Objective: The aim of this meta-analysis was to assess the diagnostic performance of EUS-FNA/B in patients with panNETs.

Methods: We conducted a computerized search of the MEDLINE and Embase databases to identify relevant articles. The primary outcomes involved grading concordance rate, diagnostic rate, and correlation coefficient (Cohen's κ) for FNA/B samples compared with surgical specimens. Secondary outcomes included sample adequacy, mean number of passes, and adverse events.

Results: Forty-five studies involving 2978 patients were finally included. The pooled concordance rate between EUS-FNA/B and surgical grading was 0.77 (95% CI: 0.73-0.80; I2 =48.2%). A significantly higher level of concordance was observed in G1 subgroup (0.88, 95% CI: 0.84-0.91), whereas the G2 subgroup revealed the lowest level of agreement (0.59, 95% CI: 0.52-0.65; P < 0.001). Pooled diagnostic rate for FNA/B sampling was 0.83 (95% CI: 0.79-0.86; I2 =63.3%). In addition, FNB outperformed FNA in terms of sample adequacy (0.93 for FNB vs. 0.81 for FNA; P =0.007) and number of needle passes required (2.53 for FNB vs. 3.32 for FNA; P =0.013). Moreover, the overall level of agreement for grading was moderate (κ=0.59, 95% CI: 0.49-0.68; I2 =84.5%). There were a limited number of adverse events that had minor influence on patient outcomes (0.03, 95% CI: 0.02-0.05; I2 =19.2%).

Conclusions: EUS-FNA/B is a reliable approach for the diagnosis and preoperative grading of panNET, with FNB demonstrating superior performance compared with FNA.