• Developing and applying artificial intelligence (AI) based image analysis for scoring PD-L1 immunohistochemistry in gastric cancer
    Share

Trial Status

Translational Research Studies

Cancer Type

Stomach Cancer

This is a translational research trial and not open for patient recruitment.

Protocol Title

Developing and applying artificial intelligence (AI) based image analysis for scoring PD-L1 immunohistochemistry in gastric cancer.

Purpose of the Study

Evolving computer technology including artificial intelligence (AI) is revolutionising workflow and business models in multiple industries. The capabilities of computers to undertake visual and spatial analysis are now beginning to be applied in the health setting. This is particularly the case in anatomical pathology which depends on classification of the presence or absence of disease processes from a tissue sample based on its visual appearance.
As more treatments become available for cancer treatment, it becomes increasingly important to make the best initial choice using an agent that is most likely to be effective. An important aspect of this decision making involves evaluation of so called ‘predictive biomarkers’ in tumour tissue. The presence or absence of predictive biomarkers informs the likelihood of success with particular treatments, and hence impacts on treatment choices (personalised therapy).
Studies using immunotherapy drugs have shown that expression of the PD-L1 protein is an important predictive biomarker, but it is technically challenging and expensive for pathologists to report it. This study aims to use AI to develop an algorithm to enable computer-based selection of which gastric cancers would be most suitable for immunotherapy treatment. After training an AI model on an initial training dataset, the study will then compare standard pathologist-based testing of the expression of the protein PD-L1 against the predicted output from the trained AI model. If the AI based method is effective, it could enable a more systematic and cheaper method to identify cohorts of gastric cancer patients suitable for immunotherapy based on PD-L1 expression.

Principal Investigator

Professor Niall Tebbutt

Contact Email

Please contact Dr Sarah Hayes (sarah@gicancer.org.au) if you have questions regarding this study.

Funding

2021 AGITG Gastric Cancer Grant

Trial Status

Translational Research Studies

Cancer Type

Stomach Cancer

If you think this clinical trial may be relevant to your patient or to discuss further, please contact the Clinical Trial team.