To dye or not to dye? Comparing real-world diagnostic performance of white light, electronic, and Lugol’s chromoendoscopy for esophageal squamous neoplasia
摘要
The optimal endoscopic technique for esophageal squamous cell carcinoma (ESCC) screening remains uncertain. This study aimed to evaluate the diagnostic performance of Lugol’s chromoendoscopy (LCE), electronic chromoendoscopy (ECE), and white light imaging (WLI) in identifying esophageal precancerous and early malignant lesions through a large-scale self-controlled study.
MethodsBetween 2019 and 2023, outpatients aged 45–69 years were enrolled at Endoscopy Center of Anyang Cancer Hospital and underwent a sequential endoscopic examination comprising WLI, ECE, and LCE with complete epidemiological, endoscopic, and pathological data. Per-lesion sensitivity, specificity, positive predictive value, and negative predictive value were evaluated for detecting esophageal intraepithelial neoplasia and above (INA), high-grade intraepithelial neoplasia and above (HGINA), and ESCC. Subgroup analyses stratified by epidemiological data were performed.
ResultsAmong the 2167 included participants, histopathology identified 53 cases of low-grade intraepithelial neoplasia (LGIN), 111 of high-grade intraepithelial neoplasia (HGIN), and 63 of ESCC. LCE showed 100% sensitivity across all lesion categories, outperforming ECE (89.3%, 94.9%, 98.5%) and WLI (68.1%, 79.1%, 87.7%) for INA, HGINA, and ESCC, with significantly better sensitivity for INA and HGINA. Specificity of LCE (74.6%-82.2%) was lower than ECE (84.5%–91.8%) and WLI (90.4%–95.8%) (all P < 0.05). Compared with the LCE-for-all strategy, in which LCE was applied to all endoscopic examinations, the two-step strategy, which restricted LCE to lesions positive on either WLI or ECE, missed 21.6% of LGIN, 6.1% of HGIN, and 1.5% of ESCC lesions. Subgroup analysis showed that senior endoscopists (> 5 years of experience) achieved significantly higher ECE sensitivity for INA detection.
ConclusionsAn LCE-for-all strategy may be preferable when the priority is to maximize detection of early precancerous lesions such as LGIN, whereas a two-step strategy may achieve adequate detection of HGINA. With adequate training and strict quality control, ECE could serve as an alternative, while WLI alone is not recommended for EC screening.
Graphical Abstract