Owkin and CancerLinQ use federated learning on real world oncology data to study lung cancer treatment

ASCO’s CancerLinQ and Owkin announced today a new research collaboration to use artificial intelligence to analyze real-world oncology data with the aim to understand why some cases of metastatic non-small cell lung cancer (NSCLC) are resistant to first-line immunotherapy treatment.

Using data from CancerLinQ® Discovery, one of the largest and most-diverse real-world-oncology databases that include de-identified data from more than 6 million patients with cancer, as well as de-identified data from a European research site, Owkin will deploy its proprietary federated learning algorithms to identify possible predictive factors that could inform the understanding of why some patients with NSCLC respond poorly to immunotherapy treatment. The study will also compare patient characteristics, treatment factors, and clinical outcomes of patients with advanced NSCLC.

Federated learning, a decentralized machine learning approach that trains machine learning models with multiple data sources, maintains privacy and ownership while allowing participants to benefit from a larger amount of data than their own. Instead of gathering data on a single server, the data remains locked on servers as the algorithms and only the predictive models travel between the servers.

Sean Khozin, MD, MPH, Chief Executive Officer of CancerLinQ, said:

Unlocking the full potential of real-world data to advance cancer care and research requires new modes of data sharing at scale and applying next-generation analytical methods such as AI to complex multimodal datasets, both of which are key features of this exciting research collaboration.

Federated learning allows a safe and secure method of sharing data and collaborating across siloes and continents as one global cancer research community focused on advancing discoveries that can improve the lives of patients with cancer.

Thomas Clozel MD, Co-founder and CEO of Owkin, said:

The next generation of medical breakthroughs will be unlocked by the application of artificial intelligence to vast amounts of rich patient data.

We are excited to use our federated learning software to safely and securely analyze CancerLinQ’s diverse patient data. We are excited to work together to make discoveries that can contribute to improving treatment for millions of lung cancer patients across the world.

Worldwide, lung cancer is the second most commonly diagnosed cancer and the leading cause of cancer deaths. NSCLC is the most common type of lung cancer in the United States, accounting for 82% of all lung cancer diagnoses.

‍

About CancerLinQ

CancerLinQ® is a real-world oncology data platform developed by ASCO that collects and aggregates longitudinal electronic health record data from oncology practices throughout the United States. CancerLinQ aims to improve the quality of patient care and accelerate discovery by securely compiling, harmonizing, analyzing, and de-identifying vast amounts of information on patient characteristics (e.g., molecular profiles, comorbidities), treatments, and long-term side effects. By using data from over six million patients in near real-time, CancerLinQ can identify trends and associations between myriad variables, thereby enabling physicians to generate new hypotheses and apply those conclusions to improve care in real-world settings. Follow CancerLinQ on Twitter or connect with us on LinkedIn.