The increasing demand on endoscopy units in Portugal requires optimization of the diagnostic yield of procedures. Routine endoscopic re-evaluation is based on limited evidence. Recent studies support an individualized approach, reserving repeat endoscopy for patients at higher risk.

To assess the incidence of malignant gastric ulcer in a cohort of patients undergoing upper gastrointestinal endoscopy and to identify clinical, endoscopic, and histological predictors of malignancy that may guide the need for endoscopic re-evaluation.

Retrospective observational study conducted in the Gastroenterology Department, including patients with gastric ulcer diagnosed on upper endoscopy between January and December 2024. Demographic, clinical, endoscopic, and histological data from index and follow-up endoscopies were collected. Descriptive and univariate analyses were performed using Microsoft Excel®.

A total of 101 patients were included (mean age 69 years; 72% male); 59% of endoscopies were elective, and biopsy was performed in 75% of index procedures. Seven malignant ulcers (6.9%) were identified, 71% diagnosed at the index endoscopy. Among 53 ulcers reassessed endoscopically, most showed mucosal healing (79%).Endoscopically suspicious ulcers represented 27% of cases, of which 26% were malignant. Among ulcers deemed benign, only 1/74 (1.3%) was later found to be malignant (late malignancy rate 2.7%).The strongest predictors of malignancy were endoscopic suspicion—particularly excavated base and/or elevated borders (OR 17.2; 95%CI 0.8–372)—and non-antral location, especially the incisura (OR 4.03; 95%CI 0.75–16.6; p=0.12), showing a trend toward increased risk. Ulcer size was also a strong and independent predictor: ulcers >20 mm were significantly associated with malignancy (OR 12.5; 95%CI 2.0–78.0; p=0.0026).Comorbidities most strongly associated with malignancy included cardiovascular disease (OR 3.22), chronic kidney disease (OR 4.64), antiplatelet therapy (OR 6.26), and smoking (OR 6.03).Helicobacter pylori at index endoscopy showed a non-significant trend toward lower malignancy risk (OR 1.8; p=0.12).A benign initial histology was a strong predictor of benign disease.The presence of multiple ulcers (≥2), including duodenal ulcer, was associated with a lower probability of malignancy (OR 0.10; 95%CI 0.005–1.76; p=0.043), supporting the observation that malignant ulcers tend to present as single lesions.All malignant cases occurred in patients aged >50 years, but age ≥50 did not reach statistical significance (OR 1.86; 95%CI 0.10–35.0; p=1.00). Sex was not associated with malignancy (OR 0.85; 95%CI 0.18–4.05).Presentation with overt bleeding at index endoscopy was not significantly associated with malignancy.

In this cohort, 7% of gastric ulcers were malignant, most diagnosed at the index endoscopy with adequate histological sampling. The diagnostic yield of routine re-evaluation after benign biopsies was low. The combination of benign histology and absence of suspicious endoscopic features was sufficient to exclude malignancy in most patients, supporting a risk-stratified approach.These findings align with current international guidelines (MAPS III, ESGE/AGA/ASGE) and support a strategy that may reduce costs, optimize resource allocation, and improve the sustainability of endoscopy services. Validation in larger, multicenter cohorts is warranted.