PURPOSE: We investigated the accuracy of FDG-PET/CT response identification following neoadjuvant or induction chemotherapy (NAIC) for invasive bladder cancer (BC) as to better select patients for radical cystectomy (RC).
METHODS: Between 2010 and 2014, 37 cT1-4N1-3 BC patients received a FDG-PET/CT before and after NAIC followed by RC. Metabolic lymph node (LN) response was evaluated according to EORTC recommendations. Additionally, primary tumor response was evaluated for 23 patients by means of delayed pelvic imaging after forced diuresis. Gold standard was response on pathologic analysis of RC specimens. Response was defined as partial response (pPR, any pathologic downstaging) or complete response (pCR, <ypT1N0). Cancer-specific survival (CSS) was correlated with pCR and metabolic CR.
RESULTS: Twenty-four cN+ patients achieved LN pCR. FDG-PET/CT identified pCR with 67% sensitivity and 46% specificity. Primary tumor response was evaluable for 17/23 patients, of whom 12 were responders. Tumor downstaging (pPR or pCR) was identified with 83% sensitivity and 80% specificity. Tumor pCR was detected with 70% sensitivity and 71% specificity. pCR of overall disease (primary tumor and LNs, n = 17) was detected with 67% sensitivity and 75% specificity. CSS was positively associated with pathologic CR (HR 0.16, p = 0.027), but not with metabolic CR (HR 0.560, p = 0.612).
CONCLUSION: FDG-PET/CT can accurately distinguish primary tumor downstaging from non-response, which suggests that response monitoring (in cN0 patients) might be used to adjust neoadjuvant treatment. pCR identification is less accurate, especially for LN metastases, which suggests that FDG-PET/CT cannot be used to select patients for RC.