survival rate was 0.46. The presence of MM only in the graft-versus-host direction or only in the host-versus-graft direction was not associated with overall mortality (hazard ratio [HR], 0.88; P = .317 and HR, 0.95; P = .670, respectively) compared with 1 bidirectional MM. This finding was consistent in both the child and adult cohorts. The presence of MM only in the graft-versus-host direction was associated with a lower incidence of nonrelapse mortality (HR, 0.65; P = .040), significant only in the child cohort. No MM category was associated with relapse. Our findings suggest that the direction of HLA MM does not have a significant impact on overall survival after UCB transplantation.

© 2013 American Society for Blood and Marrow Transplantation.

INTRODUCTION

Unrelated cord blood (UCB) has emerged as a promising alternative source of hematopoietic stem cells for adult and pediatric allogeneic hematopoietic cell transplantation [1-4], and the use of UCB transplantation (UCBT) has been rapidly increasing, particularly in the United States, Europe, and Japan. One advantage of using UCB as a hematopoietic stem cell source is that UCBT requires less stringent HLA matching compared with bone marrow or peripheral blood stem cell transplantation, making it easier to find candidate UCB units in UCB banks. One or 2 antigen/allele mismatches (MMs) in

Financial disclosure: See Acknowledgments on page 254.

* Correspondence and reprint requests: Yoshinobu Kanda, MD, Division of Hematology, Saitama Medical Center, Jichi Medical University, 1-847 Amanuma-cho, Omiya-ku, Saitama City, Saitama 330-8503, Japan. E-mail address: ycanda-tky@umin.ac.jp (Y. Kanda).

1083-8791/$ — see front matter © 2013 American Society for Blood and Marrow http://dx.doi.org/10.1016/j.bbmt.2012.09.017

the HLA-A, -B, and -DR loci between a UCB unit and recipient are acceptable without ex vivo T cell depletion methods, and the clinical outcome of transplantation using a 0-2 antigen/ allele-mismatched UCB unit was almost comparable to that from an HLA allele-matched unrelated donor [1-3].

Although the number of HLA MMs between a UCB unit and a recipient is usually counted without considering the MM direction, the effect of the immune reaction caused by HLA MM differs according to whether the MM is in the graft-versus-host (GVH) or host-versus-graft (HVG) direction. A mismatched antigen in the GVH direction can be a major target for donor T cells and can cause graft-versus-host disease (GVHD), whereas a mismatched antigen in the HVG direction can be a major target for the remaining recipient T cells and can lead to graft rejection. In related transplantation, the presence of HLA MMs in the GVH direction is associated with a higher incidence of GVHD, whereas the Transplantation.

presence of HLA MMs in the HVG direction is associated with a higher incidence of rejection [5-7]. Therefore, from a biological perspective, the impact of HLA MM should be discussed separately according to the direction of MM. However, because most patients have an equal number of MMs in the GVH and HVG directions (bidirectional MM), studying an adequate number of patients to evaluate an MM imbalance in the GVH and HVG directions has proven difficult.

The few studies that have evaluated the impact of the HLA MM direction on UCBT outcome have reported inconsistent results [8-10]. Matsuno et al. [8] reported that an HLA MM in the GVH direction was associated with lower incidence of neutrophil engraftment. In contrast, Stevens et al. [9] showed that UCBT with an MM only in the GVH direction was associated with a lower incidence of nonrelapse mortality (NRM) and overall mortality compared with UCBT with an 1 bidirectional MM, whereas UCBT with an MM only in the HVG direction was associated with a lower incidence of neutro-phil engraftment and a higher incidence of relapse.

To clarify the significance of the direction of HLA MM on transplantation outcomes, we conducted a retrospective study using national registry data in 2977 patients who underwent a single UCBT.

METHODS Data Collection

Data for 2987 patients with acute myelogenous leukemia (AML), acute lymphoblastic leukemia (ALL), myelodysplastic syndrome (MDS), and chronic myelogenous leukemia (CML) who underwent a first transplantation using a single UCB unit between January 1,1998, and December 31, 2009, were obtained from the Transplant Registry Unified Management Program (TRUMP) [11], in which all UCBTs are registered through the Japan Cord Blood Bank Network (JCBBN), a national network of all 11 cord blood banks in Japan. Ten patients lacking data on survival status or survival date were excluded. A total of 2977 patients met the criteria for study inclusion. The study design was approved by the TRUMP Data Management Committee and the Institutional Review Board of Saitama Medical Center, Jichi Medical University, where this study was organized.

Histocompatibility

Histocompatibility data for the HLA-A, -B, and -DR loci were obtained from reports collected from the institution at which the transplantation was performed or cord blood banks. HLA typing methods have been described previously [12]. To reflect current practice in Japan, HLA matching was assessed by serologic data for HLA-A, -B, and -DR loci. A secondary analysis using antigen level data for HLA-A, -B and available allele level data for HLA-DRB1 was also performed to compare our data with previously published data from the United States and Europe. HLA-DRB1 allele information was available in 84% of patients (2498 of 2977). Among these patients, 62% had the same number of MMs at HLA-DRB1 loci at either the antigen or allele level. An HLA MM in the GVH direction was defined as when the recipient's antigens or alleles were not shared by the donor, and an MM in the HVG direction was defined as when the donor's antigens or alleles were not shared by the recipient.

Endpoints

The primary study endpoint was overall survival (OS). Other endpoints assessed were relapse, NRM, neutrophil and platelet engraftment, grade II-IV or III-IV acute GVHD, and chronic GVHD. Neutrophil recovery was defined as an absolute neutrophil count exceeding 0.5 x 109/L for 3 consecutive days after UCBT. Platelet recovery was defined as an absolute platelet count exceeding 50 x 109/L without platelet transfusion. The physicians who performed transplantation at each center diagnosed and graded acute and chronic GVHD according to traditional criteria [13,14]. The incidence of acute GVHD was evaluated in patients who engrafted, and that of chronic GVHD was evaluated in patients who engrafted and survived for more than 100 days.

Statistical Analysis

The probability of OS was estimated according to the Kaplan—Meier method and the groups were compared using the log-rank test. The probabilities of relapse, NRM, neutrophil and platelet engraftment, and acute and

chronic GVHD were estimated based on cumulative incidence curves [15]. Competing events were death without relapse for relapse, relapse for NRM, death without engraftment for neutrophil and platelet engraftment, and death or relapse without GVHD for acute and chronic GVHD. The groups were compared using Gray's test [16]. The Cox proportional hazards model was used to evaluate the effect of confounding variables on OS, and the Fine and Gray proportional hazards model was used for the other endpoints [17]. Based on the report by the Center for International Blood and Marrow Transplant Research, we classified the conditioning regimens as myeloa-blative if total body irradiation >8 Gy, oral busulfan >9 mg/kg, i.v. busulfan >7.2 mg/kg, or melphalan >140 mg/m2 was used in the conditioning regimen; otherwise, the conditioning regimen was classified as reduced intensity [18]. For patients with insufficient data regarding dosages of the agents used in the conditioning regimen, we used the information on conditioning intensity (myeloablative or reduced intensity) reported by the treating clinicians. We defined AML and ALL in first or second remission, CML in first or second chronic phase or accelerated phase, and MDS with refractory anemia or refractory anemia with ringed sideroblasts as standard risk, and all other conditions as high risk.

The following possible confounding variables were considered: recipient age group (0-5 years, 6-15 years, 16-49 years, or >50 years at transplantation), matching of ABO blood type between the recipient and UCB (match or major, minor, or bidirectional MM), recipient sex, sex MM between recipient and UCB (match, male donor—female recipient, or female donor—male recipient), disease (AML, ALL, CML, or MDS), disease status before transplantation (standard or high risk), type of conditioning regimen (myeloablative or reduced intensity), type of GVHD prophylaxis (calcineurin inhibitor plus methotrexate, calcineurin inhibitor only, others), and year of transplantation (1998-2004 or 2005-2009). Factors other than HLA MM and total nucleated cell (TNC) dose category were selected in a stepwise manner from the model with a variable retention criterion of P < .05. HLA MM and TNC dose category (>10.0,5.0-9.9, 2.5-4.9,2.0-2.4, and <2.0 x 107/kg) were then added to the final model. All tests were 2-sided, and a P value <.05 was considered statistically significant. All statistical analyses were performed with Stata version 12 (StataCorp, College Station, TX) and EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface for R 2.13.0 (R Foundation for Statistical Computing, Vienna, Austria) [19]. More precisely, EZR is a modified version of R commander (version 1.6-3) designed to add statistical functions used frequently used in biostatistics.

RESULTS

Patient Characteristics

Table 1 summarizes patient and transplant characteristics. The median age of the recipients at transplantation was 41 years (range, 0-82 years), and 2300 patients (77%) were age >16 years. Diagnoses for transplantation were AML in 1606 patients, ALL in 893, CML in 135, and MDS in 343. Half of the patients had standard-risk disease. UCBT was performed between 1998 and 2004 in 1153 patients (39%) and between 2005 and 2009 in 1824 patients (61%). The combination of a calcineurin inhibitor (tacrolimus or cyclosporine) and methotrexate was used in 62% of patients, whereas a calcineurin inhibitor alone was used in 22% of patients.

Some 40% of patients received a UCB unit containing <2.5 x 107/kg TNCs, and 45% received a UCB unit containing 2.5-4.9 x 107/kg TNCs. Roughly 12% of patients received >5.0 x 107/kg TNCs, but 93% of these patients were age <16 years. Median body weight was 17 kg (range, 4-68 kg) for the children and 55 kg (range, 24-165 kg) for the adults. HLA MM was categorized as follows: HLA match in both the GVH and HVG directions (GVH 0/HVG 0 MM group; n = 273 [9%]), 1-2 antigen MMs in the GVH direction but 0 MMs in the HVG direction (GVH 1-2/HVG 0 MM group; n = 150 [5%]), 1-2 antigen MMs in the HVG direction but 0 MM in the GVH direction (GVH 0/HVG 1-2 MM group; n = 136 [5%]), 1 antigen MM in both the GVH and HVG directions at the same locus (GVH 1/HVG 1 MM group; n = 716 [24%]), 2 antigen MMs in both the GVH and HVG directions (GVH 2/hVG 2 MM group; n = 1170 [39%]), 2 antigen MMs in the GVH direction and 1 antigen MM in the HVG direction (GVH 2/HVG 1 MM group; n = 231 [8%]), 1 antigen MM in the GVH direction and

Table 1

Patient Characteristics

Characteristic Total Children ( 0-15 Years) Adults (16+ Years)

Recipient age at UCBT, years, median ( range) 41 (0-82) 5 (0-15) 49(16-82)

Recipient age at UCBT, years, n ( %)

0-9 511 17) 511 75) 0 0)

10-19 272 9) 166 25) 106 5)

20-29 287 10) 0 0) 287 12)

30-39 371 12) 0 0) 371 16)

40-49 422 14) 0 0) 422 18)

50-59 625 21) 0 0) 625 27)

>60 489 16) 0 0) 489 21)

ABO matching, n ( %)

Match 994 33) 248 37) 746 32)

Minor 815 27) 174 26) 641 28)

Major 704 24) 149 22) 555 24)

Bidirectional 458 15) 104 15) 354 15)

Missing 6 0) 2 0) 4 0)

Recipient sex, n ( %)

Female 1316 44) 305 45) 1011 44)

Male 1661 56) 372 55) 1289 56)

Donor—recipient sex match, n ( %)

Match 1157 39) 290 43) 867 38)

Male donor and female recipient 635 21) 153 23) 482 21)

Female donor and male recipient 768 26) 172 25) 596 26)

Missing 417 14) 62 9) 355 15)

Diagnosis, n %)

AML 1606 54) 234 35) 1372 60)

ALL 893 30) 391 58) 502 22)

CML 135 5) 11 2) 124 5)

MDS 343 12) 41 6) 302 13)

Disease risk at UCBT, n ( %)

Standard risk 1385 47) 423 62) 962 42)

High risk 1450 49) 226 33) 1224 53)

Missing 142 5) 28 4) 114 5)

Conditioning regimen, n %)

Myeloablative 1980 67) 585 86) 1395 61)

Reduced intensity 986 33) 86 13) 900 39)

Missing 11 0) 6 1) 5 0)

In vivo T cell depletion (ATG or alemtuzumab), n ( %)

No 2935 99) 665 98) 2270 99)

Yes 42 1) 12 2) 30 1)

GVHD prophylaxis, n ( %)

CSA only 250 8) 59 9) 191 8)

TAC only 407 14) 27 4) 380 17)

CSA + MTX 1105 37) 209 31) 896 39)

TAC + MTX 755 25) 241 36) 514 22)

CSA + MMF 104 3) 0 0) 104 5)

TAC + MMF 148 5) 6 1) 142 6)

CSA + corticosteroid 87 3) 67 10) 20 1)

TAC + corticosteroid 34 1) 26 4) 8 0)

Other 66 2) 33 5) 33 1)

Missing 21 1) 9 1) 12 1)

Year of UCBT, n( %)

1998-2004 1153 39) 389 57) 764 33)

2005-2009 1824 61) 288 43) 1536 67)

TNC dose when frozen, n ( %)

>10.0 x 107/kg 99 3) 99 15) 0 0)

5.0-9.9 x 107/kg 259 9) 234 35) 25 1)

2.5-4.9 x 107/kg 1344 45) 268 40) 1076 47)

2.0-2.4 x 107/kg 924 31) 44 6) 880 38)

<2.0 x 107/kg 275 9) 21 3) 254 11)

Missing 76 3) 11 2) 65 3)

Weight, kg, median ( range) 52 4-165) 17 4-68) 55 24-165)

HLA MM

0 MM 273 9) 144 21) 129 6)

1-2 MM/GVH only 150 5) 45 7) 105 5)

1-2 MM/rejection only 136 5) 39 6) 97 4)

1 bidirectional MM 716 24) 314 46) 402 17)

2 bidirectional MM 1170 39) 98 14) 1072 47)

2 MM: bidirectional + GVHD 231 8) 16 2) 215 9)

2 MM: bidirectional + rejection 264 9) 19 3) 245 11)

2 MM: GVHD + rejection 37 1) 2 0) 35 2)

ATG indicates antithymocyte globulin; CSA, cyclosporine; MMF, mycophenolate mofetil; MTX, methotrexate; TAC, tacrolimus; 0 MM, HLA match in both the GVH and HVG directions; 1-2 MM/GVH only, antigen MMs in the GVH direction and 0 MMs in the HVG direction; 1-2 MM/rejection only, 1 or 2 antigen MMs in the HVG direction and no MMs in the GVH direction; 1 bidirectional MM, 1 antigen MM in both the GVH and HVG directions at the same locus; 2 bidirectional MM, 2 antigen MMs in both the GVH and HVG directions; 2 MM: bidirectional + GVHD, 2 antigen MMs in the GVH direction and 1 antigen MM in the HVG direction; 2 MM: bidirectional + rejection, 1 antigen MM in the GVH direction and 2 antigen MMs in the HVG direction; 2 MM: GVHD + rejection, 1 antigen MM in the GVH direction at one locus and 1 antigen MM in the HVG direction at another locus.

2 antigen MMs in the HVG direction (GVH 1/HVG 2 MM group; n = 264 [9%]), and 1 antigen MM in the GVH direction at 1 locus and 1 antigen MM in the HVG direction at another locus (GVH 1/HVG 1 2-antigen MM group; n = 37 [1%]).

OS, Relapse, and NRM

The median follow-up period in survivors was 2.2 years (range, 0.0-11.1 years). The 2-year OS rate was 0.46 (95% confidence interval [CI], 0.44-0.48) (Figure 1). To clarify the impact of HLA MM in each vector, the GVH 1/HVG 1 MM group was considered the reference group in the multivar-iate analyses, in accordance with the approach of Stevens et al. [9], and the following hazard ratios (HRs) were adjusted for the other significant variables, including TNC dose category. The GVH 1-2/HVG 0 MM (HR, 0.88; 95% CI, 0.69-1.13; P = .317), the GVH 0/HVG 1-2 MM (HR, 0.95; 95% CI, 0.741.22; P = .670), and other groups were not associated with overall mortality compared with the GVH 1/HVG 1 MM group (Table 2 and Figure 1). The GVH 0/HVG 0 MM group

-1-1-r

12 3 4 Years after transplantation

GVH0 HVG0 GVH12 HVG0 GVH0 HVG12 GVH1 HVG1 GVH2 HVG2

~l-1-1-r

12 3 4 Years after transplantation

Figure 1. OS of total patients (A) and patients grouped according to HLA MM category (B). GVH0 HVG0, HLA match in both the GVH and HVG directions; GVH12 HVG0, antigen MMs in the GVH direction and 0 MMs in the HVG direction; GVH0 HVG12, 1 or 2 antigen MMs in the HVG direction and 0 MM in the GVH direction; GVH1 HVG1, 1 antigen MM in both the GVH and HVG directions at the same locus; GVH2 HVG2, 2 antigen MMs in both the GVH and HVG directions.

C0 O cn CN en cn m o oo o 00

r^ ©

m C0 en co cn r-v co C0 r-v 00 00

c^ <d. c^ c^

CN 00 CN en co en o o 00 s en 00 L10

o o o o © © ©

ÎN ÎN

o ^ m ^f o rs M

ffl h ¿i O O Ul aj 00 00

ÎN ^r

o o CN co co CN o cn m co

oo L10 4 en co m LO en

m r-v m oo o o 00 00 o 00 o

© © o © cn

^T M (û 01 M

O ÎN ^ aj ÎN m 00 ^ 0£ ^ ^ ^

m oa m

r-v ÎN

en cn

© T-l ^

^ m c^ c^ <d.

c^ 00 m

0i 00 en

X © o o

o_ o_ o_ o_

o m m cncncnin dodo

m o iv iv o ^

fN^^I^^iNfN

> O C

> 13 + + tj

> c o c _o s 2 S ro c fO c + Q X

o X > u c o c o u eu tj eu

O T3 T3 >

s u eu tj <u 123 123 O

s s •5 •rS ié ié

CN ÎN IS 123

o CN CN CN CN

was associated with lower overall mortality compared with the GVH 1/HVG 1 MM group (HR, 0.79; 95% CI, 0.64-0.97; P = .025); however, in both the child and adult cohorts, the association was not significant, owing in part to a lack of statistical power. The GVH 1/HVG 1 2-antigen MM group, which was represented mostly in the adult cohort, was associated with lower overall mortality compared with the GVH 1/HVG 1 MM group (HR, 0.55; 95% CI, 0.35-0.87; P = .012).

We performed an additional analysis according to the HLA matching criteria used in the United States and Europe (HLA-A and -B for antigen level and -DRB1 for allele level) (Supplemental Table 1). Consistent with the result obtained using our criteria (HLA-A, -B, and -DR for antigen level), there were no differences in the impact of the MM direction (GVH or HVG) on OS. The difference in OS between the GVH 0/HVG 0 MM and GVH 1/HVG 1 MM groups was not significant in this analysis.

The cumulative incidence rates of relapse and NRM at 2 years were 0.34 (95% CI, 0.32-0.36) and 0.26 (95% CI, 0.240.27), respectively. There was no difference in the incidence of relapse between the GVH 1/HVG 1 MM and any other MM group (Table 3 and Figure 2). The GVH 1-2/HVG 0 MM group was significantly associated with lower NRM compared with the GVH 1/HVG 1 MM group (HR, 0.65; 95% CI, 0.44-0.98; P = .040) (Table 3 and Figure 2), but only in the child cohort (child, P = .048; adult, P = .215).

Because our cohorts were mainly adults, and most adults received a TNC dose of 2.0-4.9 x 107/kg, we performed an additional analysis in the subset of adults who received a TNC dose of 2.0-2.4 x 107/kg or 2.5-4.9 x 107/kg (Supplemental Table 2). In the subset of adults who received a TNC dose of 2.0-2.4 x 107/kg, compared with the GVH 1/HVG 1 MM group, the GVH 0/HVG 0 MM group was associated with lower overall mortality (P = .027) and NRM (P = .007) and a higher incidence of relapse (P = .028), and the GVH 2/HVG 2 MM group was associated with lower overall mortality (P = .001) and NRM (P = .008). The GVH 1-2/HVG 0 MM group was significantly associated with lower NRM compared with the GVH 1/HVG 1 MM group (P = .033). In the subset of adults who received a TNC dose of 2.54.9 x 107/kg, no HLA MM group was associated with overall mortality, relapse, or NRM, except for lower overall mortality in the GVH 1/HVG 1 2-antigen MM group compared with the GVH 1/HVG 1 MM group (P = .046).

Neutrophil and Platelet Engraftment

The cumulative incidence rates of neutrophil and platelet engraftment in our study cohort were 0.76 (95% CI, 0.74-0.77) and 0.57 (95% CI, 0.55-0.59), respectively. The GVH 1-2/HVG 0 MM group was marginally associated with better neutro-phil and platelet engraftment kinetics compared with the GVH 1/HVG 1 MM group (neutrophil engraftment: HR, 1.18; 95% CI, 0.98-1.42; P = .081; platelet engraftment: HR, 1.23; 95% CI, 1.00-1.51; P = .053) (Table 4 and Figure 3). The impact on neutrophil engraftment was significant only in the adult cohort (child, P = .496; adult, P = .045).

Acute and Chronic GVHD

In all engrafted patients, the cumulative incidence rates of grade II-IV and III-IV acute GVHD were 0.45 (95% CI, 0.430.47) and 0.15 (95% CI, 0.14-0.17), respectively. The GVH 0/ HVG 0 MM group was significantly associated with a lower incidence of grade II-IV acute GVHD compared with the GVH 1/HVG 1 MM group (HR, 0.70; 95% CI, 0.54-0.90; P = .006) (Supplemental Table 3 and Figure 4), but only in the child cohort (child, P = .002; adult, P = .506). The GVH 0/HVG 0 MM group was marginally associated with a lower incidence of chronic GVHD compared with the GVH 1/HVG 1 MM group (HR, 0.72; 95% CI, 0.51-1.00; P = .050).

DISCUSSION

This nationwide retrospective study that included a large number of both pediatric and adult patients allowed us to consider an adequate number of patients who underwent UCBT with an HLA MM only in the GVH direction or only in the HVG direction, and to analyze the impact of an MM in the GVH or HVG direction on clinical outcomes after a single UCBT. Neither the GVH 1-2/HVG 0 MM group nor the GVH 0/ HVG 1-2 MM group was associated with overall mortality compared with the GVH 1/HVG 1 MM group. The point estimates of HRs of the GVH 1-2/HVG 0 MM and GVH 0/hVG 1-2 MM groups compared with the GVH 1/HVG 1 MM group were similar and both <1 (HR, 0.88 and 0.95, respectively), suggesting that HLA MMs in the GVH and HVG directions post-UCBT do not have different effects on OS. This finding does not support the conclusion of Stevens et al. [9], who recommended using UCB units with an HLA MM only in the GVH direction and avoiding units with an HLA MM only in the HVG direction.

Table 3

Relapse and NRM

HLA MM category Relapse* NRMy

Number HR95%CI P Value Number HR 95% CI P Value

0 MM 258 1-2 MM/GVH only 147 1-2 MM/rejection only 131

1 bidirectional MM 667

2 bidirectional MM 1106 2 MM: bidirectional + GVHD 217 2 MM: bidirectional + rejection 243 2 MM: GVHD + rejection 36

1.07 (0.84-1.37) 1.20 (0.90-1.59) 1.18 (0.84-1.64) 1.00

0.99 (0.83-1.19)

1.00 (0.76-1.33)

1.27 (0.99-1.63) 0.64 (0.32-1.24)

.560 .215 .338

Reference

258 147 131 667 1106 217 243 36

0.74 (0.53-1.02) 0.65 (0.44-0.98) 0.81 (0.55-1.19) 1.00

0.88 (0.72-1.07) 0.81 (0.60-1.10) 0.66 (0.49-0.91)

0.61 (0.32-1.16)

.063 .040 .292

Reference

* Other significant variables were recipient age group, 0-5 years (reference, 1.00), 6-15 years (HR, 0.61; 95% CI, 0.44-0.84; P = .002), 16-49 years (HR, 0.71; 95% CI, 0.52-0.97; P = .030), >50 years (HR, 0.72; 95% CI, 0.52-0.98; P = .040); diagnosis, AML (reference, 1.00), ALL (HR, 1.11, 95% CI, 0.94-1.30, P = .210), CML (HR, 1.33, 95% CI, 0.99-1.79, P = .059), MDS (HR, 0.67,95% CI, 0.51-0.87, P = .003); disease risk, standard risk (reference, 1.00), high risk (HR, 2.93; 95% CI, 2.54-3.39; P < .001); GVHD prophylaxis, CSA/TAC + MTX (reference, 1.00), CSA/TAC only (HR, 0.72; 95% CI, 0.61-0.86; P < .001), others (HR, 0.87; 95% CI, 0.71-1.05; P = .145).

y Other significant variables were recipient age group, 0-5 years (reference, 1.00), 6-15 years (HR, 1.44; 95% CI, 0.90-2.30; P = .128), 16-49 years (HR, 2.04; 95% CI, 1.29-3.22; P = .002), >50 years (HR, 3.52; 95% CI, 2.24-5.52; P < .001); GVHD prophylaxis, CSA/TAC + MTX (reference, 1.00), CSA/TAC only (HR, 1.90; 95% CI, 1.60-2.26; P < .001), others (HR, 1.42; 95% CI, 1.14-1.75; P = .001), year of transplantation, 1998-2004 (reference, 1.00), 2005-2009 (HR, 0.71; 95% CI, 0.61-0.83; P < .001).

Relapse

о ф о

ф 0.6 м 0.6 о.

SS 0.4 0.2 0.0

- GVH0 HVG0

-----GVH12 HVG0

......... GVH0 HVG12

GVH1 HVG1 GVH2 HVG2

12 3 4 Years after transplantation

Non-relapse mortality

- GVH0 HVG0

-----GVH12 HVG0

......... GVH0 HVG12

GVH1 HVG1 GVH2 HVG2

0 1 2 3 4 5 Years after transplantation

Figure 2. Relapse and NRM.

Several differences in patient background between the study of Stevens et al. [9] and the present study warrant clarification. The first difference is in the age distribution of

patients. Stevens et al.'s series included 907 pediatric patients age <16 years and 295 adult patients; in contrast, our series included 677 pediatric patients and 2300 adult patients, which can provide useful information for both pediatric and adult transplant physicians. Since the cell doses of UCB units in child and adult cohorts are significantly different, which may affect the impact of HLA MM, we performed stratified analyses in the child and adult cohorts. Our results consistently showed that the direction of MM had no apparent impact on overall mortality in either cohort. Consistent with the results of Stevens et al. [9], the GVH 1-2/ HVG 0 MM group was associated with lower NRM in the child cohort, but this advantage was offset by a higher incidence of relapse in this cohort. A second difference between the 2 studies is in conditioning regimens. A myeloablative regimen was used in 92% of the patients in the Stevens et al. study, compared with 67% in our study. Consequently, we performed a separate analysis in the patients who received a myeloablative regimen and confirmed that the direction of MM had no apparent impact on overall mortality (data not shown).

The third difference between the 2 studies relates to GVHD prophylaxis. Cyclosporine and steroids were used as GVHD prophylaxis in 62% of the patients in the Stevens et al. study, but in only 3% of the patients (10% of the child cohort) in our study, which might have affected outcomes. The fourth difference is in the number of patients with an HLA MM only in the GVH direction or only in the HVG direction. The Stevens et al. study included 35 patients with a GVH 1-2/HVG 0 MM and 22 patients with a GVH 0/HVG 1-2 MM in the overall mortality analysis, compared with 150 and 136 patients, respectively, in our study. Finally, the level of HLA typing used to determine the number of HLA MMs differed between the 2 studies. In the present study, MMs in HLA-DR loci were counted at the antigen level in accordance with current practice in Japan, whereas Stevens et al. counted HLA-DRB1 MMs at the allele level. Consequently, we performed an additional analysis using the same HLA matching criteria as in previous studies from the United States and Europe (HLA-A and -B for antigen level and -DRB1 for allele level), and reached a similar conclusion that an MM only in the GVH or only in the HVG direction had no impact on overall mortality.

Table 4

Neutrophil and Platelet Engraftment

HLA MM Category Neutrophil Engraftment* Platelet Engraftmenty

Number HR 95% CI P Value Number HR 95% CI P Value

0 MM 272 1.03 (0.88-1.20) .718 272 1.06 (0.88-1.27) .559

1-2 MM/GVH only 149 1.18 (0.98-1.42) .081 149 1.23 (1.00-1.51) .053

1-2 MM/rejection only 136 1.01 (0.82-1.26) .899 136 0.84 (0.66-1.07) .164

1 bidirectional MM 716 1.00 Reference 714 1.00 Reference

2 bidirectional MM 1167 0.98 (0.87-1.09) .672 1166 0.96 (0.85-1.10) .590

2 MM: bidirectional + GVHD 231 0.91 (0.76-1.08) .278 230 0.91 (0.74-1.13) .406

2 MM: bidirectional + rejection 264 0.86 (0.72-1.02) .089 264 0.98 (0.80-1.19) .816

2 MM: GVHD + rejection 37 1.40 (1.03-1.89) .030 37 2.21 (1.46-3.33) <.001

* Other significant variables were TNC category, 2.5-4.9 x 107/kg (reference, 1.00), >10.0 x 107/kg (HR, 1.76; 95% CI, 1.33-2.33; P < .001), 5.0-9.9 x 107/kg (HR, 1.26; 95% CI, 1.05-1.52; P = .015), 2.0-2.4 x 107/kg (HR, 0.87; 95% CI, 0.79-0.95; P = .003), <2.0 x 107/kg (HR, 0.82; 95% CI, 0.71-0.95; P = .007); diagnosis, AML (reference, 1.00), ALL (HR, 1.11, 95% CI, 1.00-1.22, P = .040), CML (HR, 0.87, 95% CI, 0.73-1.04, P = .124), MDS (HR, 0.88, 95% CI, 0.75-1.04, P = .129); disease risk, standard risk (reference, 1.00), high risk (HR, 0.74; 95% CI, 0.67-0.80; P < .001); GVHD prophylaxis, CSA/TAC + MTX (reference, 1.00), CSA/TAC only (HR, 1.16; 95% CI, 1.04-1.30; P = .010), others (HR, 1.09; 95% CI, 0.96-1.23; P = .169), year of transplantation, 1998-2004 (reference, 1.00), 2005-2009 (HR, 1.21; 95% CI, 1.111.33; P< .001).

y Other significant variables were TNC category, 2.5-4.9 x 107/kg (reference, 1.00), >10.0 x 107/kg (HR, 1.49; 95% CI, 1.09-2.03; P = .013), 5.0-9.9 x 107/kg (HR, 1.26; 95% CI, 1.01-1.57; P = .040), 2.0-2.4 x 107/kg (HR, 0.95; 95% CI, 0.84-1.06; P = .365), <2.0 x 107/kg (HR, 0.81; 95% CI, 0.67-0.97; P = .022); recipient age group, 0-5 years (reference, 1.00), 6-15 years (HR, 1.00; 95% CI, 0.79-1.25; P = .971), 16-49 years (HR, 0.99; 95% CI, 0.77-1.26; P = .909), >50 years (HR, 0.70; 95% CI, 0.55-0.90; P = .006); recipient sex, female (reference, 1.00), male (HR, 0.90; 95% CI, 0.82-0.99; P = .034); disease risk, standard risk (reference, 1.00), high risk (HR, 0.58; 95% CI, 0.53-0.64; P < .001); GVHD prophylaxis, CSA/TAC + MTX (reference, 1.00), CSA/TAC only (HR, 0.81; 95% CI, 0.71-0.91; P = .001), others (HR, 0.88; 95% CI, 0.76-1.01; P = .074), and year of transplantation, 1998-2004 (reference, 1.00), 2005-2009 (HR, 1.26; 95% CI, 1.14-1.40; P < .001).

Neutrophil engraftment

Grade II-IV acute GVHD

8 S 0.8

Sï "ü

■ci 2

0 ? 0.6

è a. 0.4

S ® 0.2

O) c U O)

c o — c

(1) (1) > —

■j a> JS a

M § a

GVH0 HVG0

-----GVH12 HVG0

......... GVH0 HVG12

GVH1 HVG1 ■- GVH2 HVG2

10 20 30 40 Days after transplantation

Platelet engraftment

GVH0 HVG0 GVH12 HVG0 GVH0 HVG12 GVH1 HVG1 GVH2 HVG2

0 50 100

Days after transplantation

Figure 3. Neutrophil and platelet engraftment.

- GVH0HVG0

-----GVH12 HVG0

......... GVH0 HVG12

GVH1 HVG1 ■■ GVH2 HVG2

20 40 60 80 100 Days after transplantation

Grade III-IV acute GVHD

- GVH0HVG0

-----GVH12 HVG0

......... GVH0 HVG12

GVH1 HVG1 ■■ GVH2 HVG2

20 40 60 80 Days after transplantation

Similar to Stevens et al. [9], we found a tendency for better neutrophil and platelet engraftment kinetics in the GVH 1-2/HVG 0 MM group. This finding suggests that an HLA MM in the GVH direction enhances engraftment by eradicating or suppressing the host residual immune cells responsible for the rejection or inhibition of donor cell engraftment. In contrast to our findings, Matsuno et al. [8] analyzed the impact of GVH/HVG MM on 152 patients who underwent a single UCBT in a single center, and found that the presence of a 2-antigen MM in the GVH direction was associated with slower and lower neutrophil engraftment compared with a 0- or 1-antigen MM in the GVH direction. Because Matsuno et al. used only a calcineurin inhibitor for GVHD prophylaxis in all of the patients in their cohort, we recategorized the HLA MM group according to HLA category (GVH 0-1/HVG 0-1 MM, GVH 0-1/HVG 2 MM, GVH 2/HVG 01 MM, and GVH 2/HVG 2 MM) and performed additional analyses in which patients were stratified according to GVHD prophylaxis (calcineurin inhibitor plus methotrexate, calci-neurin inhibitor only, or other). Similar to the findings of Matsuno et al., an MM in the GVH direction was significantly associated with a lower incidence of engraftment in patients who received only a calcineurin inhibitor (data not shown). In contrast, an MM in the GVH direction was associated with

C Chronic GVHD

8 0.8-

0 12 3

Years after transplantation

Figure 4. Acute and chronic GVHD.

a higher incidence of engraftment in patients who received a calcineurin inhibitor plus methotrexate. These findings suggest that the impact of HLA MM differs according to GVHD prophylaxis. A possible explanation for the different

- GVH0HVG0

-----GVH12 HVG0

........ GVH0 HVG12

GVH1 HVG1 GVH2HVG2

effects of GVHD prophylaxis on engraftment is the high incidence of hemophagocytic syndrome (HPS) and pre-engraftment immune reaction in patients who received only a calcineurin inhibitor as GVHD prophylaxis [20,21]. Takagi et al. [20] reported HPS in 20 of 119 patients who underwent UCBT with mostly tacrolimus alone as GVHD prophylaxis, resulting in a high incidence of graft failure. Less-intensive GVHD prophylaxis may enhance the immune reaction caused by donor T cells that recognize the HLA MM antigen in the GVH direction in the early phase after transplantation, which could lead to HPS or similar conditions and decrease the rate of neutrophil engraftment. These findings demonstrate the need for a prospective study using uniform GVHD prophylaxis to further evaluate the impact of HLA MM on neutrophil engraftment.

This study has several limitations. First, the patients' heterogeneous backgrounds might have produced statistical bias, although we attempted to reduce this bias by adjusting the impact in the multivariate analyses. Second, the number of subjects in each HLA MM group category was limited. Nevertheless, the number of subjects in the GVH 1-2/HVG 0 and GVH 0/HVG 1-2 MM groups was much greater than that in previous studies [8,9]. Third, we might have underestimated the degree of HLA MM, given our incomplete allelic and HLA-C antigen information; for example, the group that had only an HLA MM in the GVH direction might have included an allelic MM in the HVG direction. A potential HLA-C antigen MM or KIR ligand MM also might have affected outcomes, but we did not evaluate HLA-C in the present study. The foregoing issues might have weakened the power of this study to detect differences.

In conclusion, our findings do not support a strategy for selecting UCB donors based on the direction of the HLA MM, although GVH 1-2/HVG 0 MMs may be associated with better neutrophil engraftment, particularly when a calcineurin inhibitor plus other immunosuppressive agents, such as methotrexate, are used for GVHD prophylaxis. The impact of HLA MMs in only the GVH direction remains to be clarified further under a uniform GVHD prophylaxis regimen.

ACKNOWLEDGMENTS

We thank all of the physicians and data managers at the centers who contributed valuable data on transplantation to the JCBBN and TRUMP. We also thank the members of the Data Management Committees of JCBBN and TRUMP for their assistance. J.K. is a Research Fellow of the Japan Society for the Promotion of Science.

Financial disclosure: This work was supported in part by a Grant-in-Aid for JSPS Fellows (to J.K.).

Authorship statement: J.K. and Y.K. designed the research, organized the project, and wrote the manuscript. J.K., Y.A., and Y.K. performed the statistical analysis and analyzed the data; K.K. and T.N.-I. collected data from JCBBN; and all of the authors interpreted the data, and reviewed and approved the final manuscript.

SUPPLEMENTARY DATA

Supplementary data related to this article can be found online at http://dx.doi.org/mi016/j.bbmt.2012.09.017.

REFERENCES

1. Rocha V, Labopin M, Sanz G, et al. Transplants of umbilical-cord blood or bone marrow from unrelated donors in adults with acute leukemia. N Engl J Med. 2004;351:2276-2285.

2. 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.

3. Atsuta Y, Suzuki R, Nagamura-Inoue T, et al. Disease-specific analyses of unrelated cord blood transplantation compared with unrelated bone marrow transplantation in adult patients with acute leukemia. Blood. 2009;113:1631-1638.

4. Cohen YC, Scaradavou A, Stevens CE, et al. Factors affecting mortality following myeloablative cord blood transplantation in adults: a pooled analysis of three international registries. Bone Marrow Transplant. 2011;46:70-76.

5. Anasetti C, Amos D, Beatty PG, et al. Effect of HLA compatibility on engraftment of bone marrow transplants in patients with leukemia or lymphoma. N Engl J Med. 1989;320:197-204.

6. Anasetti C, Beatty PG, Storb R, et al. Effect of HLA incompatibility on graft-versus-host disease, relapse, and survival after marrow transplantation for patients with leukemia or lymphoma. Hum Immunol. 1990;29:79-91.

7. Kanda Y, Chiba S, Hirai H, et al. Allogeneic hematopoietic stem cell transplantation from family members other than HLA-identical siblings over the last decade (1991-2000). Blood. 2003;102:1541-1547.

8. Matsuno N, Wake A, Uchida N, et al. Impact of HLA disparity in the graft-versus-host direction on engraftment in adult patients receiving reduced-intensity cord blood transplantation. Blood. 2009; 114:1689-1695.

9. Stevens CE, Carrier C, Carpenter C, et al. HLA mismatch direction in cord blood transplantation: impact on outcome and implications for cord blood unit selection. Blood. 2011;118:3969-3978.

10. Kogler G, Enczmann J, Rocha V, et al. High-resolution HLA typing by sequencing for HLA-A, -B, -C, -DR, -DQ in 122 unrelated cord blood/ patient pair transplants hardly improves long-term clinical outcome. Bone Marrow Transplant. 2005;36:1033-1041.

11. Atsuta Y, Suzuki R, Yoshimi A, et al. Unification of hematopoietic stem cell transplantation registries in Japan and establishment of the TRUMP system. Int J Hematol. 2007;86:269-274.

12. Atsuta Y, Morishima Y, Suzuki R, et al. Comparison of unrelated cord blood transplantation and HLA-mismatched unrelated bone marrow transplantation for adults with leukemia. Biol Blood Marrow Transplant. 2012;18:780-787.

13. Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant. 1995;15:825-828.

14. Sullivan KM, Agura E, Anasetti C, et al. Chronic graft-versus-host disease and other late complications of bone marrow transplantation. Semin Hematol. 1991;28:250-259.

15. Gooley TA, Leisenring W, Crowley J, et al. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med. 1999;18:695-706.

16. Gray RJ. A class of k-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat. 1988;16:1141-1154.

17. Fine JP, Gray RJ. A proportional hazards model for subdistribution of a competing risk. J Am Stat Assoc. 1999;94:456-509.

18. Giralt S, Ballen K, Rizzo D, et al. Reduced-intensity conditioning regimen workshop: defining the dose spectrum. Report of a workshop convened by the Center for International Blood and Marrow Transplant Research. Biol Blood Marrow Transplant. 2009;15:367-369.

19. Kanda Y. Free statistical software: EZR (Easy R) on R commander. Available from: http://www.jichi.ac.jp/saitama-sct/SaitamaHP.files/ statmedEN.html. Accessed on February 1, 2012.

20. Takagi S, Masuoka K, Uchida N, et al. High incidence of haemophago-cytic syndrome following umbilical cord blood transplantation for adults. Br J Haematol. 2009;147:543-553.

21. Uchida N, Wake A, Nakano N, et al. Mycophenolate and tacrolimus for graft-versus-host disease prophylaxis for elderly after cord blood transplantation: a matched-pair comparison with tacrolimus alone. Transplantation. 2011;92:366-371.