Abstract 180P
Background
Esophageal squamous cell carcinoma (ESCC) and gastric cardia adenocarcinoma (GCA) remain major health burdens and co-occur geographically worldwide. However, early detection depends on upper gastrointestinal endoscopy, which is not feasible to implement at a population level. We aim to evaluate the feasibility and accuracy of artificial intelligence (AI)-assisted sponge cytology tests using a novel cell collection device for ESCC and GCA screening in Chinese high-risk regions.
Methods
Participants aged 50 years or older were recruited in five high-risk regions of ESCC and GCA. Cells from esophagus and esophagogastric junction were collected using a novel and minimally invasive capsule sponge, and cytology slides were scanned by a trained AI system. The qualitative (the location of abnormal cells) and quantitative outcomes (counts of total scanned cells and potentially abnormal cells) were reported. A cytological diagnosis was made by consensus. Participants scored acceptability following the procedure on a scale of 0 (least) to 10 (most acceptable). Endoscopy was performed with biopsy as needed. A composite of ESCC, GCA, high- or low-grade lesions was the primary outcome. Diagnostic accuracy, adverse events, and acceptability of cytology testing were assessed.
Results
A total of 1636 participants were enrolled and completed study procedure. No serious adverse events were documented during the cell collection process, and acceptability scores were 10 (72.9%), 9 (19.4%), 8 (3.7%), 7 (1.4%) and 6 (0.8%). 119 (7.3%) participants were diagnosed with lesions confirmed by endoscopic biopsy. The AI-assisted cytologist-confirmed cytology shown significant diagnostic value with the AUC of 0.897 (95% CI, 0.868–0.925), a sensitivity of 93.3% and specificity of 86.0%. For AI-based abnormal cell count, the AUC was 0.884 (95% CI, 0.861–0.906), with a sensitivity of 94.1% and specificity of 82.9% for detecting lesions.
Conclusions
We demonstrate the safety, acceptability and feasibility of AI-assisted sponge cytology in high-risk regions, with high accuracy for detecting ESCC, GCA and their precursor lesions. Our results pave the way for innovative etiology and early-detection research.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
National Natural Science Foundation of China.
Disclosure
All authors have declared no conflicts of interest.