New technologies are transforming cancer research. By optimizing research protocols and leveraging data more efficiently and intelligently, these tools hold the promise to improve personalized cancer care. Here, our Curious Dr. George asks Matvey B. Palchuk, MD, MS, FAMIA, VP of Informatics at TriNetX, LLC, about the capabilities of his company’s platform.

Curious Dr. George: Translational medicine has evolved to include personalized medicine and precision oncology. We have learned that all individual cancers may be unique, but that they do share some common “…omic” elements that can inform treatment decisions.

TriNetX encompasses a global network of millions of patients with real-world evidence on their responsiveness to a range of interventions for a multitude of diseases. How might the data about interventions for advanced cancers in your global network inform the best treatment options for patients with these conditions?

Dr. Palchuk: TriNetX is the global research network dedicated to optimizing clinical trial operations and real-world evidence generation. We built a data and analytics platform that is powered by an impressive collective of 170 healthcare organizations across 28 countries. A researcher utilizing our platform has instantaneous access to clinical data of tens of millions of patients. The data represents billions of observations collected in electronic health record systems, ancillary systems, cancer registries, billing and financial systems, and many more.

Users of the TriNetX platform can define patient cohorts of interest and go on to learn about their size and how that size evolved over time and what we expect it to be in the future, examine the details of clinical characteristics of the patients, assess trial feasibility, and much more. They can ask sophisticated research questions ranging from comparing outcomes to treatment pathways, burden of illness to incidence and prevalence, and more. And the platform responds in seconds—an entire panoply of actionable insight is at their fingertips, literally at the speed of thought.

One of the top considerations in creating and maintaining our platform is the protection of patient privacy, and the privacy of our member organizations. We take a conservative stance when it comes to working with clinical data and are always very diligent with how the data is handled and used.

The users of the TriNetX platform mainly focus on clinical trial optimization and population research. The capabilities of the tools we are developing lead to working with cohorts of patients and generating evidence on entire populations. However, questions similar to yours naturally arise when contemplating large sources of clinical data. We certainly face these questions in regard to the TriNetX platform—can our data asset be used to inform decisions about treatment options? Certainly, we enable doing so at population-levels all the time. But what about individual patients? Let me describe two possible scenarios:

In the first scenario, a clinician and their patient need to select a treatment option. To take advantage of the existing historical data, the first challenge is to find other patients “like this one.” It is not a trivial problem—how to choose the important characteristics? Do we focus on demographics, comorbid conditions, socio-economic status, or any number of other possible characteristics? There is early work on this topic, but much still needs to be done. Another challenge is the relative paucity of information about outcomes of care, but we won’t tackle this subject here. Of course, it is crucial that my data covers a sufficient number of patients to ensure that the sample of patients “like this one” is large enough to provide the necessary information. This is where a network like TriNetX is really powerful.

In the second scenario, let’s consider a machine-learning model capable of making a prediction about a diagnosis based on historical patterns in the data before any recognizable symptoms arise. As you know, early diagnosis leads to a significant improvement in survival. TriNetX is collaborating with a team of scientists focusing on early prediction of pancreatic cancer using such a model. When high-risk patients are found, what is the mechanism for getting in touch with them? The data in our platform is de-identified—we do not know who those patients are—and only their healthcare provider has the capability to “break the glass” and navigate the intricacies of contacting those patients.

Although not directly intended for use in improving the care of a single patient, you can see how the research-oriented clinical data can lend itself to go beyond population medicine and make a difference in individual lives. We at TriNetX are working diligently to bring this promise closer to reality.

Dr. Palchuk can be reached at matvey.palchuk@trinetx.com. TriNetX is located at at 125 Cambridgepark Drive, Suite 500, Cambridge, MA 02140.

Related Links:

Can Preclinical Data Guide Clinical Cancer Therapy?

Capturing Patients’ Real-World Experiences to Improve Cancer Research and Care

How to Beat COVID-19 with Real-Time, Real-World Data

***

Copyright: This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.