1705. Circularity Index Measurement on Preoperative Diagnostic Imaging to Distinguish Angiomyolipoma from Renal Cell Carcinoma
Authors * Denotes Presenting Author
  1. Laura Geldmaker; Mayo Clinic - Urology
  2. Lauren Alexander *; Mayo Clinic - Radiology
  3. Bryce Baird; Mayo Clinic - Urology
  4. Mikolaj Wieczoreku; Mayo Clinic - Clinical Trials and Biostatistics
  5. Colleen Ball; Mayo Clinic - Clinical Trials and Biostatistics
  6. Joseph Cernigliaro; Mayo Clinic - Radiology
  7. David Thiel; Mayo Clinic - Urology
Attempts to aid in differentiating solid renal masses through advanced CT and MR techniques, and radiomics features often require specific equipment, software or postprocessing analytics. Studies have included shape as a qualitative feature, but variable and subjective definitions make this feature difficult to standardize across studies. Kang and Park used mass area and perimeter measured on CT to calculate circularity, or how closely mass shape aligns to a perfect circle. Features that deviate from roundness (lobular or irregular margins) lower this value. This study evaluates circularity index to distinguish AML, RCC, and oncocytoma.

Materials and Methods:
An IRB-approved retrospective review of 517 patients with renal mass removed by robotic-assisted partial nephrectomy (RAPN) between February 2008 and December 2020, excluding masses > 4 cm identified 17 AMLs to case match by size on preoperative imaging (within 1 cm of AML size) and modality (CT to CT, MRI to MRI). Patient data were collected: age, sex, body mass index (BMI), Mayo Adhesive Probability (MAP) Score, and RENAL nephrotomy score. Circularity index was calculated as 4 x pi x (tumor area ÷ tumor perimeter2) using PACS measurement tool to outline mass on contiguous axial images (maximum mass perimeter in mm, area in mm2 and volume in mm3). Numeric variables summarized with sample median and range. Categorical variables were summarized with frequency and percentage. Area under the receiver operating characteristic curve (AUROCC) was used to evaluate ability of circularity index to distinguish between mass types. Wilcoxon rank sum test for comparison of circularity index between pathology types. Statistical analysis performed with R version 4.0.3 (R Foundation for Statistical Computing, Vienna Austria).

68 RAPN cases: papillary RCC (pRCC) (N=17), clear cell RCC (ccRCC) (N=17), & oncocytoma (N=17) case matched to AML. Median mass diameter: 2.0 cm AML, 2.6 cm ccRCC, 2.2 cm pRCC, & 2.2 cm oncocytoma. Median circularity index: AML = 0.833 (range: 0.660-0.947); ccRCC = 0.869 (range: 0.708-0.942), pRCC = 0.888 (range: 0.725-0.957), oncocytoma = 0.848 (range: 0.681-0.948). Circularity index AUROCC 0.678 for AML vs RCC (95% CI 0.519-0.834, p=0.04), but not oncocytoma from AML (AUROCC 0.519, range 0.322-0.716, p=0.85), oncocytoma from RCC (AUROCC 0.609, range 0.422-0.780, p=0.21), or between ccRCC & pRCC (AUROCC 0.574, range 0.384-0.768, p=0.46).

The circularity index has potential to discriminate between AML and RCC and can be calculated with tools existing in a PACS system without specific imaging equipment, scan protocol, or additional software. No significant differences were found for circularity index to discriminate oncocytoma from either AML or RCC or between ccRCC and pRCC. This quantitative measure of mass shape differences between AML and RCC may help to distinguish these solid renal masses on preoperative imaging in combination with other features or when other clear distinguishing features are lacking.