Abstract
Hematopoietic stem cell transplantation (HSCT) is a curative treatment for high‐risk childhood leukemia but may lead to chronic graft‐versus‐host disease (cGvHD), a severe long‐term complication. This study analyzed data from the LEA cohort, including 446 childhood leukemia survivors treated with allogeneic HSCT. The standardized cGvHD evaluation, using the NIH consensus criteria, was conducted 8.7 ± 0.3 years post‐HSCT. Long‐term cGvHD was reported in 21% of patients (9% mild, 7% moderate, 5% severe), primarily affecting eyes, skin, lungs, and mouth, with some cases involving multiple organs. Most patients with long‐term cGvHD were untreated (84%), while 11% received systemic and 5% local treatments. cGvHD was associated with other long‐term complications. Administration of anti‐thymocyte globulin was a significant determinant (OR 0.6, 95% CI 0.4–0.99, P = 0.045). Long‐term cGvHD showed a marked detrimental effect on the quality of life (QoL), even after adjusting for the other long‐term complications. The SF‐36 physical and psychological adjusted composite scores in patients with versus without cGvHD were 50 ± 2 versus 55 ± 1 (P = 0.01) and 38 ± 2 versus 43 ± 1 (P = 0.01), respectively. Even mild and moderate forms significantly affected the QoL, especially on psychological dimensions. These findings support standardized cGvHD evaluation and management to improve long‐term outcomes of transplanted childhood leukemia survivors.
INTRODUCTION
Chronic graft‐versus‐host disease (cGvHD) represents a complex immunological complication of allogeneic hematopoietic stem cell transplantation (HSCT), where the graft's immune response manifests against host organs. This complication is characterized by variable severity but may severely jeopardize the health status and quality of life (QoL) of survivors. While cGvHD is a well‐recognized challenge in adult leukemia survivors post‐HSCT, its prevalence and impact among childhood survivors remain less explored, with limited available data.
Pediatric HSCT recipients are reported to have a lower rate and possibly different presentations of cGvHD compared to adults. 1 , 2 , 3 A study reported the long‐term morbidity and mortality of pediatric patients who developed cGVHD after HSCT. 4 These authors showed that one‐third of the patients with cGvHD had persistent functional impairment caused by this complication. They described the involved organs and the initial therapy. However, this study has a retrospective and monocentric design and included a limited number of patients. A multicenter prospective study validated the National Institutes of Health (NIH) consensus criteria for cGvHD in children and described the most commonly involved organs. 5 However, it was limited to a short‐term evaluation of cGvHD, with a 12‐month follow‐up post‐HSCT. Furthermore, the study did not assess QoL or other transplant‐related complications. A better understanding of long‐term cGvHD, its associated risk factors, and its impact on the QoL in pediatric recipients of HSCT is crucial for enhancing diagnosis, advancing therapeutic approaches, and improving long‐term outcomes.
The primary objective of this study is to describe the long‐term prevalence of cGvHD in patients treated with HSCT for leukemia during childhood. Secondary objectives involve investigating risk factors associated with cGvHD, analyzing the association of cGvHD with other long‐term complications, and exploring its impact on the survivors' QoL. By comprehensively addressing these objectives, we aim to provide a long‐term perspective on the multifaceted nature of cGvHD and its implications for post‐HSCT care.
METHOD
The LEA program
LEA is a long‐term follow‐up program involving the childhood acute leukemia survivors treated in the French participating centers since 1980. As detailed elsewhere, participants are invited to attend the follow‐up clinic at predefined dates, starting one year after HSCT. These visits are scheduled every two years until the participants reach both 20 years of age and 10 years of complete remission, and every four years thereafter. 6 The program began in 2004 and rests on the constitution of a multicenter historical and prospective cohort, which includes both incident cases (diagnosed after the start date of the participation of the center in the LEA program) and prevalent cases (diagnosed between January 1980 and this start date).
Study design
This study was designed as a cross‐sectional evaluation of long‐term cGvHD embedded within the LEA study. HSCT survivors who attended a scheduled LEA follow‐up visit during the study period (2019–2021) were included in the study and underwent a comprehensive cGvHD assessment, as described below. Most of these patients had already attended one or more previous LEA visits, during which cGvHD had been evaluated according to standard practice by a pediatric hematologist. The study was conducted in accordance with a protocol approved by an institutional review board in accordance with the Declaration of Helsinki principles (clinicaltrials.gov identifier: NCT 01756599). All patients or their legal representatives provided a written informed consent.
Outcomes
The long‐term cGvHD data were generated by a pediatric hematologist using (1) a dedicated questionnaire and (2) the eGVHD application (UZ Leuven, Belgium) utilizing the NIH consensus criteria for cGvHD for improving the quality and standardization of the evaluation. 7 History of acute GvHD was collected in the patients' chart, and de novo cGvHD was defined as cGvHD occurring in patients without any prior history of acute GvHD. QoL was evaluated in adults using the SF‐36 questionnaire. 8 The other long‐term complications were diagnosed and recorded during the follow‐up visits in the LEA program: endocrinopathies (gonadal dysfunction, hypothyroidism, short stature), cardiomyopathy, cardiovascular risk factors or complications (obesity, metabolic syndrome, diabetes, coronary artery disease, stroke), subsequent malignant neoplasm or meningioma, pulmonary dysfunction, bone defects (osteonecrosis, osteoporosis), neurological complications, chronic kidney disease, sensory defects (cataract, hearing loss), iron overload, and alopecia. These complications are defined in the Supporting Information S1: Table 1.
Statistical analysis
Statistical analysis was performed using SPSS 20.0 software (SPSS Inc.). Quantitative variables were expressed as mean and standard error of the mean (SEM). Categorical variables were compared using the χ 2 test or Fisher's exact test. Quantitative variables were compared using the Student's t‐test or the Mann–Whitney test. The associations between cGvHD and potential risk factors were initially analyzed via univariate logistic regression. The variables associated with cGvHD in the univariate analysis with a P < 0.05 were then included in a multivariate analysis. Odds ratios are displayed with a 95% confidence interval (CI).
RESULTS
Patients and HSCT
The study included 446 patients whose characteristics are described in Table 1. Patients were treated for acute lymphoblastic leukemia (61%) or acute myeloid leukemia or another myeloid neoplasm (39%). The age at HSCT was 9.0 ± 0.3 years (mean ± SEM). At the time of HSCT, 211 patients (47%) had a history of leukemia relapse before HSCT. Myeloablative conditioning and total body irradiation were administered to 96% and 56% of patients, respectively. Anti‐thymocyte globulin and ciclosporin were administered to 53% and 94% of patients, respectively. HLA‐matched sibling donors (10/10) were the most common donor source, used in 144 patients (32%), followed by HLA‐matched unrelated donors (10/10) in 128 patients (29%). Alternative donor sources included cord blood in 101 patients (23%), comprising HLA‐matched (6/6), mismatched (3–5/6), and double cord blood units, as well as HLA‐mismatched unrelated donors (9/10) in 57 patients (13%). Haploidentical family donors with post‐transplant cyclophosphamide (PTCy) were used in 12 patients (3%). Other family donors included HLA‐matched nonsibling donors (0.7%) and HLA‐mismatched nonsibling donors (0.2%). All grafts administered to the patients in this study were unmanipulated. Among the 446 patients, 70% received hematopoietic stem cells from bone marrow, while cord blood was used in 23% and peripheral blood stem cells in 7%. Regarding acute GvHD, 37% of patients had no history of acute GvHD, while 21% experienced grade 1, 27% grade 2, 13% grade 3, and 2% grade 4 acute GvHD.
Table 1.
Description of patients and HSCT.
| Patients and HSCT characteristics (n = 446) | n (%) or m ± SEM |
|---|---|
| Female sex | 199 (45%) |
| Type of leukemia | |
| Acute lymphoblastic leukemia | 271 (61%) |
| Acute myeloid leukemia and other myeloid neoplasms | 175 (39%) |
| History of leukemia relapse before HSCT | 211 (47%) |
| Age at leukemia diagnosis (years) | 7.3 ± 0.2 |
| HSCTa | |
| Age at HSCT (years) | 9.0 ± 0.3 years |
| Second HSCT | 18 (4%) |
| Myeloablative conditioning | 398 (96%) |
| Total body irradiation | 249 (56%) |
| Anti‐thymocyte globulin | 213 (53%) |
| Ciclosporin | 409 (94%) |
| Donor and HLA matcha | |
| Family donors | |
| HLA‐matched sibling donor (10/10) | 144 (32%) |
| HLA‐matched nonsibling related donor | 3 (0.7%) |
| HLA‐mismatched nonsibling family donor (9/10) | 1 (0.2%) |
| Haploidentical family donor (with PTCy) | 12 (3%) |
| Unrelated donors | |
| HLA‐matched unrelated donor (10/10) | 128 (29%) |
| HLA‐mismatched unrelated (9/10) | 57 (13%) |
| Unrelated cord blood | |
| HLA‐matched cord blood (6/6) | 14 (3%) |
| HLA‐mismatched cord blood (5/6) | 49 (11%) |
| HLA‐mismatched cord blood (3‐4/6) | 23 (5%) |
| Double cord blood | 15 (3%) |
| Stem cell source | |
| Bone marrow | 313 (70%) |
| PBSC | 32 (7%) |
| Cord blood | 101 (23%) |
| History of acute GvHD (maximal grade) | |
| None | 164 (37%) |
| 1 | 95 (21%) |
| 2 | 119 (27%) |
| 3 | 59 (13%) |
| 4 | 9 (2%) |
| Number of long‐term evaluations | |
| 1 | 124 (28%) |
| 2 | 77 (17%) |
| 3 | 87 (20%) |
| 4 | 76 (17%) |
| ≥5 | 82 (18%) |
Abbreviations: cGvHD, chronic graft‐versus‐host disease; HLA, human leukocyte antigen; HSCT, hematopoietic stem cell transplantation; m, mean; PBSC, peripheral blood stem cells; PTCy, posttransplant cyclophosphamide; SEM, standard error of the mean.
Last HSCT in patients who received two HSCTs.
Prevalence and characteristics of long‐term cGvHD
Long‐term cGvHD data are shown in Table 2 and Figure 1. The specific long‐term evaluation of cGvHD was performed 8.7 ± 0.3 years after HSCT. The mean age at evaluation was 17.7 ± 0.4 years. At the time of the study, 324 patients (73%) did not have and never had any cGvHD, 28 patients (6%) had completely recovered from cGvHD, while 94 patients (21%) were diagnosed with cGvHD. The severity at the time of evaluation was as follows: 45% mild, 33% moderate, and 22% severe. De novo cGvHD, defined as cGvHD without any history of acute GvHD, occurred in 35 out of 94 patients (37%). Patients mostly had isolated organ involvement, most frequently affecting the skin, eyes, and mouth. Lungs were involved in 24 patients (5%). Limb tightness or contractures causing functional limitation were reported in 6 patients (1%). Liver was rarely involved (n = 5, 1%). Multiple organ defects were also reported in some patients without recurrent patterns (Figure 1).
Table 2.
Long‐term cGvHD evaluation.
| n (%) or m ± SEM | |
|---|---|
| Time between HSCTa and cGVHD evaluation (years) | 8.7 ± 0.3 |
| Age at cGVHD evaluation (years) | 17.7 ± 0.4 |
| cGvHD | 94 (21%) |
| Mild | 42 (9%) |
| Moderate | 31 (7%) |
| Severe | 21 (5%) |
| De novo cGvHDb , c | 35/94 (37%) |
| cGvHD treatment at the time of evaluationc | |
| None | 79/94 (84%) |
| Systemic | 10/94 (11%) |
| Local | 5/94 (5%) |
Abbreviations: cGvHD, chronic graft‐versus‐host disease; HSCT, hematopoietic stem cell transplantation; m, mean; SEM, standard error of the mean.
Last HSCT in patients who received two HSCTs
De novo cGvHD is defined as cGvHD occurring in patients without any history of acute GvHD
Among patients with long‐term cGvHD.
Figure 1.
Organ systems affected by long‐term cGvHD. Upset plot showing the organ system affected by long‐term cGvHD in the 94 affected patients. All associations of involved organ systems are represented by the bottom plot. The number of patients having each specific association is shown on the top bar plot. The total number of patients affected by each organ system involvement is represented on the left bar plot. cGvHD, chronic graft‐versus‐host disease. The plot was created using the R package UpSetR (version 1.4.0). GI, gastrointestinal, MSK, musculoskeletal.
Treatment of long‐term cGvHD
At the time of cGvHD evaluation, most patients suffering cGvHD were untreated (84%, n = 79). Ten patients (11%) received a systemic treatment (ruxolitinib, ciclosporin, ibrutinib, or sirolimus). Among them, eight had severe cGvHD and two had mild or moderate forms at the time of long‐term evaluation, most of them showing skin (n = 8/10, including six severe forms), eyes (n = 6/10), and lungs (n = 5/10) involvement. Five patients (5%) received a local treatment (topical corticosteroids, artificial tears, ciclosporin eye drops, autologous serum eye drops).
Risk factors for long‐term cGvHD
In the univariate analysis of risk factors for cGvHD (Table 3), several significant associations were identified. A history of leukemia relapse before HSCT was significantly associated with an increased risk of cGvHD (OR 1.7, 95% CI 1.1–2.7, P = 0.03). Age at HSCT also emerged as a risk factor (OR 1.06 per each additional year, 95% CI 1.02–1.1, P = 0.01). Patients who underwent a second HSCT had a markedly higher risk of cGvHD (OR 3.2, 95% CI 1.2–8.3, P = 0.02). Total body irradiation was also associated with a higher risk (OR 1.9, 95% CI 1.2–3.1, P = 0.01). Administration of anti‐thymocyte globulin was associated with a lower risk of cGvHD (OR 0.6, 95% CI 0.4–0.97, P = 0.037). Nonsignificant risk factors included sex (P = 0.8), type of leukemia (P = 0.2), donor type and HLA match (P = 0.1 to 0.8), stem cell source (P = 0.4), history of acute GvHD (P = 0.9), and time between HSCT and last follow‐up (P = 0.07). Similarly, neither busulfan exposure nor a history of sinusoidal obstruction syndrome showed a significant association with hepatic or pulmonary cGvHD (Supporting Information S1: Table 2). In the multivariate analysis, administration of antithymocyte globulin remained statistically associated with a reduced risk of long‐term cGvHD (OR 0.6, 95% CI 0.4–0.99, P = 0.045). Older age at HSCT showed a trend toward a higher risk of long‐term cGvHD; however, this association did not reach statistical significance in the multivariate model (OR 1.1 per additional year, 95% CI 0.99–1.1, P = 0.07). A univariate analysis of the risk of cGvHD by donor type and HLA match, restricted to patients who received only one HSCT, is shown in Figure 2. Haploidentical family donors (with PTCy) had an OR of 2.0 (95% CI: 0.5–7.6, P = 0.3). Among patients receiving cord blood, those with 6/6 HLA‐matched cord blood (n = 14) had an OR of 0.1 (95% CI: 0.01–2.2, P = 0.2), with no cases of cGvHD. In comparison, patients receiving 5/6 HLA‐matched cord blood (n = 45) had an OR of 0.7 (95% CI: 0.3–1.7, P = 0.4), 3–4/6 mismatched cord blood (n = 23) had an OR of 1.4 (95% CI: 0.5–3.6, P = 0.6), and those receiving double cord blood (n = 15) had an OR of 1.9 (95% CI: 0.6–5.7, P = 0.3).
Table 3.
Univariate and multivariate analysis of potential risk factors for long‐term cGvHD.
| Univariate | Multivariate | |||
|---|---|---|---|---|
| Potential risk factors for long‐term cGVHD | OR (95% CI) | P | Adjusted OR (95% CI) | P |
| Sex | ||||
| Female | ref | |||
| Male | 0.9 (0.6–1.5) | 0.8 | ||
| Type of leukemia | ||||
| Acute lymphoblastic leukemia | ref | |||
| Acute myeloid leukemia and other myeloid neoplasms | 0.8 (0.5–1.2) | 0.2 | ||
| History of leukemia relapse before HSCT | 1.7 (1.1–2.7) | 0.03 | 1.5 (0.9–2.6) | 0.1 |
| Age at HSCT (per each year) | 1.1 (1.02–1.1) | 0.004 | 1.1 (0.99–1.1) | 0.07 |
| Number of HSCT | ||||
| 1 | ref | ref | ||
| 2 | 3.2 (1.2–8.3) | 0.02 | 1.7 (0.5–5.5) | 0.4 |
| Total body irradiation | 1.9 (1.2–3.1) | 0.01 | 1.5 (0.9–2.6) | 0.1 |
| Anti‐thymocyte globulin | 0.6 (0.4–0.97) | 0.037 | 0.6 (0.4–0.99) | 0.045 |
| Donor and HLA matcha | ||||
| HLA‐matched sibling donor (10/10) | ref | |||
| Haploidentical family donor (with PTCy) | 1.2 (0.3–4.4) | 0.8 | ||
| HLA‐matched unrelated donor (10/10) | 0.9 (0.5–1.6) | 0.7 | ||
| HLA‐mismatched unrelated donor (9/10) | 0.7 (0.3–1.4) | 0.3 | ||
| HLA‐matched cord blood (6/6) | 0.1 (0.007–2.0) | 0.1 | ||
| HLA‐mismatched cord blood (5/6) | 0.7 (0.3–1.6) | 0.4 | ||
| HLA‐mismatched cord blood (3‐4/6) | 1.2 (0.5–3.3) | 0.7 | ||
| Double cord blood | 1.7 (0.6–5.2) | 0.3 | ||
| Stem cell source | ||||
| Bone marrow | Ref | |||
| PBSC | 0.6 (0.2–1.7) | 0.4 | ||
| Cord blood | 0.8 (0.5–1.4) | 0.4 | ||
| Prior history of acute GVHD (maximal grade) | ||||
| 0–1 | Ref | |||
| ≥2 | 1.0 (0.6–1.6) | 0.9 | ||
| Time between HSCT and cGVHD evaluation | 1.0 (1.0–1.1) | 0.07 | ||
Note: Statistically significant P values are shown in bold.
Abbreviations: cGvHD, chronic graft‐versus‐host disease; CI, confidence interval; HLA, human leukocyte antigen; HSCT, hematopoietic stem cell transplantation; OR, odds ratio; PBSC, peripheral blood stem cells; P, P‐value; PTCy, posttransplant cyclophosphamide.
HLA‐matched nonsibling family donors and HLA‐mismatched nonsibling family donors (9/10) were excluded from the analysis due to a low number of cases in these categories.
Figure 2.
Univariate analysis of donor and HLA factors as potential risk factors for long‐term cGvHD restricted to patients who received only one HSCT. HLA‐matched nonsibling family donors and HLA‐mismatched nonsibling family donors (9/10) were excluded from the analysis due to a low number of cases in these categories. Squares indicate the odds ratio value, and bars show the 95% confidence interval. HLA, human leukocyte antigen; P, P‐value; PTCy, posttransplant cyclophosphamide.
Association with other long‐term complications
Patients with cGvHD had a significantly higher number of long‐term complications compared to those without cGvHD (Table 4). Among patients without cGvHD, 10%, 18%, 18%, and 54% had 0, 1, 2, and ≥3 complications, respectively, while in patients with cGvHD, these percentages were 5%, 3%, 15%, and 77%, respectively (P < 0.001). The cumulated incidence of diabetes, cataract, pulmonary dysfunction, osteonecrosis, hypothyroidism, permanent alopecia, and underweight in patients without versus with cGvHD was 2% versus 6% (P = 0.02), 26% versus 44% (P = 0.002), 31% versus 52% (P = 0.001), 5% versus 12% (P = 0.03), 19% versus 30% (P = 0.02), 7% versus 13% (P = 0.046), and 29% versus 40% (P = 0.046), respectively. There was no statistically significant difference in the cumulated incidence of other cardiovascular risk factors and complications, anthracycline‐related cardiomyopathy, hearing loss, neurological defects, low bone mineral density, iron overload, delayed puberty, short stature, renal impairment, and subsequent malignant neoplasm or meningioma according to the cGvHD status.
Table 4.
Association between cGvHD and other long‐term complications.
| Long‐term complications | n (%*) | P | |
|---|---|---|---|
| No cGvHD | cGvHD | ||
| Total number of long‐term complications (excluding cGvHD) | < 0.001 | ||
| 0 | 34 (10%) | 5 (5%) | |
| 1 | 64 (18%) | 3 (3%) | |
| 2 | 65 (18%) | 14 (15%) | |
| ≥3 | 189 (54%) | 72 (77%) | |
| Cardiovascular risk | |||
| Obesity | 53 (17%) | 18 (20%) | 0.5 |
| Metabolic syndrome | 34 (21%) | 13 (23%) | 0.7 |
| Diabetes | 7 (2%) | 6 (6%) | 0.02 |
| Coronary artery disease | 0 (0%) | 0 (0%) | — |
| Stroke | 6 (2%) | 1 (1%) | 1 |
| Cardiomyopathy | 27 (8%) | 13 (14%) | 0.1 |
| Sensory defect | |||
| Hearing loss | 8 (2%) | 4 (4%) | 0.1 |
| Cataract | 76 (26%) | 38 (44%) | 0.002 |
| Neurological defect | 15 (4%) | 8 (9%) | 0.1 |
| Pulmonary dysfunction | 95 (31%) | 44 (52%) | 0.001 |
| Bone defect | |||
| Osteonecrosis | 19 (5%) | 11 (12%) | 0.03 |
| Osteoporosis | 32 (25%) | 14 (49%) | 0.7 |
| Iron overload | 188 (57%) | 55 (61%) | 0.5 |
| Endocrinopathy | |||
| Gonadal dysfunction | 125 (46%) | 46 (54%) | 0.2 |
| Hypothyroidism | 62 (19%) | 27 (30%) | 0.02 |
| Short stature | 101 (34%) | 30 (36%) | 0.7 |
| Chronic kidney disease | 13 (4%) | 4 (4%) | 0.8 |
| Permanent alopecia | 23 (7%) | 12 (13%) | 0.046 |
| Underweight | 92 (29%) | 36 (40%) | 0.046 |
| Subsequent malignant neoplasm or meningioma | 22 (6%) | 8 (10%) | 0.3 |
Note: Statistically significant P values are shown in bold.
Abbreviations: cGvHD, chronic graft‐versus‐host disease; P, P‐value.
Among evaluable patients.
Impact of long‐term cGvHD on the quality of life
cGvHD had a marked detrimental effect on QoL, even after adjusting for the total number of other long‐term complications (Table 5). The physical and psychological SF‐36 adjusted composite scores in patients with or without cGvHD were 50 ± 2 versus 55 ± 1 (P = 0.01) and 38 ± 2 versus 43 ± 1 (P = 0.01), respectively. Mild and moderate forms of cGvHD were also associated with impaired QoL, especially on the psychological QoL dimensions (Supporting Information S1: Table 3). The psychological adjusted composite score in patients with or without mild to moderate cGvHD was 38 ± 2 versus 43 ± 1 (P = 0.04).
Table 5.
Impact of long‐term cGvHD on quality of life.
| No cGvHD (n = 125) | cGvHD (n = 44) | Adjusted effect on QoLa | |||
|---|---|---|---|---|---|
| QoL dimension | m ± SEM | m ± SEM | Beta | P | Effect size |
| Physical abilities | 89 ± 2 | 80 ± 3 | −6.9 | 0.04 | 0.4 |
| Social life | 79 ± 2 | 66 ± 4 | −13.6 | 0.004 | 0.5 |
| Limitations due to physical condition | 85 ± 2 | 75 ± 4 | −9.2 | 0.03 | 0.4 |
| Limitations due to emotional state | 70 ± 2 | 62 ± 3 | −8.7 | 0.02 | 0.4 |
| Mental health | 69 ± 2 | 59 ± 4 | −10.1 | 0.007 | 0.5 |
| Vitality | 60 ± 2 | 49 ± 3 | −9.4 | 0.008 | 0.5 |
| Physical pain | 79 ± 2 | 65 ± 4 | −11.2 | 0.02 | 0.5 |
| Perceived health | 66 ± 2 | 50 ± 4 | −14.6 | <0.001 | 0.7 |
| Physical composite score | 55 ± 1 | 50 ± 2 | −4.1 | 0.01 | 0.6 |
| Psychological composite score | 43 ± 1 | 38 ± 2 | −5.4 | 0.01 | 0.4 |
Abbreviations: m, mean; P, P‐value; SEM, standard error of the mean.
Adjusted to the total number of long‐term complications. Quality of life (QoL) data are presented for adult patients (≥18 years of age at last evaluation) and are derived from SF‐36 questionnaires.
DISCUSSION
This study provides insights into the prevalence, risk factors, treatment, and consequences of long‐term cGvHD among childhood leukemia survivors post‐HSCT. The comprehensive assessment of cGvHD almost 10 years after HSCT in ~450 patients strikingly shows that cGvHD is more frequent, severe, and persistent than anticipated. This study also demonstrates a major and independent impact on QoL, as well as a significant association with other long‐term complications.
One of the key findings of this study is the high prevalence of long‐term cGVHD in pediatric leukemia survivors post‐HSCT. While cGVHD is often thought to affect a relatively small proportion of the pediatric population, 3 , 9 this study shows that ~20% of patients suffer long‐term cGVHD, with ~5% having severe forms. Assessment of cGvHD was performed by a single clinician at each participating center, without formal central adjudication. Prior studies have shown that central review may lead to reclassification of a substantial proportion of cases, 5 highlighting the possibility of some misclassification in our cohort. However, the use of a standardized assessment approach, including the eGVHD tool, likely mitigated this risk. This novel data complements the shorter‐term information already available. 10 , 11 , 12 , 13 Several factors may explain this high incidence. One possibility is that routine post‐HSCT care does not fully capture the broad spectrum of cGvHD manifestations in children, especially milder forms that might go undiagnosed and underreported in retrospective studies and clinical practice. The evolving landscape of HSCT practices, such as the increasing use of alternative transplants, may also be contributing to the higher cGvHD prevalence compared to older studies, as previously suggested. 14 This higher‐than‐expected prevalence has significant implications for long‐term care and survivorship planning. Routine and standardized screening for cGvHD must be integrated into long‐term post‐HSCT follow‐up care for pediatric leukemia survivors. This further emphasizes the importance of implementing dedicated long‐term follow‐up programs led by physicians specifically trained in survivorship to ensure optimal recognition and management of late complications.
The study showed a major impact on both physical and psychological QoL scores, independent of other physical complications, reinforcing the long‐term burden of the disease. The impact of cGvHD on QoL may worsen with age, as their ability to compensate for functional impairments related to cGvHD may decline, leading to an increasing burden on physical and emotional well‐being. These findings highlight the need for comprehensive, long‐term follow‐up care that not only addresses physical health but also the psychosocial aspects of survivorship. Even mild to moderate cGvHD appears to significantly impact quality of life, particularly in terms of psychological well‐being. This suggests that early detection and management of these milder forms of cGvHD may prevent deterioration of long‐term QoL.
Administration of antithymocyte globulin was independently associated with a reduced risk of long‐term cGvHD in our cohort, in line with adult randomized trials showing that anti‐thymocyte globulin‐containing conditioning regimens significantly decrease the incidence and severity of chronic GVHD in the sibling and unrelated donor setting. 15 , 16 , 17 Although most of the evidence supporting this effect comes from studies in adults receiving peripheral‐blood grafts, our findings suggest that anti‐thymocyte globulin may similarly mitigate alloreactive T‐cell responses and help reduce the long‐term burden of cGvHD in pediatric leukemia survivors. Older age at the time of HSCT has previously been described as a risk factor for the development of cGvHD. 5 , 13 We observed the same trend in our study, with a multivariate effect that was close to reaching statistical significance. This suggests that immune responses may differ with age, with older children possibly having a more pronounced alloimmune reaction from the graft. History of leukemia relapse before HSCT and the need for a second HSCT were significantly associated with an increased risk of developing cGvHD in the univariate analysis, suggesting the need for close monitoring of these patients. TBI was a significant risk factor for cGvHD in the univariate analysis, but was not confirmed in the multivariate analysis. While its precise role remains unclear in our study and the literature, 5 , 18 , 19 previous data suggest that TBI may exacerbate immune responses post‐HSCT. 20 No significant association was found between cGvHD incidence and donor type or stem cell source, suggesting that HLA mismatch and unrelated donors are less important risk factors for cGvHD than for acute GvHD, as previously suggested. 19 However, the relatively small number of patients in certain groups (e.g., HLA‐matched cord blood, for which our study suggests a lower cGvHD incidence, as previously reported 10 ), limits definitive conclusions about these subsets. Intriguingly, a history of acute GvHD did not constitute a risk factor for cGvHD in our population, contrary to previous studies. 5 , 21 , 22 This notable difference may be partly due to the design of our study, with the primary objective being the evaluation of cGVHD in long‐term survivors. The extended time (almost 10 years) since HSCT at cGVHD evaluation may have allowed the capture of long‐term forms while excluding both early fatal forms and mild chronic forms that may resolve within months or years. The heterogeneity of stem cell sources and HLA matching also limited the power of this study to accurately analyze the interactions between acute and chronic GvHD.
The study confirmed the multifaceted nature of cGvHD, with most pediatric patients experiencing isolated organ involvement, primarily affecting the skin, eyes, and mouth, while severe cases often involved multiple organs. Hepatic involvement was notably rare in this cohort, in contrast to adult populations, where liver involvement is much more frequently reported. 23 This discrepancy raises questions about potential differences in the pathophysiological mechanisms of cGvHD between children and adults, as well as the influence of cofactors such as alcohol consumption. Interestingly, Cuvelier et al. reported a ~30% incidence of liver cGvHD, whereas it was <5% in our cohort. This discrepancy may reflect differences in study design, as our study focused on long‐term follow‐up, and hepatic involvement might have responded well to systemic treatments over time. It may also be explained by differences in study populations, including a higher prevalence of nonmalignant diseases, often associated with pre‐HSCT hepatic defects, and greater use of peripheral blood stem cells. 5 Pulmonary involvement was the third most frequently affected organ in our cohort, suggesting a poor response to treatment and a tendency to persist over the long term. It remains challenging to assess in children, as conventional pulmonary function tests may not always be reliable. The study utilized the NIH consensus guidelines for cGvHD assessment via the eGVHD application, 7 , 24 which were primarily developed for adults. Although validated in children, 5 these criteria might not have fully captured the spectrum of cGvHD manifestations, particularly for organs that are more difficult to evaluate in pediatric patients, such as the lungs and genitals. Emerging imaging techniques like parametric response mapping with high‐resolution CT scans show promise in detecting early pulmonary cGvHD without relying on pulmonary function tests. 25 Additionally, the use of pediatric‐specific cGvHD biomarkers represents another promising avenue for improving diagnosis and monitoring. 5
Patients with cGvHD are at significantly higher risk for long‐term complications, including pulmonary dysfunction, osteonecrosis, cataracts, hypothyroidism, underweight, permanent alopecia, and diabetes. Some of these complications, such as pulmonary dysfunction and underweight, are directly related to the disease process, while others, like cataracts, diabetes, and osteonecrosis, may be a consequence of immunosuppressive treatments, such as corticosteroids, used to manage cGvHD. The link between cGvHD, its treatments, and osteonecrosis may be related to the higher proportion of patients with cGvHD who were treated during adolescence, which is a major risk factor for osteonecrosis, 26 , 27 as well as the prolonged use of corticosteroids for GvHD treatment. The increased rate of hypothyroidism may be partly due to the higher proportion of patients who received total body irradiation. 28 The complex interplay between dysimmunity, resulting from both cGvHD and its treatments, and the development of other complications, such as hypothyroidism, osteonecrosis, permanent alopecia, or pulmonary dysfunction, warrants further investigation.
No data on long‐term cGvHD treatment were available before this study. As a result, recent reviews and guidelines have not specifically addressed its long‐term management. 3 , 29 We show here that only ~10% of patients with long‐term cGvHD were receiving systemic treatment at the time of the evaluation. A limitation of this study is that it did not formally distinguish between active cGvHD and irreversible sequelae from prior cGvHD. Some patients with persistent manifestations but no ongoing treatment may therefore have long‐term damage rather than active disease, as recently emphasized, 30 and may not necessarily benefit from further immunosuppression. Conversely, some cGvHD manifestations may have been under‐recognized or considered too mild to warrant systemic therapy. In addition, the limited efficacy of treatment at this late phase of the disease may also contribute to the low proportion of patients under therapy. The scarcity of drugs specifically approved for cGvHD and their limited development in children present additional challenges. At an advanced stage of cGvHD evolution, fibrotic processes may predominate, with little to no efficacy of available anti‐inflammatory therapies, such as corticosteroids, ruxolitinib, or ibrutinib. While novel agents such as belumosudil and axatilimab have antifibrotic properties and may be considered in these phases, 29 they were not available at the time of this study.
This study provides novel insights into long‐term cGvHD after HSCT in children and adolescents and shows that improved recognition and treatment are urgently needed. In this population, long‐term cGvHD is frequent, often severe, and independently detrimental to the survivors' QoL. Long‐term follow‐up programs after HSCT must include systematic and standardized evaluation of cGvHD.
AUTHOR CONTRIBUTIONS
Paul Saultier: Conceptualization; investigation; writing—original draft; methodology; validation. Gérard Michel: Conceptualization; investigation; methodology; supervision; funding acquisition. Anne Sirvent: Methodology; supervision; investigation. cécile renard: investigation; writing—review and editing. Marie‐Dominique Tabone: Investigation; writing—review and editing. Guy Leverger: Investigation; writing—review and editing. André Baruchel: Investigation; writing—review and editing. Cécile Pochon: Investigation; writing—review and editing. Catherine Paillard: Investigation; writing—review and editing. Charlotte Jubert: Investigation; writing—review and editing. Stéphane Ducassou: Investigation; writing—review and editing. Marilyne Poirée: Investigation; writing—review and editing. Marion Strullu: Investigation; writing—review and editing. Mony Fahd: Investigation; writing—review and editing. Sandrine Visentin: Investigation; writing—review and editing. Arthur Stérin: Investigation; writing—review and editing. Dominique Plantaz: Investigation; writing—review and editing. Justyna Kanold: Investigation; writing—review and editing. Virginie Gandemer: Investigation; writing—review and editing. Alexandre Theron: Investigation; writing—review and editing. Nicole Raus: Investigation; writing—review and editing. Sandrine Thouvenin‐Doulet: Investigation; writing—review and editing. Carine Domenech: Investigation; writing—review and editing. Julie Berbis: Methodology; writing—review and editing; supervision; formal analysis. Pascal Auquier: Supervision; writing—review and editing; methodology; formal analysis. Jean‐Hugues Dalle: Investigation; conceptualization; funding acquisition; writing—review and editing; methodology; supervision.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.
ETHICS STATEMENT
The study was approved by the Sud Méditerranée V Ethics Committee (Opinion No. 2012‐A00984‐39), in compliance with the General Data Protection Regulation (GDPR).
FUNDING
The study was funded in part by the Bristol Myers Squibb Foundation, the French National Clinical Research Program, the French National Cancer Institute (InCA), the French National Research Agency (ANR), the Canceropole PACA, the Ligue contre le Cancer, Association Laurette Fugain, Enfants Cancers Santé, and the French Institute for Public Health Research (IRESP). All patients were included after informed consent was obtained in accordance with the Declaration of Helsinki principles.
Supporting information
20 251 128 Supplemental Material V2 HemaSphere.
ACKNOWLEDGMENTS
The study was funded in part by the Bristol Myers Squibb Foundation, the French National Clinical Research Program, the French National Cancer Institute (InCA), the French National Research Agency (ANR), the Canceropole PACA, the Ligue contre le Cancer, Association Laurette Fugain, Enfants Cancers Santé, and the French Institute for Public Health Research (IRESP). The authors thank the LEA study group for data collection, Zeinab Hamidou and Alaa Mustafa‐Shawket for the statistical analysis, the Francophone Society of Bone and Marrow Transplantation (SFGM‐TC) for its help in carrying out this study, and the patients and their families for participation in the study.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
- 1. Qayed M, Wang T, Hemmer MT, et al. Influence of age on acute and chronic GVHD in children undergoing HLA‐identical sibling bone marrow transplantation for acute leukemia: implications for prophylaxis. Biol Blood Marrow Transplant. 2018;24(3):521‐528. 10.1016/j.bbmt.2017.11.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Cuvelier GDE, Li A, Drissler S, et al. Age related differences in the biology of chronic graft‐versus‐host disease after hematopoietic stem cell transplantation. Front Immunol. 2020;11:571884. 10.3389/fimmu.2020.571884 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Sobkowiak‐Sobierajska A, Lindemans C, Sykora T, et al. Management of chronic graft‐vs.‐host disease in children and adolescents with ALL: present status and model for a personalised management plan. Front Pediatr. 2022;10:808103. 10.3389/fped.2022.808103 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Inagaki J, Moritake H, Nishikawa T, et al. Long‐term morbidity and mortality in children with chronic graft‐versus‐host disease classified by National Institutes of Health Consensus criteria after allogeneic hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2015;21(11):1973‐1980. 10.1016/j.bbmt.2015.07.025 [DOI] [PubMed] [Google Scholar]
- 5. Cuvelier GDE, Nemecek ER, Wahlstrom JT, et al. Benefits and challenges with diagnosing chronic and late acute GVHD in children using the NIH consensus criteria. Blood. 2019;134(3):304‐316. 10.1182/blood.2019000216 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Berbis J, Michel G, Baruchel A, et al. Cohort Profile: the French childhood cancer survivor study for leukaemia (L.E.A. Cohort). Int J Epidemiol. 2015;44(1):49‐57. 10.1093/ije/dyu031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Schoemans H, Goris K, Durm RV, et al. Development, preliminary usability and accuracy testing of the EBMT ‘eGVHD App’ to support GvHD assessment according to NIH criteria—a proof of concept. Bone Marrow Transplant. 2016;51(8):1062‐1065. 10.1038/bmt.2016.26 [DOI] [PubMed] [Google Scholar]
- 8. Leplège A, Ecosse E, Verdier A, Perneger TV. The French SF‐36 Health Survey. J Clin Epidemiol. 1998;51(11):1013‐1023. 10.1016/s0895-4356(98)00093-6 [DOI] [PubMed] [Google Scholar]
- 9. Baird K, Cooke K, Schultz KR. Chronic graft‐versus‐host disease (GVHD) in children. Pediatr Clin North Am. 2010;57(1):297‐322. 10.1016/j.pcl.2009.11.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Rocha V, Wagner JE, Sobocinski KA, et al. Graft‐versus‐host disease in children who have received a cord‐blood or bone marrow transplant from an HLA‐identical sibling. N Engl J Med. 2000;342(25):1846‐1854. 10.1056/NEJM200006223422501 [DOI] [PubMed] [Google Scholar]
- 11. Zecca M. Chronic graft‐versus‐host disease in children: incidence, risk factors, and impact on outcome. Blood. 2002;100(4):1192‐1200. 10.1182/blood-2001-11-0059 [DOI] [PubMed] [Google Scholar]
- 12. Bachier CR, Aggarwal SK, Hennegan K, et al. Epidemiology and treatment of chronic graft‐versus‐host disease post‐allogeneic hematopoietic cell transplantation: a US claims analysis. Transpl Cell Ther. 2021;27(6):504.e1‐504.e6. 10.1016/j.jtct.2020.12.027 [DOI] [PubMed] [Google Scholar]
- 13. Nishitani M, Graham RAT, Wang T, et al. Graft Vs Host Disease (GVHD) in pediatric hematopoietic stem cell transplant (HCT) recipients and impact on overall survival: a CIBMTR analysis. Transplant Cell Ther Off Publ Am Soc Transplant Cell Ther. 2024;30(2):S49‐S50. 10.1016/j.jtct.2023.12.084 [DOI] [Google Scholar]
- 14. Arai S, Arora M, Wang T, et al. Increasing incidence of chronic graft‐versus‐host disease in allogeneic transplantation: a report from the Center for International Blood and Marrow Transplant Research. Biol Blood Marrow Transplant. 2015;21(2):266‐274. 10.1016/j.bbmt.2014.10.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Yang X, Li D, Xie Y. Anti‐thymocyte globulin prophylaxis in patients with hematological malignancies undergoing allogeneic hematopoietic stem cell transplantation: an updated meta‐analysis. Front Oncol. 2021;11:717678. 10.3389/fonc.2021.717678 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Pidala JA, Gooley TA, Luznik L, Blazar BR. Chronic graft‐versus‐host disease: unresolved complication or ancient history? Blood. 2024;144(13):1363‐1373. 10.1182/blood.2023022735 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Kröger N, Solano C, Wolschke C, et al. Antilymphocyte globulin for prevention of chronic graft‐versus‐host disease. N Engl J Med. 2016;374(1):43‐53. 10.1056/NEJMoa1506002 [DOI] [PubMed] [Google Scholar]
- 18. Corvò R, Paoli G, Barra S, et al. Total body irradiation correlates with chronic graft versus host disease and affects prognosis of patients with acute lymphoblastic leukemia receiving an HLA identical allogeneic bone marrow transplant. Int J Radiat Oncol Biol Phys. 1999;43(3):497‐503. 10.1016/s0360-3016(98)00441-6 [DOI] [PubMed] [Google Scholar]
- 19. Flowers MED, Inamoto Y, Carpenter PA, et al. Comparative analysis of risk factors for acute graft‐versus‐host disease and for chronic graft‐versus‐host disease according to National Institutes of Health consensus criteria. Blood. 2011;117(11):3214‐3219. 10.1182/blood-2010-08-302109 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Hill GR, Crawford JM, Cooke KR, Brinson YS, Pan L, Ferrara JLM. Total body irradiation and acute graft‐versus‐host disease: the role of gastrointestinal damage and inflammatory cytokines. Blood. 1997;90(8):3204‐3213. [PubMed] [Google Scholar]
- 21. Kondo M, Kojima S, Horibe K, Kato K, Matsuyama T. Risk factors for chronic graft‐versus‐host disease after allogeneic stem cell transplantation in children. Bone Marrow Transplant. 2001;27(7):727‐730. 10.1038/sj.bmt.1702868 [DOI] [PubMed] [Google Scholar]
- 22. Tamaki M, Akahoshi Y, Inamoto Y, et al. Associations between acute and chronic graft‐versus‐host disease. Blood Adv. 2024;8(16):4250‐4261. 10.1182/bloodadvances.2024013442 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Chen CT, Liu CY, Yu YB, et al. Characteristics and risk of chronic graft‐versus‐host disease of liver in allogeneic hematopoietic stem cell transplant recipients. PLoS One. 2017;12(9):e0185210. 10.1371/journal.pone.0185210 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Jagasia MH, Greinix HT, Arora M, et al. National Institutes of Health Consensus development project on criteria for clinical trials in chronic graft‐versus‐host disease: I. The 2014 Diagnosis and Staging Working Group report. Biol Blood Marrow Transplant. 2015;21(3):389‐401.e1. 10.1016/j.bbmt.2014.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Cheng GS, Selwa KE, Hatt C, et al. Multicenter evaluation of parametric response mapping as an indicator of bronchiolitis obliterans syndrome after hematopoietic stem cell transplantation. Am J Transpl. 2020;20(8):2198‐2205. 10.1111/ajt.15814 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Girard P, Auquier P, Barlogis V, et al. Symptomatic osteonecrosis in childhood leukemia survivors: prevalence, risk factors and impact on quality of life in adulthood. Haematologica. 2013;98(7):1089‐1097. 10.3324/haematol.2012.081265 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Huault A, Michel G, Charon V, et al. Symptomatic osteonecrosis in French survivors of childhood and adolescent leukemia: a clinical and MRI study of LEA cohort. Pediatr Hematol Oncol. 2023;40(5):458‐474. 10.1080/08880018.2023.2168810 [DOI] [PubMed] [Google Scholar]
- 28. Oudin C, Auquier P, Bertrand Y, et al. Late thyroid complications in survivors of childhood acute leukemia. An L.E.A. study. Haematologica. 2016;101(6):747‐756. 10.3324/haematol.2015.140053 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Kim DDH, Popradi G, Lepic K, et al. Cell Therapy Transplant Canada (CTTC) consensus‐based guideline 2024 for management and treatment of chronic graft‐versus‐host disease and future directions for development. Curr Oncol. 2024;31(3):1426‐1444. 10.3390/curroncol31030108 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Bhatt NS, Harris AC, Gorfinkel L, et al. Pediatric transplant and cellular therapy consortium RESILIENT conference on pediatric chronic graft‐versus‐host disease survivorship after hematopoietic cell transplantation: part I. Phases of Chronic GVHD, supportive care, and systemic therapy discontinuation. Transpl Cell Ther. 2025;31(2):69. 10.1016/j.jtct.2024.12.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
20 251 128 Supplemental Material V2 HemaSphere.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.