Clinical value of 3D-FIESTA-C sequence in evaluating the arcuate ligament of the elbow joint: a retrospective study

Abstract

Background

Cubital tunnel syndrome (CuTS) is a commonly seen form of compressive neuropathy. This study aimed to assess the clinical value of 3D-FIESTA-C for evaluating the arcuate ligament in patients with CuTS.

Methods

Between October 2018 and May 2019, 22 inpatients diagnosed with CuTS were enrolled (15 males, 7 females; mean age 60.91 ± 7.38 years). All patients underwent preoperative 3D-FIESTA-C scanning of the elbow joint and subsequently received cubital tunnel release. The accuracy, sensitivity, specificity, positive predictive value, negative predictive value, and kappa value of the 3D-FIESTA-C sequence in displaying the arcuate ligament in its entirety and ulnar nerve were calculated using the intraoperative results as the gold standard, and the consistency test between this sequence and intraoperative arcuate ligament was compared. Paired t-test was used to compare the differences of arcuate ligament thickness measured by the 3D-FIESTA-C sequence and intraoperative exploration. P < 0.05 was considered statistically significant.

Results

The 3D-FIESTA-C revealed a comparatively low display rate on the site of olecranon and olecranon attachment when evaluating arcuate ligament (81.8% and 68.2%, respectively). Intraoperative exploration showed the absence of arcuate ligament and ulnar nerve dislocation in 3 patients, while the 3D-FIESTA-C revealed the absence of the arcuate ligament in 2 patients. Using ulnar neurolysis of the cubital tunnel as the gold standard, the accuracy, sensitivity, specificity, positive predictive value, negative predictive value, and Kappa of the 3D-FIESTA-C for displaying the arcuate ligament in its entirety were 81.8%, 79.0%, 100%, 100%, 42.9%, and 0.51, respectively. When using the 3D-FIESTA-C to display the arcuate ligament on the site of the ulnar nerve, the above values were 95.5%, 100%, 66.7%, 95.0%, 100%, and 0.78, respectively. There was no significant difference in arcuate ligament thickness measured by the 3D-FIESTA-C and intraoperative exploration (0.58 ± 0.15 mm Vs. 0.68 ± 0.34 mm, P = 0.37).

Conclusions

The 3D-FIESTA-C can accurately display the morphological features and positions of the arcuate ligament of the elbow joint and its relationship with the ulnar nerve. And it can also be used to conduct quantitative measurement of the thickness of the arcuate ligament.

Clinical trial number

Not applicable.

Introduction

Characterized by disturbance of ulnar nerve sensory and motor function, cubital tunnel syndrome (CuTS) is the second most common compression neuropathy of the upper limb [1]. CuTS can manifest as numbness, tingling, or pain in the ring/small fingers and dorsoulnar hand [2]. Ulnar nerve compression is the leading cause of the above symptoms. Caused by repetitive elbow pressure, chronic fatigue, or elbow trauma, hypertrophy of the arcuate ligament is an important cause of ulnar nerve compression [3]. Topography of ulnar nerve and sites of ulnar nerve compression were shown in Fig. 1 [4]. When left untreated, CuTS can lead to atrophy of the first dorsal interosseus muscle and affect fine motor function, decreasing quality of life. Conservative treatment is recommended in mild cases, and surgical treatments which consist of simple decompression, decompression with ulnar nerve anterior transposition (subcutaneous, submuscular, or intramuscular), or medial epicondylectomy may be pursued in severe CuTS. Thus, it is very important to identify the arcuate ligament and ulnar nerve during surgical treatments. The diagnosis of CuTS is usually based on clinical data combined with electrodiagnostic testing [1]. But the diagnostic sensitivity of electrodiagnostic testing in CuTS was highly variable due to the biological (i.e., low elbow temperature) and technical factors [5]. Previous study reported that the sensitivity of electrodiagnostic testing in CuTS could be increased from 33.3% to 87.5% after the application of American Association of Neuromuscular and Electrodiagnostic Medicine criteria [6]. However, this high variability of sensitivity is the reason why electrodiagnostic testing is not considered the gold standard for diagnosing CuTS, especially in early disease [7]. In short, the diagnostic value of electrodiagnostic testing was not good. The result of electrodiagnostic testing can only reflect the general function of the ulnar nerve; it is hard to locate the target lesion. Thus, there is a lack of direct morphological evidence for the etiological diagnosis of CuTS based on electrodiagnostic testing. Magnetic resonance imaging (MRI) has the advantages of high resolution of soft tissues, multi-directional, multi-parameter imaging, etc [8]. It could be used to capture morphological data about the nerve and the surrounding soft tissues [9, 10]. This technique enables detailed evaluation of peripheral nerve course and morphology, as well as precise delineation of surrounding soft‑tissue and osseous structures that may contribute to nerve compression [11]. Recently, MRI gained a relevant role in the diagnostic work-up of peripheral neuropathies [12]. Due to the orientation limitations and substantial slice thickness of two-dimensional sequences, it is challenging to visualize the complete morphology of a nerve within a single plane or to clearly depict its fine branches [13]. The increasing availability of MRI systems, along with the development of novel sequences, enables clear visualization of smaller and more delicate anatomical structures, including a variety of anatomical variations [14]. The three-dimensional double excitation balanced steady-state free precession sequence (3D-FIESTA-C) obtains high signals from tissues with large T2/T1 ratios, such as fluid, blood, and fat, and has been used to identify the anterior cruciate ligament of the knee joint. Chen et al. used 3D-FIESTA-C to visualize of the tibial nerve and its branches in the ankle canal [13]. Thus, the sequence may be helpful for the diagnosis of CuTS. The 3D-FIESTA-C sequence is expected to enable three-dimensional reconstruction of the relationship between the arcuate ligament and the ulnar nerve. This allows for visual analysis of the nerve’s surrounding anatomy, morphological changes (e.g., cross-sectional compression), and the course of the ligament through curved planar reconstruction. By displaying the structural path clearly, the technique could assist in preoperative decision-making regarding whether ligament release is required. This study applied the 3D-FIESTA-C in patients with CuTS to explore its clinical value on displaying the arcuate ligament of elbow joints.

Fig. 1.

Topography of ulnar nerve and sites of ulnar nerve compression. FCU, Flexor carpi ulnaris

Methods

Patients

This retrospective study initially enrolled 30 patients diagnosed with CuTS between October 2018 and May 2019. The clinical diagnosis was based on clinical symptoms, physical examination, and electrodiagnosis. The include criteria were as follows: (1) clinical symptoms: numbness and tingling sensations, particularly in the ring and small fingers; paresthesia; hand atrophy and weakness; clumsiness of the hand; (2) physical examination: a positive Tinel’s sign; a positive Froment’s sign; a positive sign of the elbow flexion test; (3) electrodiagnosis determined using the guidelines established by the American Association of Neuromuscular and Electrodiagnostic Medicine: conduction velocity across the elbow is less than 50 m/s, and conduction velocity from above the elbow to below the elbow is slower than that from below the elbow to the wrist by 10 m/s or more; (4) surgical candidates: conduction velocity across the elbow is less than 40 m/s. Eight patients were exclude per the exclusion criteria below: (1) Patients were unable to be diagnosed by MRI due to motion artifacts or non-cooperation (6 patients); (2) Patients with claustrophobia (1 patient); (3) Metal implants in the affected elbow and other contraindications for MR examination (1 patient). Thus, a final cohort of 22 patients was included in the study. There were 15 males and 7 females with an average age of 60.91 ± 7.38 years (range from 49 to 75 years). All patients were unilateral CuTS (left elbow: 9, right elbow: 13). Eighteen patients (81.8%) were with McGowan grade I and 4 patients (18.2%) were with McGowan grade II. All 22 patients subsequently underwent unilateral ulnar nerve release of the cubital tunnel. The current study was approved by the Ethics Committee of Tianjin Hospital, Tianjin University (No. 2023167). As the current study was retrospective in design, the ethics committee exempted the acquisition of informed consent.

Materials and equipment

A GE 3.0T MR (discovery MR 750; GE; UAS) with an eight-channel phased array knee joint coil was used to scan elbow joints. MRI scan sequence parameters of elbow joints were shown in Table 1. All patients were placed in the prone position with the head in an advanced position and maintained a posture of reaching over the head. The chest and lower jaw were supported with soft pads to maintain optimal comfort, the elbow was placed centrally in the knee coil, and soft pads were filled around the elbow to fix the elbow. The scanning range included at least 4 cm above and below the elbow joint space.

Table 1.

MRI scan sequence parameters of displaying the arcuate ligament of elbow joints

	TR (ms)	TE (ms)	Slice thickness (mm)	Gap (mm)	Slice	FOV (cm)	NEX	FA (°)	Matrix (freq. × phase)	Scan Time (s)
The 3D-FIESTA-C sequence	5.7	2.7	0.6	0	152	180 × 180	2	70	256 × 256	344

TR, Time of repetition; TE, Time of echo; Thi, Slice thickness; FOV, Field of view; NEX, Number of excitations; FA, Flip angle

Post-processing image evaluation and measurement

At the end of the scanning, the collected data of the thin-section original images of the 3D-FIESTA-C sequence were transmitted to the workstation ADW4.6 to establish a three-dimensional image mode for multiplanar reconstruction (MPR) and curved surface reconstruction (CPR) image post-processing, and coronal, sagittal, and transaxial images were first reconstructed. To visualize the arcuate ligament in its entirety, the coronal and sagittal views of the elbow joint were first observed on triplane localization, and the cross-section of the ulnar nerve was found on the reconstructed transverse images 1 to 2 cm above the elbow joint space. After determining the position of the ulnar nerve, according to the anatomical position of the arcuate ligament, move the upper and lower levels of the transverse image to find two bony landmarks of the humerus internal epicondyle and olecranon, determine the arcuate ligament from the positions of the ulnar nerve, the humerus internal epicondyle, and the olecranon, find the thinnest adjacent part of the arcuate ligament to the ulnar nerve, and the attachment point position of the arcuate ligament to the humerus internal epicondyle and olecranon, so that the whole picture of the course of the arcuate ligament can be delineated. If some ligaments are poorly visualized, select the oblique view for transaxial display, and CPR rotation shows the entire picture of the course of the arcuate ligament. Two experienced imaging physicians reconstructed the images according to the post-processing method, using a double-blind approach. They were unaware of the patients’ symptom and diagnosis. The presence of the arcuate ligament and the course of the whole picture were observed and recorded, and if there were differences, the agreement was reached after consultation. In this study, the arcuate ligament was divided into five parts according to its anatomical sites, the part on the site of attachment to the humerus internal epicondyle, site of the humerus internal epicondyle, site of the ulnar nerve, site of olecranon, and site of attachment to the olecranon. The presence of each portion of the arcuate ligament was determined and compared with that observed intraoperatively. The thickness of the arcuate ligament was measured on the site of the humerus internal epicondyle, ulnar nerve, and olecranon of the humerus. Three measurements were performed in each part to average the adjacent relationship between the arcuate ligament and the ulnar nerve. Finally, the thickness of the arcuate ligament adjacent to the ulnar nerve was compared with that during surgery as the arcuate ligament at the ulnar nerve site was closest to the ulnar nerve, which may cause ulnar nerve compression at the elbow. During surgery, the resected arcuate ligament was measured using a sterile microvascular ruler with the precision of 0.05 mm graduations. The measurements were conducted by two surgeons in a double-blind manner, and they were unaware of the results of the MRI measurements. To minimize the potential influence of peri-ligamentous adipose tissue on measurement accuracy, only the mid-portion of the ligament adjacent to the ulnar nerve was assessed. Repeated measurements were performed at this site, and the average value was calculated to enhance data reliability.

Statistical analysis

All data were statistically analyzed using SPSS 22.0 software (SPSS Inc., Chicago, IL). Categorical data were expressed as n (%). The accuracy, sensitivity, specificity, positive predictive value, negative predictive value, and kappa value of the 3D-FIESTA-C sequence in displaying the arcuate ligament in its entirety and ulnar nerve were calculated using the intraoperative results as the gold standard, and the consistency test between this sequence and intraoperative arcuate ligament was compared. Kappa values less than 0.00 was considered as no agreement, 0.00 to 0.20 were mildly credible, 0.21 to 0.40 were generally credible, 0.41 to 0.60 were moderately credible, 0.61 to 0.80 were highly credible, and greater than 0.80 were completely credible. Continuous variables were expressed as mean ± standard deviation. Paired t-test was used to compare the differences of arcuate ligament thickness measured by the 3D-FIESTA-C sequence and intraoperative exploration. P < 0.05 was considered statistically significant.

Results

In the 3D-FIESTA-C sequence, the ulnar nerve showed isointensity, and the arcuate ligament showed linear hypointensity. The course of the arcuate ligament can be delineated by MPR and CPR, and its adjacent relationship with the ulnar nerve can be observed. In two patients, a fat space was observed between the arcuate ligament and the ulnar nerve, with gap distances measuring 0.5 mm and 0.7 mm. In other patients, the arcuate ligament was in direct contact with the ulnar nerve (Fig. 2). The 3D-FIESTA-C sequence showed the absence of arcuate ligament in 2 patients, while intraoperative exploration showed the absence of arcuate ligament and ulnar nerve dislocation in 3 patients (Fig. 3). The 3D-FIESTA-C sequence revealed a low display rate on the site of the olecranon and attachment of the olecranon when evaluating the arcuate ligament (Table 2). Using ulnar neurolysis of the cubital tunnel as the gold standard, the accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of the 3D-FIESTA-C sequence for displaying the arcuate ligament in its entirety were 81.8% (95% CI: 59.2% − 94.8%), 79.0% (95% CI: 54.1% − 94.0%), 100% (95% CI: 29.2% − 100%), 100% (95% CI: 78.2% − 100%), and 42.9% (95% CI: 11.8% − 78.7%), respectively. This sequence had moderate agreement with the intraoperative display of the arcuate ligament (kappa value = 0.51). Detailed data was shown in Table 3. When using the 3D-FIESTA-C sequence to display the arcuate ligament at the ulnar nerve site, the accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were 95.5% (95% CI: 77.2% − 99.9%), 100% (95% CI: 85.4% − 100%), 66.7% (95% CI: 12.2% − 98.8%), 95.0% (95% CI: 75.4% − 99.9%), and 100% (95% CI: 22.4% − 100%), respectively. With a kappa value of 0.78, this sequence showed high agreement with intraoperative exploration when displayed at the ulnar nerve site. Detailed data was illustrated in Fig. 4; Table 4. There was no significant difference in arcuate ligament thickness, measured by the 3D-FIESTA-C sequence and intraoperative exploration (0.58 ± 0.15 mm Vs. 0.68 ± 0.34 mm, P = 0.37). Detailed data was shown in Table 5.

Fig. 2.

The arcuate ligament in its entirety and its adjacent relationship with the ulnar nerve. H, Humerus; U: Ulna. A: The arcuate ligament in its entirety and its adjacent relationship with the ulnar nerve by intraoperative exploration. B: The arcuate ligament in its entirety by the 3D-FIESTA-C sequence, the adipose tissue was observed between the arcuate ligament and the ulnar nerve

Fig. 3.

The absence of the arcuate ligament and ulnar nerve dislocation. H, Humerus; U: Ulna. A. The absence of arcuate ligament and ulnar nerve dislocation by intraoperative exploration. B. The absence of the arcuate ligament by the 3D-FIESTA-C sequence. The absence of the arcuate ligament allows for ulnar nerve subluxation, and the peripheral highsignal rim represents surrounding adipose tissue

Table 2.

The display rate of the arcuate ligament on different anatomical sites by the 3D-FIESTA-C sequence

	Site of attachment to the humerus internal epicondyle	Site of humerus internal epicondyle	Site of the ulnar nerve	Site of olecranon	Site of attachment to olecranon
The 3D-FIESTA-C sequence, n (%)	20 (90.91)	20 (90.91)	20 (90.91)	18 (81.82)	15 (68.18)

Table 3.

The consistency evaluation of 3D-FIESTA-C sequence and intraoperative exploration displaying the arcuate ligament in its entirety

The 3D-FIESTA-C sequence	Intraoperative Arcuate Ligament Status		Total
	+	-
+	15	0	15
-	4	3	7
Total	19	3	22

AL, Arcuate ligament

“+” = AL present; “-” = AL absent

Fig. 4.

The arcuate ligament in its entirety and its adjacent relationship with the ulnar nerve. H, Humerus; U: Ulna. A. The surgical knife was cutting the arcuate ligament, releasing the ligament, and exposing the ulnar nerve completely. B. After ulnar neurolysis, the indentation of the arcuate ligament over the ulnar nerve was found, causing ulnar nerve compression. C. The 3D-FIESTA-C sequence showed there was no adipose tissue between the arcuate ligament and ulnar nerve. The circle represented the cross-section of the ulnar nerve. D. The 3D-FIESTA-C sequence showed thinner linear hypointensity of the arcuate ligament and extrusion deformation of the ulnar nerve cross section

Table 4.

The consistency evaluation of 3D-FIESTA-C sequence and intraoperative exploration displaying the arcuate ligament on the site of the ulnar nerve

The 3D-FIESTA-C sequence	Intraoperative Arcuate Ligament Status		Total
	+	-
+	19	1	20
-	0	2	2
Total	19	3	22

AL, Arcuate ligament

“+” = AL present; “-” = AL absent

Table 5.

The thickness of the arcuate ligament on different anatomical sites measured by the 3D-FIESTA-C sequence and intraoperative exploration

	The 3D-FIESTA-C sequence	Ulnar neurolysis of the cubital tunnel	P-value
Site of the Site of humerus internal epicondyle (mm)	0.53 ± 0.17	-	-
Site of ulnar nerve (mm)	0.58 ± 0.15	0.68 ± 0.34	0.37
Site of olecranon (mm)	0.73 ± 0.18	-	-

Discussion

Cubital tunnel syndrome is a common compressive neuropathy in the general population. Hypertrophy of the arcuate ligament is an important cause of ulnar nerve compression [3]. We found 3D-FIESTA-C could accurately display the changes of the arcuate ligament of the elbow joint and its adjacent to the ulnar nerve and could be used to measure the thickness of the arcuate ligament.

With the clinical application of new sequences and techniques for MRI scanning, subtle anatomical structures and high-resolution imaging findings have been gradually developed. As balance steady-state free precession, the 3D-FIESTA-C, also known as SD-CISS, is a brand-new sequence introduced by GE in recent years. Encoded in the 3D space, the 3D-FIESTA-C sequence is characterized by high resolution. A thin layer can ensure high-quality reformation images on any plane. Thus, some delicate anatomic structures could be displayed clearly. This sequence is mainly used to assess the central nervous system at present [15]. Some studies have applied the 3D-FIESTA-C sequence to evaluate the knee joints [16, 17]. In contrast, few studies used the 3D-FIESTA-C sequence to display elbow joints. We applied the 3D-FIESTA-C sequence in patients with CuTS to explore its clinical value. When setting the parameters in this study, TR/TE/FA was 5.7 ms/2.7 ms/70 °, respectively. The TE of 2.7 ms was chosen to limit the signal loss due to susceptibility effects and minimize black-edge streak artifacts between the fat and water interface. With the FA of 70 °, T1 weighting was reduced, the ulnar nerve showed isointensity, and the arcuate ligament showed hypointensity. An isotropic voxel size of 0.6 × 0.6 × 0.6 mm images was obtained. At the same time, the slice thickness was ultrathin. Thus, the thickness of the arcuate ligament of the elbow joint could also be better measured.

Calfee et al. found the possible rate of occurrence of hypermobility of the ulnar nerve at the cubital tunnel in the general population was 37% [18]. The ulnar nerve dislocation was reported in about 20% of healthy individuals and 28% of carcasses [19]. The arcuate ligament restricts ulnar nerve dislocation, and its absence is associated with ulnar nerve dislocation. In the current study, we found the 3D-FIESTA-C sequence showed the display rate of arcuate ligament in the medial humeral condyle attachment, medial humeral condyle, ulnar nerve, olecranon, and olecranon attachment was 90.9%, 90.9%, 90.9%, 81.8%, and 68.1%, respectively. During ulnar neurolysis of the cubital tunnel, we found the absence of arcuate ligament and ulnar nerve dislocation occurred in 3 patients. At the same time, two of the three patients were found to have the absence of arcuate ligament by the 3D-FIESTA-C sequence. As for the other false-positive patient, we mistakenly regarded the transverse supracondylar ligament as the arcuate ligament. The transverse supracondylar ligament is a transverse variant ligament located between the medial epicondyle of the humerus and the olecranon, which can also prevent ulnar nerve dislocation and compress the ulnar nerve when this ligament is hypertrophic and hyperplastic. Therefore, attention should be paid to this variant transverse supracondylar ligament when searching the arcuate ligament on 3D-FIESTA-C sequences and rotating different angles on CPR images to observe the arcuate ligament. Because the arcuate ligament is more hypertrophic and hyperplastic in the olecranon and poorly demarcated from the surrounding fascial tissue, there was a lower display rate of the arcuate ligament on the site of olecranon and attachment of olecranon.

Using the ulnar neurolysis of the cubital tunnel as the gold standard, we found the 3D-FIESTA-C sequence had a higher positive predictive value and a lower negative predictive value when displaying the arcuate ligament in its entirety. The arcuate ligament was curved in shape. It was divided into five parts according to its anatomical sites. With the curved surface reconstruction, at least one part of the five parts will be adjacent to the ulnar nerve. Therefore, the possibility of predicting the absence of the arcuate ligament was lower. Although the arcuate ligament and ulnar nerve dislocation are absent, the incidence is relatively small, which coincides with the anatomy of the arcuate ligament. And the sensitivity was 79.0%, showing a lower false positive rate. The 3D-FIESTA-C sequences showed moderate agreement with intraoperative exploration in displaying the arcuate ligament in its entirety (Kappa value = 0.51). Regarding the 3D-FIESTA-C sequence displaying the arcuate ligament at the ulnar nerve site, the accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were 95.5%, 100%, 66.7%, 95.0%, and 100%, respectively. This sequence showed high agreement with intraoperative exploration when displayed at the ulnar nerve site (Kappa value = 0.78). It can accurately evaluate the display of arcuate ligament at the ulnar nerve site and its adjacent relationship with the ulnar nerve. At present, we have not found any studies using MRI to display the arcuate ligament. Alina et al. applied 3-Tesla MRI to detect the compression points in ulnar neuropathy at the elbow, and they found MRI indicated only fair-to-moderate agreement with intraoperative findings [20]. But MRI may provide valuable information in ambiguous patients or relapses.

Under chronic elbow strain, the fibrous bands of the arcuate ligament become thinned by stretching, and the ulnar nerve is damaged by repeated squeezing and friction, followed by hypertrophy and hyperplasia. And taking into account the influence of peri ligamentous adipose tissue, it was likely to yield greater thickness values during surgery than in vivo measurements. To minimize the potential influence of peri ligamentous adipose tissue on measurement accuracy, the mid portion of the ligament adjacent to the ulnar nerve was assessed. At the same time, repeated measurements were performed at this site, and the average value was calculated to enhance data reliability. Used the 3D-FIESTA-C sequences to measure the thickness of the arcuate ligament adjacent to the ulnar nerve, no significant difference was found compared to intraoperative exploration (0.58 ± 0.15 mm Vs. 0.68 ± 0.34 mm). It validated the quantitative measurement capability of the 3D-FIESTA-C sequence in arcuate ligament thickness. The arcuate ligament is composed of collagen fibers, and hydrogen atoms are fixed on peptides, thus forming a dense mesh and are not involved in magnetic resonance imaging, so MRI shows low signal intensity. Moreover, the arcuate ligament is too thin, and the visual accuracy of measurements can be erroneous for low signal.

The limitation of this study was that the scan time of the 3D-FIESTA-C sequence was slightly longer, which may cause patients’ non-cooperation. Moreover, the current study was a single-center, retrospective analysis, and included only surgical patients, so selection bias could not be excluded. There may also be false positives. The transverse supracondylar ligament may be mistakenly identified as the arcuate ligament. Therefore, attention should be paid to this variant transverse supracondylar ligament when searching the arcuate ligament on 3D-FIESTA-C sequences and rotating different angles on CPR images to observe the arcuate ligament. Lastly, due to the small sample size, lack of comparison with alternative MRI sequences, and absence of postoperative outcome correlation, the results need to be validated through a larger prospective study.

Conclusion

In conclusion, the 3D-FIESTA-C sequence has the advantages of three-dimensional, high-resolution, and noninvasive imaging, which can clearly show the arcuate ligament in its entirety and the adjacent relationship with the ulnar nerve and can quantitatively measure the thickness of the arcuate ligament. Thus, the 3D-FIESTA-C sequence shows important clinical value for the diagnosis of CuTS and provides a good reference value for the treatment of CuTS.

Author contributions

LG contributed to the study conception and design. Material preparation, data collection and literature search were performed by YC, ZH, and KZ. Statistical analysis and data interpretation were performed by YC and ZH. The first draft of the manuscript was written by YC and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

The current study was funded by the Science and technology project of Tianjin Health Commission (TJWJ2023QN054).

Data availability

The data materials were obtained from reasonable request to the corresponding author. All data and materials were unpublished.

Declarations

Ethics approval and consent to participate

The current study was approved by the Ethics Committee of Tianjin Hospital, Tianjin University (No. 2023167), and conducted in accordance with the principles of the Declaration of Helsinki. As the current study was retrospective in design, the ethics committee exempted the acquisition of informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.