GenomOncology, a Cleveland-based healthcare technology company founded in 2012, works to provide actionable insights for clinicians and researchers via genomic analysis tools and services. Institution-tailored implementations of their GO Precision Medicine Portfolio enable the streamlined development, validation, and eventual clinical adoption of cancer and hereditary disease tests. These targeted tests unlock detailed, individualized patient information that guides the treatment decisions made by clinicians and, ultimately, improves patient outcomes.
The Medgadget team recently had an opportunity to speak over the phone with Dr. Ben Salisbury, Vice President of New Products and Marketing with GenomOncology, about their technology and current trajectory. An abridged transcript is provided below.
Zach Kaufman, Medgadget: As we start off here, is there anything specific about GenomOncology’s offerings or behind-the-scenes processes that you would like our readership to know?
Dr. Ben Salisbury, GenomOncology: I guess if I had one goal here, it would be to clarify to the world that our tools and services are for more than just interpretation and reporting. GenomOncology enables precision medicine. That’s our guiding vision. We are about supporting the whole world of precision medicine.
So one aspect of that is certainly in the lab with the production of clinical testing. Beyond that, though, we offer decision support — how do you help the recipients of lab test results, the clinicians themselves, make use of the information they’re given? For each specific patient, oncologists and medical geneticists might suggest going in a given therapeutic direction or considering enrollment in a clinical trial. We help them interpret the relevant data to make those decisions confidently.
As an institution gathers more data by testing and treating patients, that data accumulates in our system, which is all securely behind the client’s firewall. From there, it can be queried to provide a greater level of understanding of how precision medicine is being practiced at the institution, and also to foster discovery. What can we learn about how patients are responding to different treatments? Who’s more likely to succeed or not on a given therapeutic approach?
And, as a next step, we will soon be able to gather and use that information across institutions that are willing to share data through a network. That’s in development at this point, and right now data all belong to a single instigating institution. However, we’re currently developing a federated approach where data can be anonymously shared to whatever comfort and interest level individual institutions will allow.
Medgadget: Do the potential member institutions of such a federated database, via preliminary feedback, seem to be open to sharing their data?
Dr. Salisbury: Absolutely. Most of them are really eager for it due to the nature of precision medicine. We’re trying to draw conclusions by looking at individuals and small groups of individuals who may resemble one another. That means that, even within an institution that looks at hundreds of patients each year, when you subdivide that patient population by their diagnoses and other clinical factors or treatments you pursue, you end up with very few individuals who can be bucketed together and considered as a group. So it’s really only by sharing data across multiple institutions that you have sufficient data to make confident conclusions about these differences.
The one piece that’s missing, then, in this virtuous cycle, is to take those discoveries and put them into clinical practice, with the goal of ultimately improving the tests that are offered. That involves confirming the discoveries and validating the assays and testing procedures that will be used in a diagnostic lab to provide the updated information to patients going forward. That’s really what we’re after.
Medgadget: I imagine that the issue of where this database lives would be a sensitive one for member institutions. Is a solution fully mapped out? In such a solution, will the data be accessible publicly, controlled by GenomOncology, or subject to some sort of hybrid arrangement?
Dr. Salisbury: The primary data live locally at the source institutions and are not centralized. That’s why I’ve been very deliberate about referring to a federated solution. As for access, it’s possible of course for institutions, within their own rights, to publish results and suggest improvements to guidelines. These suggestions usually stem from published results or formal clinical trials. So, in that sense, public access to data is generally going to be dictated by the client institutions.
Medgadget: Switching gears a bit: How does GenomOncology distinguish itself from other companies in the genomic analysis and clinical decision support spaces?
Dr. Salisbury: Certain things distinguish GenomOncology. One of them is that our GO Precision Medicine Portfolio’s software is somewhat tailored to each client. So each lab that works with us on the lab side gets its own customized implementation. With that in mind, we tend to develop very rapidly and our agile software development approach is very client-driven. Our solutions are not off-the-shelf, although they are all instantiated from the same code, and we provide a lot of services that go along with them. A portion of that is the customization and configuration, but also a large aspect of helping labs with validation is service-based. We work with many labs to bring up these assays in their labs, many of which are new to this space, not having run tests using NGS, or next-generation sequencing.
Additionally, our technology works securely and quickly. The system is installed locally behind a client’s firewall. Many other interpretation software offerings are cloud-based these days, which is really not palatable to a large fraction of the medical community. Working as closely with the customers as we have, also, we come to understand their workflow. Knowing how they conduct a test so that the system logically leads them from step to step, from order to quality control to variant review to report drafting, allows for a streamlined process.
Another aspect of what I think we do particularly well is really pay attention to the underlying biology of what we’re dealing with. Our systems do not rely on text matching or over-simplified representations of data. We have the ability, then, to understand the multiple ways that variants could be named categorically at the DNA level, the protein level, or the structural level. This is a really important aspect of working in genetics because naming conventions can vary so drastically. If you want to be successful in making clinical trial representations or interpreting guidelines, your systems have to understand the different language that can be used in their documentation.
A final differentiator is that we don’t only focus on NGS. One of our most recently announced efforts with Thomas Jefferson University Hospitals is the development of an integrated system for testing, interpreting, and reporting of NGS in combination with karyotyping and FISH assays. This is something that’s generating a good deal of attention for us. What we’re doing is bringing the established testing and putting it together with the latest generation of testing. Each has its own advantages and the combined information can have clinical ramifications.
Medgadget: Along that same line of thought, are there specific implementations or uses of your technology at certain institutions that you would like the broader scientific community to be noticing?
Dr. Salisbury: Thomas Jefferson, around that multi-assay integrated testing, is certainly telling of our client diversity. As I mentioned, in addition to oncology, we’re also in inherited disease. Our flagship customer there is Children’s National Medical Center in Washington D.C., which is working with us exclusively in inherited disease. For them, we’ve created an ordering portal that allows them to rapidly select a set of genes to examine for each patient so that each receives his or her own specialized gene list that will be examined based on the patient’s symptoms.
I spoke moments ago about working with labs that are just getting started with NGS, and that’s true — we’ve worked with many labs that are brand new to it. We’ve also, though, been working with large, established labs. Just recently we announced that UCLA Health had implemented its tumor testing using our platform, and Ohio State has been working with us as well. Many of these labs are associated with a medical institution or university, but there are also third party reference labs involved.
More generally, and, I think, more notably: Our initial entry was in the pathology lab dealing with cancer. However, we have since been rapidly expanding out from oncology into inherited disease, from solid tumor into heme, and working not just with the lab but also at the level of the hospital, the researcher, and the practicing clinician.
Medgadget: We’ve talked in depth about the direction GenomOncology is heading in the future with your services. Looking to the recent past, instead, as you bring more and more institutions online — are you learning anything from the speed of or complications in the system onboarding process, or from initial client feedback?
Dr. Salisbury: At the lab level, every implementation has its set of technical issues getting off of the ground. That’s one of the reasons we often help guide labs through the process of development on the wet lab side. That includes planning validation studies, identifying what samples are needed, and determining how to prepare a validation report to satisfy the different regulatory bodies such as CAP and CLIA.
One of the complexities involved, especially on the tumor side of things, is bringing together different types of variants across multiple genes and tracking down specific recommendations from the various medical societies as to appropriate treatments. That’s not simple to do on your own, but our software makes it easy. Our systems tell you exactly what you, as a decision-maker, need to know about the variants that they detect. That’s how we hope to improve clinical processes and, when all is said and done, patient outcomes.
Company page: GenomOncology…