May 11C-Choline PET/CT predict tumor response after radiotherapy on lymph node recurrence of prostate cancer patients?

513 Objectives The aim was to investigate the predictive role of 11C-Choline PET/CT (PET/CT) parameters in patients (pts) previously treated with helical tomotherapy (HTT) on lymph node (LN) relapses. Methods This retrospective study include 68 pts with PCa (mean age: 68 yrs; range: 51-81 yrs) with biochemical recurrence after radical treatment (median PSA: 2.42 ng/ml; range: 0.61-27.56 ng/ml) who underwent PET/CT from January 2005 to January 2013. Best cut-off values of PET parameters (SUVmax, SUVmean, MTV with a threshold of 40%-50%-60%) discriminating between patients with/without relapses were assessed by ROC analysis: OS, lRFS, cRFS and bRFS were considered. Univariate and multivariate Cox regression analysis including the most predictive PET parameters adjusted for standard clinical parameters (age, PSA, Gleason score at initial diagnosis, primary treatment, PET positive LN site) were performed. Results PET/CT showed pathologic LNs at pelvic level (n=4), at abdominal level (n=5), at both pelvic and abdominal (n=13), at abdominal and/or pelvic and/or other sites (n=46). The 2-year OS, lRFS, cRFS, bRFS were 86.7%, 91.4%, 51.5% and 40.0%, respectively. The most significant best cut-off, based on the AUC of the ROC curves, was MTV60%>0.64cc that confirmed its independent predictive role in multivariate analysis. A two variable model including MTV60% and PET positive LN site (inside vs outside pelvis) may well predict the risk of clinical relapses (p Conclusions PET parameters are predictive of tumor response of PCa pts treated with HTT at LN relapses. This information may be useful in assessing the comparative effectiveness of various conventional and emerging treatment strategies, and to determine their ability to stratify pts in clinical trials.