CHICAGO – Tempus has ventured into a new market segment with its recent introduction of Edge, a platform that provides pathologists with online access to artificial intelligence models to identify samples with potentially significant biomarkers with a single hematoxylin and eosin-stained slide. The platform is meant to help physicians decide whether to order genomic sequencing that might indicate eligibility for targeted therapies and clinical trials.
At the same time, the Chicago-based molecular laboratory and informatics firm said that it would build a network of pathology laboratories across the US to contribute data to improve the firm's AI models, again with the goal of identifying patients for targeted therapies and clinical trials. This network, called NAPA, is analogous to Tempus Integrated Molecular Evaluation (TIME) Trial, a 3-year-old network that provides similar services for cancer patients.
NAPA will also help Tempus develop, test, and refine new algorithms. "As we develop an algorithm, we look to test that with our partners in the NAPA network to figure out if, in fact, it's good enough to have clinical use," said Tempus Founder and CEO Eric Lefkofsky. "If it is, we roll it out nationally or internationally."
The company introduced Edge March 21 with a model portfolio called Tempus Histogenomics, which will include digital pathology algorithms to predict the potential of biomarkers including microsatellite instability and homologous recombination deficiency status from an H&E slide in patients whose biopsy samples are typically not sequenced.
"At its heart, Edge is all about beginning the journey to figure out how to help pathologists use technology and use AI to generate greater insights," Lefkofsky said.
Typically, a patient being evaluated for cancer would have a biopsy and the tissue would go to pathology for review under a microscope. Tempus and its partners digitize slides so the firm can apply its AI to automate the process of finding malignancies.
Edge hopes to address the issue of training machines to examine the morphology and structure of biopsied cells, then predict whether the patient has a genetic mutation, according to Lefkofsky. "Edge is all about using technology to help pathologists … find all kinds of diagnostic insights right there when they're doing their work instead of having to get them a week or two or three later," he said.
"You're trying to make sure the right people get genetic testing. You're trying to make sure that you don't lose weeks when you have a high degree of probability that a patient should be on a particular therapy," he added.
Edge is designed to reduce the number of open-ended questions about pathology results, while NAPA is a network for sharing de-identified data across institutions and deploying Tempus informatics tools. Both are also meant to accelerate the transition to digital pathology, according to Lefkofsky.
Because it has already automated its workflows, PathNet Labs, a multistate network of digital pathology practices, is an early adopter of Edge and member of the NAPA network.
PathNet CEO Matthew Leavitt said that the Tempus technology enables his company to suggest molecular tests to a patient in real time rather than having to wait weeks for a lab to send back manually produced pathology results.
"We get that result in a time frame that enables us to then go back to the patient and say, 'Hey, not only do we know you have prostate cancer, but we know that you have a prostate cancer that is likely to respond [to a targeted therapy]. Talk to the folks at Tempus in this clinical trial network, and they'll hook you up with a clinical trial,'" if the patient qualifies after having genetic testing, Leavitt said.
This real-time knowledge also bridges what Leavitt called a "huge disconnect between the oncology community and the surgical community," in which a surgeon might perform unnecessary invasive procedures because they don't know that there is a genetic marker suggesting that an approved or investigational drug might address a specific tumor.
"It changes everything to the patient," said Leavitt, brother of former US Health and Human Services Secretary Mike Leavitt and a practicing pathologist.
Leavitt also founded Lumea, a company dedicated to digitizing pathology workflows. "You can't get to computational pathology unless you're digital," Leavitt said.
PathNet uses Lumea technology, though the latter does not offer computational analysis services like Tempus does.
He said that Tempus and Lumea are already synchronizing their technology platforms to support PathNet, which is preparing to go live on Edge in the next few months with a single pathologist in Arkansas. "Then we'll push this out to the whole PathNet network as we get the kinks worked out," Leavitt said.
A major goal in 2022 for PathNet is to use Edge to identify prostate cancer patients with microsatellite instability-high tumors who are likely to benefit from PARP inhibitors.
"MSI-high tumors right now require multiple expensive [immunohistochemistry or molecular sequencing] tests, and it's rare enough that it's not really very practical to do that on everyone," Leavitt explained. "If we can identify those patients, then we will accelerate tremendously the clinical trials for PARP inhibitors and be able to offer [trials] to patients in rural communities and in underserved communities that we provide diagnostic care for."
Lefkofsky said that he envisions a future where pathology has the same kind of decision support as frontline clinicians. "There will come a day when pathologists are armed with incredibly powerful tools and have access to vast amounts of information in real time so that right from the moment when somebody comes in … those pathologists are able to make incredibly precise determinations ... of what that patient has and what they should consider doing next," he said.