Long-Term Morbidity and Mortality in Children with Chronic
Graft-versus-Host Disease Classi
ﬁed by National Institutes of
Health Consensus Criteria after Allogeneic Hematopoietic Stem
, Hiroshi Moritake2
, Takuro Nishikawa3
, Nobuyuki Hyakuna4
, Masahiko Okada5
, Kozo Nagai7
, Yuko Honda8
, Maiko Shimomura9
, Reiji Fukano1
, Koichiro Kurauchi1
, Shinji Tanioka1
, Jun Okamura1 1Department of Pediatrics, National Kyushu Cancer Center, Fukuoka, Japan
2Division of Pediatrics, Faculty of Medicine, University of Miyazaki, Miyazaki, Japan
3Department of Pediatrics, Kagoshima University Graduate School of Medical and Dental of Sciences, Kagoshima, Japan 4Division of Pediatrics, Center of Bone Marrow Transplantation, Hospital of University of the Ryukyus, Okinawa, Japan 5Department of Pediatrics, Nagasaki University School of Medicine, Nagasaki, Japan
6Division of General Pediatrics and Emergency Medicine, Department of Pediatrics, Oita University Faculty of Medicine, Oita, Japan 7Department of Pediatrics, Faculty of Medicine, Saga University, Saga, Japan
8Department of Pediatrics, School of Medicine, University of Occupational and Environmental Health, Kitakyushu, Japan 9Department of Pediatrics, Yamaguchi University Hospital, Yamaguchi, Japan
Article history: Received 3 June 2015 Accepted 27 July 2015 Key Words: Chronic graft-versus-host disease Children Nonrelapse mortality Functional impairment Performance status National Institutes of Health consensus criteria
a b s t r a c t
We report the long-term morbidity and mortality of 105 pediatric patients who developed chronic graft-versus-host disease (cGVHD) after allogeneic hematopoietic stem cell transplantation (HSCT). According to the consensus criteria of the National Institutes of Health, the global severity of cGVHD was mild in 26 pa-tients (25%), moderate in 30 papa-tients (29%), and severe in 49 papa-tients (47%). Papa-tients with severe cGVHD had a signiﬁcantly lower cumulative incidence of cGVHD remission and higher probability of continuing cGVHD at 8 years from cGVHD diagnosis compared with those with mild or moderate cGVHD. The 10-year cumulative incidence of nonrelapse mortality in severe cGVHD patients was signiﬁcantly higher and the probability of disease-free survival was signiﬁcantly lower than those among patients with mild and moderate cGVHD. Of the 59 patients who survived for more than 5 years, 20 (34%) (4 with moderate and 16 with severe cGVHD) had persistent functional impairment caused by cGVHD with a Karnofsky/Lansky performance score of 90% in 3 patients, 80% in 4 patients, and below 70% in 13 patients at the time of relapse, death, or last follow-up. Better therapeutic strategies are needed to lower the incidence of severe cGVHD, considering the longer life expectancy of pediatric HSCT survivors.
Ó 2015 American Society for Blood and Marrow Transplantation.
Chronic graft-versus-host disease (cGVHD) is the major cause of nonrelapse mortality (NRM) in patients who survive for more than 2 years after allogeneic hematopoietic stem cell transplantation (HSCT) [1,2]. A signiﬁcantly increased risk of NRM among patients with cGVHD has been reported
to persist for at least 15 years after HSCT. The major cause of death is life-threatening infection resulting from long-lasting immunosuppressive treatment (IST), followed by progressive organ failure due to cGVHD involvement[1,4,5]. In addition, cGVHD can lead to substantial deﬁcits in quality of life (QOL), such as respiratory distress, joint contractures, loss of sight, and sexual inactivity[4-6]. Prolonged treatment with steroids, which are mainly used in theﬁrst-line treat-ment of cGVHD, may further compromise QOL. Several studies demonstrated that cGVHD has a signiﬁcant associa-tion with the adverse health status of HSCT survivors[3,5,6]. For pediatric HSCT recipients, considering their long life expectancy and developmental process, the decreased
Financial disclosure: See Acknowledgments on page 1979.
* Correspondence and reprint requests: Jiro Inagaki, MD, Department of Pediatrics, National Kyushu Cancer Center, 3-1-1, Notame, Minami-ku, Fukuoka, 811-1395, Japan.
E-mail address:email@example.com(J. Inagaki).
1083-8791/Ó 2015 American Society for Blood and Marrow Transplantation.
Biology of Blood and
Marrow Transplantationj o u r n a l h o m e p a g e : w w w . b b m t . o r g
physical, and functional status due to cGVHD is a cause for concern because these debilitating sequelae are, in many cases, persistent or irreversible. A number of studies have evaluated the risk factors for the development of cGVHD after HSCT in adult and children [7-12]or for survival and NRM for adult patients who developed cGVHD [13-17]. However, only a few studies have reported the long-term consequences of cGVHD in children, including the probabil-ity of resolution of cGVHD or the probabilprobabil-ity of discontinu-ation of IST[10,14], and little is known about the prevalence of long-lasting symptoms of cGVHD and the physical status in pediatric HSCT survivors.
The present study aimed to investigate the long-term morbidity and mortality of children with cGVHD, focusing on the prevalence of functional impairment in the involved organs and performance status, affecting their QOL and mortality.
Another aim of the present study was to assess the prognostic implication of the consensus criteria of the Na-tional Institutes of Health (NIH)for children with cGVHD. Since the NIH criteria were initially proposed in 2005, several studies have evaluated the utility of the criteria for predicting the prognosis of patients with cGVHD, mostly adult HSCT recipients [15,18-20]. To date, little is known about the applicability of the NIH criteria in pediatric HSCT recipients
. Therefore, we retrospectively reclassiﬁed the severity of cGVHD according to the NIH criteria for pediatric patients with cGVHD included in the present study and analyzed the long-term outcomes, including relapse, NRM, survival, and outcome of cGVHD.
PATIENTS AND METHODS
Patients and Transplantation Procedures
Data of patients who developed cGVHD after allogeneic HSCT conducted at the Department of Pediatrics, National Kyushu Cancer Center between 1995 and 2012 were retrospectively analyzed. All patients or their guardians provided written informed consent to all aspects of the HSCT procedure. This study was approved by the institutional review board of the National Kyushu Cancer Center and was conducted in accordance with the Declara-tion of Helsinki. All patients achieved neutrophil engraftment and survived for at least 100 days. We reviewed the department’s HSCT database, in which patients’ data for procedures and outcomes of HSCT were recorded using a uniform data sheet. Follow-up data regarding the status of under-lying disease and physical health status, including cGVHD, transplantation-related complications, late effects, and performance scores, were updated at least annually or as appropriate according to patients’ visits to our clinic. The information related to cGVHD includes onset day, affected organs, type of onset (de novo, quiescent, or progressive), stage (limited or extensive), and treatments. We also reviewed the medical records of each patient to obtain data regarding the clinical course and outcome of cGVHD, including response to treatments, functional status of involved organs, duration of IST, and adverse events. Data of the patients who had not visited our clinic for more than 1 year were obtained using a questionnaire created for use in the present study from referring hospitals that provided primary care for patients.
A total of 105 patients were reclassiﬁed as cases of classic cGVHD or overlap syndrome, as outlined in the NIH criteria. Patient characteristics are summarized inTable 1. The patients consisted of 57 boys and 48 girls with a median age of 9 (range, 0 to 18) years at the time of HSCT. The most common diagnosis was acute lymphoblastic leukemia (48%), followed by acute myeloid leukemia (18%). Nine patients (9%) had nonmalignant disease. Sixty-two patients (59%) were assigned to the low-risk group for disease recurrence, including acute leukemia inﬁrst or second remission, chronic myeloid leukemia inﬁrst chronic phase, myelodysplastic syndrome without excess blasts, non-Hodgkin lymphoma inﬁrst or second remission, and nonmalignant diseases. The high-risk group included all other diseases and stages. A myeloablative conditioning regimen was used in 94 patients (90%) and the remaining 11 (10%) received a reduced-intensity conditioning regimen. Bone marrow (BM) was the most common stem cell source (83%) followed by cord blood (10%) and peripheral blood (PB) stem cells (8%). Thirty-one patients (30%) underwent HSCT from HLA-matched related do-nors, 51 (49%) from matched unrelated dodo-nors, and 23 (21%) from
mismatched donors. Seventy-six patients (73%) had prior grades II to IV acute GVHD (aGVHD) and 5 patients received donor lymphocyte infusions before developing cGVHD.
Deﬁnitions and Statistical Analysis
The severity of cGVHD was reclassiﬁed according to the NIH criteria. In these criteria, each organ site involved is scored according to a 4-point scale (0 to 3), in which with a score of 0 indicates no symptoms, a score of 1 indicates no clinically signiﬁcant functional impairment, a score of 2 indicates clinically signiﬁcant but not severe impairment, and a score of 3 indicates severe impairment. The overall clinical severity of cGVHD was also evaluated using the global scoring classiﬁed in the criteria as mild, moderate, or severe according to the number of organs affected and the severity in each affected organ. Mild cGVHD involved 1 or 2 organs, with a maximum score of 1. Moderate cGVHD involved at least 1 organ with a maximum score of 2, or 3 or more organs with a maximum score of 1. Severe cGVHD indicated major disability with a score of 3 in any organ. A lung score of 1 was considered moderate and a score of 2 or greater was considered severe cGVHD. In the present study, cGVHD remission was deﬁned as the achievement of score 0 or 1 (no clinically signiﬁcant func-tional impairment) in all affected organs without systemic IST. To the contrary, continuing cGVHD was deﬁned as at least 1 sustained symptom with clinically signiﬁcant impairment (score of 2 or 3, or score of 1 to 3 in the lungs), including both active disease and past tissue damage, or continuation of systemic IST.
Disease recurrence was deﬁned as any evidence of malignancy regardless of morphological, cytogenetic, or molecular relapse after achievement of Table 1
Characteristic Value Age, median (range), yr 9 (0-18) Sex Male 57 (54) Female 48 (46) Diagnosis ALL 50 (48) AML 19 (18) CML 8 (8) MDS/JMML 14 (13) NHL 5 (5) AA 6 (6) WAS 3 (3) Disease risk Low 62 (59) High 43 (41) Donor HLA-matched related 31 (30) HLA-mismatched related 11 (10) HLA-matched unrelated 51 (49) HLA-mismatched unrelated 12 (11) Stem cell source
BM 87 (83) CB 10 (10) PB 8 (8) Conditioning regimen Myeloablative 94 (90) Reduced-intensity 11 (10) GVHD prophylaxis MTX 6 (6) CsA sMTX 31 (30) Tac sMTX 68 (65)
Prior acute GVHD grade
0 16 (15)
I 13 (12)
II 41 (39)
III-IV 35 (33)
Follow-up period of survivors from diagnosis of chronic GVHD, median (range), mo
96 (12-228) ALL indicates acute lymphoblastic leukemia; AML, acute myeloid leukemia; CML, chronic myeloid leukemia; MDS, myelodysplastic syndrome; JMML, juvenile myelomonocytic leukemia; NHL, non-Hodgkin lymphoma; AA, aplastic anemia; WAS, Wiscott-Aldrich syndrome; CB, cord blood; MTX, methotrexate; CsA, cyclosporine A; sMTX, short-term methotrexate; Tac, tacrolimus.
complete remission of malignant diseases, or the emergence of>5% recip-ient chimerism after the achievement of complete donor chimerism for nonmalignant diseases. NRM was deﬁned as death from any cause without evidence of disease recurrence. Disease-free survival (DFS) was deﬁned as the time from the day of cGVHD diagnosis to the day of the last follow-up or ﬁrst event, including disease recurrence and NRM. The probability of DFS was estimated by the Kaplan-Meier method. Univariate analysis of DFS was performed using the log-rank test. The cumulative incidence of disease recurrence and NRM were estimated on the basis of the cumulative inci-dence method treating each recurrent disease and NRM as a competing risk for the other. The groups were compared using Gray’s test. Variables considered were patient’s age, sex, disease risk group, donor type, stem cell source, HLA disparity, conditioning regimen, use of total body irradiation (TBI) 10 Gy in conditioning regimen, time of cGVHD onset, presence of IST just before cGVHD, number of involved organs, number of treatment agents for cGVHD, global score by NIH criteria, presence of lung involvement, liver involvement, thrombocytopenia, previous II to IV acute GVHD, and cyto-megalovirus (CMV) antigenemia. The factors that were found to be signiﬁ-cant at a P value of less than .05 were entered into the multivariate analysis. Multivariate analysis was performed using the Cox proportional hazard regression model for DFS and the method of Fine and Gray for disease recurrence and NRM. The cumulative incidence of cGVHD remission was calculated, with disease recurrence and NRM considered as a competing risk. The cumulative incidence of disease recurrence as well as NRM and cGVHD remission were deﬁned as the time from the day of cGVHD diagnosis to the day of each event. The probability of continuing cGVHD was estimated using the Kaplan-Meier method calculating as the time from the day of cGVHD diagnosis to the day of the last follow-up or theﬁrst event, including disease recurrence, NRM, and cGVHD remission. Statistical analyses were performed using EZR software (Saitama Medical Center, Jichi Medical Uni-versity) and a P value of less than .05 was considered to be statistically signiﬁcant.
The clinical features of cGVHD are summarized inTable 2. The median time from HSCT to the onset of cGVHD was 133 days (range, 58 to 617). The global severity of cGVHD was mild in 26 patients (25%), moderate in 30 patients (29%), and severe in 49 patients (47%), according to NIH criteria. His-tological diagnosis of cGVHD was determined in 26 patients using specimens of skin in 5 patients, liver in 6, oral mucosa in 4, gastrointestinal (GI) in 9, and lung in 2. The most frequently affected organs were the skin (29%) and mouth (24%), followed by liver (15%), GI tract (9%), and eyes (8%). Other manifestations included thrombocytopenia (5%), eosinophilia (1%), and serositis (1%). Fifty-ﬁve patients (52%) had involvement of 3 or more organs. Sixty-three patients (60%) had received calcineurin inhibitor (CNI) with or without a steroid immediately before the onset of chronic GVHD, whereas 26 patients (25%) had not received any IST. After onset of cGVHD, 29 patients did not receive any new or additional systemic IST because of mild or moderate cGVHD requiring only local therapy in 25 patients, parents’ refusal of systemic IST in 2, and unknown reasons in 2. On the other hand, initial systemic IST was administered to 76 patients. An increase in the dose of steroids or CNI was also considered as initial systemic IST if patients were already being treated with these agents. Fifty-two patients (68%) received steroids (prednisone or methylprednisolone) with or without other agents, and 24 patients (32%) received a CNI, weekly low-dose methotrexate, or mycophenolate mofetil as initial IST. The reasons for not using steroids as initial IST in these 24 patients were as follows: long-lasting administration of steroids until the onset of cGVHD in 13 patients, develop-ment of cGVHD during tapering of the dose of CNI in 5, parents’ refusal of steroid administration in 2, and unknown in 4. In 19 patients (18%), 4 or more agents were used through the treatment of cGVHD.
A total of 18 patients (17%), all of who had malignant disease, experienced disease recurrence at a median dura-tion of 298 days (range, 66 to 2995 days) after the diagnosis of cGVHD. Eight (44%) patients had mild cGVHD by NIH global score, 6 (33%) had moderate disease, and 4 (22%) had severe cGVHD. The estimated cumulative incidence of relapse at 5 years for the entire cohort was 16.3% (95% con-ﬁdence interval [CI], 10.0 to 24.1) and the incidence in pa-tients with severe cGVHD (8.3%; 95% CI, 2.6 to 18.3) was signiﬁcantly lower compared with those with mild and moderate cGVHD (23.4%; 95% CI, 13.2 to 35.2) (P ¼ .027) (Figure 1A). In multivariate analysis, TBI 10 Gyecontaining conditioning regimen and mild cGVHD were associated with higher cumulative incidence of relapse, whereas positive CMV antigenemia after HSCT and severe cGVHD were factors associated with a lower incidence of relapse (Table 3).
Overall, 22 patients (21%) died of nonrelapse causes at a median of 382 days (range, 26 to 5975 days) after the onset of cGVHD. Eighteen (82%) patients had severe cGVHD, 3 (14%) had moderate, and 1 (5%) had mild cGVHD. The causes of death among these patients are shown inTable 4. The major causes of death were fatal infections associated with IST, followed by bronchiolitis obliterans (BO). The estimated
Characteristics of cGVHD
Characteristic Value Global score (NIH)
Mild 26 (25) Moderate 30 (29) Severe 49 (47) Involved organ Skin 84 (29) Mouth 69 (24) Liver 44 (15) GI tract 25 (9) Eyes 22 (8) Lungs 14 (5)
Joint and fascia 7 (2)
Others 20 (7)
Number of involved organs
1 or 2 50 (48)
3 55 (52)
IST just before onset of chronic GVHD
None 26 (25)
CNI 41 (39)
CNIþ steroid MTX/MMF 27 (26) Steroid MTX 10 (10)
MTX 1 (1)
Initial IST for chronic GVHD*
None 29 (28) CNI 10 (10) CNIþ steroid 4 (4) Steroid 46 (44) Steroidþ MTX 2 (2) MTX 13 (12) MMF 1 (1)
Number of agents used for IST of cGVHD
0-1 37 (35)
2-3 49 (47)
4 19 (18)
Time from HSCT to onset of chronic GVHD, median (range), d
133 (58-617) MMF indicates mycophenolate mofetil.
Data presented are n (%), unless otherwise indicated.
*An increase in the dose of CNI or steroid in patients who were already being treated with these agents was considered initial IST.
cumulative incidences of NRM for the entire cohort at 5, 10, and 15 years were 15.8% (95% CI, 9.4 to 23.7), 18.1% (95% CI, 11.2 to 26.4), and 24.4% (95% CI, 15.3 to 34.6), respectively. The cumulative incidence of NRM at 10 years was signi ﬁ-cantly higher in patients with severe cGVHD (34.5%, 95% CI, 21.0 to 48.5) compared with that in those with mild cGVHD (3.8%; 95% CI, .3 to 16.9) or moderate cGVHD (4.4%; 95% CI, .3 to 19.1) (P¼ .0012) (Figure 1B). NRM of patients with mod-erate and severe cGVHD continued to increase even after 10 years of the onset of cGVHD. In multivariate analysis, factors associated with higher NRM were severe cGVHD, cGVHD in the lungs, 3 or more organs affected by cGVHD, PB as the stem cell source, and positive CMV antigenemia after HSCT (Table 3).
The median follow-up of survivors was 96 months (range, 12 to 228 months) after diagnosis of cGVHD. In the entire cohort, the probabilities of DFS at 5, 10, and 15 years were 69.2% (95% CI, 59.0 to 77.3), 66.6% (95% CI, 56.2 to 75.1), and 58.5% (95% CI, 45.6 to 69.3), respectively. Ten-year DFS of patients with mild, moderate, and severe cGVHD was 71.2% (95% CI, 48.6 to 85.2), 78.8% (95% CI, 58.6 to 89.9) and 57.3% (95% CI, 41.8 to 70.2), respectively (Figure 1C). The survival curve of patients with mild cGVHD reached a plateau at 5 years after the diagnosis of cGVHD, whereas that of pa-tients with moderate and severe cGVHD continued to decrease and the 17-year DFS was 64.7% (95% CI, 38.7 to 81.9) and 34.7% (95% CI, 9.2 to 62.6), respectively. The DFS of pa-tients with severe cGVHD was signiﬁcantly lower compared with those with mild and moderate cGVHD (P ¼ .037). In multivariate analysis, TBI >10 Gyecontaining conditioning regimen and cGVHD in the liver and lungs were each asso-ciated with lower DFS (Table 3).
Outcomes of Chronic GVHD
Except for the patients who relapsed or died, 44 patients achieved cGVHD remission at a median of 415 days (range, 30 to 2337 days) from diagnosis of cGVHD. Of the 44 patients, 17 (39%) patients had mild cGVHD, 18 (41%) had moderate dis-ease, and only 9 (20%) had severe cGVHD. In the entire cohort, the cumulative incidence of cGVHD remission at 8 years was 44.6% (95% CI, 34.4 to 54.3). The cumulative incidence of remission at 8 years in patients with severe cGVHD was signiﬁcantly lower (21.1%; 95% CI, 10.1 to 34.8) than that of patients with mild cGVHD (65.4%; 95% CI, 41.8 to 81.3) or moderate cGVHD (64.6%; 95% CI, 42.9 to 79.8) (P< .001) (Figure 1D). On the other hand, the probability of continuing cGVHD at 8 years was 20.9% (95% CI, 13.3 to 29.7) in the entire cohort; 3.8% (95% CI, .3 to 16.4) in patients with mild cGVHD, 11.4% (95% CI, 2.9 to 26.3) in patients with moderate disease, and 36.1% (95% CI, 21.9 to 50.4) in patients with severe cGVHD (P< .001) (Figure 2).
Prognosis and Status of cGVHD for Long-term Survivors At the time of the present study, 59 patients survived for more than 5 years from diagnosis of cGVHD. Among them, relapse occurred in 1 patient (2%) and NRM occurred in 7 patients (12%). The causes of NRM were BO (n¼ 3), bacterial pneumonia (n ¼ 1), second malignant neoplasm (n ¼ 1), suicide (n¼ 1), and accident (n ¼ 1). At the time of relapse, death, or last follow-up, 39 patients (66%) were in remission of cGVHD, with a Karnofsky/Lansky performance score of 100% in 28 patients, 90% in 9 patients, and 0% in 2 deceased patients. On the other hand, 20 patients (34%), including 4 with moderate cGVHD and 16 with severe cGVHD, had continuing cGVHD with a Karnofsky/Lansky performance score of 90% in 3 patients, 80% in 4 patients, and below 70% in 13 patients; the organs with persistent functional
Figure 1. Outcomes of patients with cGVHD according to the NIH global severity. (A) Cumulative incidence of disease recurrence. (B) Cumulative incidence of NRM. (C) Probability of DFS. (D) Cumulative incidence of cGVHD remission.
impairment with a score of 2 or more by the NIH criteria were the skin (n¼ 11), eyes (n ¼ 4), fascia (n ¼ 4), mouth (n¼ 2), liver (n ¼ 2), GI tract (n ¼ 1), and lungs (score of 1 or more) (n¼ 7).
As has been reported in a number of previous studies, long-term survivors of pediatric HSCT have signiﬁcantly higher risk of late effects and adverse health status, including physical, psychological, and cognitive problems, compared with cancer survivors who did not undergo HSCT[22-25]or the general population[22,26,27]. Most of these studies have demonstrated that cGVHD had a signiﬁcant impact on many aspects of health status resulting in a worse health-related
QOL. One large study of long-term health-related out-comes in survivors of childhood cancer from the Bone Marrow Transplant Survivor Study and Childhood Cancer Survivor Study showed that allogeneic HSCT recipients with active cGVHD were at greater risk for poor general health, functional impairment, activity impairment, and pain compared with the survivors treated with conventional therapy.
One of the main objectives of the present study was to investigate the prevalence of morbidity due to long-lasting symptoms of cGVHD in pediatric HSCT survivors. We evalu-ated the outcomes of cGVHD on the basis of the functional status of the involved organs according to NIH score, ie, cGVHD remission was deﬁned as the achievement of a score of 0 or 1 (no clinically signiﬁcant functional impairment) in all affected organs, because we were primarily concerned with the debilitating sequelae that might affect QOL. In the entire study cohort, the cumulative incidence of cGVHD remission was 44.6% and the probability of continuing cGVHD was 20.9% at 8 years after diagnosis of cGVHD. These results were similar to those described in 2 large retrospec-tive studies on children from the Center for International Blood and Marrow Transplant Researchand the Italian Association for Pediatric Hematology and Oncology. The results of the present study suggest that most surviving pa-tients who developed mild and moderate cGVHD eventually achieved cGVHD remission, whereas only approximately one third of the surviving patients who developed severe cGVHD achieved remission, even after 8 years from diagnosis of cGVHD. For long-term survivors with continuing cGVHD, the major impaired organs were the skin, followed by lungs, eyes, and fascia, and their performance scores at the last follow-up were poor. Most of these functional deﬁcits will continue for the rest of the patients’ lives and may compro-mise their QOL. Besides these sequelae of involved organs by cGVHD, we should also consider the risk of a worsening of health status caused by prolonged IST, such as diabetes mellitus, osteoporosis, and recurrent infections. Compre-hensive and continuous support, including medical care and welfare, will be necessary.
As described for adult cGVHD patients[15,19,20], the re-sults of the present study demonstrated that the NIH global score could predict the risk of mortality even in children with cGVHD. Consistent with the lower cumulative incidence of cGVHD remission, the cumulative incidence of NRM was signiﬁcantly higher in patients with severe cGVHD. Most causes of death were associated with cGVHD, including BO and fatal infections during long-term IST. The risk of NRM continued to increase even 10 years after the diagnosis of cGVHD, as reported in a previous study. Therefore, pa-tients with severe cGVHD showed lower probability of DFS compared with those with mild and moderate cGVHD. Considering the higher mortality as well as poorer QOL, ef-forts to lessen the risk of developing severe cGVHD are necessary. The effectiveness of post-transplantation cyclo-phosphamide and antithymocyte globulinecontaining pretransplantation conditioning regimenhas been re-ported as prophylaxis for acute and cGVHD in various situ-ations of HSCT, ranging from HLA-matched related BM transplantation to haploidentical PB stem cell trans-plantation. Therefore, decreased risk of developing severe cGVHD is expected in future.
Several factors associated with increased risk of NRM were previously identiﬁed among adult and children with cGVHD, such as low performance score, thrombocytopenia,
Multivariate Analysis of Risk Factors for Disease Recurrence, NRM, and DFS in Children with cGVHD Risk Factor HR 95% CI P Disease recurrence TBI 10 Gy No 1 .01 Yes 4.86 1.4-16.5 CMV antigenemia Negative 1 .01 Positive .2 .05-.7 NIH grade
Mild and moderate 1 .001 Severe .07 .01-.4
Stem cell source
BM and CB 1 .01 PB 1.99 1.2-3.4 CMV antigenemia Negative 1 .04 Positive 2.8 1.0-7.6 NIH grade
Mild and moderate 1 .02 Severe 3.07 1.2-7.8
No. of involved organ
1 or 2 1 .02 3 3.98 1.2-13.0 Lung involvement No 1 .03 Yes 2.85 1.1-7.5 DFS TBI 10 Gy No 1 .005 Yes 2.93 1.4-6.2 Liver involvement No 1 .004 Yes 2.69 1.4-5.3 Lung involvement No 1 <.001 Yes 4.4 2.0-9.6 HR indicates hazard ratio.
Causes of Nonrelapse Death for Patients with cGVHD
Cause Entire study cohort (n¼ 105) Invasive fungal infection 6
Bacterial pneumonia 2 Necrotizing soft tissue infection 1
Idiopathic interstitial pneumonia 2 Intracranial hemorrhage 2 Second malignant neoplasm 2
elevated bilirubin, extensive skin involvement, progressive onset of cGVHD, and poor response to ﬁrst-line treatment
[13,14,17]. In the present study, in the multivariate analysis, 5 risk factors (severe cGVHD, 3 or more affected organs, lung involvement, PB as the stem cell source, and positive CMV antigenemia after HSCT) were identiﬁed. It is reasonable that 3 or more affected organs and lung involvement were asso-ciated with higher NRM because these factors are closely associated with the severity of cGVHD. PB stem cell trans-plantation was also identiﬁed as a risk factor for higher NRM,
as reported from previous studies in children with cGVHD from the Center for International Blood and Marrow Trans-plant Research. Over the past decade, the number of PB stem cell transplantations from both related and unrelated donors has increased and has been reported to be associated with the higher incidence and severe manifestations of cGVHD and with prolonged IST, compared with BM trans-plantations, mainly in adult patients without signiﬁcant differences in survival[7,8,16,31,32]. However, it is uncertain whether the higher incidence of severe cGVHD after PB stem
Figure 2. Frequency of continuing cGVHD and other events. (A) All events that occurred in patients with mild cGVHD. The probability of continuing cGVHD at 8 years after diagnosis of cGVHD was 3.8% (95% CI, .3 to 16.4). (B) All events in patients with moderate cGVHD. The probability of continuing cGVHD at 8 years was 11.4% (95% CI, 2.9 to 26.3). (C) All events in patients with severe cGVHD. The probability of continuing cGVHD at 8 years was 36.1% (95% CI, 21.9 to 50.4).
cell transplantation will reduce survival with longer follow-up in children.
Positive CMV antigenemia after HSCT was identiﬁed as a risk factor for higher NRM and lower incidence of relapse in the present study. The inﬂuence of CMV reactivation after HSCT on HSCT outcomes in patients with hematological malignancy has been reported [33-35]. A previous study demonstrated that developing CMV antigenemia after HSCT is a strong and independent predictor of a reduced risk of leukemic relapse and superior survival in adult patients with acute myeloid leukemia, whereas another study showed no survival beneﬁt because CMV reactivation was associated with increased NRM . There may be similar effects of CMV reactivation on the antileukemic effect and on the toxicity after HSCT, even in the cohort limited to the patients who developed cGVHD.
The limitations of the present study include its retro-spective design, use of a single-center experience, small number of patients included, and heterogeneity of trans-plantation factors, such as stem cell source, donor type, and HLA disparity. In addition, the possibility of ambiguities in the diagnosis of cGVHD cannot be eliminated because of the diagnostic bias of the attending physicians, particularly for patients diagnosed before 2005, and the relatively low number of histologically conﬁrmed diagnoses. The evalua-tion of treatment response and cGVHD outcomes also depend on the subjectivity of the attending physicians without the performance of objective measurements. On the other hand, the strength of the present study is that the observational period was long enough to indicate the late morbidity and mortality among children with cGVHD. In addition, we analyzed the detailed outcomes of cGVHD, including the functional status of involved organs, length of IST exposure, and performance score in each patient with close follow-up, which may be the beneﬁt of a single-center study.
A number of previous studies have shown that cGVHD was associated with a lower incidence of relapse; however, data on the survival beneﬁt of cGVHD are conﬂicting mainly because of higher NRM in patients with extensive/severe cGVHD for hematological malignancies. With longer follow-up periods, the undesirable effects of persistent cGVHD will be more evident, particularly in pediatric HSCT survivors. In addition, effective salvage treatments for steroid-refractory cGVHD remain to be established. Therefore, developing better therapeutic strategies, including drug-mediated methods and graft manipulation, is needed to lower the incidence of severe cGVHD.
The authors thank all patients who received trans-plantation at our department and the physicians and nursing staff of the transplantation team for providing excellent treatment and care for patients.
Financial disclosure: The authors have no relevant ﬁnan-cial conﬂicts of interest.
Conﬂicts of interest: The authors have no conﬂicts of in-terest to declare.
Authorship statement: J.I. managed the patients, designed this study, analyzed the data, and wrote the manuscript; H.M., T.N., N.H., M.O., S.S., K.N., Y.H., and M.S. managed the patients and collected the data. R.F., M.N., K.K., S.T., and J.O. managed the patients and analyzed the data. All authors reviewed and approved theﬁnal version of the manuscript.
1. Lee SJ, Vogelsang G, Flowers MED. Chronic graft-versus-host disease. Biol Blood Marrow Transplant. 2003;9:215-233.
2. Wingard JR, Majhail NS, Brazauskas R, et al. Long-term survival and late deaths after allogeneic hematopoietic cell transplantation. J Clin Oncol. 2011;29:2230-2239.
3. Bhatia S, Francisco L, Carter A, et al. Late mortality after allogeneic hematopoietic cell transplantation and functional status of long-term survivors: report from the Bone Marrow Transplant Survivor Study. Blood. 2007;110:3784-3792.
4. Vogelsang GB. How I treat chronic graft-versus-host disease. Blood. 2001;97:1196-1201.
5. Filipovich AH, Weisdorf D, Pavletic S, et al. National Institutes of Health consensus development project on criteria for clinical trials in chronic graft-versus-host disease: I. Diagnosis and staging working group report. Biol Blood Marrow Transplant. 2005;11:945-956.
6. Fraser CJ, Bhatia S, Ness K, et al. Impact of chronic graft-versus-host disease on the health status of hematopoietic cell transplantation survivors: a report from the Bone Marrow Transplant Survivor Study. Blood. 2006;108:2867-2873.
7. Flowers ME, Inamoto Y, Carpenter PA, et al. Comparative analysis of risk factors for acute versus-host disease and for chronic graft-versus-host disease according to National Institutes of Health consensus criteria. Blood. 2011;117:3214-3219.
8. Eapen M, Logan BR, Confer DL, et al. Peripheral blood grafts from un-related donors are associated with increased acute and chronic graft-versus-host disease without improved survival. Biol Blood Marrow Transplant. 2007;13:1461-1468.
9. Ozawa S, Nakaseko C, Nishimura M, et al., Japan Marrow Donor Pro-gram. Chronic graft-versus-host disease after allogeneic bone marrow transplantation from an unrelated donor: incidence, risk factors and association with relapse. A report from the Japan Marrow Donor Pro-gram. Br J Haematol. 2007;137:142-151.
10.Zecca M, Prete A, Rondelli R, et al., AIEOP-BMT Group. Italian Associ-ation for Pediatric Hematology and Oncology-Bone Marrow Transplant. Chronic graft-versus-host disease in children: incidence, risk factors, and impact on outcome. Blood. 2002;100:1192-1200.
11.Diaz MA, Vicent MG, Gonzalez ME, et al., Spanish Group for Allogeneic Peripheral Blood Transplantation. Risk assessment and outcome of chronic graft-versus-host disease after allogeneic peripheral blood progenitor cell transplantation in pediatric patients. Bone Marrow Transplant. 2004;34:433-438.
12.Alsultan A, Giller RH, Gao D, et al. GVHD after unrelated cord blood transplant in children: characteristics, severity, risk factors and inﬂu-ence on outcome. Bone Marrow Transplant. 2011;46:668-675. 13.Lee SJ, Klein JP, Barrett AJ, et al. Severity of chronic graft-versus-host
disease: association with treatment-related mortality and relapse. Blood. 2002;100:406-414.
14.Jacobsohn DA, Arora M, Klein JP, et al. Risk factors associated with increased nonrelapse mortality and with poor overall survival in children with chronic graft-versus-host disease. Blood. 2011;118: 4472-4479.
15.Inamoto Y, Martin PJ, Storer BE, et al., Chronic GVHD Consortium. As-sociation of severity of organ involvement with mortality and recur-rent malignancy in patients with chronic graft-versus-host disease. Haematologica. 2014;99:1618-1623.
16.Stewart BL, Storer B, Storek J, et al. Duration of immunosuppressive treatment for chronic graft-versus-host disease. Blood. 2004;104: 3501-3506.
17.Pavletic SZ, Smith LM, Bishop MR, et al. Prognostic factors of chronic graft-versus-host disease after allogeneic blood stem-cell trans-plantation. Am J Hematol. 2005;78:265-274.
18.Vigorito AC, Campregher PV, Storer BE, et al., National Institutes of Health. Evaluation of NIH consensus criteria for classiﬁcation of late acute and chronic GVHD. Blood. 2009;114:702-708.
19.Cho BS, Min CK, Eom KS, et al. Feasibility of NIH consensus criteria for chronic graft-versus-host disease. Leukemia. 2009;23:78-84. 20.Pidala J, Kim J, Anasetti C, et al. The global severity of chronic
graft-versus-host disease, determined by National Institutes of Health consensus criteria, is associated with overall survival and non-relapse mortality. Haematologica. 2011;96:1678-1684.
21.Lee JW, Lee DH, Jang PS, et al. Prognostic implications of the NIH consensus criteria in children with chronic graft-versus-host disease. Yonsei Med J. 2011;52:779-786.
22.Armenian SH, Sun CL, Kawashima T, et al. Long-term health-related outcomes in survivors of childhood cancer treated with HSCT versus conventional therapy: a report from the Bone Marrow Transplant Survivor Study (BMTSS) and Childhood Cancer Survivor Study (CCSS). Blood. 2011;118:1413-1420.
23.Bresters D, van Gils IC, Kollen WJ, et al. High burden of late effects after haematopoietic stem cell transplantation in childhood: a single-centre study. Bone Marrow Transplant. 2010;45:79-85.
24.Berbis J, Michel G, Chastagner P, et al. A French cohort of childhood leukemia survivors: impact of hematopoietic stem cell transplantation
on health status and quality of life. Biol Blood Marrow Transplant. 2013; 19:1065-1072.
25. Michel G, Bordigoni P, Simeoni MC, et al. Health status and quality of life in long-term survivors of childhood leukaemia: the impact of haematopoietic stem cell transplantation. Bone Marrow Transplant. 2007;40:897-904.
26. Hovi L, Kurimo M, Taskinen M, et al. Suboptimal long-term physical performance in children and young adults after pediatric allo-SCT. Bone Marrow Transplant. 2010;45:738-745.
27. Sanders JE, Hoffmeister PA, Storer BE, et al. The quality of life of adult survivors of childhood hematopoietic cell transplant. Bone Marrow Transplant. 2010;45:746-754.
28. Parsons SK, Phipps S, Sung L, et al. NCI, NHLBI/PBMTC First Inter-national Conference on Late Effects after Pediatric Hematopoietic Cell Transplantation: health-related quality of life, functional, and neu-rocognitive outcomes. Biol Blood Marrow Transplant. 2012;18: 162-171.
29. Al-Homsi AS, Roy TS, Cole K, et al. Post-transplant high-dose cyclo-phosphamide for the prevention of graft-versus-host disease. Biol Blood Marrow Transplant. 2015;21:604-611.
30. Storek J, Mohty M, Boelens JJ. Rabbit anti-T cell globulin in allogeneic hematopoietic cell transplantation. Biol Blood Marrow Transplant. 2015; 21:959-970.
31. Anasetti C, Logan BR, Lee SJ, et al., Blood and Marrow Transplant Clinical Trials Network. Peripheral-blood stem cells versus bone marrow from unrelated donors. N Engl J Med. 2012;367:1487-1496. 32. Friedrichs B, Tichelli A, Bacigalupo A, et al. Long-term outcome and late
effects in patients transplanted with mobilised blood or bone marrow: a randomised trial. Lancet Oncol. 2010;11:331-338.
33. Elmaagacli AH, Steckel NK, Koldehoff M, et al. Early human cytomeg-alovirus replication after transplantation is associated with a decreased relapse risk: evidence for a putative virus-versus-leukemia effect in acute myeloid leukemia patients. Blood. 2011;118:1402-1412. 34. Green ML, Leisenring WM, Xie H, et al. CMV reactivation after
alloge-neic HCT and relapse risk: evidence for early protection in acute myeloid leukemia. Blood. 2013;122:1316-1324.
35. Jeljeli M, Guérin-El Khourouj V, Porcher R, et al. Relationship between cytomegalovirus (CMV) reactivation, CMV-driven immunity, overall immune recovery and graft-versus-leukaemia effect in children. Br J Haematol. 2014;166:229-239.