Assessing local invasion in cervical cancer: the diagnostic performance of transrectal contrast-enhanced ultrasonography
Abstract
Objectives
To evaluate the diagnostic performance of transrectal ultrasound (TR-US) combined with contrast-enhanced TR-US (TR-CEUS) against MRI and histopathology for assessing cervical cancer local invasion, and to validate the “rim sign” as a biomarker for parametrial exclusion.
Methods
A retrospective study of 69 cervical cancer patients (54 surgical, 15 neoadjuvant therapy recipients) undergoing TR-US, TR-CEUS, and MRI was conducted. Tumour dimensions, vaginal/parametrial invasion, and the “rim sign” were evaluated. Diagnostic agreement was assessed using intraclass correlation coefficients (ICC) and Cohen’ s κ, while the specificity and negative predictive value (NPV) of the “rim sign” were calculated.
Results
TR-CEUS improved TR-US reliability in tumour measurement (ICC = 0.751 to 0.784), achieving MRI-comparable accuracy (ICC = 0.804 vs. histopathology). Combined TR-US/TR-CEUS enhanced interobserver consistency for vaginal infiltration (κ = 0.621 to 0.694) and parametrial invasion (κ = 0.579 to 0.678) compared to TR-US alone. TR-US/TR-CEUS showed good agreement with histopathology (κ = 0.672) and MRI (κ = 0.789) for parametrial assessment than TR-US (κ = 0.563/0.679). The “rim sign” demonstrated 95.8% specificity and 92.0% NPV for parametrial exclusion.
Conclusions
TR-US/TR-CEUS achieves diagnostic accuracy comparable to MRI while improving observer consistency. The “rim sign” serves as a high-specificity biomarker for preoperative parametrial exclusion, offering a cost-effective alternative to MRI.
Advances in knowledge
This study provides quantitative evidence of TR-CEUS improving diagnostic reproducibility for cervical cancer invasion. The “rim sign” is identified as a novel imaging biomarker with high specificity for parametrial preservation.