Accurate loco-regional staging is essential in the management of gastric cancer, guiding decisions regarding surgical resection, neoadjuvant therapy, and endoscopic interventions. Endoscopic ultrasound (EUS) is widely used for evaluating tumour depth (T stage) and regional lymph node involvement (N stage), but reported diagnostic performance varies across studies. This systematic review and meta-analysis aimed to synthesize evidence on the accuracy of EUS for staging gastric cancer.

A comprehensive literature search was performed in MEDLINE, Embase and Cochrane Library until May 27, 2025. We identified primary studies reporting the diagnostic performance of conventional EUS (linear or radial) for gastric cancer staging compared with histopathology as the reference standard. We excluded studies using miniprobe or non-conventional EUS techniques, as well as patients who received neoadjuvant radiochemotherapy. Pooled sensitivity, specificity, PPV, NPV, accuracy, overstaging and understaging for T and N staging were calculated using a bivariate random-effects model. Heterogeneity was assessed with the I² statistic.

Nineteen studies comprising 2,982 patients were included for T-stage evaluation, and 15 studies comprising 1,272 patients for N-stage evaluation. Of the nine retrospective and ten prospective studies, most used a radial probe (n=16) with only two using linear and one using both. The accuracy rate of EUS for overall T stage was 72% with 14% overstaging and 13% understaging. For early gastric cancer (T1), diagnostic performance showed high specificity (0.95; 95% CI 0.93–0.97) but modest sensitivity (0.76; 95% CI 0.66–0.85), with an accuracy of 0.89 (95% CI 0.86–0.92), with 24% overstaging. T2 lesions showed lower sensitivity (0.64; 95% CI 0.53–0.75) and specificity (0.89; 95% CI 0.85–0.92), respectively, with 25% overstaging and 8% understaging. T3 disease had the highest rate of accurate T staging at 81%, while T4 staging had the lowest at 57%. For nodal disease, diagnostic performance was moderate with accuracy of 0.73 (95% CI 0.67–0.78), sensitivity 0.69 (95% CI 0.59–0.78) and specificity 0.82 (95% CI 0.74–0.88), with overstaging and understaging of rates of 8% and 20%, respectively.

Conventional EUS provides robust diagnostic performance for T staging in patients with gastric cancer. This is particularly useful to select patients with early gastric cancer who are candidates for endoscopic or surgical resection. However, there remains a significant risk of overstaging, which would suggest the role of EUS as a tool within a multimodal staging strategy. Performance of nodal staging is moderate, with a modest risk of missing nodal disease. These highlight the potential value of adjunctive modalities, such as miniprobe or contrast-enhanced EUS, or techniques such as EUS-guided fine-needle biopsy to enhance diagnostic yield.