Influence of Histologic Criteria and Confounding Factors (CFs) in Staging Equivocal Cases for Microscopic Perivesical Tissue Invasion (MPVI or pT3a): An Interobserver Study among Expert GU Pathologists
Vijayalakshmi Ananthanarayanan, Yi Pan, Maria Kocherginsky, Maria Tretiakova, Mahul B Amin, Liang Cheng, Jonathan I Epstein, David J Grignon, Donna E Hansel, Rafael E Jimenez, Jesse K McKenney, Rodolfo Montironi, Esther Oliva, Adeboye O Osunkoya, Priya Rao, Victor E Reuter, Jae Y Ro, Steven S Shen, John R Srigley, Toyonori Tsuzuki, Jorge L Yao, Tatjana Antic, Michael Haber, Jerome B Taxy, Gladell P Paner. University of Chicago, Chicago, IL; Emory University, Atlanta, GA; Cedars-Sinai Medical Center, Los Angeles, CA; Indiana University, Indianapolis, IN; Johns Hopkins Hospital, Baltimore, MD; Cleveland Clinic, Cleveland, OH; Mayo Clinic, Rochester, MN; Polytechnic University of the Marche Region, Ancona, Italy; Massachusetts General Hospital, Boston, MA; MD Anderson Cancer Center, Houston, TX; Memorial Sloan-Kettering Cancer Center, New York, NY; Methodist Hospital, Houston, TX; McMaster University, Hamilton, ON, Canada; Nagoya Daini Red Cross Hospital, Nagoya, Japan; University of Rochester, Rochester, NY
Background: Current oncology guidelines consider giving adjuvant chemotherapy for bladder cancer with at least MPVI (>pT3a). The boundary of muscularis propria (MP) vs. perivesical tissue (PVT) is commonly ill defined and may influence the interpretation of MPVI.
Design: 20 sets of images including 19 equivocal for MPVI with CFs were sent to 17 expert GU pathologists. CFs included “histoanatomic" (HA) defined by irregular MP-PVT border and “tumor-related” (TR) such as fibrosis/desmoplasia, dense inflammation, tumor at edge of outermost MP bundle, and lymphovascular invasion (LVI). Cases were grouped by CFs: I - HA (7/19), II - HA+TR (7/19) and III - TR (5/19) factors.
Results: Criteria used for MPVI varied among pathologists: A (3/17); B (9/17); C (4/17).
In 4/7 group I cases most A and C users systematically disagreed, and in none (0/7) A and C agreed and B disagreed. Median pairwise agreement for all (regardless of criteria and case groupings) was fair (κ 0.281). However, for B and group I cases only, median agreement was substantial (κ 0.696). For groups I and II, median agreements by criteria were: A κ 0.588, B κ 0.423, and C κ 0.512, which showed the negative effects of TR factors. For group III, median agreement for all was fair (κ 0.286). LVI only at MP-PVT was not called as MPVI by 87.5% pathologists.
Conclusions: Three main approaches are applied in defining MP-PVT boundary and agreement for MPVI is increased when similar criterion is applied. Future studies using these different approaches may identify the most clinically relevant criteria for MPVI.
Category: Genitourinary (including renal tumors)
Monday, March 4, 2013 9:30 AM
Poster Session I Stowell-Orbison/Surgical Pathology/Autopsy Awards Poster Session # 128, Monday Morning