Premium
Contrast‐Enhanced Ultrasound Classification of Previously Indeterminate Renal Lesions
Author(s) -
Zarzour Jessica G.,
Lockhart Mark E.,
West Janelle,
Turner Eric,
Jackson Bradford E.,
Thomas John V.,
Robbin Michelle L.
Publication year - 2017
Publication title -
journal of ultrasound in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.574
H-Index - 91
eISSN - 1550-9613
pISSN - 0278-4297
DOI - 10.1002/jum.14208
Subject(s) - medicine , indeterminate , confidence interval , contrast (vision) , magnetic resonance imaging , radiology , contrast enhanced ultrasound , ultrasound , lesion , retrospective cohort study , nuclear medicine , predictive value of tests , predictive value , pathology , mathematics , artificial intelligence , pure mathematics , computer science
Objectives To determine the utility of contrast‐enhanced ultrasound (US) for characterizing renal lesions that were indeterminate on prior imaging. Methods This Institutional Review Board–approved retrospective diagnostic accuracy study evaluated all patients who underwent renal contrast‐enhanced US examinations from 2006 to 2015 at our tertiary care hospital. We compared the number of lesions definitively characterized by contrast‐enhanced US with the indeterminate lesions by prior imaging. The accuracy of contrast‐enhanced US was compared with the final diagnosis by histologic examination and follow‐up (mean, 3.63 years). Accuracy and agreement estimates were compared with the exact binomial distribution to assess statistical significance. Results A total of 134 lesions were evaluated with contrast‐enhanced US, and 106 were indeterminate by preceding computed tomography, magnetic resonance imaging, or US. Only the largest lesion per patient was included in analysis. A total of 95.7% (90 of 94) of the previously indeterminate lesions were successfully classified with contrast‐enhanced US. The sensitivity was 100% (20 of 20; 95% confidence interval [CI], 83%–100%; P < .0001); specificity was 85.7% (18 of 21; 95% CI, 62%–97%; P = .0026); positive predictive value was 87.0% (20 of 23; 95% CI, 66%–97%; P = .0005); negative predictive value was 100% (18 of 18; 95% CI, 81%–100%; P < .001); and accuracy was 90.2% (37 of 41; 95% CI, 80%–98%; P < .0001). Conclusions Contrast‐enhanced US has a high likelihood of definitively classifying a renal lesion that is indeterminate by computed tomography, magnetic resonance imaging, or conventional US.