EVALUATION OF DIGITAL PATHOLOGY IN IMPROVING CANCER DIAGNOSTICS AND TREATMENT OUTCOMES IN RESOURCE LIMITED SETTINGS: A CASE OF INTERNATIONAL CANCER INSTITUTE IN KENYA

Background:

Despite the exponential increase in the global cancer burden, there is a markedly unmatched diagnostic capacity among Low- and Middle-Income Countries (LMICS), especially in Sub-Saharan Africa (SSA) in averting these trends. Proper cancer diagnosis is premised on quality histo-cytopathology which is hinged on good histology practices. Moreover, the current oncology practice advocates for quality diagnostics and efficient systems in improving treatment outcomes. 

Digital pathology integrated with Artificial intelligence (AI) tools, is an essential component of telemedicine which aims to improve access to pathology diagnostic services both locally and internationally. The diagnostic accuracy is (97%) sensitive and (99%) specific. It provides digital slide banks for research, platform for specialist boards, diagnosis and case conferencing. Besides, it improves access to pathology services in resource constrained settings and reduces reporting turn-around time. We aimed to evaluate the utilization of digital pathology platform in improving Cancer diagnostics and treatment outcomes in resource limited settings. 

Materials and Methods 

International Cancer Institute (ICI), a leading cancer care, training and research organization in Sub-Saharan Africa, championed telepathology using DP 200 scanner in collaboration with ROCHE to reduce the burden of cancer. Approximately, 400 cases have been scanned into the DP 200 Scanner database between April 2021 to January 2022.

Cases are available to all assigned pathologists and can be retrieved for discussion and education purposes. To date, fifteen consultant pathologists have been trained on the u-Path platform with 4 (26.6%) of these trained on the use of the DP200 scanner. Three trained ICI lab staff are competent in scanning and dispatching cases. Four cases of HER2 scores were used during the training (HER2 negative, positive +1, positive +2 and positive +3 cases) scored on traditional microscopic methods by the same pathologist. The same results were reproduced by each participant throughout the training. It takes 30 to 45 minutes for a pathologist to report a case. However, it was determined that it takes 15-20 minutes for the pathologist to generate a complete histopathology report using the digital pathology platform. Tumor board discussions held twice weekly were used to evaluate the effectiveness of digital pathology. Comparative statistical analysis between the conventional histology techniques and digital pathology was done. 

Results

Four patient cases have been used in tumor board discussions with image displays done for two (50%) out of the four. From a survey conducted in the tumor board session using an evaluation form, 316 (43.7%) out of 723 attendees found the inclusion of digital whole slide images for case discussions very important while 339 (47%) reported high knowledge gain and better quality of discussions when slide images were displayed. 

Conclusion: 

Digital pathology is of great value in learning and a big boost to the current traditional diagnostic practice as it reduces diagnostic turn-around time.