Introduction

Endoscopic submucosal dissection (ESD) is an effective treatment for superficial esophageal cancer, but postoperative stenosis remains a significant concern. Particularly, probability of the stenosis is less investigated in cases with a 50%-75% circumferential mucosal defect.

Aims & Methods

This study aimed to identify influential factors for postoperative stenosis after the moderate circumferential resection in esophageal ESD. This multi-center retrospective observational study analyzed data from 513 patients across 19 institutions who underwent esophageal ESD from April 2018 to March 2022 and showed 50% to 75% of the circumferential mucosal defect. We firstly overviewed the current status on the postoperative stenosis, the steroid use, and postoperative adverse events in these patients. Subsequently, we identified influential factors for postoperative stenosis. Patients were categorized into stenosis and no stenosis groups, and univariate and multivariate analyses were performed for each parameter. We further investigated influential factors of stenosis in two cohorts: steroid and no steroid groups. Finally, the parameters were compared between the steroid and no steroid groups to assess the tendency of the steroid use in general practice.

Results

Postoperative stenosis was observed in 28 cases (5%). Univariate analysis comparing the stenosis group (n = 28) and no stenosis group (n = 485) revealed significant differences in the prior endoscopic treatment near the lesion (p = 0.02), cervical esophageal location (p = 0.04), defect circumference (p =0.01), and procedure duration (p = 0.02). There was no significant difference in the steroid use between two groups. Multivariate analyses identified history of prior endoscopic treatment close to the lesion (p = 0.01, odds ratio [OR], 5.10), cervical esophageal lesions (p = 0.03, OR, 3.97), large circumferential defect (p <0.01, OR, 3.78), and longer procedure duration (p <0.01, OR, 3.83) as independent risk factors for stenosis formation. Among the steroid group, significant associations of the postoperative stenosis were revealed with the prior endoscopic treatment near the lesion (p = 0.02), cervical esophageal location (p = 0.04), and >5/8 of defect circumference (p = 0.02). In the no steroid group, only the procedure duration (p = 0.01) showed a significant difference between the stenosis and non-stenosis groups. Comparison between the steroid (n = 223) and no steroid groups (n = 290) revealed significant differences in the prior endoscopic treatment near the lesion (p <0.01), cervical esophageal lesions (p <0.01), and resected specimen dimensions (p <0.01). The stenosis rates were 7% in the steroid group and 4% in the no steroid group, without a significant difference. No significant difference in adverse events was observed between the two groups.

Conclusion

In esophageal ESD with a 50%-75% circumferential mucosal defect, the steroid was used in over 40% of patients, and postoperative stenosis still occurred in 5%. Because wider circumference, location of the cervical esophagus, prior endoscopic treatment close to the lesion, and prolonged operative time were identified as independent risk factors on the stenosis, steroid therapy should be preferred in these cases. Particularly, lesions with a history of endoscopic treatment, on cervical esophagus, and with >5/8 of defect circumference should require more aggressive countermeasures than currently applied.

Introduction

Chronic liver disease remains one of the most common causes for admission in the Gastroenterology department, with high morbidity and mortality rates. As quality-of-care gains traction in our clinical practice, chronic liver disease, particularly when regarding decompensated cirrhotic patients in the emergency department, remains without a standardized approach. Additionally, timely and adequate management within the first 24 hours of admission reduces morbidity and mortality in decompensated cirrhosis. In the United Kingdom, a standardized cirrhosis care bundle as been effectively applied.

Aims & Methods

We conducted a retrospective review to determine baseline management of patients with decompensated cirrhosis attending two Portuguese hospital centers, in 2021 and 2022.

Results

We included a total of 148 patients from 2 hospital centers, 141 were male (95.3%) and had a median age of 62 years at admission (IQR 15). Median Child-Pugh and Model for End-stage Liver Disease (MELD) -Na scores at admission were 9 (IQR 3) and 18 (IQR 10) points, respectively.
The most common decompensations were gastrointestinal bleeding in 33.1% of patients, ascites in 30.8% and hepatic encephalopathy in 25.6%. Sixty-five patients had more than 1 decompensation (43.9%). It was their first liver disease decompensation in 55 patients (37.2%). Regarding etiology, alcoholic liver disease was the most common, in 134 patients (90.5%). Median duration of admission was 8 days (IQR 6). The in-hospital mortality rate was 4.7% (7 patients) and mortality rate at 6 weeks after admission was 8.8% (13 patients). Forty-six patients were readmitted up to 3 months (31.1%).
In patients with ascites, 29 did not undergo paracentesis (33.3%). In patients with recent history of alcohol abuse, 5 did not receive thiamine (7.5%) and 6 did not undergo alcohol abstinence syndrome prevention therapy (9%). In suspected infection, 4 out of 38 patients did not receive empirical antibiotics in the first 24 hours of admission (10.5%). In acute renal lesion, 2 patients did not stop diuretics and/or nephrotoxic drugs (5.6%) and 4 did not receive fluids and/or albumin (12.1%). In 3 patients there was no diuresis monitoring (7.5%). At admission for gastrointestinal bleeding, 3 patients received a nonrestrictive approach regarding transfusions (5.5%), 18 did not undergo vasopressor therapy (32.1%) and in 5 prophylactic antibiotics were not administered (8.9%). Prophylactic lactulose was used in 25 patients (45.5%). Only 1 patient did not undergo upper endoscopy before 12 hours of admission (1.8%). In hepatic encephalopathy, the precipitating factor was recognized in 29 patients (69%), with infection as the most commonly identified (55.2%). Venous Thromboembolism prophylaxis was not administered (45.3%) in 67 patients and 88 patients did not undergo a nutritional evaluation during admission (59.5%).

Conclusion

We were able to identify several areas in need of improvement: one-third of patients with ascites did not undergo diagnostic ascitic taps, nearly on-third of patients did not start vasopressor therapy at admission and more than half had no nutritional evaluation.
Training and best practice policies are a priority in Gastroenterology nowadays, and the identification of poor performance areas is relevant for the betterment of our clinical practice. An adapted bundle for prospective application is underway in our centers as we hope to reduce overall hospital stay and mortality in decompensated cirrhosis.

References

Dyson JK, et al. Implementation of a ‘care bundle’ improves the management of patients admitted to hospital with decompensated cirrhosis. Aliment Pharmacol Ther 2016; 44: 1030–1038, doi:10.1111/apt.13806
McPherson S, et al. Decompensated Cirrhosis Care Bundle - First 24 Hours. Disponível em: https://www.bsg.org.uk/wp-content/uploads/2019/12/BSG-BASL-Decompensated-Cirrhosis-Care-Bundle-First-24-Hours.pdf