Graft-versus-host disease (GVHD) is a significant complication of hematopoietic stem cell transplantation (HSCT), representing a leading cause of both morbidity and mortality.1,2 In pediatric patients, HSCT is most commonly performed to treat hematologic malignancies, although its application for non-malignant conditions ‒ including primary immunodeficiencies, bone marrow failure syndromes, and hereditary metabolic disorders ‒ is growing.3 Conditioning regimens ‒ administered with or without total-body irradiation (TBI) ‒ are essential to suppress the recipient’s immune system and facilitate engraftment. Donor sources may include family members or unrelated donors, with hematopoietic progenitors derived from bone marrow, umbilical cord blood (UCB), or mobilized peripheral blood stem cells (PBSC).2

GVHD is divided into acute and chronic forms, which differ in their clinical presentation, risk factors, and long-term consequences. These two forms can occur independently of each other.1 The skin is the most affected organ in both.4,5 Distinction between acute and chronic GVHD is defined by specific clinical characteristics.1

Manifestations of cutaneous chronic GVHD (ccGVHD) are diverse.6 According to the National Institutes of Health (NIH) Consensus Development Project, the presence of one of the following manifestations alone is sufficient to diagnose ccGVHD: poikiloderma, lichen planus-like eruptions, sclerodermoid features, morphea-like features, and lichen sclerosus-like features.7

Skin biopsy is indicated when clinical manifestations of ccGVHD are inconclusive. Histopathologically, ccGVHD can be classified into epidermal, sclerotic, or combination disease. Epidermal ccGVHD predominantly shows a lichen planus/interface dermatitis pattern, whereas sclerotic ccGVHD is characterized by dermal and subcutaneous sclerosis with a lichen sclerosus/morphea-like pattern. Minimal pathologic features of active GVHD ‒ including apoptosis in the epidermal basal layer, vacuolar change, and lymphocyte satellitosis ‒ are common in all biopsies; however, most specimens also exhibit findings overlapping with autoimmune and inflammatory skin diseases.8

Previous studies in the pediatric population have shown a prevalence of ccGVHD between 18%‒32% in recipients of HSCT.2,9 Early diagnosis and treatment of ccGVHD is essential, due to its high morbidity and deterioration in quality of life. Studies in the pediatric population are scarce, with few patients and a retrospective design. The objective of this study is to assess the risk factors of ccGVHD in children undergoing HSCT at a children’s hospital in Chile.

This retrospective cohort study included all children under 18-years of age undergoing HSCT between 2007 and 2017 with a clinical follow-up period of at least 2-years, at the Luis Calvo Mackenna Children’s Hospital. This study was approved by the Ethics Committee of Research Protocols of the Faculty of Medicine at the University of Chile, Santiago, Chile (136-2019).

In all patients, the diagnosis of ccGVHD and acGVHD was established by pediatric dermatologists, based on clinical manifestations defined according to the standardized criteria of the NIH Consensus Development Project.7 In patients presenting with nonspecific clinical manifestations that complicated the differential diagnosis, a skin biopsy was performed.

Recorded data included patient demographics, diagnoses that required HSCT, conditioning with TBI, and donor characteristics (sex, graft source and type of donor). Presence of ccGVHD and its characteristics: clinical type of ccGVHD, time interval between HSCT and onset of ccGVHD, presence of prior or concomitant acGVHD, lung involvement, and mortality.

The exclusion criterion was a personal history of autoimmune skin disease prior to the transplant.

This study represents a large series of ccGVHD in children in Chile. In our study, 17.3% of the children developed ccGVHD, with a median onset of 8-months. Based on these findings, we recommend active surveillance for ccGVHD for at least 15-months following transplantation. An Israeli study of 112 children undergoing HSCT showed an incidence of ccGVHD of 18%, with a mean time interval from transplantation of 4.9 ± 3.95 months.2 A publication from Kuwait in 14-patients showed a median time between HSCT and onset of ccGVHD of 7.5-months.10 An Argentinian study in patients under 20-years of age showed an incidence of ccGVHD of 32%.9 These differences in incidence can be explained by differences in protocols, graft sources, and the type of donors used in each reference center.2,9

Alvarado et al. conducted a histopathological analysis of ccGVHD in 49 adult patients. Clinically, 67.3% had epidermal ccGVHD, 2.1% sclerotic ccGVHD, and 30.6% combination disease. Epidermal ccGVHD predominantly showed a lichen planus–like/interface dermatitis pattern with basal keratinocyte apoptosis (94%) and vacuolar alteration (92.5%), whereas sclerotic ccGVHD showed a higher proportion of thickening of the papillary and reticular collagen and the subcutaneous septae (up to 68.8%) with a lichen sclerosus/morphea-like pattern. Subclinical sclerosis was noted in several epidermal ccGVHD lesions, suggesting potential prognostic implications.8 In our cohort, the analyzed biopsies showed findings consistent with those reported in this study, highlighting the importance of comprehensive histologic reporting. Additional studies are warranted to better characterize the histopathologic features of ccGVHD in pediatric patients.

Our cohort of patients with ccGVHD was older at HSCT compared to patients without ccGVHD. Similarly, in the Israeli study, the group with ccGVHD had a significantly older age at transplantation than patients without ccGVHD (9.5 ± 5.2 years vs. 5.4 ± 4.8 years).2 However, a Spanish study of 50 children with acute and chronic cutaneous GVHD did not find significant differences between the groups with and without ccGVHD.11 The higher risk of chronic GVHD in older patients might reflect a role of age-related atrophy of the thymus, which results in a loss of thymic negative selection post HSCT.12,13

UCB transplantation has become more prevalent in recent years in the pediatric population, especially when a related matched bone marrow is unavailable.1 In our study, the most common graft source was UCB with 52.67%. Research in pediatric patients has found that the UCB source is associated with a lower risk of acute and chronic GVHD in comparison with the bone marrow source.14,15 In our study, bone marrow source had a 4.29 times higher risk in the univariate analysis and a 7.28 times higher risk in the multivariate analysis of developing ccGVHD compared to UCB source. In a meta-analysis in 1453 children, the rate of chronic GVHD in the unrelated UCB group was significantly lower than the unrelated bone marrow group (OR = 0.36, p < 0.0001).14 One possible explanation is that transplanted UCB stem cells are immunologically naive, reducing their capacity to induce chronic GVHD.14,15

Patients with ccGVHD had a significantly higher proportion of history of prior or concomitant acGVHD compared to patients without ccGVHD (80.77% vs. 58.06%), with an HR of 2.76 in the univariate analyses. Therefore, patients with acGVHD should be considered at high risk for developing ccGVHD. 66% of patients with ccGVHD had a history of acGVHD in the Spanish study,11 86% in the Kuwait publication,10 and 100% in the Argentinian study.9 None of these 3 studies compared the frequency of acGVHD in patients with and without ccGVHD. The evidence shows that acute GVHD damages the thymus, resulting in a loss of thymic negative selection, leading to chronic GVHD.1,13

In our study, TBI as a conditioning regimen had a 3.53 times higher risk of developing ccGVHD in the multivariate analysis. The Spanish publication showed that children with ccGVHD had a significantly higher proportion of TBI versus patients without ccGVHD (66% vs. 33%).11 TBI administered prior to HSCT results in higher toxicity and tissue damage, establishing a pro-inflammatory response, possibly explaining our results.2 Therefore, it is pertinent to consider patients who will undergo TBI as being at higher risk of developing ccGVHD.

The related donor had a 3.58 times higher risk of developing ccGVHD than the unrelated donor in univariate analysis. This may be explained by the source of the graft, being 96% from bone marrow in the related group in comparison to 81.9% from UCB in the unrelated group. As previously discussed, bone marrow source is a known risk factor for ccGVHD compared with UCB.14,15

Unlike other studies, we did not find any difference in the risk of ccGVHD comparing the sex of the donor, malignant disease as the cause of HSCT, and lung involvement.2,11

One limitation of this study is its retrospective, single-center design, which may limit external validity. Additionally, the diversity of conditioning regimens, GVHD prophylaxis, and treatment protocols introduces heterogeneity that could further compromise generalizability. However, the selected hospital was the only public pediatric HSCT center in the country, serving as the reference institution for nearly 80% of the public health system’s population. This setting enabled comprehensive and consistent monitoring; every patient was followed for at least 24-months, with no exclusions due to insufficient follow-up.

This study demonstrates a 17.3% incidence of ccGVHD in pediatric recipients of HSCT, with a median onset of 8-months post-transplant. These findings underscore the importance of sustained vigilance, and we therefore recommend standardized monitoring for ccGVHD up to at least 15-months after HSCT. Our results highlight the need for individualized risk stratification and tailored surveillance protocols in pediatric centers performing HSCT.

While several studies exist with larger cohorts ‒ encompassing all age groups or focused specifically on adults ‒ data exclusive to the pediatric population remain limited. This report should be regarded as a preliminary, single-center study and underscores the need for prospective, multicenter collaborations both nationally and internationally. Such efforts would broaden our understanding of the disease’s influencing factors and support the development of future guidelines for prevention and management.

Teresa Dossi: 0009-0008-6838-2881

Camila Downey: 0000-0002-1624-1170

Daniela Krämer: 0000-0003-3022-3187

Natalia González: 0009-0000-8903-3770

Javier Fernández: 0000-0003-0397-2043

Paula Muñoz: 000-0003-2676-7464

Daniela Carvajal: Writing of the manuscript or critical review of important intellectual content; critical review of the literature; final approval of the final version of the manuscript.

Teresa Dossi: Data collection, analysis and interpretation; critical review of the literature; final approval of the final version of the manuscript.

Camila Downey: Data collection, analysis and interpretation; critical review of the literature; final approval of the final version of the manuscript.

Daniela Krämer: Data collection, analysis and interpretation; critical review of the literature; final approval of the final version of the manuscript.

Natalia González: Data collection, analysis and interpretation; critical review of the literature; final approval of the final version of the manuscript.

Javier Fernández: Data collection, or analysis and interpretation of data; statistical analysis; final approval of the final version of the manuscript.

Paula Muñoz: The study concept and design; data collection, or analysis and interpretation of data; data collection, analysis and interpretation; effective participation in the research guidance; critical review of the literature; final approval of the final version of the manuscript.

During the preparation of this work, the author(s) used ChatGPT in order to improve language and readability. After using this tool/service, the author(s) reviewed and edited the content as needed and take (s) full responsibility for the content of the publication.

None declared.

The entire dataset supporting the results of this study was published in this article.