Choi response criteria for prediction of clinical outcome in patients with metastatic renal cell cancer treated with sunitinib

5044 Background: Sunitinib (SU) has been approved for the treatment of metastatic renal cell cancer (mRCC). Since SU can induce extensive necrosis, RECIST may be inappropriate for tumor response evaluation. We evaluated whether the new Choi criteria (J Clin Oncol. 2007;25:1753–1759) are of additional value to predict outcome in mRCC patients (pts) treated with SU. Methods: 56 mRCC pts treated with SU were included. Imaging data consisted of thoracic and abdominal helical CT scans at baseline, after a median of 2 months and were repeated during treatment. For Choi criteria the longest diameter was ≥15 mm. Density of these lesions was determined in Hounsfield units. According to Choi criteria partial response (PR) was defined as ≥10% decrease in size or ≥15% decrease in density, while progressive disease (PD) was defined as ≥10% increase in size without meeting PR criteria by density. Progression-free survival (PFS) and overall survival (OS) were calculated according to Kaplan-Meier. Log rank test was used to test the statistical difference between survival curves. Results: For RECIST and Choi criteria, respectively, 230 and 156 tumor lesions were eligible. At first evaluation, according to RECIST 7 pts had PR, 39 stable disease (SD), and 10 PD, while according to Choi criteria 33 pts had PR, 8 SD and 15 PD. The median tumor density decreased significantly (Wilcoxon Signed Ranks test, p ≤ 0.001). At first evaluation in patients with PR, Choi criteria had a significantly better predictive value for PFS and OS (p < 0.001 and p < 0.001, respectively) than RECIST (p = 0.384 and p = 0.392, respectively). When best response during treatment was analyzed according to RECIST, the predictive value of RECIST increased for both PFS and OS (p = 0.004 and p = 0.002, respectively). For clinical benefit (PR+SD), the predictive value of RECIST and Choi criteria for PFS and OS were comparable (both p < 0.001). Conclusions: Choi criteria can be easily applied on contrast-enhanced CT scans. RECIST and Choi criteria have similar predictive value for outcome in pts with clinical benefit. Choi criteria, however, are more useful to early define a large pt population with favourable clinical outcome. [Table: see text]