Abstract
AbstractAmong renal cell carcinomas (RCCs), clear cell RCCs (ccRCCs) is a highly aggressive class characterized by highly invasive clinical course and poor survival. The presence of a sarcomatoid component implicates an even poorer prognosis. Spatial measures can establish features that could be routinely used in clinical practice to stratify important RCC outcomes. Therefore, we tested the effectiveness of spatial features as predictors of survival differentiation in 58 Grade 2 ccRCCs. We developed a machine learning model using the acquired spatial features to predict survival in Grade 2 ccRCCs which we termed “SurvCal”. The receiver operating characteristic accuracy from the derived model was 0.812. Subsequent feature analysis identified the spatial model fitting intensity parameter Gamma and the pair correlation function as critical features in distinguishing the classes. In the light of the increasing digitization of pathology routines, our results demonstrate the importance of spatial point pattern features as determinants of ccRCC survival outcomes and phenotypes.
Publisher
Cold Spring Harbor Laboratory
Reference20 articles.
1. Ferlay, J. , Soerjomataram, I. , Ervik, M. , Dikshit, R. , Eser, S. , Mathers, C. , Rebelo, M. , Parkin, D. , Forman, D. , Bray, F. : Globocan 2012 v1. 0. Cancer incidence and mortality worldwide: IARC CancerBase 11 (2013)
2. Rising Incidence of Small Renal Masses: A Need to Reassess Treatment Effect
3. Risk Group Assessment and Clinical Outcome Algorithm to Predict the Natural History of Patients With Surgically Resected Renal Cell Carcinoma
4. An Outcome Prediction Model for Patients with Clear Cell Renal Cell Carcinoma Treated with Radical Nephrectomy Based on Tumor Stage, Size, Grade and Necrosis: The Ssign Score
5. Sarcomatoid renal cell carcinoma: biology, natural history and management;Nature Reviews Urology,2020