By Rob Preston | June 2021
Immunotherapies, invented about a decade ago to treat certain forms of cancer, can still cost up to $100,000 per patient—but they have only a 10% to 40% chance of actually working, depending on the cancer under treatment. For that reason, insurers tend to use these therapies only when a patient is in a late stage of the disease.
It would be a markedly different situation if clinicians could identify which cancer patients aren’t likely to respond or respond well to such therapies, based on the makeups of their individual immune systems. Such information could spare them the false hope and enormous expense of a gruelling, potentially toxic series of treatments, while enabling their doctors to focus instead on providing them with more productive care alternatives. Meanwhile, those patients identified as likely to respond to the therapies could proceed much earlier in their illness with far more confidence.
That’s one of the potentially positive outcomes of the “clinical-grade analytical platform” developed by GMDx Genomics, a Melbourne, Australia-based company that’s carving out a critical niche in the fast-moving field of personalized medicine. After years of R&D led by cofounder and chief scientific officer Robyn Lindley, GMDx is now starting to work with healthcare systems in Australia, the United States, and other countries to use its platform to monitor the Innate Immune Fitness (IIF) of individual patients.
Each IIF profile is derived from an in-depth analysis, powered by Oracle Cloud technology, of about 40,000 metrics generated from a comprehensive sequencing of a patient’s genome. Depending on the stage at which the GMDx analysis is applied to each patient and the combinations of genome data that get analyzed, applications developed on GMDx’s patented IIF technology function in three main ways:
1. to determine which patients are the most and least likely to respond to certain kinds of treatments, as in the immunotherapy example cited above.
“In oncology over the years, healthcare providers have transitioned from treating the cancer with toxic chemotherapy—where they killed everything that’s bad but unfortunately caused a lot of collateral damage—to more targeted therapies around specific organs,” notes GMDx CEO Bernie Romanin, a 40-year industry veteran. “The biggest challenge is that immunotherapies, when they work, are very effective, but when they don’t work, they come unfortunately with very, very adverse side effects.”
GMDx also reasons that if it can show increased success rates with certain treatments, the medical industry’s payers—national healthcare programs and private insurers, which hold most of the power in the industry—are more likely to get onboard.
2. to chart the “progression signatures” of patients already diagnosed with cancer and other diseases, to determine the likelihood of the disease returning after treatment and thus dictate the aggressiveness of follow-up care.
If patients’ IIF signatures indicate that their cancer isn’t likely to return for some time, their clinicians will treat them differently from patients whose cancer is likely to progress more rapidly. In the latter case, the clinician might then recommend that the patient come back every two or three months for a scan, or the clinician might prescribe a different post-surgical follow-up regime that includes not just alternative therapeutic treatments, but also changes to lifestyle and diet.
3. to detect the onset of cancer and other diseases, in order to treat them before they advance or spread.
For example, GMDx is partnering with the Garvan Institute of Medical Research, based outside Sydney, to identify individuals most likely to contract sarcoma, a particularly brutal form of cancer, before signs are evident. Their research has revealed that patients later diagnosed with the disease show a “dysregulation” of a particular group of mutagenic enzymes early on. GMDx’s IIF analytics can help doctors monitor for these enzymes, enabling them to intervene with drug stabilizers—even precancer.
“All of a sudden we’re talking about a diagnostic that is not only affordable, but also provides almost exponentially more information than we’ve ever been able to use before.”
Bernie Romanin, CEO, GMDx Genomics
“This is really the first time that we’re taking medicine to an individual profile from a population profile,” Romanin says about precision medicine broadly. “So often medicine says: This is how to treat everybody who’s got metastatic renal cell carcinoma or some other disease. But the reality is that everybody’s different. We think we’re on the cusp of being able to show why.”
Genomic profiling for cancer and beyondTo date, GMDx’s IIF profiling has focused mostly on early detection and treatment of different cancers, since genomics is furthest along in oncology. But the company is also doing work in the area of cognitive-impairment diseases, such as Alzheimer’s and Parkinson’s.
GMDx derives its IIF profiles from the sequencing of each individual’s whole genome—about 25,000 genes in all. Third-party labs can now complete such sequencing for a fee of about $500 per person, compared with about $10,000 as recently as a decade ago and $100 million (yes, per person) going back two decades.
The cost of genome sequencing continues to drop, to the point where it could soon become part of the standard lab tests doctors order up to check patients’ cholesterol, triglycerides, hemoglobin A1C, and other levels. “All of a sudden we’re talking about a diagnostic that is not only affordable, but also provides almost exponentially more information than we’ve ever been able to use before,” Romanin says.
Oracle ‘engine room’
GMDx built its core technology on a platform developed by partner Applied Precision Medicine. In association with that platform, the company uses Oracle Cloud Infrastructure (OCI) server and storage instances for the intensive data processing required to create each IIF profile. The company can scale those OCI instances up and down cost-effectively during the multistage processing.
GMDx also uses the OCI-based Oracle Autonomous Data Warehouse to capture and analyze the mostly genomic data it accumulates—10 to 12 terabytes in aggregate so far, estimates CTO Richard Rendell. Two sets of built-in AI algorithms, which sift through that data to identify relevant genetic biomarkers and patterns, cross-check each other to ensure accuracy, he says.
Rendell, who says GMDx will eventually hold about a terabyte of data per individual it profiles, says a big draw of Oracle’s cloud database is that it allows the company to run a lean IT shop, since the autonomous software provisions, manages, and patches itself.
Early on GMDx partnered with other cloud providers, “who were easy enough to work with,” but their infrastructure wasn’t as price-elastic as Oracle’s, he says. What’s more, Rendell says, those other providers couldn’t offer GMDx a pathway to integrate its platform with full-stack health applications, a pathway Oracle offers via its Oracle Life Sciences global business unit. And given the massive amounts of sensitive data GMDx collects from and must distribute to partners, the company places great value on Oracle’s expertise to help it address its data security and regulatory compliance requirements, he says.
“One of the critical points here is that this is a truly innovative platform technology, an engine room that is very much Oracle-oriented,” CEO Romanin says. “That engine room is almost agnostic to the clinical application. Once you’ve got that core platform in place, we can construct and apply a set of analytics that is disease-specific. And we can replicate this platform, make mirror images that can be hosted in any facility, any laboratory.”
Romanin also acknowledges that no single healthcare or health technology company is going to have every piece of information needed to diagnose and treat patient
s with pinpoint precision. “What GMDx is doing is complementary to the genomics being done now,” he says. “We’re adding a layer of value that’s filling in the missing gap.”