Vascular Tumor Burden as a New Quantitative CT Biomarker for Predicting Metastatic RCC Response to Antiangiogenic Therapy

Purpose To quantify initial changes in the vascular tumor burden (VTB), a measure of the area of vascularized tumor on computed tomographic (CT) images, and predict tumor response to antiangiogenic therapy in patients with metastatic renal cell carcinoma (RCC). Materials and Methods For this institutional review board-approved HIPAA-compliant secondary analysis of a prospective phase III trial, adult patients with digital CT images and metastatic clear-cell RCC treated with sunitinib were included (n = 275). Target lesions were selected by using Response Evaluation Criteria in Solid Tumors (RECIST) 1.1 guidelines, and the CT images obtained after one cycle of sunitinib therapy were evaluated in comparison with baseline images. Tumor-response software was created to quantify tumor metrics (length, area, VTB, and mean attenuation) and automate response assessment. Progression-free survival (PFS) in responders and nonresponders according to VTB criteria was compared by using the Cox proportional hazard ratio (HR). The intraclass correlation coefficient (ICC) was used to assess interobserver agreement among three readers evaluating 28 randomly selected patients. Results VTB criteria nonresponders (n = 120) according to the initial posttherapy CT study were 5.7 times more likely to experience progression of disease (HR = 5.70; 95% confidence interval [CI]: 4.07, 7.97; P < .001) than responders (n = 155). There was not a statistically significant difference in PFS between RECIST nonresponders (n = 255) and responders (n = 20; HR = 1.54; 95% CI: 0.85, 2.77; P = .148). In a patient-level analysis, interobserver agreement was very good for assessing percentage change in length, area, and VTB (ICC = 0.82 [95% CI: 0.67, 0.91], 0.89 [95% CI: 0.79, 0.94], and 0.88 [95% CI: 0.79, 0.94], respectively) but was very poor for assessing percentage change in mean attenuation (ICC = 0.17 [95% CI: -0.05, 0.45]). Conclusion A quantitative CT imaging biomarker designed to measure initial changes in the VTB separated patients into responders and nonresponders, each with significantly different PFS, and showed very good interobserver agreement in patients with metastatic RCC treated with sunitinib. © RSNA, 2016 Online supplemental material is available for this article.