AUA 2020: Planned Secondary Analysis of PURE-01: Role of FDG-PET/CT in Evaluating Lymph Node Involvement of Patients with Muscle-Invasive Bladder Cancer Receiving Neoadjuvant Pembrolizumab and Radical Cystectomy

(UroToday.com) In patients with muscle-invasive bladder cancer, pre-operative staging is typically performed with computed tomography. Prior work has demonstrated that in patients without clinical evidence of nodal disease based on conventional imaging, the use of fluorodeoxyglucose-positron emission tomography (FDG-PET)/computed tomography (CT) prior to neoadjuvant chemotherapy and radical cystectomy (RC) or RC alone has limited utility. In a podium presentation at the American Urologic Association (AUA) 2020 Virtual Annual Meeting, Laura Marandino, MD, Andrea Necchi, MD, and colleagues examined a secondary analysis of their prior study specifically aimed at evaluating the role of PET/CT in lymph node (LN) involvement assessment.


To do so, the authors utilized a secondary analysis of patients enrolled in the PURE-01 study: Neoadjuvant Pembrolizumab for Muscle-invasive Urothelial Bladder Carcinoma. To briefly summarize, in PURE-01 (NCT02736266), patients received 3 courses of 200 mg pembrolizumab, every 3 weeks, prior to RC. Imaging assessment was performed using a thorax-abdomen CT scan and with PET/CT scan during screening and after treatment, before RC. Imaging review and analysis was performed by two experienced nuclear medicine physicians blinded to clinical information with semiquantitative and volumetric analysis performed.

AUA2020_PURE-01.png



For each patient with identified increased activity in abdominopelvic lymph nodes, the SUVmax and the short-axis size of the most intense lymph node were recorded. All patients underwent templated pelvic LN dissection (LND) with packeted node submission at the time of radical cystectomy.

In this analysis, the authors assessed PET/TC diagnostic ability using sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and accuracy with surgical pathology considered the gold standard.

Between February 2017 and June 2019, 114 patients were enrolled and treated in PURE-01. 11 of these patients received additional chemotherapy post-pembrolizumab and were excluded from the analyses, resulting in 103 total evaluable patients for whom there were a total of 206 PET/CT scans.

Six pts (5.8%) had positive nodes at baseline PET/CT with a mean SUVmax of 2.75 and a mean short axis length of 6.2 mm. Following neoadjuvant pembrolizumab, eight patients (7.8%) had positive LN at PET/CT with a mean SUVmax of 4.21 and a mean short axis length of 7.2mm.

The rate of pathologic LN positive (pN+) disease was 15.5% (16 patients). The performance of PET/CT in predicting pN+ disease is indicated in the table below:

AUA2020_SUVmax.png

In short, PET/CT has relatively poor sensitivity but quite high specificity. While predictive values are sensitive to disease prevalence, in this study population positive predictive value was 75% while negative predictive value was 89.5%.

4 of 6 patients (66.7%) with baseline FDG uptake were found to have nodal positive disease at the time of LND while 12 of 97 (12.4%) with no baseline FDG uptake had node-positive disease (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%).

Presented by: Laura Marandino, MD, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy; Andrea Necchi, MD, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy

Written by: Christopher J.D. Wallis, MD, PhD, Urologic Oncology Fellow, Vanderbilt University Medical Center, Nashville, TN, USA, Twitter: @WallisCJD, at the 2020 American Urological Association (AUA) Annual Meeting, Virtual Experience #AUA20, June 27- 28, 2020.
email news signup