Out-of-sample prediction estimation to accurately predict response to immunotherapy drugs in metastatic melanoma

Anti-PD1 based immunotherapy has revolutionised cancer treatment in recent years. These drugs have been approved for many cancer types, and are now front-line treatment for metastatic melanoma, which has the highest response rate of any cancer type. Despite this, approximately half of metastatic melanoma patients fail to respond to therapy. In addition these treatments are expensive and can cause toxicity. Valid and accurate assessment of immunotherapy response prediction is essential for clinical decision making but has still remained a serious challenge.

This project aims to take comprehensive approach where both clinicopathological features, and, genomics expression variables are explored simultaneously to derive the most accurate prediction of response to immunotherapy treatment. The data analytics methods that will be used (with potential methodology extension) will deal with the “large p, small n” challenge that necessarily also induce computational complexity.

Main Applicants

Serigne Lo, University of Sydney

Maria-Pia Victoria-Feser, University of Geneva