Background: FDG-PET/CT has limited utility in clinical N0 (cN0) patients (pts) with MIBC who receive neoadjuvant chemotherapy and RC or RC alone (Dason, AUA19). Methods: In PURE-01 (NCT02736266), 3 courses of 200 mg pembrolizumab, q3 weeks, were administered prior to RC. Pts were assessed with thorax-abdomen CT scan and with PET/CT scan during screening and before RC. Imaging review and analysis was internally performed. For each pt with lymph node (LN) increased uptake in abdomino-pelvic area, the SUVmax and the short-axis size of the most intense LN were recorded. All pts underwent extended pelvic LN dissection (LND) with packeted node submission. Results: From 02/17 to 06/2019, 103 total evaluable pts (206 PET/TC scans) were enrolled ad treated. Six pts (5.8%) had LN uptake at baseline PET/CT: mean SUVmax=2.75; mean short axis: 6.2 mm. Eight pts (7.8%) had LN uptake at PET/CT post-pembrolizumab: mean SUVmax=4.21; mean short axis: 7.2mm. The rate of pathologic LN positive (pN+) disease was 15.5% (16 pts). The performance of post-pembrolizumab PET/CT in predicting pN+ disease is indicated in the Table. In total, 4/6 pts (66.7%) with baseline FDG uptake revealed as pN+ vs 12/97 (12.4%) with no baseline FDG uptakes (p=0.005). A total of 39 pts (37.9%) developed inflammatory FDG-uptakes post-pembrolizumab in several target organs/regions: top 5 sites were thyroid (N=21, 61.8%), stomach and mediastinum (13 pts each, 12.6%), lung (N=10, 9.7%), other lymph nodes (N=4, 3.9%). These changes were clinically evident (signs/symptoms or laboratory changes) in 15 pts (38.5%). Conclusions: Criteria for eligibility of cN0 pts to single-agent neoadjuvant pembrolizumab trials may be enhanced with PET/CT use. Three cycles of pembrolizumab determined profound inflammatory changes, whose long-term impact on safety is still to be determined. Clinical trial information: NCT02736266.

Abbreviations: CI: confidence interval; NPV: negative predictive value; PPV: positive predictive value