Clinical and pathological features and treatment of nine cases of gastrointestinal composite gangliocytoma/neuroma and neuroendocrine tumor

Abstract

BACKGROUND

Composite gangliocytoma/neuroma and neuroendocrine tumor (CoGNET), previously known as gangliocytic paraganglioma, is a rare neuroendocrine neoplasm with metastatic potential. CoGNET occurs most frequently in the duodenum, presents with non-specific symptoms, is usually detected incidentally on endoscopy or cross-sectional imaging, and requires histopathological examination as the sole means of definitive diagnosis. To date, clinical research on CoGNET is limited to isolated case reports in China, with a notable lack of more cases to systematically characterize its full spectrum of clinicopathological features, treatment outcomes, and long-term prognosis.

AIM

To investigate the clinicopathological features and treatment of gastrointestinal CoGNET.

METHODS

This retrospective case series was approved by the ethics review committee of a tertiary care hospital in China. Nine patients with pathologically confirmed gastrointestinal CoGNET from 2019 to 2024 were enrolled, and cases with incomplete clinical data were excluded. Clinicopathological data including demographics, clinical manifestations, endoscopic and imaging findings, treatment regimens and pathological results were collected. All patients were followed up via outpatient visits and telephone interviews.

RESULTS

Nine patients with pathologically confirmed CoGNET were included, with a mean age of 51.2 years and a male-to-female ratio of 5:4. Eight lesions were located in the duodenum, and one was the first reported primary rectal CoGNET. All lesions achieved en bloc resection (4 surgical, 5 endoscopic), with one case of regional lymph node metastasis. All lesions presented the typical triphasic cellular differentiation, and no tumor recurrence, metastasis or death occurred during a mean follow-up period of 41 months.

CONCLUSION

CoGNET presents an indolent clinical course and favorable prognosis, and endoscopic resection is effective for eligible patients at low risk of lymph node metastasis.

Key Words: Composite gangliocytoma/neuroma and neuroendocrine tumor; Clinicopathological features; Endoscopic therapy; Immunohistochemistry; Prognosis

Core Tip: Composite gangliocytoma/neuroma and neuroendocrine tumor (CoGNET) is an extremely rare neuroendocrine neoplasm. This study reports the largest case series of gastrointestinal CoGNET in China to date, including the world’s first reported case of primary rectal CoGNET as we know. We systematically analyzed the clinicopathological, endoscopic, and radiological characteristics of 9 pathologically confirmed CoGNET cases. Our findings validated the safety and long-term efficacy of endoscopic resection, and confirmed an excellent prognosis for all patients, including the case with lymph node metastasis. This work provides pivotal real-world evidence to optimize clinical diagnosis and treatment strategies for this rare disease.

INTRODUCTION

Composite gangliocytoma/neuroma and neuroendocrine tumor (CoGNET) is a rare neuroendocrine neoplasm that most commonly arises in the descending duodenum, with less frequent primary sites including the respiratory tract, spinal cord, jejunum, esophagus, and appendix[1]. Initially reported as ganglioneuroma by Dahl et al[2], it was renamed gangliocytic paraganglioma by Kepes et al[3], before receiving its current official designation - CoGNET - in the 2022 the 5^th edition of the World Health Organization Tumor Classification[4].

Clinically, patients with CoGNET may present with non-specific symptoms including abdominal pain, gastrointestinal bleeding, dysphagia, nausea, and diarrhea, while a subset of affected individuals are asymptomatic, with lesions identified incidentally on endoscopic or radiological imaging; definitive diagnosis is established via histopathological examination. The defining pathological feature of CoGNET is the presence of three distinct cellular components: Epithelioid neuroendocrine cells, spindle-shaped Schwann cells, and ganglion-like cells. While CoGNET is generally considered to follow an indolent clinical course with a favorable prognosis, a comprehensive literature review of 153 relevant studies reported lymph node metastasis in 25 (10%) of 254 included patients, and distant metastasis in 6 (2%) of cases[1].

To date, published studies on CoGNET conducted in China have been largely restricted to sporadic case reports, and systematic analyses of large case series remain scarce. Here, we retrospectively analyzed the clinicopathological data of nine patients with CoGNET, including the first globally reported case of primary rectal CoGNET. We systematically characterized the clinicopathological features, radiological and endoscopic manifestations, treatment strategies, and clinical outcomes of this cohort, with the aim of providing a practical reference for clinicians in the diagnosis and therapeutic decision-making for this rare neoplasm.

MATERIALS AND METHODS

Study design

This was a single-center, retrospective case series study conducted at the First Affiliated Hospital of Zhejiang University. The study protocol was approved by the Ethics Review Committee of the First Affiliated Hospital of Zhejiang University, and the requirement for written informed consent was waived due to the retrospective, non-interventional nature of the study. All procedures were conducted in compliance with the principles of the Declaration of Helsinki.

Study population

We consecutively enrolled patients with pathologically confirmed CoGNET who underwent surgical or endoscopic resection at our center between January 2019 and December 2024. The diagnosis of CoGNET was established in accordance with the 2022 the 5^th edition of the World Health Organization Classification of Tumors of the Digestive System, based on the presence of the characteristic triphasic histological pattern and confirmatory immunohistochemical (IHC) staining. Exclusion criteria were as follows: (1) Incomplete clinical, pathological, or follow-up data; (2) Preoperative receipt of neoadjuvant antitumor therapy; and (3) Secondary metastatic lesions of non-gastrointestinal primary CoGNET. A total of nine eligible patients were included in the final analysis.

Data collection

Clinical data were retrospectively extracted from the hospital electronic medical record system, including: (1) Baseline demographic characteristics (age, sex, comorbidities, past medical history); (2) Clinical manifestations at initial presentation; (3) Preoperative imaging findings [including contrast-enhanced computed tomography (CT), Figure 1]; (4) Endoscopic and endoscopic ultrasonography (EUS) features (Figure 2); (5) Treatment modalities and perioperative outcomes; (6) Histopathological and IHC characteristics; and (7) Follow-up and survival data.

Figure 1 Contrast-enhanced abdominal computed tomography reveals a small nodular soft tissue density lesion in the horizontal part of the duodenum, measuring approximately 1.5 cm × 1.2 cm, which demonstrates marked enhancement after contrast administration (arrow). A: Unenhanced axial view; B: Contrast-enhanced axial view; C: Contrast-enhanced coronal view; D: Contrast-enhanced sagittal view.

Figure 2 Endoscopic image and endoscopic ultrasonography image of composite gangliocytoma/neuroma and neuroendocrine tumor (A and B in the duodenum, C and D in the rectum). A: White light endoscopy reveals a smooth-surfaced submucosal protrusion in the descending part of the duodenum, adjacent to the orifice of the major duodenal papilla; B: Endoscopic ultrasonography reveals a hypoechoic lesion arising from the muscularis propria (arrow), with heterogeneous internal echoes and a maximum cross-sectional dimension of approximately 1.64 cm × 1.54 cm. C: White light endoscopy reveals a submucosal protrusion with erythematous surface at 10 cm from the anal verge in the rectum; D: Endoscopic ultrasonography reveals a moderate to low echoic mass arising from the deep mucosa with submucosal involvement, well-defined borders, a measured cross-sectional dimension of approximately 1.35 cm × 0.91 cm, and an intact muscularis propria and adventitia.

Pathological evaluation

All resected tumor specimens were fixed in 10% neutral buffered formalin, embedded in paraffin, and serially sectioned. Hematoxylin and eosin (HE) staining was performed for routine histological assessment. All slides were initially reviewed by two independent board-certified surgical pathologists, and a senior gastrointestinal pathologist with over 15 years of experience performed a blinded central review to confirm the final diagnosis of CoGNET.

IHC staining was performed on formalin-fixed paraffin-embedded sections using a standard polymer detection system. The IHC marker panel included pan-cytokeratin (CK-pan), chromogranin A (CgA), synaptophysin (Syn), S-100 protein, cluster of differentiation 56 (CD56), SOX10, and Ki-67. The staining results were interpreted by two independent pathologists, with discrepancies resolved via consensus review. Representative HE and IHC staining images of CoGNET are presented in Figures 3 and 4.

Figure 3 Pathological and immunohistochemical features of composite gangliocytoma/neuroma and neuroendocrine tumor (from the patient shown in the previous computed tomography images). A: Pathological features of the lesion under low-power magnification (hematoxylin and eosin); B: High-power view showing spindle-shaped tumor cells arranged in fascicular sheets, nested pseudoglandular epithelioid tumor cells, and scattered ganglion-like cells with distinct nuclear membranes and prominent nucleoli (hematoxylin and eosin); C: Immunohistochemical (IHC) staining for pan-cytokeratin: Cytoplasmic positivity in epithelioid cells, with negative expression in spindle cells and ganglion-like cells; D: IHC staining for S-100: Nuclear positivity in spindle-shaped Schwann cells, with negative expression in epithelioid cells; E: IHC staining for SOX10: Nuclear positivity in spindle cells, with negative expression in epithelioid cells (in the lower right corner); F: IHC staining for synaptophysin: Diffuse cytoplasmic positivity in epithelioid cells and ganglion-like cells.

Figure 4 Pathological and immunohistochemical features of rectal composite gangliocytoma/neuroma and neuroendocrine tumor. A: Pathological features of the lesion under low-power magnification (hematoxylin and eosin); B: High-power view showing intermingled growth of spindle cells and epithelioid cells with scattered ganglion cells (hematoxylin and eosin); C: Immunohistochemical (IHC) staining for chromogranin A: Strong cytoplasmic positivity in ganglion-like cells with focal weak positivity in epithelioid cells; D: IHC staining for S-100: Cytoplasmic and nuclear positivity in spindle cells, with negative expression in epithelioid and ganglion-like cells; E: IHC staining for Ki-67: Low proliferative index; F: IHC staining for synaptophysin: Focal cytoplasmic positivity in epithelioid and ganglion-like cells.

Follow-up

All included patients were followed up postoperatively via regular outpatient review and standardized telephone follow-up. The follow-up deadline was October 1, 2025. The primary follow-up endpoint was recurrence-free survival, defined as the time from curative resection to the first occurrence of local recurrence, regional lymph node metastasis, or distant metastasis. Secondary endpoints included overall survival (defined as the time from resection to all-cause mortality or the last follow-up).

Statistical analysis

Statistical analyses were performed using R version 4.3.3. Categorical and continuous variables were expressed as n (%) and mean (SD) or median (interquartile range), respectively. The difference in lesion size between groups was compared using the independent-samples t-test.

RESULTS

General clinical characteristics

A total of nine consecutive patients with nine pathologically confirmed gastrointestinal CoGNET were enrolled in this retrospective study. The mean patient age was 51.2 years (range, 23-72 years), with a male-to-female ratio of 5:4. None of the patients had a prior history of malignant neoplasms. Five patients presented with clinical symptoms: Abdominal pain in three cases and abdominal distension in two cases. The remaining four patients were asymptomatic, with lesions detected incidentally via endoscopic or radiological examinations (Table 1). There was no significant difference in tumor size between asymptomatic patients (with an average size of 2.08 ± 1.37 cm) and symptomatic patients (with an average size of 2.18 ± 1.06 cm) (P = 0.91).

Table 1 Clinicopathologic characteristics, treatment and follow-up of 9 composite gangliocytoma/neuroma and neuroendocrine tumors.

No.	Sex	Age (year)	Clinical symptoms	Site	Size (cm)	Incisal margin	Metastasis	Recurrence	Treatment	Follow-up (month)
1	M	66	Abdominal distension	Descending part of duodenum	4.5	N	N	N	ESD	10
2	M	56	Abdominal distension	Descending part of duodenum	1.4	N	N	N	ERCP	14
3	M	54	N	Descending part of duodenum	2	N	N	N	Partial duodenectomy	36
4	F	72	Abdominal pain	Descending part of duodenum	1.8	N	Lymph node	N	Pancreaticoduodenectomy	40
5	F	40	Abdominal pain	Horizontal part of the duodenum	1.5	N	N	N	Local resection of ampullary tumor	41
6	F	23	Abdominal pain	Bulb-descending junction of the duodenum	1.2	N	N	N	EMR	53
7	M	52	N	Descending part of duodenum	1.7	N	N	N	ERCP	69
8	M	38	N	Horizontal part of the duodenum	3.7	N	N	N	Pancreaticoduodenectomy	69
9	F	60	N	Rectum	1.3	N	N	N	ESD	40

M: Male; F: Female; N: Negative; ESD: Endoscopic submucosal dissection; ERCP: Cholangiopancreatography-guided papillectomy; EMR: Endoscopic mucosal resection.

Lesion distribution, endoscopic and radiological findings

Among the nine lesions, six were located in the descending duodenum, one in the horizontal duodenum, one in the duodenal bulb-descending junction, and the remaining one lesion arose in the rectum (Table 1).

Complete endoscopic evaluation was performed in eight patients; the remaining one received external endoscopy without repeated testing at our center. Endoscopically, lesions manifested as submucosal protrusions in four cases, mucosal protrusions in two cases, and enlarged duodenal papillae in two cases. Six patients underwent EUS, which revealed hypoechoic lesions in five cases and a single isoechoic lesion in one case. Preoperative differential diagnoses based on EUS included neuroendocrine neoplasms, gastrointestinal stromal tumors, and cystadenomas. Seven patients underwent preoperative contrast-enhanced CT, which demonstrated well-circumscribed, round or nodular soft-tissue lesions with contrast enhancement in all cases: Mild-to-moderate enhancement in four cases and intense enhancement in the remaining three cases (Table 1).

Treatment modalities

Among the nine patients, four underwent surgical treatment, including two cases of pancreaticoduodenectomy (Whipple procedure), one case of partial duodenectomy, and one case of local resection of ampullary tumor. Five patients underwent endoscopic treatment, including two cases of endoscopic retrograde cholangiopancreatography-guided papillectomy, two cases of endoscopic submucosal dissection (ESD), and one case of endoscopic mucosal resection. En bloc resection was achieved for all tumors (Table 1).

Histopathological and IHC findings

Histopathological evaluation revealed a mean maximum tumor diameter of 2.1 cm (range, 1.2-4.5 cm). Most lesions were confined to the submucosa, with invasion into the muscularis propria noted in three cases. No definite lymphovascular invasion or perineural invasion was identified on microscopic examination in any of the nine cases. Notably, one patient who underwent radical pancreatoduodenectomy had a single perienteric lymph node with confirmed metastasis; pathological examination verified negative surgical margins for this case, with systematic lymph node dissection conducted in the peripancreatic, perigastric and perienteric regions. The metastatic focus was composed purely of epithelioid cells, with weak immunoreactivity for CK-pan, diffuse positivity for CgA, and absent staining for S-100 protein.

All lesions, including the rectal CoGNET, displayed the characteristic triphasic differentiation pattern of CoGNET, comprising three distinct cellular components: Epithelioid neuroendocrine cells, spindle-shaped Schwann cells, and ganglion-like cells. Immunohistochemically, the Ki-67 proliferation index was 3% in three cases, 2% in two cases, and 1% in the remaining four cases. Among the three patients with a Ki-67 index of 3%, two were treated with endoscopic resection and one underwent surgical resection. All lesions showed diffuse positivity for CgA, Syn, and S-100 protein, while seven cases were concomitantly positive for CK-pan and CD56. The rectal CoGNET was negative for CK-pan and CD56, whereas most duodenal lesions showed positive expression of these markers (Table 2).

Table 2 Immunohistochemical findings of 9 composite gangliocytoma/neuroma and neuroendocrine tumors.

No.	Ki-67 (%)	CK-pan	CgA	CD56	Syn	S-100	SOX10
1	1	P	P	P	P	P	P
2	3	P	P	N	P	P	P
3	1	P	P	P	P	P	P
4	1	P	P	P	P	P	P
5	3	P	P	P	P	P	P
6	3	N	P	P	P	P	N
7	2	P	P	P	P	P	N
8	2	P	P	P	P	P	N
9	1	N	P	N	P	P	N

Ki-67 indices were determined based on a 500-cell count in an expression hotspot. CK-pan: Pan-cytokeratin; CgA: Positive expression of chromogranin A; Syn: Synaptophysin; SOX10: SRY-box transcription factor 10; P: Positive; N: Negative.

Follow-up and prognosis

Clinical follow-up was completed in all nine patients, with a mean follow-up duration of 41 months (range, 10-69 months); seven patients (75%) were followed up for over 3 years. No patient, including the individual with lymph node metastasis, received adjuvant postoperative antitumor therapy. No locoregional recurrence, metachronous neoplasms, or distant metastasis were observed during follow-up. All patients were alive at the last follow-up, with no major procedure-related adverse events documented (Table 1).

DISCUSSION

CoGNET is a rare subtype of neuroendocrine neoplasm, and to date, published studies on CoGNET conducted in China have been largely restricted to sporadic case reports. In this study, we retrospectively analyzed the largest and most comprehensive single-center case series of CoGNET in China to date, with a systematic characterization of its clinical, radiological, and histopathological features, treatment strategies, and long-term prognosis.

CoGNET exhibits a strong predilection for the duodenum, with nearly half of cases arising in the descending duodenum (periampullary region), and frequent involvement of the major duodenal papilla[5]. In our cohort, apart from duodenal primary lesions, we identified an extremely rare case of primary rectal CoGNET, which, to the best of our knowledge, represents the first globally reported case of primary rectal CoGNET. Clinically, patients with CoGNET may present with non-specific symptoms including abdominal pain, abdominal distension, jaundice, and gastrointestinal bleeding, while a substantial proportion of affected individuals are asymptomatic, with lesions detected incidentally via CT, magnetic resonance imaging (MRI), or upper gastrointestinal endoscopy[6]. Consistent with prior literature, the median age at diagnosis in our cohort was approximately 50 years, with a slight male predominance; previously reported studies have described a male-to-female ratio of approximately 1.5:1, compared with 1.25:1 in our study population.

Endoscopically, duodenal CoGNET most commonly manifests as submucosal lesions or an enlarged duodenal papilla, while EUS typically reveals hypoechoic or isoechoic nodules. As a result, CoGNET is frequently misdiagnosed preoperatively as conventional well-differentiated neuroendocrine neoplasms, gastrointestinal stromal tumors, or cystadenomas. Conventional cross-sectional imaging with CT and MRI also has limited accuracy in differentiating CoGNET from other duodenal neoplasms. Encouragingly, prior studies have demonstrated high ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) avidity in CoGNET lesions, suggesting that ¹⁸F-FDG positron emission tomography (PET)/CT may serve as a valuable modality for the detection of CoGNET and its differentiation from conventional neuroendocrine neoplasms[7]. The retrospective design of this study leads to heterogeneity in preoperative imaging examinations, which limits the diagnostic reproducibility and inter-case comparability. For the clinical diagnosis and research of CoGNET in the future, a standardized imaging assessment protocol integrating EUS and contrast-enhanced CT should be established to improve the standardization and accuracy of preoperative imaging evaluation for this rare neoplasm.

Histopathologically, the defining feature of CoGNET is its characteristic triphasic cellular composition, comprising three distinct cell lineages: (1) Nested epithelioid neuroendocrine cells, morphologically analogous to well-differentiated neuroendocrine neoplasms, with immunoreactivity for CgA, cytokeratin, and Syn; (2) Spindle-shaped Schwann cells, typically surrounding the epithelial cell nests, resembling the spindle cells of peripheral nerve sheath tumors, and expressing neuron-specific enolase, S-100, and SOX10; and (3) Scattered ganglion-like cells with amphophilic cytoplasm and prominent nucleoli, showing positive staining for CgA, CD56, and Syn. However, emerging evidence has shown that despite their morphological resemblance to mature ganglion cells, these ganglion-like cells exhibit positive cytokeratin expression and absent staining for SOX10 and GATA3, casting doubt on their true ganglionic differentiation[8]. Compounding the diagnostic challenge, the neuroendocrine component of CoGNET shares significant morphological and immunophenotypic overlap with somatostatin-expressing neuroendocrine neoplasms[8], frequently leading to misdiagnosis. Accordingly, the combined assessment of the unique triphasic architecture, alongside the co-expression profile of pancreatic polypeptide and somatostatin, is critical for the accurate differential diagnosis of CoGNET[9]. Further studies have revealed that ganglion-like cells in CoGNET consistently express ARX (aristaless related homeobox), cytokeratin, pancreatic polypeptide, somatostatin, Syn, and islet-1, mirroring the immunophenotype of the epithelioid neuroendocrine cell nests, while lacking expression of the mature ganglion cell differentiation marker PHOX2B. These findings collectively indicate that the immunophenotypic profile of ganglion-like cells is more closely aligned with neoplastic neuroendocrine cells, rather than true ganglion cells[10]. However, specific molecular markers including ARX and islet-1 were not analyzed in our study due to its retrospective design and the rarity of this tumor. Future studies with larger cohorts are warranted to evaluate these markers to further characterize the neuroendocrine differentiation of CoGNET.

In our cohort, lymph node metastasis was observed in one of nine (11.1%) patients, which is highly consistent with the previously reported lymph node metastasis rate of 10%-20% in the literature[5,11,12]. In line with our findings, most published studies have reported no disease recurrence or progression in patients with lymph node metastasis, even in the absence of reoperation or adjuvant therapy[12]. While no distant metastasis was identified in our study population, prior studies have documented liver metastasis in 1.1% of CoGNET patients[11], along with a fatal case of CoGNET secondary to widespread liver metastasis[13]. These data highlight that although CoGNET generally follows an indolent clinical course, it harbors inherent malignant potential with rare but clinically significant metastatic risk.

Currently, there is no standardized consensus on the optimal first-line treatment strategy for CoGNET. Nevertheless, given the well-documented metastatic risk, complete resection (surgical or endoscopic) remains the cornerstone of curative treatment for CoGNET[14]. Endoscopic resection offers a minimally invasive, cost-effective treatment option with a favorable safety profile and rapid postoperative recovery, and has emerged as a reliable therapeutic approach for carefully selected CoGNET patients. In our cohort, even the largest lesion (4.5 cm × 2 cm × 1.2 cm) was successfully resected en bloc via ESD, with uneventful postoperative recovery and no evidence of recurrence or metastasis during long-term follow-up. The rectal CoGNET was completely resected via ESD with negative surgical margins, and no metastasis was detected, and the patient had a favorable prognosis with no recurrence after 40 months of follow-up, which was consistent with the duodenal CoGNET. However, it is important to note that endoscopic resection cannot achieve regional lymph node dissection or staging. Therefore, surgical resection should be strongly considered for patients with high risk of lymph node metastasis, as stratified by preoperative radiological imaging and EUS. Previous studies have identified tumor size > 2 cm and invasion beyond the submucosa as independent risk factors for lymph node metastasis in CoGNET[11,15], and individualized treatment selection is particularly critical for this high-risk subgroup.

This study has inherent limitations including a small sample size (n = 9) and a single-center retrospective design, which may restrict the generalizability of the findings, especially for the novel primary rectal CoGNET subtype reported here for the first time. Establishing standardized treatment selection criteria is important for CoGNET to optimize clinical decision-making and improve the standardization of its clinical management. However, the extreme rarity of CoGNET leads to a lack of large-sample clinical data at present, making it difficult to formulate rigid, universal standardized criteria for endoscopic and surgical treatment selection. Tumor size alone cannot be used as an absolute indicator for treatment decision-making of CoGNET, and clinical treatment selection should be based on a comprehensive individualized assessment integrating multiple factors including tumor invasion depth, lesion location, and preoperative EUS/CT features. Multicenter collaboration and larger-sample registry studies are urgently needed in the future to accumulate more clinical evidence. For future research, in-depth molecular analytical studies of CoGNET, including next-generation sequencing and transcriptomic analysis, will help identify the driver mutations and core signaling pathways underlying tumorigenesis in CoGNET and elucidate the molecular heterogeneity between duodenal and rectal subtypes. In addition, establishing a standardized, operable diagnostic workflow for CoGNET by integrating endoscopic, radiological and histopathological criteria will be further verified and improved via multicenter clinical practice to enhance diagnostic consistency and accuracy, thereby providing a reliable diagnostic basis for the standardized clinical management of this rare neoplasm.

CONCLUSION

This study includes the largest case series of duodenal CoGNET in China to date, as well as the first known reported case of primary rectal CoGNET. The characteristic triphasic cellular components in histopathology are the core basis for the diagnosis and differential diagnosis of CoGNET. CoGNET has a favorable overall prognosis, and endoscopic resection is a safe and effective treatment modality, while surgical treatment is recommended for patients with potential lymph node metastasis. Further large-scale and long-term follow-up studies are needed to verify our results.

ACKNOWLEDGEMENTS

The authors would like to thank the Ethics Review Committee of the First Affiliated Hospital of Zhejiang University for the approval of this study. We also express our sincere gratitude to the colleagues from the Department of Pathology, Department of Gastroenterology and Endoscopy Center, and Department of Gastrointestinal Surgery of the First Affiliated Hospital of Zhejiang University for their technical support and assistance in this study. Finally, we thank all the patients and their families who participated in this study for their cooperation and support.