Khaled Y Elbanna1, Hyun-Jung Jang1, Tae Kyoung Kim1, Korosh Khalili1, Luís S Guimarães1, Mostafa Atri2,3. 1. Toronto Joint Department of Medical Imaging, University Health Network, Sinai Health System and Women's College Hospital, University of Toronto, Toronto, ON, Canada. 2. Toronto Joint Department of Medical Imaging, University Health Network, Sinai Health System and Women's College Hospital, University of Toronto, Toronto, ON, Canada. Mostafa.Atri@uhn.ca. 3. Department of Medical Imaging, Toronto General Hospital, University of Toronto, 585 University Avenue, Toronto, ON M5G 2N2, Canada. Mostafa.Atri@uhn.ca.
Abstract
OBJECTIVES: To investigate accuracy of contrast-enhanced ultrasound (CEUS) to characterize indeterminate small solid renal masses (sSRMs), excluding lipid-rich AMLs, and cystic renal masses (CRMs) according to the proposed Bosniak Classification 2019 MATERIALS AND METHODS: CEUS of pathology-proven CRMs and sSRMs (without definite enhancement or macroscopic fat on CT/MRI), and CRMs with ≥18 months follow-up were retrospectively reviewed. Two radiologists blindly categorized CRMs according to new Bosniak Classification on CT/MRI. On CEUS, two other radiologists evaluated arterial-phase enhancement of sSRMs relative to renal cortex and categorized CRMs following new Bosniak Classification. Fisher's exact/chi-squared test was used to compare categorical variables, and Cohen κ statistics for inter-observer agreement RESULTS: A total of 237 patients had 241 lesions: 161 pathology-proven sSRMs (122 malignant and 39 benign), 29 pathology-proven CRMs, 51 CRMs with adequate follow-up. Arterial-phase enhancement < renal cortex predicted malignancy with specificity of 97.4% (38/39) (CI 85.6-99.9%), and positive predictive value (PPV) of 98.2% (54/55) (CI 90.4-99.9%). Inter-observer kappa was 0.95. In pathology-proven CRMS, sensitivity of CEUS vs CT/MRI was 100% (15/15) (CI 79.6-100%) vs 60% (9/15) (CI 35.8-80.1%) (p value = .002) and negative predictive value (NPV) 100% (2/2) (CI 17.8-100%) vs 25% (2/8 ) (CI 4.4-59.1%) (p value < 0.0001), with similar specificity (50%) and PPV- 88.2% (15/17) (CI 65.7-97.9%) vs 81.8% (9/11) (CI 52.3-96.8%) ( p value = 0.586). Bosniak Classification inter-observer kappa was 0.92 for CEUS vs 0.68 for CT/MRI (p value = 0.009). CONCLUSION: In our cohort, CEUS had high specificity and PPV to diagnose RCC in sSRMs excluding lipid-rich AML. CEUS had significantly higher sensitivity/NPV to diagnose malignancy in CRMs as compared to CT/MRI. KEY POINTS: • Once lipid-rich AML is excluded by the other modalities, sSRM arterial phase hypo-enhancement relative to renal cortex on CEUS yielded high specificity (97.4%) and PPV (98.2%) to diagnose RCC. • When applying the proposed Bosniak Classification 2019, CEUS showed higher sensitivity compared to CT/MRI (100% vs 60%), p value=.0024, in the stratification of cystic renal masses to diagnose malignancy. • CEUS may reduce the number of CT/MRI Bosniak IIF lesions by assigning them to either II or III/IV categories.
OBJECTIVES: To investigate accuracy of contrast-enhanced ultrasound (CEUS) to characterize indeterminate small solid renal masses (sSRMs), excluding lipid-rich AMLs, and cystic renal masses (CRMs) according to the proposed Bosniak Classification 2019 MATERIALS AND METHODS:CEUS of pathology-proven CRMs and sSRMs (without definite enhancement or macroscopic fat on CT/MRI), and CRMs with ≥18 months follow-up were retrospectively reviewed. Two radiologists blindly categorized CRMs according to new Bosniak Classification on CT/MRI. On CEUS, two other radiologists evaluated arterial-phase enhancement of sSRMs relative to renal cortex and categorized CRMs following new Bosniak Classification. Fisher's exact/chi-squared test was used to compare categorical variables, and Cohen κ statistics for inter-observer agreement RESULTS: A total of 237 patients had 241 lesions: 161 pathology-proven sSRMs (122 malignant and 39 benign), 29 pathology-proven CRMs, 51 CRMs with adequate follow-up. Arterial-phase enhancement < renal cortex predicted malignancy with specificity of 97.4% (38/39) (CI 85.6-99.9%), and positive predictive value (PPV) of 98.2% (54/55) (CI 90.4-99.9%). Inter-observer kappa was 0.95. In pathology-proven CRMS, sensitivity of CEUS vs CT/MRI was 100% (15/15) (CI 79.6-100%) vs 60% (9/15) (CI 35.8-80.1%) (p value = .002) and negative predictive value (NPV) 100% (2/2) (CI 17.8-100%) vs 25% (2/8 ) (CI 4.4-59.1%) (p value < 0.0001), with similar specificity (50%) and PPV- 88.2% (15/17) (CI 65.7-97.9%) vs 81.8% (9/11) (CI 52.3-96.8%) ( p value = 0.586). Bosniak Classification inter-observer kappa was 0.92 for CEUS vs 0.68 for CT/MRI (p value = 0.009). CONCLUSION: In our cohort, CEUS had high specificity and PPV to diagnose RCC in sSRMs excluding lipid-rich AML. CEUS had significantly higher sensitivity/NPV to diagnose malignancy in CRMs as compared to CT/MRI. KEY POINTS: • Once lipid-rich AML is excluded by the other modalities, sSRM arterial phase hypo-enhancement relative to renal cortex on CEUS yielded high specificity (97.4%) and PPV (98.2%) to diagnose RCC. • When applying the proposed Bosniak Classification 2019, CEUS showed higher sensitivity compared to CT/MRI (100% vs 60%), p value=.0024, in the stratification of cystic renal masses to diagnose malignancy. • CEUS may reduce the number of CT/MRI Bosniak IIF lesions by assigning them to either II or III/IV categories.
Authors: Stuart G Silverman; Ivan Pedrosa; James H Ellis; Nicole M Hindman; Nicola Schieda; Andrew D Smith; Erick M Remer; Atul B Shinagare; Nicole E Curci; Steven S Raman; Shane A Wells; Samuel D Kaffenberger; Zhen J Wang; Hersh Chandarana; Matthew S Davenport Journal: Radiology Date: 2019-06-18 Impact factor: 11.105