Extended endoscopic transnasal orbital decompression for dysthyroid optic neuropathy: efficacy and prognostic factors

Abstract

Background

Dysthyroid optic neuropathy (DON) is a vision-threatening complication of Graves’ ophthalmopathy, and insufficient orbital decompression may result in persistent or irreversible visual impairment. While endoscopic transnasal orbital decompression (ETOD) has become a mainstream surgical approach for orbital decompression, the optimal extent of ETOD remains incompletely defined, particularly regarding reproducible anatomical landmarks and safe surgical boundaries.

Methods

This retrospective study evaluated patients with DON who underwent an extended ETOD procedure at a tertiary medical center. The surgical technique involved decompression of the medial, inferior, and superior orbital walls, with standardized anterior, posterior, superior, and inferior anatomical landmarks. Clinical characteristics, imaging findings, treatment outcomes were reviewed before surgery and during postoperative follow-up. Factors associated with improvement in visual acuity and exophthalmos were further explored.

Results

Most affected eyes showed postoperative improvement in visual acuity, accompanied by reductions in exophthalmos, intraocular pressure, and lagophthalmos. The extended procedure did not result in major surgical complications or new-onset diplopia. Younger age, poorer preoperative visual acuity, and shorter duration of visual impairment were associated with greater visual recovery, while more severe baseline exophthalmos was associated with greater reduction in exophthalmos. By contrast, diabetes appeared to be associated with less exophthalmos improvement.

Conclusion

Extended ETOD appears to be a safe and effective surgical strategy for DON. By defining clear anatomical boundaries, this standardized decompression approach may provide clinicians with effective and reproducible surgical guidance and thereby improving patients’ prognosis and psychosocial well-being.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12886-026-05106-y.

Introduction

Dysthyroid optic neuropathy (DON) is a vision-threatening complication of Graves’ ophthalmopathy (GO), affecting approximately 3–5% of patients with GO [1]. Progressive expansion of orbital fat and hypertrophy of the extraocular muscles can result in compression of the optic nerve at the orbital apex, leading to substantial visual impairment and exophthalmos. Despite its relatively low incidence, DON imposes a considerable burden on patients’ quality of life (QoL) and psychosocial well-being [2, 3].

High dose intravenous glucocorticoid therapy (IVGC) remains the first-line treatment for GO and DON patients according to 2022 Consensus Statement by the American Thyroid Association and the European Thyroid Association [4]. However, a considerable proportion of patients with DON are either ineligible for intravenous glucocorticoid therapy or respond inadequately, thereby requiring urgent orbital decompression [4, 5]. The primary objective of orbital decompression surgery is to reduce or redistribute orbital contents, thereby relieving optic nerve compression, decreasing exophthalmos, restoring visual function, and improving patients’ QoL and psychosocial well-being [6, 7].

Traditional decompression techniques include open approaches, transconjunctival surgery and transblepharal surgery. In the Recent years, with advances in endoscopic techniques, endoscopic transnasal orbital decompression (ETOD), first introduced by Kennedy et al., has emerged as an important alternative [8]. Compared with conventional approaches, ETOD offers effective decompression, reduced surgical morbidity and favorable cosmetic outcomes, and therefore has become a mainstream surgical option for DON [9, 10]. Nevertheless, substantial heterogeneity exists among reported ETOD techniques, particularly with respect to the extent of orbital wall decompression, which has ranged from isolated medial wall decompression to inferomedial wall decompression or combined medial-lateral wall approaches [11–13]. Previous studies have reported improvements in visual acuity ranging from 0.11 to 0.40 measured by pinhole best corrected visual acuity (p-BCVA) in DON patients treated with ETOD [14, 15]. Given that DON represents an ophthalmic emergency in which insufficient decompression may result in irreversible visual loss, maximizing the extent of decompression while maintaining surgical safety is important for improving outcomes and prognosis. However, the optimal extent of decompression and the prognostic factors associated with surgical efficacy remain incompletely understood.

In this study, we describe an extended ETOD technique for DON, involving combined decompression of the medial, inferior, and superior orbital walls, and report its clinical and follow-up outcomes in patients treated at a tertiary medical center in Southern China. In addition, we further investigate the prognostic factors associated with postoperative visual acuity and degree of exophthalmos, with the goal of informing clinical decision-making and optimizing surgical strategies for the management of DON.

Method

Research subjects

This study was approved by the Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University (2025 − 101). 50 DON patients who underwent extended ETOD at the First Affiliated Hospital of Sun Yat-sen University from January 2021 to July 2024 were initially screened. The exclusion criteria included: (1) Previous history of ETOD; (2) Incomplete clinical data or imaging records; (3) Loss to follow-up; and (4) Concomitant malignancy. After exclusion, 67 orbits from 42 patients were included for further analysis. Demographic, hematological, thyroid function, bone metabolism, and imaging data were collected for all eligible patients. Written informed consent was waived due to the deidentified data retrospectively collected.

Patients receiving IVGC for at least 1 week (cumulative dose: 4.5–6.0 g methylprednisolone) without significant clinical improvement were selected for ETOD. Postoperative management included prophylactic antibiotics, glucocorticoids, mouse nerve growth factor, and other supportive treatments as clinically indicated.

Previous studies reported that most DON patients achieve stable clinical outcomes within 6 months after ETOD. Patients enrolled in the study were followed up and underwent p-BCVA assessments using standardized decimal charts at 1 week, 1 month, 3 months, 6 months, and 1 year after surgery. Other outcomes were measured at 1 year after the patients’ conditions had stabilized. Degree of exophthalmos was measured on CT or MRI scans as the distance from the corneal apex to a line connecting the anterior projections of the bilateral zygomaticofrontal bones. Extraocular muscle belly width (MBW), including the horizontal MBW of lateral rectus (H-MBW-LR) and medial rectus (H-MBW-MR), the coronal MBW of inferior rectus (C-MBW-IR), superior rectus and levator palpebrae superioris complex (C-MBW-SR/LPS), lateral rectus (C-MBW-LR), and medial rectus (C-MBW-MR), were measured perpendicular to the muscle long axis at the maximal cross-sectional area.

Surgical procedure of extended ETOD

All surgical procedures were performed by the same experienced surgeon (K.Z., 20 years of experience in ETOD surgery). The paranasal sinuses on the affected side were completely opened. The uncinate process and ethmoidal bulla were resected, while the mucosa was carefully preserved and kept moist with saline for subsequent reconstruction. The sphenoethmoidal partition was fully opened to allow clear visualization of the optic canal. The medial orbital wall mucosa was dissected and elevated, exposing the lamina papyracea, followed by gentle elevation of the periorbita to reveal the orbital fascia. The maxillary sinus ostium was maximally widened (Fig. 1a), and a Draf IIa frontal sinusotomy was performed (Fig. 1b).

Fig. 1.

Intraoperative images demonstrating the surgical steps and operative field of ETOD. (a) Maxillary sinus*; (b) Frontal sinus*; (c) Optic canal; (d) Lamina papyracea; (e) Bone removal; (f) Orbital fascia; (g) Incision of the orbital fascia; (h) Partial intra-orbital fat excision; (i) Repositioned mucosal flap. (MS: maxillary sinus, FS: frontal sinus, OC: optic canal, CP: carotid prominence, LP: lamina papyracea, PMF: pedicled mucosal flap, EBS: ethmoid bulla stump, OF: orbital fascia, ONS: optic nerve sheath, IF: intra-orbital fat, MFG: mucosal free graft). *The two images were captured under a 70-degree endoscope

Bone fragments within the ethmoidal cells were removed, and the sphenoethmoidal partition was further enlarged to achieve complete exposure of the optic canal (Fig. 1c). The medial orbital wall mucosa was dissected to expose the lamina papyracea (Fig. 1d). Using a drill, the bone of the optic canal and medial orbital wall was removed in a posterior-to-anterior direction. Once the bone was sufficiently thinned, a probe was used to gently elevate the periorbita (Fig. 1e), thereby exposing the orbital fascia (Fig. 1f).

Orbital decompression was performed by removing bone within the following anatomical boundaries: (1) posteriorly to the orbital opening of the optic canal at the orbital apex (Fig. 2a); (2) superiorly to the frontoethmoidal suture, which served as the superior boundary landmark (Fig. 2b); (3) inferiorly from the superior margin of the maxillary sinus ostium to the superior boundary of the pterygopalatine fossa (Fig. 2c); and (4) anteriorly to the floor of the ethmoidal infundibulum (Fig. 2d). The medial periorbital fascia were exposed. The periorbital fascia was then incised using a sickle knife at approximately 5 mm from the superior and inferior orbital margins (Fig. 1g), allowing orbital fat to herniate through the incision. Partial intra-orbital fat was removed using endoscopic suction forceps (Fig. 1h). Gentle external pressure on the globe facilitated further fat herniation.

Fig. 2.

Defined decompression boundaries of ETOD in DON patients. (a) Posterior Decompression Extent. (b) Superior Decompression Extent. (c) Inferior Decompression Extent. (d) Anterior Decompression Extent. (e) Extent of decompression surgery in the sagittal anatomical diagram. (f) Coronal CT image of the decompression boundaries. (ONS: optic nerve sheath, FS: frontoethmoidal suture, MS: maxillary sinus, EBS: ethmoid bulla stump, OC: optic canal, LB: lacrimal bone)

After decompression, the medial orbital wall mucosal flap was repositioned, and the preserved mucosa of the uncinate process and ethmoidal bulla was used to cover the exposed medial orbital surface (Fig. 1i). Resorbable high–molecular-weight sealant membrane was applied to secure the mucosal flap. Sodium hyaluronate and mouse nerve growth factor (NGF) were injected into the surgical cavity [16], and gelatin sponges were placed for hemostasis. The detailed surgical procedure can be found in the video file provided in the Supplementary Materials.

Statistical analysis

All the statistical analyses were performed using R software (version 4.5.2; R Foundation for Statistical Computing, Vienna, Austria) and GraphPad Prism software (version 10.0.0; GraphPad Software, Boston, Massachusetts, USA). Categorical variables were presented as number and percentage, and continuous variables presented as median and interquartile range. When comparing pre- and postoperative characteristics, categorical variables were compared using paired chi-square tests, while continuous variables were compared using the Wilcoxon signed-rank test. Univariate and multivariate logistic regression analysis was applied to investigate the prognostic factors. In all analyses, a value of p < 0.05 was considered to be statistically significant.

Results

Baseline characteristics of DON patients

A total of 42 patients were included in the study, with 22 of them being male (52.4%). The median age was 53 years (IQR: 44–58 years). Hypertension (12/42, 28.6%), smoking (9/42, 28.6%) and alcohol exposure (12/42, 28.6%) were relatively common, whereas diabetes was less frequent (12/42, 28.6%). Laboratory examination showed that TRAb remained at relatively elevated levels in the cohort, while thyroid hormones remained within a relatively controlled range, supporting the autoimmune background of the disease.

For the baseline characteristics of affected eyes, the median visual acuity was 0.15 (IQR: 0.03–0.35), and the median degree of exophthalmos was 21.5 mm (IQR: 20.0–23.5 mm), suggesting substantial vision impairment and orbital involvement. The median duration of ocular symptoms was 8 months (IQR: 6–12), while the median duration of vision loss was 6 months (IQR: 3–11 months), indicating a relatively prolonged disease course before surgery. All affected eyes showed characteristic GO-related spindle-shaped hypertrophy of extraocular muscles in images, the muscle belly was thickened while the tendon insertion remained normal. Other detailed information was presented in Tables 1 and 2.

Table 1.

The baseline characteristics of patients. GO-QOL, graves’ ophtalmopathy quality of life questionnaire; TRAb, thyrotrophin receptor antibody; TSH, thyroid stimulating hormone; FT3, free triiodothyronine; FT4, free thyroxine; TGAb, thyroglobulin antibodies; TPOAb, thyroid peroxidase autoantibody; iPTH, intact parathyroid hormone

Characteristics	Value (N=42)
Sex, male, n(%)	22 (52.4%)
Age, median (IQR), years	53 (44, 58)
GO-QOL visual function	27 (21, 35)
GO-QOL appearance	31 (24, 40)
Medical History
Hypertension, n(%)	12 (28.6%)
Diabetes, n(%)	9 (11.4%)
Smoke, n(%)	13 (31.0%)
Alcohol, n(%)	14 (33.3%)
History of Radiotherapy, n(%)	1 (2.4%)
History of Iodine-131 Treatment, n(%)	5 (11.9%)
Laboratory Results
Thyroglobulin, median (IQR), ng/mL	72.92 (20.54, 216.00)
TRAb, median (IQR), IU/L	15.66 (4.62, 28.76)
TSH, median (IQR), uIU/mL	0.44 (0.02, 3,60)
FT3, median (IQR), pmol/L	4.95 (4.36, 5.50)
FT4, median (IQR), pmol/L	10.2 (8.96, 12.415)
TT3, median (IQR), nmol/L	1.53 (1.35, 1.81)
TT4, median (IQR), nmol/L	108.55 (89.29, 131.20)
TGAb, median (IQR), IU/mL	0.12 (0, 0.80)
TPOAb, median (IQR), IU/mL	1.60 (0.36, 19.70)
iPTH, median (IQR), pg/mL	34.70 (27.40, 49.70)
Urine iodine, median (IQR), ug/L	95.00 (56.25, 136.00)
Calcitonin, median (IQR), ng/mL	0.60 (0.50, 1.32)

Table 2.

The baseline characteristics of affected eyes. IOP, intraorcular pressure; CAS, clinical activity score

Characteristics	Value (N = 67)
Eye side, right, n(%)	35 (52.2%)
Visual acuity (decimal notation), median (IQR)	0.15 (0.03,0.35)
Exophthalmos, median (IQR), mm	21.5 (20,23.5)
IOP, median (IQR), mmHg	21 (14, 24)
Duration of ocular symptoms, median (IQR), months	8 (6,12)
Duration of vision impairment, median (IQR), months	6 (3,11)
CAS, median (IQR)	4 (4,5)
Lagophthalmos, n(%)	41 (65.1%)
Blepharoptosis surgery, median (IQR), n(%)	4 (6.0%)

Surgical outcomes

Postoperative follow-up showed that most patients experienced marked improvement of visual acuity after surgery, with visual acuity stabilizing after 3 months postoperatively (Fig. 3). Among the patients, 62 (88.1%) eyes showed improvement, 3 (4.5%) remained stable, and 2 (7.5%) experienced decline in visual acuity. The median visual acuity increased from 0.15 (IQR: 0.03–0.35) to 0.6 (IQR: 0.30–0.80). The median degree of exophthalmos decreased from 21.5 mm (IQR: 20.0–23.5 mm) to 19.1 mm (IQR: 16.6–21.3 mm). 41 eyes (65.1%) demonstrated preoperative lagophthalmos, and the median degree of lagophthalmos decreased from 1.62 mm (IQR: 0–2.15 mm) to 1.0 mm (IQR: 0–1.55 mm). Moreover, significant improvements were also observed in patients’ QoL, including visual function and appearance, demonstrating the effectiveness of this surgical approach (Table 3).

Fig. 3.

Changes of visual acuity after extended ETOD (up to 1 year follow-up). p-BCVA, pinhole best corrected visual acuity

Table 3.

Analysis and comparation of pre- and post-operative outcomes. IOP, intraorcular pressure; GO-QOL, Graves’ Ophtalmopathy Quality of Life Questionnaire; H-MBW-LR, horizontal belly width of rectus lateralis; H-MBW-MR, horizontal belly width of rectus medialis; C-MBW-IR, the coronal belly width of rectus inferior; C-MBW-SR, the coronal belly width of rectus superior; C-MBW-LR, the coronal belly width of rectus lateralis; C-MBW-MR, the coronal belly width of rectus medialis

	Pre-operation	Follow-up of 1 year	p
Visual acuity (decimal notation)	0.15 (0.03,0.35)	0.60 (0.50,0.80)	< 0.001
IOP (IQR), mmHg	21 (14, 24)	17 (11, 21)	0.018
Exophthalmos (IQR), mm	21.50 (20.00, 23.50)	19.10 (16.60, 21.30)	0.037
Lagophthalmos (IQR), mm	1.62(0, 2.15)	1.10 (0, 1.55)	0.013
Diplopia	32 (47.8%)	30 (44.8%)	0.861
Limited ocular motility	58 (86.6%)	58 (86.6%)	1
GO-QOL visual function	27 (21, 35)	63 (55, 75)	< 0.001
GO-QOL appearance	31 (24, 40)	57 (48, 65)	< 0.001
H-MBW-LR (IQR), mm	5.30 (4.10,6.45)	4.70 (4.00,6.40)	0.902
H-MBW-MR (IQR), mm	8.10 (6.40,9.20)	8.00 (6.05,9.30)	0.984
C-MBW-SR (IQR), mm	8.50 (7.05,9.75)	8.50 (7.35,9.70)	0.391
C-MBW-MR (IQR), mm	8.00 (7.20,8.60)	8.40 (6.30,9.10)	0.843
C-MBW-IR (IQR), mm	7.70 (6.45,8.85)	7.60 (6.70,9.80)	0.605
C-MBW-LR (IQR), mm	6.00 (5.05,7.60)	6.30 (5.20,7.90)	0.514

Two patients (4.8%) with preoperative diplopia showed postoperative improvement in diplopia symptoms, whereas another 30 patients (71.4%) had persistent diplopia without notable clinical change. However, quantitative ocular alignment measurements were not systematically available. A large proportion of affected eyes (58/67, 86.6%) exhibited limited ocular motility preoperatively, with no significant improvement after surgical decompression. No clinically documented new-onset diplopia or ocular motility disorder was observed during follow-up. Details of other outcome measures were provided in Table 3. No severe complications were observed, including massive hemorrhage, hypoglobus, or cerebrospinal fluid leakage.

Prognostic factors for postoperative improvements in visual acuity and exophthalmos

Given that visual impairment and exophthalmos were the major factors affecting social well-being in patients with DON, which were also the primary reasons for undergoing ETOD, we further explored the prognostic factors associated with the treatment outcomes of the two manifestations. After univariate analysis, we identified 7 characteristics correlated with postoperative visual acuity improvement, including age, preoperative visual acuity, duration of vision impairment, limited ocular motility, preoperative IOP, thyroglobulin and diabetes. We further evaluated the independent effects of these factors using multivariable linear regression analysis and found that age, preoperative visual acuity and duration of vision impairment were negatively associated with visual acuity improvement (Fig. 4a).

Fig. 4.

Forest plot of factors associated with postoperative reduction of visual acuity (a) and exophthalmos (b). IOP, intraocular pressure; TSH, thyroid stimulating hormone

On the other hand, regarding improvement in exophthalmos, univariable analysis showed that it was associated with age, diabetes, preoperative degree of exophthalmos and neutrophil count. After further selection using multivariable regression analysis, we found that preoperative degree of exophthalmos was positively associated with exophthalmos improvement, while diabetes had a negative effect (Fig. 4b).

Discussion

As a severe complication of GO, DON profoundly impairs visual function and facial appearance, thereby significantly affecting patients’ psychosocial well-being [2, 3]. In DON, enlargement of extraocular muscles and expansion of orbital fat leads to orbital apex crowding, resulting in compressive optic neuropathy, vision loss, and exophthalmos. With disease progression, persistent optic nerve compression may cause progressive or even irreversible nerve injury and visual impairment, underscoring the importance of timely intervention.

Although high-dose IVGC remain the first-line treatment for GO and DON, a proportion of patients respond inadequately and require urgent orbital decompression [4, 5]. A recent systematic review by Xu et al. found that surgical decompression may be superior to intravenous steroids for improving visual acuity, reducing exophthalmos, and lowering DON recurrence [17]. To date, various ETOD approaches have been described, most commonly involving the medial and inferior walls, with or without additional lateral wall decompression via external approaches. However, the optimal extent of decompression and the anatomical landmarks defining the safe limits, particularly for each orbital wall, have not yet been fully standardized. In the present study, we proposed an extended ETOD strategy and provided a detailed description of the key anatomical boundaries: (1) superiorly extend to the frontoethmoidal suture, (2) inferiorly from the upper edge of the maxillary sinus ostium to the superior boundary of the pterygopalatine fossa, (3) posteriorly extend to the opening of the optic canal at the orbital apex, and (4) anteriorly extend to the floor of the ethmoidal infundibulum. These landmarks may help make decompression more reproducible and facilitate a wider yet controlled decompression. Importantly, this extended approach was associated with favorable visual and exophthalmos outcomes without an apparent increase in major complications such as cerebrospinal fluid leakage.

Improvement in visual acuity and reduction in exophthalmos are the primary goals of ETOD in DON patients [6, 7]. In the present study, age, duration of visual impairment, and preoperative visual acuity were identified as factors associated with postoperative visual improvement. Older age and longer duration of visual impairment were associated with less visual recovery, which may reflect reduced reversibility of optic nerve dysfunction following prolonged compression [18]. Moreover, although better preoperative visual acuity has been associated with superior postoperative outcomes in previous studies, this differs from the magnitude of visual improvement [19, 20]. In our cohort, patients with worse baseline vision demonstrated greater postoperative gain, likely reflecting a larger reversible component of optic nerve dysfunction and a ceiling effect in patients with better baseline vision.

Regarding exophthalmos, greater postoperative improvement was associated with more severe baseline exophthalmos, whereas diabetes was associated with less improvement. The association with baseline exophthalmos is expected, as patients with more marked globe protrusion may have greater potential for posterior displacement after decompression [21]. Although previous studies have reported a lower prevalence of exophthalmos in GO patients with diabetes, diabetic patients appear to have more severe extraocular muscle involvement and a higher risk of DON [22, 23]. This suggests that diabetes may be associated with a muscle-predominant, apex-crowding phenotype rather than a fat-predominant proptotic phenotype, which may partly explain the more limited postoperative reduction in exophthalmos observed in our cohort.

Several limitations should be acknowledged. First, this was a retrospective study and is therefore subject to selection bias and residual confounding. For example, since all included patients received preoperative IVGC and postoperative NGF, these treatment factors were not variable across the cohort, and their independent effects on postoperative outcomes could not be evaluated as predictors in the regression analysis. Therefore, their potential contribution to visual recovery cannot be excluded, representing an unavoidable source of residual confounding in the present study. Second, the sample size was relatively limited, which may affect the stability of the regression analysis. Third, we did not include a direct comparison with conventional decompression techniques. Therefore, we cannot determine whether this extended ETOD approach provides additional benefits in visual recovery or exophthalmos reduction compared with established decompression techniques. In addition, NGF was administered postoperatively as an adjunctive neurotrophic treatment in our institutional protocol. Although NGF has been investigated as a potential neuroprotective agent in optic nerve injury-related conditions, its efficacy in DON remains unproven. Finally, the lack of external validation may limit the generalizability of our findings. Further prospective studies with larger cohorts are needed to validate these findings and to refine the optimal decompression strategy for DON.

Conclusion

In the present study, we described an extended ETOD technique for the treatment of DON and detailed its surgical procedure and anatomical boundaries, aiming to provide significant improvements in visual acuity, exophthalmos, and quality of life. Younger age, worse preoperative visual acuity, and shorter duration of vision impairment were associated with greater visual improvement, whereas severe baseline exophthalmos predicted a larger postoperative reduction in exophthalmos. Taken together, this study provides a refined and reproducible extended ETOD strategy with potential clinical value for improving surgical management and outcome prediction in DON.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

DON

Dysthyroid optic neuropathy

GO

Graves’ ophthalmopathy

ExM

Exophthalmos measurements

ETOD

Endoscopic transnasal orbital decompression

IVGC

Intravenous glucocorticoid therapy

p-BCVA

pinhole best corrected visual acuity

MS

Maxillary sinus

FS

Frontal sinus

OC

Optic canal

CP

Carotid prominence

LP

Lamina papyracea

PMF

Pedicle mucosal flap

EBS

Ethmoid bulla stump

OF

Orbital fascia

ONS

Optic nerve sheath

MFG

Mucosal free graft

IOF

Inferior orbital fissure

CAS

Clinical activity score, CT, computed tomography

MRI

Magnetic resonance imaging

MBW

Muscle belly width

H-MBW-LR

Horizontal belly width of rectus lateralis

H-MBW-MR

Horizontal belly width of rectus medialis

C-MBW-IR

The coronal belly width of rectus inferior

C-MBW-SR/LPS

The coronal belly width of rectus superior

C-MBW-LR

The coronal belly width of rectus lateralis

C-MBW-MR

The coronal belly width of rectus medialis IOP, intraocular pressure

Author contributions

S.W. and H.L. contributed equally to study design, data acquisition and analysis, interpretation of results, and manuscript drafting. G.X., Y.F., Y.L., J.S., and W.G. contributed to data collection, clinical assessment and manuscript revision. K.Z. conceived and supervised the study, interpreted the data, and revised the manuscript. All authors approved the final manuscript and agree to be accountable for their contributions.

Funding

Not applicable.

Data availability

The datasets analysed during the current study are not publicly available due to confidentiality reasons but are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki and was approved by Ethics Committee for Research and Publication of Sun Yat-sen University (2025 − 101). Written informed consent was waived due to the use of deidentified patient data that was retrospectively collected.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.