Efficacy of combined use of magnifying narrow-band imaging/blue laser imaging and Lugol chromoendoscopy for tumor extent delineation in esophageal squamous cell carcinoma.

Multiple Lugol-voiding lesions (LVLs) on Lugol chromoendoscopy can predict the development of metachronous multiple cancers in the esophagus and the head and neck regions. However, Lugol chromoendoscopy sometimes causes adverse events such as chest pain and discomfort. We therefore investigated the endoscopic findings on narrow band imaging (NBI) or blue laser imaging (BLI) that correspond to the presence of multiple LVLs in patients with esophageal squamous cell carcinoma. First, we investigated the NBI/BLI findings corresponding to individual small LVLs (one-to-one correspondence). Second, we investigated the association between the grade of multiple LVLs and the five endoscopic findings, including multiple foci of dilated vessels (MDV), multiple small brownish areas without microvascular irregularity, and a nonuniform color tone. One-to-one correspondence of endoscopic findings was analyzed in 106 small LVLs. The main findings matched with small LVLs were a focus of dilated vessels (44 lesions), a small brownish area (17 lesions), and a small brownish area with a focus of dilated vessels (19 lesions). The relationship between multiple LVLs and each finding assessed by NBI/BLI was assessed in 155 patients. Multivariate logistic regression indicated that the presence of MDV was the only finding independently associated with multiple LVLs (P<0.01). The presence of MDV in the noncancerous background esophageal mucosa was significantly associated with multiple LVLs. This pilot study demonstrates that MDV has the potential to be a new risk factor for the development of metachronous multiple esophageal squamous cell carcinoma.

The relationship between esophageal squamous cell carcinoma (ESCC) and Lugol-voiding lesions (LVLs) in patients with head and neck squamous cell carcinoma (HNSCC) is unclear. To investigate the characteristics of ESCC and the relationship between ESCC and LVLs in patients with HNSCC. Between 2003 and 2006, 157 patients with primary HNSCC underwent Lugol chromoendoscopy at the Hiroshima University Hospital, Hiroshima, Japan. Of the patients, 135 were followed up for more than 6 months. We retrospectively analyzed the incidence of synchronous and metachronous ESCC and cumulative proportions of patients without metachronous ESCC with or without multiple LVLs. Synchronous and metachronous ESCC were detected in 17 of 157 (10.8%) and 9 of 135 (6.7%) patients, respectively. The incidence of synchronous and metachronous ESCC was significantly higher in patients with LVLs compared with the incidence in those without LVLs [13 of 32 (40.6%) vs. 4 of 125 (3.2%), P<0.0001 and 8 of 19 (42.1%) vs. 1 of 116 (0.9%), P<0.0001, respectively]. Cumulative proportions of patients without metachronous ESCC were significantly lower in patients with multiple LVLs compared with that in those without multiple LVLs (P<0.0001). Patients who had HNSCC, especially those with multiple LVLs in the esophagus, should be followed with close endoscopic observation with Lugol chromoendoscopy.

Endoscopic surveillance is mandatory for patients who underwent endoscopic resection (ER) for esophageal squamous cell carcinoma (ESCC) to detect multiple ESCCs. This study aimed to evaluate whether the current surveillance system is appropriate from the viewpoint of the clinical course after ER. This was a retrospective study conducted at a multicenter, high-volume center. Patients who underwent ER without additional therapy for ESCC with T1a pathology were included. The primary endpoints were the 3- and 5-year cumulative incidence of ESCC with invasion depth of the muscularis mucosa or deeper (index ESCC). A total of 378 patients were included in this analysis. The median observation period was 67.3 (range 24.1-104.3) months. In total, 375 patients underwent surveillance endoscopy with narrow band imaging (NBI), followed by Lugol chromoendoscopy (LCE). The numbers of patients with surveillance intervals of 6 months, 12 months, and irregular were 180, 133, and 65, respectively. There were 275 cases of multiple ESCCs, including 11 index ESCCs. The 3- and 5-year cumulative incidence rates of index ESCCs were 1.3% and 2.7%, respectively. Multiple ESCCs were diagnosed with higher Lugol voiding lesion grades and during 6-month rather than 12-month or irregular surveillance (p < 0.001). However, cox proportional hazards analysis showed no significant factors associated with the cumulative incidence of index ESCCs. No ESCC-related death occurred during the study period. Surveillance endoscopy with NBI and LCE at intervals of 6-12 months can be acceptable, given the low cumulative incidence of index ESCCs and favorable prognosis.

Multicentric squamous dysplasia in the esophagus can be visualized by Lugol chromoendoscopy as multiple Lugol-voiding lesions (LVLs). Narrow-band imaging combined with magnifying endoscopy (NBI-ME) facilitates the detection of superficial squamous cell carcinoma within the head and neck region (HNSCC). We investigated risk factors for superficial HNSCC in patients with esophageal squamous cell carcinoma (ESCC). Case-control study. We studied 71 patients with synchronous or former ESCC. All patients underwent screening of the head and neck by NBI-ME and Lugol chromoendoscopy of the esophageal mucosa. The history of tobacco and alcohol use was documented. Genetic polymorphisms of aldehyde dehydrogenase type 2 (ALDH2) were identified by the sequence-specific primer polymerase chain reaction. Clinical factors related to superficial HNSCC were analyzed. All patients with superficial HNSCC were drinkers. On univariate analysis, multiple LVLs (odds ratio [OR], 56.92; 95% confidence interval [CI] 6.93-467.38; P < .001), ALDH2-2 allele (OR, 14.48; 95% CI, 1.8-116.56; P = .01), current smoker (OR, 4.25; 95% CI, 1.44-12.57; P = .009), and smoking index ≥ 1,000 (OR, 3.45; 95% CI, 1.19-9.99; P = .02) were associated with superficial HNSCC. On multivariate analysis, multiple LVLs (OR, 61.12; 95% CI, 5.4-691.64; P = .001), ALDH2-2 allele (OR, 16.19; 95% CI, 1.15-228.06; P = .04), and current smoker (OR, 8.02; 95% CI, 1.09-59.22; P = .04) were associated with superficial HNSCC. Patients with ESCC, particularly drinkers, current smokers, and those with the ALDH2-2 allele and multiple LVLs, have an increased risk of superficial HNSCC.

Multicentric squamous dysplasia of the esophagus is characterized by multiple Lugol-voiding lesions (LVLs) on Lugol chromoendoscopy. Multiple LVLs are associated with a very high risk of multiple cancers arising in the esophagus as well as the head and neck. To gain insight into the pathogenesis of multiple LVLs of the esophageal mucosa, we studied risk factors for the development of such lesions in 76 patients who had a current or previous diagnosis of esophageal squamous cell carcinoma. All patients underwent Lugol chromoendoscopy of the esophageal mucosa. The history of tobacco and alcohol use was documented. Polymorphisms of the aldehyde dehydrogenase type 2 (ALDH2) gene were identified by polymerase chain reaction using sequence-specific primers. Clinical factors related to multiple LVLs were analyzed. All patients with multiple LVLs were drinkers. On univariate analysis, male sex (odds ratio [OR] 15, 95% confidence interval [CI] 1.84-122.45: P = 0.011), presence of the ALDH2-2 allele (OR 4.5, 95% CI 1.55-13.24: P = 0.006), and smoking index ≥1000 (OR 2.6, 95% CI 1.02-6.6: P = 0.045) were associated with multiple LVLs. On multivariate analysis, male sex (OR 10.02, 95% CI 1.13-88.44: P = 0.038) and presence of the ALDH2-2 allele (OR 4.56, 95% CI 1.4-14.82: P = 0.012) were associated with multiple LVLs. Among drinkers, a daily alcohol intake of ≥100 g pure ethanol with the ALDH2-2 allele (OR 17.5, 95% CI 1.97-155.59: P = 0.01) and a daily alcohol intake of <100 g pure ethanol with the ALDH2-2 allele (OR 8.85, 95% CI 1.68-46.69: P = 0.01) more strongly correlated with multiple LVLs than did a daily alcohol intake of <100 g pure ethanol without the ALDH2-2 allele, whereas a daily alcohol intake of ≥100 g pure ethanol without the ALDH2-2 allele (OR 4.0, 95% CI 0.54-29.81: P = 0.18) did not. In conclusion, male sex and the ALDH2-2 allele are associated with an increased risk for multiple LVLs of the esophageal mucosa in patients with esophageal squamous cell carcinoma. Among drinkers with the ALDH2-2 allele, the risk of multiple LVLs increased in parallel to the daily alcohol intake.

Previously, we identified that rs1229984 in ADH1B, rs671 in ALDH2, and smoking status were independently associated with the risk of developing metachronous squamous cell carcinoma (SCC) after endoscopic resection (ER) for esophageal SCC (ESCC). However, this analysis included cases with short-term follow-up. In the present study, we investigated the environmental and genetic factors associated with developing metachronous SCC using long-term follow-up observation after ER for ESCC. One hundred and thirty ESCC patients who underwent treatment with ER were followed up using endoscopy for ≥ 30months. We investigated the incidence of, and genetic/environmental factors associated with, metachronous SCC development after ER for ESCC. We also analyzed the potential risk factors for multiple metachronous SCC development using Cox's proportional hazards model. Moreover, we constructed a risk model for the development of metachronous SCC after ER for ESCC. Male, rs1229984, rs671, alcohol consumption (> 20g/day), smoking, and multiple Lugol-voiding lesions (LVLs) significantly affected the incidence of multiple metachronous SCCs. Multiple Cox proportional analysis revealed that rs1229984, rs671, alcohol consumption, smoking, and multiple LVLs were independently associated with the risk of developing metachronous SCC. Patients who had ≤ 2 risk factors did not develop metachronous SCC, and the risk of developing metachronous SCC in patients with ≥ 3 risk factors was significantly higher than in patients with ≤ 2 risk factors. The risk model using these 5 genetic and environmental factors is useful as an indication for multiple metachronous development in ESCC patients.

Patients with squamous-cell carcinoma in the head and neck (HNSCC) often develop second primary esophageal squamous-cell carcinomas (ESCC). In addition, widespread epithelial oncogenic alterations are also frequently observed in the esophagus and can be made visible as multiple Lugol-voiding lesions (multiple LVL) by Lugol chromoendoscopy. Multiple occurrences of neoplastic change in the upper aerodigestive tract have been explained by the concept of 'field cancerization', usually associated with repeated exposure to carcinogens such as alcohol and cigarette smoke. However, the etiology of second ESCC in HNSCC patients remains unclear and acetaldehyde, the first metabolite of ethanol, has been implicated as the ultimate carcinogen in alcohol-related carcinogenesis. We first investigated the relation between second ESCC and multiple LVL in 78 HNSCC patients. Multiple LVL and second ESCC were observed in 29 (37%) and 21 (27%) patients, respectively. All of the second ESCC were accompanied by multiple LVL. This may indicate that episodes of multiple LVL are precursors for second ESCC. We then examined the association of multiple LVL with the patients' characteristics, including genetic polymorphisms of the alcohol metabolizing enzymes, alcohol dehydrogenase type 3 (ADH3) and aldehyde dehydrogenase type 2 (ALDH2). We also investigated acetaldehyde concentrations in the breath of 52 of the 78 patients. All the patients with multiple LVL were both drinkers and smokers. Multivariable logistic analysis showed that the inactive ALDH2 allele (ALDH2-2) was the strongest contributing factor for the development of multiple LVL (odds ratio 17.6; 95% confidence intervals 4.7-65.3). After alcohol ingestion, acetaldehyde in the breath was elevated to a significantly higher level in all patients with the ALDH2-2 allele than in those without it. The high levels of breath acetaldehyde were significantly modified by the slow-metabolizing ADH3-2 allele. These results reveal strong evidence for a gene-environmental interaction between the ALDH2-2 allele and alcohol consumption, for the risk of developing multiple LVL, resulting in the development of second ESCC in patients with HNSCC. Ultimately, increased local acetaldehyde exposure thus appears to be a critical determinant of the phenomenon of 'field cancerization'.

Inactivated alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH) are related to esophageal carcinogenesis. We aimed to clarify the clinical features associated with the alcohol-degrading enzyme genotypes, ADH1B and ALDH2. We also investigated the risk factors for metachronous esophageal squamous cell carcinoma (ESCC) and head and neck SCC (HNSCC). We conducted a single-center, retrospective study including patients with ESCC treated by endoscopic resection. Patients were recruited between October 2020 and September 2021. Buccal mucosal swabs were obtained from them to analyze the genetic polymorphisms affecting ADH (ADH1B) and ALDH (ALDH2) activity. Patients were categorized into three groups: both inactivated = double-inactivated group; inactivated ADH1B or ALDH2 = single-inactivated group; and both activated = activated group. Among the 297 enrolled patients, patients in the double-inactivated group were significantly younger (P < 0.001) and 60% of them were ≤ 50years old. This group also had more ESCCs located in the upper esophagus (P < 0.001) and more simultaneous multiple ESCCs (P = 0.044). More than half of the patients had multiple Lugol-voiding lesions (LVLs) (P < 0.001) and heavy alcohol consumers (P = 0.012). Metachronous ESCC and HNSCC were more common in the double-inactivated group (P < 0.001, P = 0.001). Multivariate analysis identified located in the upper esophagus, multiple LVLs and history of HNSCC as risk factors for metachronous ESCC. Activation patterns of alcohol-metabolizing enzymes were related to age at ESCC onset, lesion location, and metachronous ESCC and HNSCC. Different approaches to the prophylaxis and treatment of esophageal cancer should be considered, depending on the enzyme activity pattern.

Patients with early esophageal squamous cell carcinoma (ESCC) may develop multiple second primary ESCC and cancers in other organs even after curative endoscopic resection (ER). We investigated whether administration of chemoradiotherapy (CRT) after ER decreases the incidence of second primary cancers. We conducted a post hoc analysis of the prospective study. Among the registered 170 patients with clinical submucosal ESCC, 74 underwent ER alone, and 96 underwent ER followed by CRT (ER + CRT) because of pathological results of submucosal or lympho-vascular invasion. We compared the incidence of second primary cancers in esophagus and in other organs between two treatment groups. A univariate analysis was performed to investigate the related risk factors. All patients were followed up with esophagogastroduodenoscopy and CT every 4months for the first 3years and every 6months thereafter. Sixty-one ESCC were detected in 32 patients, and the 3-year cumulative incidence of multiple ESCCs was not different between ER + CRT and ER alone (10.4% vs. 13.5%). Sixty-three second primary cancers in other organs were detected in 45 patients, and there was no difference in the cumulative incidence between two groups. The risk factors for multiple ESCCs were high alcohol consumption and grade C multiple Lugol-voiding lesions. Heavy drinker or patients with grade C multiple Lugol-voiding lesion rather than CRT were at risk for second primary ESCC. CRT after ER did not decrease the cumulative incidence of second primary ESCC nor cancers in other organs comparing with ER alone.

BACKGROUNDMetachronous multiple esophageal squamous cell carcinomas (ESCCs) may occur in some patients after endoscopic resection. Multiple dysplastic lesions in the esophagus increase risk of multiple squamous cell carcinomas (SCCs). Endoscopic imaging technology such as narrow band imaging (NBI), can detect early SCC. Lugol chromoendoscopy is also the conventional standard technique for detecting superficial ESCC. However, little is known about the interval from the first SCC to the metachronous SCC. Effective methods to prevent multiple metachronous SCCs are needed in survivors of esophageal SCC.CASE SUMMARYA 56-year-old man showed a slightly elevated reddish area in the middle thoracic esophagus at 30 cm from the incisors on gastroscopy for routine examination. Esophageal mucosa lesion was about 2.5 cm. NBI and magnifying gastroscopy confirmed intra-epithelial papillary loop type B-1 according to the Japan Esophageal Society Classification. Lugol chromoendoscopy was used to evaluate the dysplastic squamous epithelium in the esophagus. Biopsy pathology revealed severe dysplastic squamous epithelium. Computed tomography showed no lymph node metastasis. His complete blood test and tumor markers were within reference values. He had no history of alcohol consumption and smoking. Mucosal lesion was dissected by endoscopic submucosal dissection (ESD). Postoperative pathological results showed moderately differentiated squamous carcinoma. No cancer thrombus was seen in the vasculature, and the surrounding cut edge was not involved. The patient underwent radiotherapy within 2 months after ESD. The multiple Lugol-voiding lesions disappeared, and enhanced chest computed tomography revealed no lymph node metastasis.CONCLUSIONThis is the first case of multiple dysplastic lesions of esophagus cured by radiotherapy. Radiotherapy after minimally invasive endoscopic treatment might be a safe and effective optional therapeutic strategy to prevent metachronous multiple esophageal SCCs.

Background: Multiple development of dysplastic squamous epithelium and squamous cell carcinoma (SCC) in the esophagus and head/neck region was associated with alcohol and cigarette use and have been explained by field cancerization phenomenon. Second primary cancer developed after curative treatment have harmful effect on patients’ quality of life and survival, however, until recently there have been no evidence-based assessment of the effects of these consumption and cessation on field cancerization. Methods: We conducted a multicentre cohort study at 16 hospitals in Japan. Patients with superficial esophageal SCC (ESCCs) who treated by endoscopic resection were registered. Dysplastic squamous epithelium in the background esophageal mucosa was evaluated by Lugol chromoendoscopy and was classified into 3 groups based on the number of Lugol-voiding lesions (LVLs) per endoscopic view: grade A, 0; grade B, 1–9; or grade C, ³10 LVLs. Lifestyle surveys were conducted using a self-administered questionnaire. Patients were counselled on the need for ceasing alcohol and cigarette consumption by physicians, and were evaluated endoscopical surveillance every 6 months. This study conforms to the STROBE guidelines Findings: Between September 12, 2005 and April 9, 2010, 331 patients with superficial ESCCs were registered. Pooled data from 1,022 healthy subjects were used for comparison. Grades of LVL were positively associated with alcohol drinking intensity, flushing reactions, smoking, high-temperature food and negatively with eating green/yellow vegetables and fruit. Secondary ESCCs and HNSCCs were significantly more prevalent in the grade C LVL group (cumulative 5-y incidences 47.1%, 95% confidence interval [CI], 38.0 to 57.2, and 13.3%, 95% CI, 8.1 to 21.5, respectively). Ceasing alcohol and cigarette use significantly reduced the development of secondary ESCCs (adjusted hazard ratios 0.47, 95% CI, 0.26 to 0.85 and 0.49, 95% CI, 0.26 to 0.91, respectively) Interpretation: Alcohol and cigarette use were closely associated with field cancerization and cessation of them significantly reduced the risk of development of second primary cancer. Funding Statement: Grant from National Cancer Center Research and Development Fund 36 by the Ministry of Health, Labour and Welfare of Japan. Declaration of Interests: The authors declared no conflict of interest. Ethics Approval Statement: This study was approved by the institutional review board at each hospital, and we obtained written informed consent from all patients. (UMIN Clinical Trials Registry ID:UMI01676).

Aim. Little is known about the usefulness of narrow band imaging (NBI) for surveillance of patients after chemoradiotherapy for esophageal neoplasia. Its usefulness in detecting esophageal squamous cell carcinoma (SCC) or high-grade intraepithelial neoplasia (HGIN) in these patients was retrospectively compared to Lugol chromoendoscopy. Patients and Methods. We assessed the diagnostic ability of NBI with magnification based on the biopsy specimens obtained from iodine-unstained lesions. Seventy-two iodine-unstained lesions were biopsied and consecutively enrolled for this study. The lesions were divided into NBI positive and NBI negative. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of NBI with magnification and PPV of Lugol chromoendoscopy was calculated using histological assessment as a gold standard. Results. Forty-six endoscopic examinations using NBI with magnification followed by Lugol chromoendoscopy were performed to 28 patients. The prevalence of SCC and HGIN was 21.4%. Sensitivity, specificity, PPV, NPV, and accuracy of NBI were 100.0%, 98.5%, 85.7%, 100%, and 98.6%, respectively. On the contrary, PPV of Lugol chromoendoscopy were 8.3%. Compared to Lugol chromoendoscopy, NBI with magnification showed equal sensitivity and significantly higher PPV (P < 0.0001). Conclusion. NBI with magnification would be able to pick up esophageal neoplasia more efficiently than Lugol chromoendoscopy in patients after chemoradiotherapy.

Background and study aims Recent evidence suggests that lugol chromoendoscopy (LCE) and narrow-band imaging (NBI) have comparable sensitivity for detection of superficial esophageal squamous cell carcinoma (SCC). However, LCE is time-consuming and associated with side effects. The aim of this study was to compare the effectiveness of NBI and LCE in defining resection margins of esophageal SCC.Patients and methods This was a retrospective observational cohort study of patients with esophageal SCC and dysplasia who underwent en-bloc resection between 1999 and 2017 at the Cliniques universitaires Saint-Luc, Brussels. Two groups were defined: 1) inspection with NBI only; and 2) inspection with LCE (with or without NBI). The primary endpoint was complete lateral resection rate. Multivariate regression was used to adjust for potential confounders.Results A total of 102 patients with 132 lesions were included. Lesions were inspected with LCE in 52 % (n = 68) and with NBI only in 48 % (n = 64). Lesions 0-IIa were more frequent in the NBI group (37 %) and 0-IIb (60 %) in LCE. Lesion location, size, and histology and resection technique (endoscopic submucosal dissection in 122/132 cases, 92 %) were similar between the groups. The rate of complete lateral resection for invasive carcinoma was 90 % in LCE group and 94 % in NBI group (P = 0.498) and 65 % and 67 % (P = 0.813), respectively, for dysplasia complete lateral resection. These results remained non-significant after adjusting for potential confounders.Conclusions Mucosal inspection and delineation of tumors with lugol chromoendoscopy before endoscopic resection of esophageal squamous cell lesions was not associated with increased complete lateral resection rate when compared to NBI.

Lugol chromoendoscopy is useful for the detection of early esophageal squamous cell cancer (ESCC). Multiple lugol-voiding lesions (LVLs) on lugol chromoendoscopy are associated with a very high risk of multiple cancers arising in the esophagus. Due to the widespread use of narrow band image technology in many institutions, esophageal cancer without LVLs in the background esophagus is sometimes detected. This retrospective study aims to clarify the clinical characteristic of esophageal cancer without LVLs inthe background esophagus. A total of 191 consecutive patients with 204 ESCCs had undergone endoscopic submucosal dissection (ESD) from 2011 and 2014. Amongst these lesions, the number of LVLs in the background esophagus per endoscopic view was counted excluding main lesion, and the grading was divided into no LVLs ESCC (nL-ESCC) group and LVLs ESCC (L-ESCC) group. This study evaluated the clinical characteristics and the cumulative incidence of metachronous ESCC after ESD in both groups. Thirty-six patients with 36 lesions and 155 patients with 168 lesions were separated into the nL-ESCC group and L-ESCC group, respectively. On multivariate analysis, the nL-ESCC group was found to be more common in females, who were non-drinkers, or with erosive esophagitis. During follow-up periods, the cumulative incidence of metachronous ESCC at 3-years was 14.4% and 0.00% in the L-ESCC and nL-ESCC groups, respectively (P<0.01). Our study showed that esophageal cancer without LVLs in the background esophagus was mostly occurred in females, who were non-drinkers, or with erosive esophagitis, which are uncommon features of ESCC.

Objective: To construct the diagnostic model of superficial esophageal squamous cell carcinoma (ESCC) and precancerous lesions in endoscopic images based on the YOLOv5l model by using deep learning method of artificial intelligence to improve the diagnosis of early ESCC and precancerous lesions under endoscopy. Methods: 13, 009 endoscopic esophageal images of white light imaging (WLI), narrow band imaging (NBI) and lugol chromoendoscopy (LCE) were collected from June 2019 to July 2021 from 1, 126 patients at the Cancer Hospital, Chinese Academy of Medical Sciences, including low-grade intraepithelial neoplasia, high-grade intraepithelial neoplasia, ESCC limited to the mucosal layer, benign esophageal lesions and normal esophagus. By computerized random function method, the images were divided into a training set (11, 547 images from 1, 025 patients) and a validation set (1, 462 images from 101 patients). The YOLOv5l model was trained and constructed with the training set, and the model was validated with the validation set, while the validation set was diagnosed by two senior and two junior endoscopists, respectively, to compare the diagnostic results of YOLOv5l model and those of the endoscopists. Results: In the validation set, the accuracy, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of the YOLOv5l model in diagnosing early ESCC and precancerous lesions in the WLI, NBI and LCE modes were 96.9%, 87.9%, 98.3%, 88.8%, 98.1%, and 98.6%, 89.3%, 99.5%, 94.4%, 98.2%, and 93.0%, 77.5%, 98.0%, 92.6%, 93.1%, respectively. The accuracy in the NBI model was higher than that in the WLI model (P<0.05) and lower than that in the LCE model (P<0.05). The diagnostic accuracies of YOLOv5l model in the WLI, NBI and LCE modes for the early ESCC and precancerous lesions were similar to those of the 2 senior endoscopists (96.9%, 98.8%, 94.3%, and 97.5%, 99.6%, 91.9%, respectively; P>0.05), but significantly higher than those of the 2 junior endoscopists (84.7%, 92.9%, 81.6% and 88.3%, 91.9%, 81.2%, respectively; P<0.05). Conclusion: The constructed YOLOv5l model has high accuracy in diagnosing early ESCC and precancerous lesions in endoscopic WLI, NBI and LCE modes, which can assist junior endoscopists to improve diagnosis and reduce missed diagnoses.