Chronic graft-versus-host disease following umbilical cord blood transplantation: retrospective survey involving 1072 patients in Japan

Chronic graft-versus-host disease (GVHD) is a significant concern in allogeneic hematopoietic stem cell transplantation (HSCT). Many studies have been published on clinical features of chronic GVHD following bone marrow transplantation (BMT) and peripheral blood stem cell transplantation (PBSCT). In contrast, we have limited information on chronic GVHD following cord blood transplantation (CBT)

Informed consent was obtained in accordance with the Declaration of Helsinki. According to local policy, the study was approved by Japan Cord Blood Bank Network. We had no direct contact with human subjects during our study; data on patients who underwent CBT were obtained from the Japan Cord Blood Bank Network. The primary end point of this study was to investigate the incidence of any chronic GVHD after CBT. Cord blood units are provided with written informed consent. Between June 1997 and August 2006, 2713 cord blood transplant recipients were registered with the Japan Cord Blood Bank Network. All recipients received a single cord blood unit. We included those with hematologic malignancies who underwent CBT without T-cell depletion. We excluded patients with a history of any type of allogeneic HSCT prior to CBT. A total of 2015 patients met the criteria. Of those, we excluded 943 with disease progression, death without progression, and graft failure within 100 days after transplantation. In this study, we retrospectively investigated clinical features of chronic GVHD in the remaining 1072 patients. Cytomegalovirus-seropositive cord blood unit was not provided to transplantation centers. Acute and chronic GVHD were diagnosed and graded according to standard criteria.^1,2  GVHD that developed after day 100 was defined as chronic GVHD. If the mode of presentation for chronic GVHD was progressive, the date of onset was defined as day 100. SAS version 9.1.3 (SAS Institute, Cary, NC) was used for all statistical analyses.

The median age of the 1072 patients was 33 years (range, 0-79 years) and the median patient body weight was 50.0 kg (range, 4.0-96.1 kg). The median follow up of the surviving patients was 18.1 months (range, 3.3-110.1 months). Of the 1072 patients, 492 (46%) developed grades II to IV acute GVHD within 100 days of transplantation. Chronic GVHD was diagnosed in 312, and the cumulative incidence of chronic GVHD 2 years after transplantation was 28% (95% confidence interval [CI], 25%-31%; Figure 1). Of the 299 patients in whom the information on types of chronic GVHD was available, 184 and 115 patients developed limited and extensive diseases, respectively.

The onset of chronic GVHD was quiescent (n = 225), de novo (n = 48), progressive (n = 18), and unknown (n = 8). Information on involved organs of chronic GVHD was available in 145 patients. Their involved organs were skin (n = 94), oral cavity (n = 35), liver (n = 32), lung (n = 24), gastrointestinal tract (n = 20), and eye (n = 16). Of 224 patients from whom information on both the treatment of chronic GVHD and its response was available, 93 patients (42%) received systemic immunosuppressive agents including corticosteroids (n = 63), calcineurin inhibitors (n = 15), both (n = 13), and others (n = 2). Of those 93 patients, 64 patients (68%) responded to the treatment. Chronic GVHD was refractory in the remaining 29, and fatal in 5.

Multivariate analysis identified prognostic factors of chronic GVHD: higher patient body weight (hazard ratio [HR], 1.01; 95% CI, 1.01-1.02; P = .001), higher number of mismatched antigens for graft-versus-host direction (HR, 1.31; 95% CI, 1.07-1.60; P = .01), reduced-intensity preparative regimen (HR, 0.69; 95% CI, 1.11-3.28, P = .02), use of mycophenolate mofetil in GVHD prophylaxis (HR, 1.91; 95% CI, 1.11-3.28, P = .02), and development of grades II to IV acute GVHD (HR, 1.51; 95% CI, 1.18-1.93; P = .001; Table S1, available on the Blood website; see the Supplemental Materials link at the top of the online article).

Of the 1072 patients, 307 died during follow up due to disease progression (n = 189), chronic GVHD (n = 29), infection (n = 23), and others (n = 66). Prognostic factors of transplantation outcomes are shown in Table 1. Overall survival (OS) and event-free survival (EFS) at 3 years after CBT were 63% (95% CI, 36%-90%) and 54% (95% CI, 27%-80%), respectively.

The present study is the first larger report addressing features of chronic GVHD in CBT. The incidence of chronic GVHD after CBT was 28% in our study, which was comparable with the previous smaller reports on CBT conducted in Western countries^3-6  and was lower than that after BMT from unrelated donors (UR-BMT) in Japanese patients.^7,8  Since the genetic backgrounds of the Japanese people are relatively homogeneous compared with those of the Western people, the incidences of chronic GVHD following UR-BMT are reportedly low in Japanese patients; our results in CBT contrast with such previous investigations.^7,8  The present study suggests that the mechanism of GVHD development is possibly affected by factors other than the genetic backgrounds of the patients' ethnicities. There are few previous studies on treatment responses of chronic GVHD after CBT. The present study suggests a higher response rate and lower mortality in chronic GVHD patients after CBT than that after UR-BMT and PBSCT.^9-16  The low overall incidence despite HLA mismatches, relatively low need for systemic immunosuppression, and positive survival impact indicate milder nature of chronic GVHD after CBT. This may be related to the immature lymphocytes in cord blood that are immunologically tolerant,^17,18  although the basic mechanism is unknown and awaits further investigations.

The present study demonstrated that chronic GVHD after CBT improved EFS and reduced the risk of disease progression. Although graft-versus-malignancy (GVM) effects of chronic GVHD after CBT have not been reported, our results were comparable with previous studies in BMT and PBSCT.^2,19,20  Chronic GVHD following CBT probably has clinically significant GVM effects. Interestingly, chronic GVHD after CBT improved OS as well. These findings contrasted with the unclear effects of chronic GVHD on OS improvement in conventional allogeneic HSCT.^2,19,20  The observations are probably associated with the possible favorable impact of chronic GVHD after CBT on nonrelapse mortality.

The present study provided useful information on chronic GVHD after CBT, although some issues remain to be discussed. First, since our study was retrospective, unrecognized bias may have affected the results. Second, we used the classical diagnostic criteria of chronic GVHD²  and did not distinguish the late onset acute GVHD that occurs after day 100 from the classic chronic GVHD as the National Institutes of Health consensus project proposed.²¹  Third, suggestions by our study on future strategies for cord unit selection and GVHD prophylaxis would be informative. Our multivariate analysis suggested an association of mycophenolate mofetil with development of chronic GVHD, whereas methotrexate was associated with better nonrelapse mortality. Although the mechanism is uncertain, clinicians may prefer to use methotrexate as GVHD prophylaxis. Since GVM effects of chronic GVHD were suggested, risk stratification may be possible in cord blood selection. Using cord blood with higher HLA mismatches for patients with high-risk primary diseases might be permissible. Finally, the applicability of our results to patients with higher body weight with heterogeneous genetic backgrounds in Western countries remains to be discussed. All of our patients received single cord blood units, whereas the majority of cord blood transplants are double cord blood units in Western countries. However, the incidence of chronic GVHD is lower and response to immunosuppressants is better in CBT than in UR-BMT in Western countries²² ; hence, the milder nature with lower incidence in chronic GVHD following CBT is probably common among Japanese and Westerns patients. The features of chronic GVHD following CBT in Western adult patients require further investigations.

In conclusion, the present study demonstrated that the incidence of chronic GVHD after CBT is probably lower than BMT or PBSCT, and is associated with improved survival. More intensive studies on the biology of chronic GVHD in CBT are warranted for our understanding and better management of chronic GVHD.

We are grateful to Dr Takanori Teshima at the Center for Cellular and Molecular Medicine of Kyushu University Hospital (Fukuoka, Japan) for providing useful information.

This work was supported by a Research Grant on Tissue Engineering (H17-14) from the Ministry of Health, Labor and Welfare of Japan.

Contribution: H.N., S.M., and M.K. designed the research and wrote the paper; T.Y. analyzed data; T.M., K.Y., N.M., E.K., Y. Kodama, T.K., H.S., Y. Kouzai, M.O., Y.O., R.K., M.I., and S. Takahashi performed research; S. Kai, K.K., T.I.-N., and S. Taniguchi managed the data; and S. Kato took leadership of the authors.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Hiroto Narimatsu, Division of Exploratory Research, Institute of Medical Science, the University of Tokyo, 4-6-1, Shirokanedai, Minato-ku 180-8639, Tokyo, Japan; e-mail: narimt54@med.nagoya-u.ac.jp.