Study
            
                PLCO
                (Learn more about this study)
            
 
            
            
                Project ID
                
                    
                        PLCOI-1413
                    
                
            
            
                Initial CDAS Request Approval
                Dec 11, 2023
            
            Title
            Validation of cancer cell morphological diversity-based prognostic biomarkers
            
                Summary
                Intratumor heterogeneity (ITH) is a universal phenomenon observed in all cancer types. It is well established that ITH drives disease progression and therapeutic resistance, which leads to poor survival outcomes in patients. ITH at the genetic level has been extensively investigated and can be measured using whole-genome sequencing or single-cell molecular profiling technologies. The practical application is challenging, however, due to several issues including the need for high-quality tissue, complexity, and cost. In a recent study published in the JNCI, we developed a computational approach for quantitative evaluation of cancer cell morphological diversity in routine hematoxylin and eosin (H&E)-stained histopathology images. We integrated two distinct measures of ITH: inter-cellular diversity, and intra-cellular heterogeneity. We further proposed a cancer cell diversity score and evaluated the prognostic significance across four different tumor types. Here, we propose to validate the cancer cell morphological diversity-based prognostic biomarkers using prospectively collected PLCO data.
            
            
                Aims
                Aim 1. We will apply machine learning techniques to compute the cancer cell morphological diversity scores from digitized H&E-stained slides.
Aim 2. We will evaluate the prognostic value of the diversity scores by associating with outcomes in multiple cancer types from PLCO patient cohorts.
Collaborators
                
                Xiyue Wang, Stanford University