e15047 Background: The oral mTOR inhibitor, everolimus, affects tumor growth by targeting cellular metabolic proliferation pathways and delays RCC progression. Yet most patients (pts) will not respond. Preclinical evidence suggests that elevated tumor glucose uptake as quantified by FDG-PET may predict anti-tumor activity. METHODS In this multicenter study, mRCC pts refractory to VEGF inhibition were treated with standard dose everolimus. FDG-PET scans were obtained at baseline and at 1 month; serial CT scans were obtained at baseline and every 8 weeks. Maximum standardized uptake value (SUVmax) of the most FDG avid lesion, the average SUVmax of all measured lesions and the corresponding one-month relative change were examined for association with 8-week change in tumor size. 60 pts were planned to provide 81% power (α=0.1) to detect a difference between a mean 20% vs. 10% increase in RECIST based tumor burden measurement in pts with low (SUVmax≤4) vs. high (SUVmax>4) FDG uptake. RESULTS 63 pts were enrolled; 51 were evaluable for primary endpoint of which 48 had both PET scans. Pt characteristics included: 46 (73%) clear cell histology, median age 61 (32 - 80), median number of prior systemic therapies 1.5 (1 - 4). Median pre- and post-treatment average SUVmax were 7 (1 -17.9) and 4.2 (1 - 13.9), respectively. Median relative change in average SUVmax was -21.2% (-59.8% - 0%). RECIST-based measurements demonstrated an average change in tumor burden of 0.3% (-32.0% - 35.9%) at 8 weeks, with 2 patients (4%) experiencing a PR. Among SUV measures considered, relative change in average SUVmax was the best predictor of change in tumor burden (all evaluable p=0.013; clear cell subtype p=0.031), but the overall correlation was modest (ρ = 0.36). Baseline SUV characteristics and low vs. high SUV classification were not associated with tumor response. Evaluation of progression and overall survival is pending. CONCLUSIONS Systemic therapy with everolimus substantially decreased SUVs on follow-up PET scans in mRCC pts, but these changes were only modestly correlated with changes in tumor size. Correlation with survival metrics will be assessed. Clinical use of FDG-PET based biomarkers is challenged by high variability.