PhD Defense: Improving prostate cancer detection and grading using novel digital technologies

This dissertation examines the variation in Gleason grading for prostate cancer and its impact on patient treatment. The Gleason grading is determined by a pathologist based on a prostate biopsy, and it is used for risk stratification, and the appropriate treatment strategy thereafter. The study reveals significant differences in how pathologists and laboratories assign these grades, which affect the treatment patients receive. For example, patients from labs that tend to grade higher often undergo more aggressive treatments, while those from lower-grading labs receive less intensive treatment strategies. This variation can influence patient outcomes.

The study also explores interventions to reduce these inter-laboratory and inter-observer variation. An online e-learning module for pathologists led to some improvement, especially for those who deviated the most from the standard. However, even after this training, variation remained. Feedback reports sent to labs did not bring significant improvements in grading consistency.

Additionally, the dissertation discusses the role of artificial intelligence (AI) in pathology. AI has the potential to improve cancer detection and grading, while saving time and money. In this research, pathologists using AI to detect prostate cancer needed less additional immunohistochemical stains, hereby reducing costs, without compromising patient safety. However, implementing AI still faces challenges, including technical and legal barriers. While AI shows promise in enhancing diagnostic accuracy, further research is needed before it can be widely adopted in everyday clinical practice.