Endoscopy International Open最新文献

Background and study aims: Advanced endoscopic resection methods are recommended for removing rectal neuroendocrine tumors (NETs) < 10 mm, but there is no consensus on optimal endoscopic technique. This study aimed to determine whether endoscopic mucosal resection with ligation (EMR-L) is superior to endoscopic submucosal dissection (ESD) in terms of efficacy and safety, focusing on achieving adequate vertical margin distance.

Patients and methods: This dual-center cohort study included consecutive patients with rectal NETs ≤ 10 mm. Adequate vertical margins were exploratively defined as a margin distance exceeding 115 μm, the 25th percentile threshold. Propensity score matching (PSM) was applied to compare outcomes between EMR-L and ESD.

Results: The study included 204 rectal NETs from 186 patients, with 61 lesions in each group after PSM. Compared with ESD, the EMR-L group demonstrated a higher R0 resection rate (98.4% vs. 83.6%, P = 0.021), a greater proportion of adequate vertical margins (80.3% vs. 62.3%, P = 0.030), and a lower rate of positive vertical margins (1.6% vs. 13.1%, P = 0.041). Subgroup analysis indicated that lesions ≤ 5 mm without prior biopsy or central depression derived greater benefit from EMR-L. Furthermore, EMR-L was associated with significantly shorter procedure times (median 5.0 vs. 19 minutes) and a markedly lower overall complication rate (8.2% vs. 29.5%, P = 0.005), particularly perforation (3.3% vs. 16.4%, P = 0.028).

Conclusions: EMR-L outperforms ESD for small rectal NETs by achieving higher R0 and better vertical margins in shorter times, while minimizing risk of perforation.

Background and study aims: Endoscopic papillectomy (EP) is an effective endoscopic modality for managing ampullary lesions. This study aimed to evaluate predictors for recurrence and adverse events (AEs) in patients who underwent EP for major papilla lesions.

Patients and methods: This was a retrospective analysis of all patients who underwent endoscopic snare papillectomy for a major papilla lesion between January 2006 and December 2021. We assessed multiple patient- and procedure-related variables to identify risk factors related to post-EP AEs and lesion recurrence using both univariate and multivariate analysis. In addition, we compared baseline characteristics and outcomes in patients with familial adenomatous polyposis (FAP) vs. sporadic ampullary lesions (SALs).

Results: Fifty-one patients (11 FAP) were included in the final analysis. Recurrence was seen in 17 patients (37.0%) among those who followed up after technical success. Complete histological (R0) resection was the only factor associated with no recurrence following EP (odds ratio [OR] 5.4, 95% confidence interval [CI] 1.4-20.8, P = 0.014). PEP was associated with delayed bleeding (OR 7.5, 95% CI 1.2-46.1, P = 0.03). In FAP vs. SAL, lesion size was smaller in FAP: 10 mm (interquartile range 6-15 mm) vs. 15 mm (12-21 mm, P = 0.03) and the en-bloc resection rate was higher (100 vs. 67.5%, P = 0.01). Although the recurrence rate was higher in FAP vs. SAL (55.6 vs. 32.4%, P = 0.2), this was not statistically significant. Rates of AEs were similar.

Conclusions: R0 resection was associated with reduced risk of recurrence whereas delayed bleeding after EP is associated with an increased risk of developing PEP. EP is safe and effective for removing ampullary lesions irrespective of lesion type.

Background and study aims: Endoscopic resection for duodenal gastrointestinal stromal tumors (GISTs) is still considered a great challenge with a high risk of complications. This study aimed to evaluate effectiveness and safety of endoscopic resection for duodenal GIST.

Patients and methods: Between June 2013 and August 2024, we performed a retrospective study of patients with duodenal GISTs who underwent endoscopic resection at Zhongshan Hospital. Data on patient characteristics, clinical outcome, and follow-up were collected.

Results: A total of 73 patients with duodenal GISTs were enrolled, including 31 patients who underwent endoscopic submucosal dissection (ESD) and 42 who underwent endoscopic full-thickness resection (EFTR). Mean lesion size was 1.2 ± 0.5 cm and 1.9 ± 0.9 cm, respectively. En bloc resection rates were 96.8% and 95.2%, respectively. Rates of R0 resection were 45.2% and 42.9%, respectively. Rates of R1 resection were 54.8% and 57.1%, respectively. No patient transferred to open surgery. Postoperative adverse events included delayed bleeding (1 case), delayed perforation (1 case), delayed wall edema (2 cases), hydrothorax (1 case), and retroperitoneal infection (1 case). Mean hospital stays were 4.1 ± 2.8 days and 6.2 ± 4.9 days, respectively. No metastasis or duodenal stenosis were detected during the follow-up period (64.8 ± 43.6 months and 61.3 ± 40.2 months, respectively). Local recurrence occurred in one patient with high recurrence risk at 56 months after EFTR.

Conclusions: ESD and EFTR are safe, minimally invasive treatments for duodenal GISTs. Moreover, the EFTR technique may have advantages of completely resecting lesions originating from the deep muscularis propria layer, particularly lesions with extraluminal growth.

Background and study aims: Gastroesophageal reflux (GERD) can occur in a significant number of achalasia patients undergoing post-peroral endoscopic myotomy (POEM). POEM with endoscopic fundoplication (POEM-F) is a new endoscopic technique to treat post- POEM GERD. We conducted a multicenter cohort study to compare outcomes between POEM and POEM-F.

Patients and methods: We included patients who underwent POEM or POEM-F from six tertiary centers. Primary outcomes were: 1) clinical success, defined as post-procedure cessation or reduction of proton-pump inhibitors; 2) post procedure percent time pH < 4 and DeMeester score; and 3) post-procedure Eckardt scores. Secondary outcomes included adverse events, procedure time, and hospital stay duration.

Results: Sixty-four patients were included: 31 patients underwent POEM-F (mean age 51, 48% male), whereas 33 patients POEM (mean age 56, 58% male). Technical success was 100% in both groups. POEM-F patients achieved reduction/cessation in proton pump inhibitor use in 25 of 31 patients (80%); POEM patients in eight of 33 (24%) ( P ≤ 0.00001). Percent time with pH < 4 was significantly lower in the POEM-F group (2.75 +/- 2.53 vs 9.3 +/- 3.6 min, P ≤ 0.0001). Post-procedure DeMeester scores were lower (< 14.7) in POEM-F (mean 9.6) than POEM (mean 15.8) ( P ≤ 0.0011). There were three mucosal injuries in the POEM-F group and four in the POEM group. Hospital stay duration and procedure time did not differ between groups. In the POEM group, four patients required Dor fundoplication. No repeat interventions were recommended for the POEM- F group at 6-month follow-up.

Conclusions: POEM-F achieves statistically significant improvement in post- POEM GERD compared with POEM alone.

Background and study aims: Direct visualization endoscopic retrograde appendicitis therapy (ERAT), an advanced technique building upon conventional ERAT, represents a novel endoscopic approach for managing acute uncomplicated appendicitis. This study aimed to assess clinical efficacy and safety of employing cholangioscope-guided endoscopic intervention as a therapeutic approach.

Patients and methods: A retrospective analysis was conducted on 656 patients presenting with suspected acute appendicitis between February 2024 and November 2024. To minimize baseline differences, propensity score matching was applied, resulting in a final study population of 34 patients undergoing direct visualization ERAT and 68 patients treated with laparoscopic appendectomy (LA). Key outcome measures included technical and clinical success rates, operative time, time to postoperative pain resolution, length of hospital stay, recurrence rate, incidence of adverse events (AEs), and overall patient satisfaction.

Results: The technical success rate was 97.06% (33/34) in the ERAT group and 100% in the LA group ( P = 0.333), while clinical success was achieved in 94.12% (32/34) of ERAT cases compared with 100% in the LA cohort ( P = 0.109). Notably, ERAT was associated with a significantly shorter operative time (37 vs 50 minutes; P < 0.001) and more rapid postoperative pain relief ( P = 0.001), with a greater proportion of patients reporting complete symptom resolution within 2 days of the procedure. There were no significant differences between the two groups in terms of AEs, length of hospital stay, or patient satisfaction. During follow-up, a recurrence of appendicitis was observed in one ERAT patient (2.94%, 1/34).

Conclusions: Direct visualization ERAT demonstrated high feasibility and effectiveness as a diagnostic and therapeutic modality for acute uncomplicated appendicitis, offering a promising alternative to conventional approaches.

Background and study aims: Whether routine capnography monitoring during gastrointestinal endoscopy sedation can reduce occurrence of hypoxia is controversial. Older patients are more prone to hypoxia. This study aimed to determine the effect of additional capnography monitoring on incidence of hypoxia in older patients undergoing gastrointestinal endoscopy under propofol sedation.

Patients and methods: A multicenter, randomized, single-blind, two-arm, parallel-group, controlled with an active comparator, interventional superiority clinical trial was performed at three teaching hospitals in China between September 1, 2021, and September 1, 2022. This study compared additional capnography monitoring (intervention group) and standard monitoring (control group) among older patients (aged 65-79 years) undergoing gastrointestinal endoscopy under propofol sedation. The primary outcome was incidence of hypoxia (75%-89% for < 60s). Secondary outcomes were incidence of subclinical hypoxia (90%-94%), incidence of severe hypoxia (< 75% for any duration or 75%-89% for ≥ 60s), and other adverse events (AEs).

Results: Data from 1777 participants (888 intervention, 889 control group) were analyzed. Additional capnography monitoring reduced incidence of hypoxia in older patients from 19% to 12% ( P < 0.001). Incidence of subclinical hypoxia in the additional capnographymonitoring group was 23% and in the standard monitoring group was 15% ( P < 0.001). There was no significant difference in incidence of severe hypoxia ( P = 0.070) and other AEs between the two groups ( P = 0.374).

Conclusions: Additional capnography monitoring during gastrointestinal endoscopy for older patients who were sedated with propofol reduces incidence of hypoxia.

Background and study aims: Current general-purpose artificial intelligence (AI) large language models (LLMs) demonstrate limited efficacy in clinical medicine, often constrained to question-answering, documentation, and literature summarization roles. We developed GastroGPT, a proof-of-concept specialty-specific, multi-task, clinical LLM, and evaluated its performance against leading general-purpose LLMs across key gastroenterology tasks and diverse case scenarios.

Methods: In this structured analysis, GastroGPT was compared with three state-of-the-art general-purpose LLMs (LLM-A: GPT-4, LLM-B: Bard, LLM-C: Claude). Models were assessed on seven clinical tasks and overall performance across 10 simulated gastroenterology cases varying in complexity, frequency, and patient demographics. Standardized prompts facilitated structured comparisons. A blinded expert panel rated model outputs per task on a 10-point Likert scale, judging clinical utility. Comprehensive statistical analyses were conducted.

Results: A total of 2,240 expert ratings were obtained. GastroGPT achieved significantly higher mean overall scores (8.1 ± 1.8) compared with GPT-4 (5.2 ± 3.0), Bard (5.7 ± 3.3), and Claude (7.0 ± 2.7) (all P < 0.001). It outperformed comparators in six of seven tasks ( P < 0.05), except follow-up planning. GastroGPT demonstrated superior score consistency (variance 34.95) versus general models (97.4-260.35) ( P < 0.001). Its performance remained consistent across case complexities and frequencies, unlike the comparators ( P < 0.001). Multivariate analysis revealed that model type significantly predicted performance ( P < 0.001).

Conclusions: This study pioneered development and comparison of a specialty-specific, clinically-oriented AI model to general-purpose LLMs. GastroGPT demonstrated superior utility overall and on key gastroenterology tasks, highlighting the potential for tailored, task-focused AI models in medicine.