Background. Due to improvements in diagnostic imaging and the popularity of bariatric surgery, gastric lipoma (GL), a rare benign mesenchymal tumor, is becoming more widely recognized. Though open surgery has been the conventional method of management, the therapeutic scenario has changed considerably in the past ten years. Objective. This scoping review intends to delineate the clinical features, diagnostic findings, and the management trends for gastric lipomas that were documented between 2016 and 2025. Methods. A systematic search was carried out across many databases in accordance with PRISMA-ScR criteria. Eligible studies included case reports and series focusing on benign GLs in adult patients. Data extraction focused on demographics, symptomatology, tumor size, diagnostic modalities, and surgical outcomes. Results. A total of 33 articles describing 35 cases were included. For cases with available demographic data, the median age was 63 years (range: 17–85), with a 2.5:1 male predominance. The most common presentation was gastrointestinal bleeding (68.6%), followed by gastric outlet obstruction (20.0%). Larger tumor size was generally related to symptomatic presentation in the reviewed literature; notably, all reported “giant” lesions (>4 cm) were symptomatic. The most common diagnostic methods were esophagogastroduodenoscopy (91.4%) and CT (80.0%). There was an apparent shift toward less invasive treatments, such as laparoscopic (20.0%), robotic (5.7%), and endoscopic (17.1%) procedures, all of which had successful results. Conclusion. In published cases, gastric lipomas >4 cm were predominantly symptomatic and necessitated intervention. There is a clear trend toward minimally invasive and parenchyma-sparing techniques. The absence of long-term follow-up data, however, continues to be a major knowledge gap. Furthermore, the development of prospective multi-institutional registries is crucial to better define the natural history and intervention thresholds for this disorder, since asymptomatic giant lesions are probably underestimated because of publication bias.

Šis darbas apsaugotas Creative Commons priskyrimo 4.0 viešąja licencija.