Case study: TLS status
Authors
Case study: TLS status
Authors
About Owkin
Owkin is an AI biotechnology company that uses AI to find the right treatment for every patient. We combine the best of human and artificial intelligence to answer the research questions shared by biopharma and academic researchers. By closing the translational gap between complex biology and new treatments, we bring new diagnostics and drugs to patients sooner.
We use AI to identify new treatments, de-risk and accelerate clinical trials and build diagnostic tools. Using federated learning, a pioneering collaborative AI framework, Owkin enables partners to unlock valuable insights from siloed datasets while protecting patient privacy and securing proprietary data.
Owkin was co-founded by Thomas Clozel MD, a former assistant professor in clinical onco-hematology, and Gilles Wainrib, a pioneer in the field of machine learning in biology, in 2016. Owkin has raised over $300 million and became a unicorn through investments from leading biopharma companies (Sanofi and BMS) and venture funds (Fidelity, GV and BPI, among others).
Prediction of patient’s TLS status from H&E whole slide images in pan-cancer cohort
Context
Tertiary lymphoid structures (TLSs) are formations at sites with persistent inflammatory stimulation, including tumors. These ectopic lymphoid organs mainly consist of chemo-attracting B cells, T cells, and supporting dendritic cells (DCs).
TLS presence in the tumor compartment is considered a novel biomarker to stratify the overall survival risk of untreated cancer patients and as a marker of efficient immunotherapies for patients with solid tumors.
Methods
Based on a pan-cancer cohort of 289 WSI, Owkin has developed a deep learning method to predict the presence of TLS from routine digitized H&E WSI.
The model scores regions of 112x112μm² by their relevance for TLS status and aggregates scores to make the final prediction.
Results
This model is able to predict TLS status at the patient level with an AUC of 0.91 in 5-fold cross validation on the pan-cancer cohort. Moreover the model is robust to the transfer on an external cohort of sarcoma patients and achieves an AUC = 0.89.
Impact
- Select high-value subgroups of patients that are most likely to respond to the ICI being tested, improving the statistical power of a trial.
- Better selection of trial participants will improve success rates across trial phases and ultimately faster regulatory approval and more precise marketing.