Morphometric Analysis in the Distinction between Invasive and In Situ Endocervical Adenocarcinoma
Carlos Parra-Herran, Tyna Hope, Sharon Nofech-Mozes, Taha Rashed, Guangming Han, Elzbieta Slodkowska, Fang-I Lu, Nadia Ismiil, Mahmoud Khalifa, Gina Clarke, Martin Yaffe, Matthew Cesari. Sunnybrook Health Sciences Centre, Toronto, ON, Canada; Centre for Imaging Technology Commercialization, Toronto, ON, Canada
Background: Morphologic distinction between in situ (ACIS) and invasive (INVA) endocervical adenocarcinoma, while having important clinical implications, can be challenging. As diagnostic digital pathology gains momentum, we aimed to identify quantifiable histologic features by digital image analysis that can differentiate between ACIS and INVA, which may be used as an adjunct diagnostic tool.
Design: Scanned images of 7 cases of ACIS and 7 cases of INVA in cone or hysterectomy specimens were included. Representative glands (2-4) were selected from each case and analyzed using Mercator Version 1.0 (ExploraNova, La Rochelle, France). The system detects structures based on optical density and analyzes them as closed objects. We collected morphometric variables in glands and periglandular stroma. Non-parametric Wilcoxon Mann Whitney test was used to evaluate the difference between both groups.
|Mean nuclear form factor (MNFF)||SNF/NC||0.18|
|% Nuclear surface area||SNA/TSA*100||0.75|
|STROMA||Nuclear density (ND)||NC/TSA||0.85|
|Mean nuclear form factor (MNFF)||SNF/NC||0.009|
|% Nuclear surface area||SNA/TSA*100||0.48|
|% Collagen surface area||SCoA/TSA*100||0.02|