A Consensus Study on Invasiveness of Bladder Cancer Using Virtual Microscopy and Heatmaps (pT1 European Network of Uropathology (ENUP) Study)
Eva Comperat, Lars Egevad, Antonio Lopez-Beltran, Ferran Algaba, Mahul Amin, Jonathan Epstein, Hans Hamberg, Christina Hulsbergen-Van de Kaa, Glen Kristiansen, Rodolfo Montironi, Chin-Chen Pan, Jenna Sykes, Kilian Treurniet, Philippe Camparo, Theo van der Kwast. Hopital La Pitié Sampetriere, Université Paris VI, Paris, France; Karolinska University, Stockholm, Sweden; Faculty of Medicine, University of Cordoba, Cordoba, Spain; Fundacio Puigvert, Barcelona, Spain; Cedars-Sinai Medical Center, Los Angeles, CA; Johns Hopkins Hospital, Baltimore, MD; Uppsala University Hospital, Uppsala, Sweden; Radboud University Nijmegen Medical Centre, Nijmegen, Netherlands; University Hospital Bonn, Bonn, Germany; Polytechnic University of the Marche Region, Ancona, Italy; Taipei Veterans General Hospital, Taipei, Taiwan; Ontario Cancer Institute, Toronto, Canada; Leiden University Medical Center, Leiden, Netherlands; Laboratoire de Pathologie, Amiens, France; University Health Network, Toronto, Canada
Background: Distinction between pTa and pT1 non-muscle invasive bladder cancer (NMIBC) is of big clinical relevance. Literature demonstrated up- and downstaging of NMIBC after central pathological review. The ENUP initiated a study to improve correct staging of NMIBC.
Design: Twenty-five cases of NMIBC initially reported as pT1 were selected by 3 genitourinary pathologists(GUP) based on potential uncertainty for stromal invasion. All slides were digitized and reviewed independently by a panel of eight GUP. GUP had to stage NMIBC and annotate invasive areas on digital slides if present. Annotations were reviewed and heatmaps showing prevalence of stromal invasion were constructed for each case. Kappa statistics was performed to determine agreement among the 8 GUP.
Results: Full agreement was obtained in 11 of 25 cases (44%) and majority consensus (6-7 of 8) in 18 / 25 cases (72%). Weighted kappas for interobserver reproducibilitywere 0.42 - 0.6, mean 0.49. There were 7 discordant cases (28 %). The constructed heatmaps were used to identify areas contentious with regard to invasion. After review of the 7 discordant cases by the three expert GUP consensus was obtained for 6 of the 7 cases.
Conclusions: ]The interobserver agreement among the 8 GUP was moderate (mean kappa 0.49), consensus among three expert GUP could be reached in almost all cases. Heatmaps allowed us to identify cases that can be used as a teaching set of images for standardization of histologic criteria for invasiveness of NMIBC.
Category: Genitourinary (including renal tumors)
Wednesday, March 6, 2013 9:30 AM
Poster Session V # 148, Wednesday Morning