Liquid Biopsy for Molecular Residual Disease Detection and Postoperative Surveillance in Gastric Cancer: Current Evidence and Future Directions

Gastric cancer remains a major cause of cancer mortality worldwide, mainly due to its frequent diagnosis at advanced stages and the high probability of recurrence even after curative treatment. Conventional postoperative follow-up is mainly based on imaging studies, endoscopy and serological tumor markers, such as carcinoembryonic antigen (CEA), carbohydrate antigen 19-9 (CA19-9), and carbohydrate antigen 72-4 (CA72-4)This narrative review was based on a structured literature search of PubMed, Scopus, Web of Science, MEDLINE, the Cochrane Library, and ClinicalTrials.gov from database inception through June 2026, with 75 studies included in the final narrative synthesis. Among these analytes, circulating tumor DNA (ctDNA)currently provides the most mature data for the detection of molecular residual disease and postoperative risk stratification. Although it allows for the early detection of recurrent or residual disease prior to imaging confirmation, postoperative ctDNA has shown significant prognostic value in many studies; however, routine use as a basis for decision-making in treatment planning is still considered investigational and will require future prospective clinical validation. Tumor-informed ctDNA approaches offer high analytical specificity and sensitivity in low-burden disease settings. In contrast, tumor-agnostic approaches, such as methylation analysis and fragmentomics, may improve the scalability of the method. However, they require further validation in the postoperative setting. At the same time, emerging data indicate that extracellular vesicles, exosomal RNA and peritoneal lavage analytes can provide complementary biological information, especially in cases of peritoneal dissemination. Despite the significant prospects, the use of liquid biopsy in guiding the treatment of gastric cancer remains under investigation. This is because even today there are limitations. Characteristic are the low ctDNA excretion and the anatomical heterogeneity of the disease, as well as clonal hematopoiesis. Limitations also include the lack of standardization as well as the cost and the need for prospective clinical studies. This review summarizes the biological basis of molecular residual disease, liquid biopsy technologies, ctDNA data, the concept of molecular recurrence, and the future prospects of multi-analytic and artificial intelligence (AI)-assisted surveillance models in gastric cancer.