Superficial esophageal squamous cell carcinoma (SESCC) is a malignant lesion confined to the mucosa or submucosa regardless of the presence or absence of lymph node metastasis (1, 2). Endoscopic submucosal dissection (ESD) is widely used to treat SESCC (3). Indications for ESD have been expanded to cases in which the risk of lymph node metastasis is assumed to be low. Even if the pathological invasion depth is classified as T1b (tumour invasion of the submucosa), ESD can be performed if the invasion is limited to SM1 (submucosal invasion <200 μm from the muscularis mucosa) (4–6), and it is not suitable for submucosal invasion ≥200 μm (7–9).

Therefore, accurate determination of the depth of invasion (T stage) of esophageal cancer is crucial in determining treatment strategies.

Conventional endoscopic ultrasonography (EUS-C) is a technique that can provide clear cross-sectional images of the gastrointestinal (GI) tract wall. It helps diagnose the depth of submucosal tumours or assess cancer invasion, which is why it has been used for T staging of esophageal cancer (10, 11). Although previous studies have shown the clinical efficacy of EUS in T staging of esophageal cancer (12), the results have been largely variable (13). EUS-C cannot satisfactorily distinguish between T1a and T1b stages of esophageal squamous cell carcinoma (11, 14). Specifically, it is difficult to distinguish between T1a (tumour invading the lamina propria and muscularis mucosae) and T1b lesions by EUS because of the thin border between the mucosa and submucosa and overlap in echogenicity (15). Therefore, it is necessary to develop reliable and less invasive methods to distinguish between the T1a and T1b stages of SESCC.

Submucosal saline injection (SSI) is a necessary step in the ESD treatment of early esophageal cancer (16). SSI can increase the thickness of the submucosal layer to prevent gastrointestinal perforation caused by damage to the muscularis propria during treatment (17). And physiological saline is a good transmission medium for sound waves and can be used as an echo contrast enhancer for ultrasound transmission (18, 19).

Therefore, we evaluated whether SSI can be used to improve the accuracy of EUS in distinguishing between T1a and T1b stages of SESCC and analyse potential factors that interfere with staging results.

Accurate staging is an important basis for selecting treatment options for esophageal cancer (21). Patients with SESCC diagnosed at early stages have a higher survival rate than those diagnosed at advanced stages (22).

With the development of high-definition endoscopy and endoscopic staining technology, the detection rate of early esophageal cancer continues to increase, which is contributing to reducing the mortality rate associated with esophageal cancer (23). Radical esophagectomy and super-minimally invasive resection under digestive endoscopy are the main treatments for early-stage esophageal cancer (24–26). However, the choice of treatment depends on the staging, i.e., T1a or T1b (25, 26). Surgery is usually recommended for T1b cancer cases, whereas ESD is an option for T1a cancer cases (24, 26). Unlike radical esophagectomy, ESD preserves the integrity of organs and physiological structures, and patients have a better quality of life after treatment (27, 28). Therefore, it is crucial to accurately distinguish between T1a and T1b, and EUS is the preferred method for staging.

Although EUS has been used to assess the depth of invasion in early esophageal cancer cases, the reported diagnostic accuracy ranges from 60 to 80% (11, 14). The staging accuracy of EUS-C by two experienced endoscopists was 68.5% in our study, which is consistent with previous reports. However, after SSI, the accuracy of EUS staging for early-stage SESCC increased to 84.8%, suggesting improved staging accuracy for early-stage esophageal cancer.

Some studies have postulated that EUS has no significant advantage over traditional endoscopy and magnification endoscopy in predicting the depth of invasion of early esophageal cancer (29).

The accuracy of EUS is reported to vary greatly depending on the experience of the endoscopist, tumour location, macroscopic type of tumour, tumour size, presence or absence of ulceration, and tumour differentiation type (11).

Lesions with severe inflammation or ulceration or in cases where multiple biopsies have been performed, there is evidence of submucosal fibrosis, and hypoechoic lesions observed on EUS resemble tumour infiltration. For lesions in the upper third of the oesophagus, the accuracy of EUS may be reduced due to differences in esophageal wall thickness, poor esophageal lumen relaxation, and air interference. Furthermore, due to the tubular structure of the oesophagus and frequent peristalsis, it is difficult to fill the oesophagus with degassed water and position the EUS probe near the lesion. In addition, larger tumour size is a risk factor for incorrect judgment of invasion depth (30). Undifferentiated tumour cells may have the potential to undergo micro-infiltration in the submucosa of the esophageal wall, and EUS cannot visualize these micro-invasions, thereby underestimating tumour staging (31).

In this study, 15 patients with stage T1a cancer were over-staged as T1b in EUS-C examination. Although it was recommended that patients choose diagnostic ESD treatment, some patients received traditional surgical treatment, and their postoperative quality of life was significantly reduced. Subsequent EUS-SSI staging significantly reduced over-staging (25.0% with EUS-C to 9.8% with EUS-SSI). However, there were four patients whose preoperative and postoperative pathologies were incompatible, ranging from T1b to T1a. Further analysis showed that the lesion was located in the upper one-third of the oesophagus. Multiple preoperative endoscopic examinations and moderate-to-poor differentiation are risk factors for EUS-SSI over-staging. Multiple biopsies under endoscopy lead to scar-like hyperplasia and adhesion between the mucosa and submucosa. With EUS visualization, the boundaries can be unclear and hypoechoic changes between the mucosa and submucosa may be observed. These findings affect the judgment of the invasion depth.

Limitations of EUS-SSI include a longer examination time than EUS-C and the possibility of greater discomfort to the patient. These drawbacks can be addressed by using sedatives.

All enrolled patients underwent esophagectomy or ESD within 1 week, and all recovered well. There was no difficulty in endoscopy dissection due to poor submucosal injection lift. All lesions were completely removed. SSI may elicit an inflammatory response and cause fibroplasia in the submucosa, but it is unknown whether SSI makes ESD treatment more difficult if the patient cannot receive treatment soon. More cases and longer follow-ups are needed for further confirmation.

In this study, submucosal injection improved the accuracy of EUS staging of early-stage, i.e., T1a and T1b stages of SESCC. Lesions located in the upper oesophagus and multiple repeated biopsies are risk factors for inaccurate staging. This study has limitations, including its single-center, non-randomized design, which may introduce selection bias and confounding factors. The small sample size might also limit statistical power. Future multicenter RCTs with larger samples are needed to further validate the advantages of EUS-SSI, incorporating broader outcome measures for a more robust evaluation.