Long-Term Outcome of Gastric Mild-Moderate Dysplasia: A Real-World Clinical Experience

Role of submucosal injection in radiofrequency ablation of gastric low-grade dysplasia: Effects on symptoms and outcomes

BACKGROUND

To date, there is still a lack of standardized management strategies for gastric low-grade dysplasia (LGD), which is a direct neoplastic precancerous lesion and requires specifically superficial destruction. Radiofrequency ablation (RFA) is expected to be an effective method for gastric LGD, but post-RFA pain may affect patients' satisfaction and compliance. The current study aimed to evaluate the value of a submucosal injection prior to RFA (SI-RFA) for postoperative pain and treatment outcomes.

METHODS

Between October 2014 and July 2021, gastric LGDs without risk factors (size >2 cm, unclear boundary, and abnormal microsurface and microvascularity) undergoing regular RFA and SI-RFA were retrospectively analyzed. Postoperative pain scores, wound healing, and clinical efficacy were compared. Propensity score matching, stratified analysis, and multivariable logistic regression were performed to control the confounding variables.

RESULTS

One hundred and ninety-seven gastric LGDs in 151 patients received regular RFA. Forty-nine gastric LGDs in 36 patients received SI-RFA. Thirty-six pairs of patients were selected for the assessment of postoperative pain by propensity score matching. Compared to regular RFA, SI-RFA significantly decreased the degree and duration of postoperative pain (OR, 0.32; 95% CI, 0.13-0.84; P = 0.020), improved wound healing rate (80.0% [36/45] vs. 58.9% [89/151], P = 0.012), increased the complete ablation rate (91.8% [45/49] vs. 86.3% [170/197], χ2 = 1.094, P = 0.295), but correlated with higher rates of local recurrence and progression (25.6% [10/39] vs. 13.2% [18/136], χ2 = 3.471, P = 0.062; 8.3% [3/36] vs. 0.9% [1/116], P = 0.042). The multivariable logistic regression model confirmed that submucosal injection was associated with local recurrence (OR, 2.93; 95% CI, 1.13-7.58; P = 0.027).

CONCLUSIONS

Submucosal injections prior to RFA may reduce postoperative pain and scar formation while ensuring complete ablation of gastric LGD. However, local recurrence and progression should be considered seriously.

A deep learning and natural language processing-based system for automatic identification and surveillance of high-risk patients undergoing upper endoscopy: A multicenter study

BACKGROUND

Timely identification and regular surveillance of patients at high risk are crucial for early diagnosis of upper gastrointestinal cancer. However, traditional manual surveillance method is time-consuming, and current surveillance rate is below 50%. Here, we aimed to develop a surveillance system named ENDOANGEL-AS (automatic surveillance) for automatic identification and surveillance of high-risk patients.

METHODS

7874 patients from Renmin Hospital of Wuhan University between May 1 and July 31, 2021 were used as the training set, 6762 patients between August 1 and October 31, 2021 as the internal test set, and 7570 patients from two other hospitals between August 1 and October 31, 2021 as the external test sets. We first extracted descriptions of abnormalities from endoscopic and pathological reports based on natural language processing techniques to identify individuals. Then patients were classified at nine risk levels according to endoscopic and pathological findings, and a deep learning model was trained to identify demarcation line (DL) in gastric low-grade intraepithelial neoplasia (LGIN) using 1561 white-light still images for risk stratification of gastric LGIN. Finally, patients undergoing upper endoscopy were classified and assigned one of ten surveillance intervals according to guidelines. The performance of ENDOANGEL-AS was evaluated and compared with physicians.

FINDINGS

Patient identification module achieved an accuracy of 100% and 99.91% in internal and external test sets, respectively. Risk level classification module achieved an accuracy of 100% and 99.85% in the internal and external test sets, respectively. DL identification module achieved an accuracy of 87.88%. ENDOANGEL-AS on surveillance interval assignment achieved an accuracy of 99.23% and 99.67% in internal and external test sets, respectively. ENDOANGEL-AS had significantly higher accuracy compared with physicians (99.00% vs 38.87%, p < 0.001). The accuracy (63.67%, p < 0.001) of endoscopists with the assistance of ENDOANGEL-AS was significantly improved.

INTERPRETATION

We established a surveillance system that can automatically identify patients and assign surveillance intervals with high accuracy and good transferability.

FUNDING

This work was partly supported by a grant from the Hubei Province Major Science and Technology Innovation Project (2018-916-000-008) and the Fundamental Research Funds for the Central Universities (2042021kf0084).