Y Jiao, Y Zhao, Z Chen, L Yang, Q Fu, Z Wang, F Tao, Y Du, M Sun, W Lu, S Qi, H Xiong

Abstract

Objective: This study investigated the role of gastrointestinal (GI) endoscopy in diagnosing acute GI graft-versus-host disease (GVHD) in paediatric patients following allogeneic haematopoietic stem cell transplantation (allo-HSCT). Methods: We clinically analysed 35 paediatric patients suspected of having GI-GVHD after allo-HSCT who underwent gastrointestinal endoscopy between January 2019 and December 2023. We evaluated their endoscopic findings, biopsy histopathological diagnoses, and their implications for clinical management. Results: During the study period, 73 out of 256 patients who underwent allo-HSCT developed GI symptoms, and 35 (14%) patients required GI endoscopy. The primary endoscopic manifestations in this study included mucosal congestion, oedema, erosion, ulceration, exfoliation, and active bleeding. Tissue biopsy histopathology was not performed in seven patients, and four (14%) showed no significant histologic abnormalities. GI-GVHD was diagnosed histologically in 24 patients (86%), of whom four had GI-GVHD with cytomegalovirus (CMV) enterocolitis, one had GI-GVHD with Epstein-Barr virus (EBV) enterocolitis, and 15 had GI-GVHD with transplant-associated thrombotic microangiopathy (TA-TMA). Seventeen patients were started on anti-GI-GVHD treatment according to clinical criteria before endoscopy, and the diagnosis of clinical GI-GVHD was confirmed by biopsy in 15 patients. Conclusions: In our experience, endoscopic diagnosis of GI-GVHD in paediatric patients is both feasible and safe. Patients with clinically suspected intestinal GVHD should undergo timely and complete GI endoscopy for a definitive diagnosis.

Introduction

With the continuous improvement of allogeneic haematopoietic stem cell transplantation (allo-HSCT), the overall survival (OS) rate of treated paediatric patients has significantly increased.¹ Consequently, allo-HSCT became a crucial therapeutic approach for paediatric malignant haematological diseases, bone marrow failure disorders, immunodeficiency diseases, and inherited metabolic disorders.² Graft-versus-host disease (GVHD) is an important and serious complication of allo-HSCT,^3,4 particularly intestinal grades III~ IV GVHD.

Endoscopic findings are used to enhance diagnostic confidence and exclude other potential etiologies. Gastrointestinal (GI) endoscopy and histopathology examination of biopsies are effective methods for diagnosing intestinal GVHD.⁵ Intestinal GVHD does not present specific clinical features. Its clinical manifestations are similar to those of other complications that may occur after allo-HSCT, including GI infections caused by Clostridium difficile or cytomegalovirus (CMV) re-activation and transplantation-associated thrombotic microangiopathy (TA-TMA), immunosuppressive medications such as mycophenolate mofetil (MMF), and adverse reactions to antibiotics, making it difficult to diagnose based on clinical symptoms alone. It should be noted that TA-TMA involves endothelial damage in multiple organs, and the intestine is one such target organ of TA-TMA, making intestinal TMA (iTMA) a manifestation of TA-TMA in the intestine, and iTMA is an ischaemic small bowel colitis caused by microangiopathy and can accompany GI-GVHD. If not treated in time, GVHD might lead to treatment failure and possibly to the death of the patient. Identifying intestinal GVHD accurately remains a challenge. The continuously developing electronic digestive track endoscopy has gradually been applied to paediatric patients with GI symptoms after allo-HSCT in recent years. However, there are also few studies on the specific endoscopic features of paediatric acute GI-GVHD. Therefore, this study analyzed the clinical data of 35 paediatric patients who underwent GI endoscopy due to suspected GI-GVHD after allo-HSCT.

The purpose of the present study was to elucidate the characteristic endoscopic findings in paediatric GI-GVHD, with the aim of improving the diagnostic accuracy of intestinal GVHD.

Materials and Methods

Patients
Among the 256 patients who underwent allo-HSCT in Wuhan Children's Hospital, PRC, from January 2019 to December 2023, 73 (28.5%) were clinically diagnosed with GI-GVHD based on Mount Sinai Acute Graft-versus-Host Disease International Consortium (MAGIC) criteria.^6,7

Inclusion criteria: (1) Age <18 years old; (2) Paediatric patients who underwent gastroenteroscopy after allo-HSCT from January 2019 to December 2023; (3) Exclusion of paediatric patients with only pathological manifestations of TMA: thrombosis in the microvasculature of biopsy tissues; (4) Bacterial and viral infections confirmed by stool routine tests and cultures were excluded.

Their median time from endoscopy to transplantation was 134 (30-570) days, and 12 of them had a clinical diagnosis of degree III-IV intestinal GVHD. In this study, only paediatric patients with histological pathological diagnosis consistent with GI-GVHD were enrolled, and only the first data consistent with GI-GVHD were enrolled if patients with 2-3 colonoscopies were examined. Data retrieved from the electronic medical record system included patient demographic characteristics, primary disease, donor type, graft source, conditioning regimen, and GVHD prophylaxis regimen.

Table 1 summarises the characteristics of these 35 patients, 18 males and 17 females, with a median age of 5.7 (0.75-17) years. Nineteen patients had malignant haematological disorders (including 14 cases of acute myeloid leukaemia, 2 cases of acute lymphoblastic leukaemia, 2 cases of myelodysplastic syndrome), Fifteen patients had benign haematological disorders (including 10 cases of SAA, 2 cases of phagocytic syndromes, 1 case of thalassaemia, 1 case of Fanconi anaemia, the 1 case of haemorrhagic platelet insufficiency), and 1 other disease (a chronic granulomatous disease). The diagnosis of clinical gastrointestinal GVHD was confirmed by endoscopic manifestations and histopathological evaluation in 24 paediatric patients, and data were collected on their clinical manifestations of GI-GVHD, other organs involved in the GVHD, endoscopic manifestations, gastrointestinal biopsies, GI-GVHD treatment, and whether or not they had changed their GI-GVHD management and other data. All included patients have consented to participation before any diagnostic work or treatment was performed, and all should have undergone the same diagnostic protocol. This retrospective single-centre cohort study at the Medical Ethics Committee of Wuhan Children's Hospital, Tongji Medical College, Huazhong University of Science and Technology according to the regulations of the local ethics committee (2023R022-F01), and written informed consent was obtained from all participants and/or their guardians. Consent was obtained from the patient's guardian for the publication of all patient gastrointestinal endoscopic images and information. No organs/tissues were procured from prisoners in this study.

Diagnostic and Grading Criterias of GI-GVHD
The choice of gastroduodenoscopy and colonoscopy was decided on the basis of the patient's clinical symptoms and multidisciplinary discussion. The preconditions needed to perform the procedures included a platelet count above 50×10⁹/L and normal coagulation function.

Grading of enteroscopic mucosal damage was based on enteroscopic mucosal manifestations, with the main endoscopic manifestations being mucosal congestion, mucosal oedema, mucosal erosion, mucosal ulceration, active bleeding and mucosal detachment. Regarding the endoscopic findings of intestinal GVHD, the Cruz-Correa et al8 intestinal mucosal grading criteria were used (Table 2).

Pathological biopsy diagnosis of GI-GVHD was based on the presence of crypt cell degeneration or crypt cell apoptosis and the presence or absence of crypt loss, and all patients underwent immunohistochemical testing of mucosal biopsies and molecular pathology testing to exclude viral infection. Histological grading of biopsies obtained endoscopically was performed by a blinded pathologist applying the modified system for colonic graft-versus-host disease described by Lerner et al (Table 3).⁹ GI-GVHD was recorded if the endoscopy showed positive histological evidence, including the presence of features such as apoptosis and crypt destruction. The results of GI endoscopy could not be used to exclude GI-GVHD. The first purpose of the endoscopy was to determine whether the patient's subsequent treatment needed to be adapted after the endoscopy, and, if no evidence of GI-GVHD was found, to promptly reduce or discontinue the use of immunosuppressants and methylprednisolone, and vice versa. The second purpose of the endoscopy was to determine whether GI endoscopy confirmed or denied the clinician's clinical suspicion of GI-GVHD.

For hepatic acute GVHD (aGVHD) grading, it was mainly assessed by clinical biochemical indicators. Hepatic grade I was defined as total bilirubin <34.2 μmol/L, while grade II was defined as total bilirubin ≥34.2 μmol/L. For skin aGVHD grading, skin grade I was defined as a rash area <25%; skin grade II was defined as a rash area of 25% to 50%; skin grade III was defined as a rash area >50% with generalised erythema; skin grade IV was defined as generalised erythroderma with maculopapular formation.¹⁰

Diagnosis of Combined Viral Enteritis
All patients underwent faecal microbiological examination at the onset of diarrhoea symptoms and polymerase chain reaction (PCR) of blood and mucosal biopsies to detect viral infection. CMV haemorrhagic disease was defined as the presence of peripheral blood CMV greater than 400 copies/ml as detected by real-time quantitative polymerase chain reaction or the presence of CMV infection confirmed by histopathological examination in patients after allo-HSCT. Colonoscopy revealed the presence of deep ulcers and a few atypical cells in the intestinal wall tissue.¹¹ Histopathological examination confirmed the presence of viral inclusion bodies, and immunohistochemical examination was positive for CMV antigen in some cases.¹² In cases of EBV infection, biopsy molecular pathology suggested EBV-encoded RNA (EBER) positivity.

Diagnostic Criteria of iTMA and GI-GVHD
Once GI symptoms were present, an iTMA and/or GI-GVHD diagnosis was made based on physical examination, laboratory data, and colonoscopy findings. TA-TMA was diagnosed according to the criteria proposed by the Hematopoietic Stem Cell Application Group, the Chinese Society of Hematology, Chinese Medical Association.¹³ TA-TMA diagnosis was made if at least three of the seven laboratory or clinical indicators were met or if tissue biopsy showed evidence of microthrombosis. The laboratory indicators were LDH above the upper normal limit, proteinuria, hypertension, new-onset thrombocytopenia, decreased haemoglobin, evidence of microangiopathy, and plasma sC5b-9 above the upper normal limit (252 ng/mL).¹⁴ Pathohistological features of iTMA included:¹⁵ (1) microthrombus deposition in mucosal vessels; (2) perivascular mucosal haemorrhage; (3) swelling of vascular endothelial cells; (4) endothelial cell stripping; (5) endothelial cell apoptosis; (6) erythrocyte fragmentation in the capillary lumen; (7) fibrin deposition or fibrinoid necrosis in the lumen; (8) complete mucosal detachment; (9) loss of the glands. Clinical diagnosis of GI-GVHD followed the Mount Sinai Acute Graft-versus-Host Disease International Consortium (MAGIC) criteria.¹⁶ The pathological diagnostic criteria for GI-GVHD were graded by pathologists blinded to the patient's data using the modified colon GVHD grading system described by Lerner et al.⁹ Histological biopsy results consistent with GI-GVHD and iTMA confirmed a combined diagnosis.

Treatment for GI-GVHD
The first-line treatment of GI-GVHD comprised methylprednisolone. If there was no response to the first-line treatment, the combination of second-line treatment can be considered, including the addition of tacrolimus, ruxolitinib, baliximab, etc. Currently, there was no standard dosing regimen for the second-line treatment, which is mainly based on the specific conditions of the child and the experience of the doctor. For patients with combined TMA, clinicians should immediately reduce or discontinue calcium-modulated phosphatase inhibitors, and add MMF, rituximab, eculizumab, low-molecular heparin calcium or plasma exchange, as appropriate. Immunosuppressive drugs were either reduced, or their blood concentration was adjusted in patients with viral enteritis. These patients were given ganciclovir and anti-CMV immunoglobulin. Changes in treatment were defined as initiation, dose escalation, dose reduction, or discontinuation of anti-GI-GVHD therapy, as well as other medication changes triggered by histopathologic findings.

Results

1. Clinical, endoscopic, and histopathologic manifestations of GI-GVHD
Of the 256 patients with allo-HSCT performed between 2019 and 2023, 73 paediatric patients (29%) were clinically diagnosed with GI-GVHD, of whom 35 (48%) had GI endoscopy completed. The characteristics of these 35 patients are summarised in Table 2. Twenty-eight of the 35 patients underwent endoscopic tissue biopsy pathology, and Table 3 details the endoscopic and histopathological grading, treatment, and prognosis of GVHD.

The main clinical symptoms of 35 patients with GI-GVHD included diarrhoea (n=14), haematochezia (n=11), haematemesis and haematochezia (n=2), abdominal pain (n=3), abdominal pain (n=6), and nausea with vomiting (n=2). Among the 15 patients with GI-GVHD complicated by TA-TMA, microthrombosis was observed in all cases on histopathological biopsy. Thirteen patients had sC5b-9 >252 ng/mL. Two patients lacked sC5b-9 data because plasma sC5b-9 testing had not yet been performed in our center when they were treated. Thirteen patients had markedly elevated LDH, six had proteinuria, a progressive decreases in platelets in 9 cases and in haemoglobin in 13 cases, and six had age-specific hypertension.

Gastro-duodenoscopy only was carried out in 2 patients, colonoscopy only in 10 patients, and combined gastro-duodenoscopy and colonoscopy in 23 cases. In this study, we summarised 35 paediatric patients with clinical manifestations of intestinal rejection, with abnormal changes in the mucosa of the gastric sinus (96%) in the vast majority of patients at gastroscopy, as well as distribution of lesions in the mucosa of the body of the stomach (52%), the fundus of the stomach (20%), and the duodenum (24%). At colonoscopy 14 patients did not have access to the end of the ileum due to difficulty in insertion of the colonoscope, so it was not possible to determine whether there were mucosal lesions at the end of the ileum in these 14 patients. There were abnormal mucosal changes in the colorectum in 85% (28/33) of the patients, and abnormal mucosa at the end of the ileum in 42% (8/19) of the patients. Most of the lesions were scattered and some were distributed diffusely. The main lesions were mucosal oedema, florid and patchy congestion, lamellar erosions, ulceration, and mucosal exfoliation; one patient had old bleeding and two patients had active bleeding. Mucosal erosions and ulcers appeared to be mostly seen in the colorectal mucosa, mostly scattered shallow ulcers predominated, with deep and massive ulcers in the stomach of one of the patients. The main gastroduodenoscopic manifestations included mucosal congestion (n=25), oedema (n=15), erosion (n=16), ulceration (n=3), and old bleeding spots (n=1). The main colonoscopic manifestations included mucosal congestion (n=10), oedema (n=20), erosion (n=15), ulceration (n=14), and exfoliation (n=3), and active bleeding (n=2). Examples of endoscopic findings of the Gastrointestinal tract are shown in Figure 1. One patient, who underwent colonoscopy only, presented no organic damage. The grading of gastroduodenorectoscopic and colonoscopic manifestations according to Cruz-Correa⁸ was 6 cases of grade I, 6 cases of grade II, 12 cases of grade III and 10 cases of grade IV.

Figure 1 Examples for endoscopic findings of Gastrointestinal tract: (A) gastric ulcer; (B) colonic mucosal oedema; (C) Colonic epithelial oedema, partial absence of mucosal epithelium; (D) Cecal ulcers that bleed when touched.

Based on the histopathological findings of GI endoscopy, the clinical diagnosis of GI-GVHD could not be confirmed by histological biopsy as histopathological records were incomplete in seven patients. The clinical diagnosis of GI-GVHD was confirmed histologically in 24 patients, representing 69% of patients who underwent endoscopy and 9% of all transplant recipients. The Lerner's⁹ histologic grades were grade I in five patients, grade II in eight, grade III in nine, and grade IV in two. The majority of patients (20/24) showed histological changes only in the lower GI tract, and four patients showed histological changes in both the upper and lower GI tracts. Biopsies reflect the entire histological spectrum of intestinal GVHD, including normal mucosa, crypt epithelial cell apoptosis, crypt destruction and loss, and mucosal denudation (Figure 2).

Figure 2 Histology of acute graft-versus-host disease of the colon. (A) and (B) Crypt epithelial cell apoptosis; (C) Crypt abscess; (D) Distortion of the crypt fossa structure; (E) Destruction of crypt fossa structure; (F) Suspected microthrombus deposits in the lumen of the vessel.

Four patients had nonspecific histological manifestations that did not correlate with the clinical diagnosis of GI-GVHD, but all had GI clinical manifestations. Patient 6 presented with subxiphoid abdominal pain and acid reflux. Gastroduodenoscopy revealed marked congestion and edema of the gastric body and sinuses with striated erosions. Patient 7 was infected with norovirus. Colonoscopy showed edema of intestinal mucosa, absence of vascular network, villous colonic mucosa, and flaky-like erythematous erosion in the descending colon and rectum. Histopathology seemed normal. Patient 12 showed fine particles over the entire duodenal mucosa under gastroduodenoscopy, and colon oedema, erosion, hyperemia, and bloody secretions under colonoscopy. Biopsy immunohistochemistry was consistent with CMV enteritis, and the microthrombosis visible in the blood vessels was consistent with TMA. In Patient 13, gastroduodenoscopy showed a huge ulcer of 2.0 × 2.5 cm on the anterior wall of the stomach’s greater curvature, oedema and erosion of the surrounding mucosa, and scattered erosion on the posterior wall. No organic lesions were observed on the mucosa below the transverse colon, and histopathology seemed normal.

2. Endoscopic pathological changes in patients with GI-GVHD complicated with viral enteritis or TA-TMA
Among the 24 patients with histopathologically confirmed GI-GVHD, the intestinal mucosa of 15 had histopathologic changes consistent with GVHD, including crypt structure disruption, mucosal crypt number reduction, apoptosis of crypt epithelial cells, crypt abscesses, and TMA-related changes such as microthrombus deposition in the mucosal vasculature also visible in endoscopy. Ten patients had cytomegalovirus anemia and three paediatric patients had combined cytomegalovirus enteritis. Heterogeneous large cells were seen in the interstitium of patient 15's cecum, which was positive for CMV antigen in combination with immunohistochemistry. A fecal examination in Patient 22 detected cytomegalovirus DNA (4.681×10⁹ copies/mL). Patient 23 showed a few atypical cells during histopathological examination of the intestinal wall tissue at the end of the ileum. Immunohistochemistry was positive for CMV antigen. Deep ulcers in Patient 26 were visible under endoscopy. Histopathology detected giant cells with viral inclusion bodies in the gastric sinus, ileocecal valve. Immunohistochemistry was positive for CMV antigen. Patient 28 was diagnosed with EBV infection as molecular pathology was positive for EBER in the antrum, cecum, and transverse colon. Histological results confirmed GI-GVHD in all these viral infection cases.

3. GI-GVHD treatment and prognosis
No acute adverse events occurred after the endoscopic procedure, and all patients tolerated the endoscopic examination and GI biopsy.

Seventeen patients started anti-GVHD therapy according to clinical criteria before endoscopy, and biopsy confirmed the diagnosis of clinical GI-GVHD in 15 (88%) patients, who continued to receive anti-GVHD therapy. Another 4 patients with negative biopsies tapered off anti-GVHD medication. Seventeen of 28 patients changed treatment because of biopsy results, 4 patients developed hormone dependence, 4 patients developed hormone resistance. Ten patients were treated with a clear diagnosis of intestinal GVHD after colonoscopy and intensified anti-GVHD therapy such as methylprednisolone dosing, addition of immunosuppressants or biologics. Six patients with mild GI-GVHD improved with first-line anti-GVHD therapy. Of the 35 patients, 26 improved with treatment. Nine patients had died, one due to recurrence of AML, five of TMA, three of septic shock.

Discussion

Despite the great advances in paediatric HSCT in the past decades, one of the toughest challenges is how to improve the quality of survival of paediatric HSCT recipients and to reduce the associated complications. Paediatric patients post-allo-HSCT are prone to complications such as intestinal GVHD, viral enteritis, TA-TMA, and antibiotic-associated diarrhoea, which have been difficult to identify based on clinical presentation alone.17 GI endoscopy is one of the most important means of diagnosing and treating GI diseases, and it is also suitable for the diagnosis and differential diagnosis of GI-GVHD because of its tolerability and safety. Although GI endoscopy can provide biopsies for histopathological examination, enteroscopic biopsies cannot be performed safely in some patients with GI-GVHD, as most patients with GI-GVHD are associated with severe thrombocytopenia and a marked tendency to bleed. Thirty-five patients included in this study had a median platelet (PLT) value of 119×109/L at the time of colonoscopy, and nine of them were transfused with platelets prior to endoscopy. Oh et al¹⁸ demonstrated that the risk of bleeding from endoscopic biopsy was significantly elevated when blood PLT was <50×10⁹/L. All patients in this study tolerated endoscopy and gastrointestinal biopsy, and there were no acute adverse events after endoscopic procedures.

There is a correlation between the histological diagnosis of the upper and lower GI tract. GI-GVHD is not only lower GI tract GVHD but also whole GI tract GVHD, and our test also showed that GI-GVHD is present in the upper GI tract. In our study we summarised 35 patients with clinical presentation of intestinal symptoms, out of which 4 patients had upper GI GVHD, for example, things like anorexia, nausea and vomiting may be the upper GI tract GVHD. For atypical clinical manifestations, therefore, both gastroscopy and colonoscopy need to be performed. For all patients who underwent complete colonoscopy are mainly diarrhoea with watery stools, which may be accompanied by abdominal cramping pain and, in severe cases, GI bleeding. Of the 35 patients who developed GI-GVHD in this study, 13 developed moderate to severe lower gastrointestinal bleeding, of whom 4 died. GI-GVHD combined with GI bleeding often indicates a poor prognosis and are an independent predictor of transplant-related death, with a mortality rate approaching 40% in such cases.^19,20

In this study, a possible association was found between endoscopic presentation, such as mucosal erosions, ulcers, and mucosal exfoliation, and grade 3-4 GI-GVHD. Of the 24 paediatric patients whose diagnosis of clinical GI-GVHD was confirmed by histological biopsy, erosions and ulcers or even mucosal exfoliation were seen in the mucosa of 18 patients, and 11 patients had a histological grading of grade III-IV. However, due to the insufficient sample size of the study, a histological grading of grade III-IV could not be further analyzed statistically. This was confirmed by Kreft et al.²¹ However, this correlation has not been established at this time.²² However, almost all cases of GI-GVHD with histological grade 4 showed macroscopic changes consistently observed on endoscopic biopsy. Furthermore, 7 of these 24 patients (29%) with histologically confirmed GI-GVHD had a histological grade was lower than the clinical grade, and 8 (33%) had a histological grade was lower than the endoscopic grade. Notably, some patients exhibited severe clinical manifestations but had only mild pathological damage. With increasing clinical experience, clinicians could proactively administer anti-GVHD treatment in severe cases, resulting in improved treatment outcomes.

On the other hand, endoscopically, the mucosa of 2 patients in this study showed normal or slightly abnormal findings, while histologically, mucosal damage up to grade 3 GI-GVHD could be seen. This could be due to subtle changes in monocyte apoptosis, disruption of isolated crypts, or tiny erosions, which are consistent with the histological diagnosis of GVHD but cannot be detected by conventional endoscopy.²³ Some studies have shown that the concordance between the macroscopic and histological findings in acute GVHD was as low as 38.9% and that a difference of two grades or more was observed in 28% of the cases.^24,25 In this study, the agreement between macroscopic and histological manifestations was 39.29%, and differences of two or more grades were observed in 21.43% of cases. Therefore, endoscopy biopsies of normal and abnormal mucosal tissues are essential for a reliable diagnosis. The GI mucosa might appear normal in mild GI-GVHD cases, while in moderate cases, it might present with diffuse oedema and congestion. In severe cases, GI-GVHD might lead to ulcers in the stomach, small intestine, and colon and large mucosal shed areas.²⁶

In our study, the incidence of intestinal GVHD was 29% (73/256), and we believe that the prevalence of GVHD in patients with GI symptoms after allo-HSCT is high, but not all biopsies show histological manifestations of GI-GVHD, and even if histological features of gastrointestinal bacterial or viral infections are found, although negative histological manifestations cannot exclude GI-GVHD, positive histological manifestations can further confirm the clinical diagnosis of GI-GVHD and guide targeted treatment. Based on clinical diagnostic criteria of GI-GVHD, 17 patients were started on anti-GVHD therapy prior to gastrointestinal endoscopy, and biopsies confirmed the diagnosis of clinical gastrointestinal GVHD in 15 patients (88%), and 10 patients intensified their anti-GVHD therapy after the diagnosis was clarified by colonoscopy. Methylprednisolone or immunosuppressants had to be gradually reduced or discontinued in patients with negative histopathological results and improved GI symptoms to avoid unnecessary long-term immunosuppression. Cautious management was needed for those with continued symptoms but negative biopsy results, as the lesion might be in the GI tract but was not detected by endoscopy. Also, the low status of patients with severe intestinal GVHD and thrombocytopenia may limit the invasiveness of endoscopy, so biopsies are usually selected only from obvious sites of mucosal lesions. Due to the possible inconsistent distribution of GI-GVHD lesions in the digestive tract, the choice of gastroscopy or colonoscopy and the depth that colonoscopy should reach have been debated in recent years.^27,28

This study involved 24 patients with histologically confirmed GI-GVHD. Among them, 4 cases were complicated by CMV enteritis, 1 case by EBV enteritis, and 15 cases by TA-TMA. The early clinical manifestations were all similar and difficult to differentiate. Timely gastrointestinal endoscopy with biopsy pathology can assist clinicians in promptly adjusting treatment strategies. Currently, there are limited reports on the endoscopic findings of GI-GVHD, iTMA, and intestinal CMV infection in paediatric patients following allo-HSCT. Furthermore, no standardised criteria for the diagnosis of iTMA have been established. Gastrointestinal endoscopy with histopathological biopsy remains the gold standard for diagnosing both GI-GVHD and iTMA in paediatric patients.

The main limitation of this study is its retrospective nature, not all patients with a clinical suspicion or diagnosis of GI-GVHD underwent GI endoscopy, which prevented us from determining its true sensitivity. In addition, this study had some limitations as the number of patients was small, and the study lacked data comparison, making it impossible to obtain statistically significant results. Nevertheless, our study demonstrates the importance of GI endoscopy and histologic biopsy for the diagnosis of intestinal complications in paediatric patients with HSCT. However, multicenter prospective clinical studies with large samples are needed next to further improve the diagnosis and treatment of GI-GVHD through GI endoscopy.

Declaration of Interest

• Data availability statement: The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
• Funding statement: This study was funded by Wuhan Clinical Medical Research Project (2021FE001, 2022FEYJS006,WX21Z48).
• Ethics approval statement: Ethical approval was obtained from the Ethics Committee of Wuhan Children's Hospital (2023R022-F01)
• Patient consent statement/Informed consent statement: The study was done after agreement from the local ethics committee and written informed consent was obtained from all participants and/or their guardians.
• Permission to reproduce material from other sources: Not applicable
• Clinical trial registration: Not applicable

References

1. Gratwohl A, Baldomero H, Aljurf M, et al. Hematopoietic stem cell transplantation: a global perspective. JAMA 2010;303:1617-24.

2. Thomas E, Storb R, Clift R A, et al. Bone-marrow transplantation (first of two parts). N Engl J Med 1975;292:832-4.

3. Ferrara JL, Levine JE, Reddy P, Holler E. Graft-versus-host disease. Lancet 2009,373:1550-61.

4. MacMillan M L, Holtan S G, Rashidi A, et al. Pediatric acute GVHD: clinical phenotype and response to upfront steroids. Bone Marrow Transplant 2020;55:165-71.

5. Dignan FL, Clark A, Amrolia P, et al. Diagnosis and management of acute graft-versus-host disease. Br J Haematol 2012;158:30-45.

6. Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant 1995;15:825-8.

7. Martensson T, Mellgren K, Toporski J, et al. Clinical relevance of endoscopy with histopathological assessment in children with suspected gastrointestinal graft-versus-host disease. Clin Transplant 2020;34:e13867.

8. Cruz-Correa M, Poonawala A, Abraham SC, et al. Endoscopic findings predict the histologic diagnosis in gastrointestinal graft-versus-host disease. Endoscopy 2002;34:808-13.

9. Lerner KG, Kao GF, Storb R, et al. Histopathology of graft-vs.-host reaction (GvHR) in human recipients of marrow from HL-A-matched sibling donors. Transplant Proc 1974,6:367-71.

10. Stem Cell Application Group, Chinese Society of Hematology, Chinese Medical Association. [Chinese consensus of allogeneic hematopoietic stem cell transplantation for hematological disease (Ⅲ) -acute graft-versus-host disease (2020)]. Zhonghua Xue Ye Xue Za Zhi 2020;41:529-36.

11. Bhutani D, Dyson G, Manasa R, et al. Incidence, risk factors, and outcome of cytomegalovirus viremia and gastroenteritis in patients with gastrointestinal graft-versus-host disease. Biol Blood Marrow Transplant 2015;21:159-64.

12. He JD, Liu YL, Wang ZF, Liu DH, Chen H, Chen YH. Colonoscopy in the diagnosis of intestinal graft versus host disease and cytomegalovirus enteritis following allogeneic haematopoietic stem cell transplantation. Chin Med J (Engl) 2008;121:1285-9.

13. Hematopoietic Stem Cell Application Group, Chinese Society of Hematology, Chinese Medical Association. [Chinese consensus on the diagnosis and management of transplant-associated thrombotic microangiopathy (2021)]. Zhonghua Xue Ye Xue Za Zhi 2021;42:177-84.

14. Horvath O, Kallay K, Csuka D, et al. Early Increase in Complement Terminal Pathway Activation Marker sC5b-9 Is Predictive for the Development of Thrombotic Microangiopathy after Stem Cell Transplantation. Biol Blood Marrow Transplant 2018;24:989-96.

15. El-Bietar J, Warren M, Dandoy C, et al. Histologic Features of Intestinal Thrombotic Microangiopathy in Pediatric and Young Adult Patients after Hematopoietic Stem Cell Transplantation. Biol Blood Marrow Transplant 2015;21:1994-2001.

16. Martensson T, Mellgren K, Toporski J, et al. Clinical relevance of endoscopy with histopathological assessment in children with suspected gastrointestinal graft-versus-host disease. Clin Transplant 2020;34:e13867.

17. Yamada-Fujiwara M, Miyamura K, Fujiwara T, et al. Diagnosis of intestinal graft-versus-host disease and thrombotic microangiopathy after allogeneic stem cell transplantation. Tohoku J Exp Med 2012;227:31-7.

18. Oh HJ, Park JM, Yoon S B, et al. Bleeding After Endoscopic Procedures in Patients With Chronic Hematologic Thrombocytopenia. Dig Dis Sci 2017;62:746-54.

19. Wachsmuth LP, Patterson MT, Eckhaus MA, et al. Post-transplantation cyclophosphamide prevents graft-versus-host disease by inducing alloreactive T cell dysfunction and suppression. J Clin Invest 2019;129:2357-73.

20. Schwartz JM, Wolford JL, Thornquist MD, et al. Severe gastrointestinal bleeding after hematopoietic cell transplantation, 1987-1997: incidence, causes, and outcome. Am J Gastroenterol 2001;96:385-93.

21. Kreft A, Schulze L, Ries I, Schindeldecker M, Neumann H. Histological diagnosis of acute graft-versus-host disease in different sites of the upper gastrointestinal tract with correlation to endoscopic findings. Leuk Lymphoma 2023;64:433-9.

22. Hiejima E, Nakase H, Matsuura M, et al. Diagnostic accuracy of endoscopic features of pediatric acute gastrointestinal graft-versus-host disease. Dig Endosc 2016;28:548-55.

23. Kreft A, Mottok A, Mesteri I, et al. Consensus diagnostic histopathological criteria for acute gastrointestinal graft versus host disease improve interobserver reproducibility. Virchows Arch 2015,467:255-63.

24. Kreisel W, Dahlberg M, Bertz H, et al. Endoscopic diagnosis of acute intestinal GVHD following allogeneic hematopoietic SCT: a retrospective analysis in 175 patients. Bone Marrow Transplant 2012;47:430-8.

25. Velasco-Guardado A, Lopez-Corral L, Alvarez-Delgado A, et al. Endoscopic evaluation and histological findings in graft-versus-host disease. Rev Esp Enferm Dig 2012;104:310-4.

26. Naymagon S, Naymagon L, Wong S Y, et al. Acute graft-versus-host disease of the gut: considerations for the gastroenterologist. Nat Rev Gastroenterol Hepatol 2017;14:711-26.

27. Kus OS, Kirmizi A, Aras N, et al. Patterns and Grading of Gastrointestinal Graft-Versus-Host Disease: A Clinicopathologic Correlation Study. Turk J Gastroenterol 2023;34:516-24.

28. Sharaf RN, Shergill AK, Odze RD, et al. Endoscopic mucosal tissue sampling. Gastrointest Endosc 2013;78:216-24.