CEUS-guided percutaneous core needle biopsy for peritoneal lesions: A 10-year retrospective study on initial...

To evaluate the diagnostic yield of contrast-enhanced ultrasound (CEUS)-guided percutaneous core needle biopsy (CNB) for peritoneal lesions and to identify factors associated with the initial diagnosis accuracy.

A total of 253 consecutive patients with peritoneal lesions who underwent CEUS-guided percutaneous CNB from January 2015 to March 2025 were retrospectively enrolled. Based on whether the initial pathological diagnosis matched the final diagnosis, patients were classified into an accurate initial diagnosis group (n = 239) and an inaccurate initial diagnosis group (n = 14). The final diagnosis was determined by integrating histopathological findings, multi-modality imaging, and clinical follow-up data. Baseline, conventional ultrasound (US), CEUS, and CNB-related characteristics were compared. The overall technical success rate, initial diagnostic accuracy, and complications of CEUS-guided CNB were evaluated. Factors affecting diagnostic accuracy were identified via univariate and multivariate analyses. Within subgroups stratified by the median maximum anteroposterior diameter (MAD), CNB-related characteristics were compared.

The study achieved 100% (253/253) technical success, with an initial diagnostic accuracy of 94.5% (239/253). The overall complication rate was 9.1% (23/253), predominantly minor, with no visceral organ injury or needle tract seeding occurred. A prior history of cancer and a trans-ascitic approach were independent factors affecting initial accuracy (P < 0.05). Comparative analysis revealed statistically significant differences between the larger (MAD > 21 mm) and smaller (MAD ≤ 21 mm) lesion subgroups in five parameters: number of cores, diagnosis, lesion type, echo pattern, and arterial phase enhancement pattern (P < 0.05).

CEUS-guided percutaneous CNB is a safe, feasible, and highly accurate technique for diagnosing peritoneal lesions. Attention to prior tumor history and the selection of a non‑trans‑ascitic approach are important for improving initial diagnostic accuracy.