Introduction

Although an effective and relatively safe procedure, bleeding complications often occur after gastric endoscopic submucosal dissection (ESD). We aimed to comprehensively evaluate the risk factors of post-procedural bleeding following gastric ESD.

Aims & Methods

This study was designed as a nationwide, population-based, retrospective study utilizing the Korean National Health Insurance Service database, comprising patients who underwent gastric ESD for gastric cancer or adenoma from November 2011 to December 2022. Operational definitions were created to identify the target population and post-procedural bleeding using the diagnosis and procedure code, which underwent validation. We established operational definitions to identify the target population and post-procedural bleeding by utilizing diagnosis and procedure codes, which underwent validation. The risk factors analyzed included age, sex, hospital volume, socioeconomic status (SES), histology, lesion numbers, comorbidities, and medication usage.

Results

During the study period, there were 191,932 procedures conducted on 175,779 patients, resulting in 9921 cases of bleeding events. Several factors were associated with increased post-ESD bleeding risk. Following multiple logistic regression, relevant factors with adjusted odds ratios (aORs) were identified. These factors included young age (aOR: 0.990, 95% CI 0.987-0.992), male sex (aOR: 1.377, 95% CI 1.311-1.448), low-volume center compared to high-volume center (aOR: 1.289, 95% CI 1.022-1.621), medical aid compared to the top 25% SES (aOR: 1.307, 95% CI 1.188-1.440), higher Charlson comorbidity index (aOR: 1.040, 95% CI 1.026-1.054), gastric cancer (aOR: 1.369, 95% CI 1.307-1.434), and multiple lesions (aOR: 1.144, 95% CI 1.032-1.267). Furthermore, bleeding risk was significantly higher in groups with several comorbidities and antithrombotic agent users, particularly showing Notably high aORs in groups with renal failure (aOR: 2.086, 95% CI 1.909-2.278), warfarin (aOR: 2.641, 95% CI 2.219-3.143), direct oral anticoagulants (aOR: 2.519, 95% CI 2.220-2.859), and antiplatelet combinations (aOR: 2.073, 95% CI 1.783-2.410).

Conclusion

This study revealed various factors associated with increased bleeding risk after gastric ESD using a large cohort, which could be an important issue in terms of risk stratification and personalized treatment.

Introduction

Post-procedural bleeding (PPB) is the most common adverse event after gastric endoscopic submucosal dissection (ESD). The BEST-J score was created to predict PPB risk following gastric ESD.

Aims & Methods

We aimed to evaluate predictive factors of PPB and validate the BEST-J score in a large European cohort. We conducted a retrospective cohort study including consecutive gastric epithelial lesions removed by ESD between January of 2005 and June of 2024. Incomplete ESD and patients with missing information that prevented the score calculation were excluded. PPB was defined as presence of compatible clinical picture in the 28 days following the procedure. Univariate and multivariate analysis were used to identify predictive factors of PPB, and the diagnostic ability of the BEST-J score was calculated in this cohort.

Results

From 990 ESD, 874 were included in the final analysis. PPB occurred in 53 patients (6.1%), with a median time after procedure of 2 days (P25-P75: 1-7). PPB was significantly superior in patients with higher risk as evaluated by the BEST-J score (PPB rate of 3.7%, 3.9%, 19.2% and 35.7% for low, intermediate, high and very-high risk, respectively). The area under the curve for PPB prediction using the BEST-J score was 0.684 (IC95% 0.598-0.679). For a score ≥ 3 (high and very-high risk), the sensitivity and specificity for PPB detection were 47.2% (IC95% 34.2-60.5) e 88.7% (IC95% 86.4-90.7), respectively. In this sample, the following predictive factors for PPB were identified: ischemic cardiopathy [adjusted Odds Ratio (aOR) 4.351; IC95% 1.660-11.404], medication with warfarin (aOR 18.339; IC95% 4.541-74.064) or direct oral anticoagulants (aOR 7.596; IC95% 2.666-21.641) and lesion size > 20mm (aOR 2.563; IC95% 1.246-5.272).

Conclusion

The BEST-J score presented moderate discrimination for PPB following gastric ESD in our cohort. The identified factors may assist the decision of implementing prophylactic measures in high-risk patients and the creation of more precise prediction models.

References

Vasconcelos AC, et al. Endoscopic Resection of Early Gastric Cancer and Pre-malignant Gastric Lesions. Cancers, 2023 Jun 7;15(12):3084. doi: 10.3390/cancers15123084.
Hatta W, et al. Prediction model of bleeding after endoscopic submucosal dissection for early gastric cancer: BEST-J score. Gut, 2021 Mar;70(3):476-484. doi: 10.1136/gutjnl-2019-319926

Introduction

For gastric neoplasms, endoscopic submucosal dissection (ESD) is a well-known therapeutic option. Our study aimed to investigate the influence of procedural volume on the outcome of gastric ESD.

Aims & Methods

Patients who received ESD for gastric cancer or adenoma from November 2011 to December 2017 were identified using the Korean National Health Insurance Service database. The diagnostic and procedure codes were combined to develop operational definitions, which were validated using individual hospital medical record data. Outcomes included bleeding, perforation, pneumonia, 30-day mortality, composite outcome comprising all these adverse outcomes, and additional resection. Hospital volume was categorized into four groups: very high-, high-, low-, very low-volume hospitals (VHVH, HVH, LVH, VLVH). The outcomes of ESD were compared in relation to hospital volume.

Results

A total of 89,780 patients underwent 95,411 procedures during the research period. There were 5,607 composite events, which included 5,098 bleeding, 601 perforation, and 712 pneumonia cases, respectively. Additional resection occurred within 180days in 7,900 cases. There were significant differences in ESD-related adverse outcomes among the four hospital volume categories. Multiple logistic regression revealed that VHVH, HVH, and LVH were associated with lower risk of a composite outcome, when compared to VLVH (OR, 0.627, 95% CI, 0.445-0.885, p=0.008; OR, 0.542, 95% CI, 0.406-0.722, p<0.001; OR, 0.697, 95% CI, 0.548-0.886, p=0.003). Similar tendencies were also shown for bleeding, perforation, and pneumonia, although this was not evident in additional resection.

Conclusion

The procedural volume at an institutional level was closely associated with clinical outcomes, such as bleeding, perforation, pneumonia, and a composite outcome, in patients with ESD for gastric cancer or adenoma. The outcomes of gastric ESD should be closely and systemically monitored in a nationwide level, to ensure the quality of procedures and the safety of patients.

Introduction

Delayed bleeding after gastric endoscopic submucosal dissection (ESD) is one of the most frequent complications (up to 7%). The BEST-J score is a risk predictive model for delayed bleeding following gastric ESD. We aim to assess potential risk factors associated with bleeding and validate the BEST-J score in a population of patients undergoing ESD.

Aims & Methods

Prospective cohort study, single centre, of patients who underwent gastric ESD between 2017 and 2023. The characteristics of the lesion were evaluated, and logistic regression of risk factors associated with bleeding after ESD was performed. The BEST-J score was calculated for each patient and the discriminative ability was evaluated through the area under the receiver operating characteristic curve (AUC-ROC).

Results

Included 101 gastric ESDs (63.4% female), with a mean age of 69.7 ± 8 years. 17.8% of patients were on antiplatelet therapy and 13.9% were on anticoagulant therapy. Among the analyzed lesions: 43.6% measured between 0-20mm, 56.4% were ≥ 20mm; 13.9% were sessile, 60.4% flat, 14.9% depressed and 10.9% subepithelial; 22.8% were located in the cardia, fundus, or body, 17.8% in the incisura angularis and 59.4% in the antrum. The most frequent histological results were: 38.6% low-grade dysplasia, 26.7% high-grade dysplasia and 19.8% intramucosal adenocarcinoma. The mean procedure duration was 87±43 minutes. Delayed bleeding after ESD occurred in 5.9% (n=6) of cases. In the univariate analysis of risk factors, the lesion size ≥ 20mm showed statistical significance (10.5% vs. 0%; p=0.034), but not the location in the antrum (6.7% vs. 4.9%; p=0.5), lesion nature (7.1% vs. 8.2% vs. 0% vs. 0%; p=0.5) or procedure duration ≥60 minutes (5% vs. 9.5%; p=0.3). The BEST-J score predicted the risk of delayed bleeding after gastric ESD with an AUC of 0.70 (95% CI: 0.53-0.87; p=0.1).

Conclusion

The use of bleeding predictive scores may be useful in identifying high-risk patients; in our series, the BEST-J score was not very effective, with only lesion size ≥ 20mm being a relevant factor for bleeding after ESD.

Introduction

Delayed bleeding (DB) is a possible adverse event following gastric endoscopic submucosal dissection (ESD). The Bleeding after ESD Trend from Japan (BEST-J) score was created as a risk prediction model for DB following gastric ESD, but is yet to be validated in non-Asiatic populations.

Aims & Methods

We aimed to apply and validate the BEST-J score on a European sample.
We have conducted a longitudinal study of all consecutive patients undergoing gastric ESD on a European Endoscopic Unit. DB was defined as a hemorrhage with clinical symptoms and confirmed by emergency endoscopy from the time of completion of ESD to 28 days after ESD.
BEST-J score (use of vitamin K antagonists +4, direct oral anticoagulants +4, aspirin or P2Y12R antagonists +2; withdrawing antithrombotic drugs -1; chronic kidney disease +3; multiple lesions +1; lesion ≥30mm +1; lesion in lower-third of the stomach +1) was calculated in each patient and confronted with the outcome (DB).

Results

Final sample included 161 patients, 102 (63.4%) male, with a mean age of 68±8 years. From these, 10 (6.2%) presented DB following ESD, with a median time to bleeding of 7 days (IQR 6.8).
According to BEST-J score, bleeding risk was low (0-1 points) in 111 (68.9%), intermediate (2 points) in 29 (18.0%), high (3-4 points) in 16 (9.9%) and very high (≥5 points) in 5 (3.1%) patients.
BEST-J score presented an excellent accuracy predicting DB in our sample, with an AUC=0.907 (95%CI=0.801-1.000; p<0.001). The optimal cut-off value to predict DB was a BEST-J score ≥3, which matches the cut-off value for high-risk of bleeding in the original investigation. This value had a sensitivity of 90% and specificity of 92%.

Conclusion

The BEST-J score still presents excellent accuracy in risk stratification for post-ESD bleeding in European individuals. Thus, this score may help to guide which patients benefit the most from prophylactic therapies following gastric ESD in this setting.