Magnetically controlled capsule gastroscopy and serology for screening precancerous high-risk events of gastric cancer: a cross-sectional study (SIGES)

<h2>Abstract</h2><h3>Background</h3> Gastric cancer is a common malignancy in China, while screening and early detection of its precancerous high-risk events might help surveillance of high-risk subpopulation and improve the population survival. Magnetically controlled capsule gastroscopy (MCCG) emerged as a novel, non-invasive, and better compliant diagnostic method for screening gastric cancer and precancerous high-risk conditions, such as atrophic gastritis, gastric ulcer, and gastric polyp. The effectiveness of MCCG and the feasibility of sequential serology-MCCG protocol need to be preliminarily assessed in the aspect of screening precancerous high-risk events of gastric cancer. <h3>Methods</h3> This cross-sectional study collected health check-up observations (18-75 years old) at the Health Management Center, West China Hospital of Sichuan University between 2018 and 2020. Demographic data were retrieved including sex, age, ethnicity, education level, body mass index, smoking, alcohol drinking, and family history of cancers. Those observations had undergone MCCG and relevant serologic examinations including pepsinogen-I (PG-I), pepsinogen-II (PG-II), PG-I/-II ratio (PGR), gastrin-17 (G-17), carcinoembryonic antigen (CEA), and carbohydrate antigen 19-9 (CA19-9). Major findings from MCCG concerned tumor and high-risk precancerous events, i.e. chronic atrophic gastritis (CAG), gastric ulcer, and gastric polyp. The primary outcome measure was all high-risk events, while the second outcome measures were any of the above events. The detection rate of MCCG was estimated, and the predictive strengthen of serology on MCCG findings was analyzed. <h3>Findings</h3> A total of 1,432 eligible healthy check-up persons were included. MCCG reported none tumor and 114 cases with any precancerous high-risk event. The overall detection rate was 79.6‰ (95% CI 66.7‰-94.8‰) for all high-risk events, while the specific detection rates were 15.4‰ (95% CI 10.1‰-23.2‰) for CAG, 14.7‰ (95% CI 9.6‰-22.4‰) for gastric ulcers, and 55.2‰ (44.5‰-68.3‰) for gastric polyp, respectively. Compared between MCCG-positive (all high-risk events) and MCCG-negative groups, the baselines were generally comparable, with the only exception of age. The elder persons appeared significantly higher rates of high-risk events (P<0.001), particularly increased as ≥60 years old. All the levels of serologic examinations were not significantly different between MCCG-positive and MCCG-negative groups. However, regarding CAG outcome dataset, the levels PG-I (median 58.9 ug/L, IQR 44.3 ug/L-85.2 ug/L, P=0.016) and PGR (median 6.2, IQR 3.4-9.5, P=0.017) were significantly lower in the CAG group, as well as G-17 (P=0.009). The serology couldn't well predict in all high-risk events dataset, but serologic CAG performed great predictive strength for MCCG CAG (adjusted diagnostic odds ratio [aDOR]=10.40, 95% CI 2.08-51.98; SPE=98.6%; LR+=8.85). Additionally, CA19-9 seropositivity predicted MCCG gastric ulcer (aDOR=7.86, 95% CI 1.58-39.08; SPE=97.8%; LR+=4.72). Seropositivity of G-17 and CEA couldn't predict any high-risk events. <h3>Interpretation</h3> MCCG might perform reasonable capacity of detecting precancerous high-risk events of gastric cancer. MCCG is an alternative for screening high-risk candidates of gastric cancer for the sake of personal compliance, especially for the elders. Serologic CAG and CA19-9 seropositivity could be considered as preposed indications for MCCG. Better sequential serology-MCCG protocol needs further investigation and optimization in the cost-effective aspect.