NEW YORK (GenomeWeb) – Having shifted its focus over the last few years from diagnostic test development to providing patient subtyping services for drug development, GeneCentric is now working to expand its subtype classification approach to other cancer types.
The company's most recent effort, in bladder cancer, was presented earlier this month at the annual meeting of the American Society for Clinical Oncology in Chicago.
Investigators from the company and collaborators from the University of North Carolina reported in a poster presentation that using the company's Cancer Subtyping Platform (CSP), they were able to define a 60-gene expression signature that distinguished four disease subtypes.
According to GeneCentric founder and CEO Myla Lai-Goldman, while there are other bladder cancer subtyping methods, previous signatures have required data on several thousand genes, which limits their potential to be translated to the clinic and their applicability to the drug development process.
GeneCentric's approach, meanwhile, is a methodology for gleaning smaller, more manageable genomic profiles from large public datasets.
"Gene expression subtypes are nothing new," Lai-Goldman said. "The proprietary aspect of what we do is the ability to convert [a large, complex signature] into under-100-gene assays that are applicable to the clinic." She cited Nanostring's US Food and Drug Administration-cleared Prosigna test as an example.
Although focused now on developing universal classifiers for use in drug development and as companion diagnostics, the company's approach is different from the responder subtyping that is becoming more widely integrated into pharmaceutical trials.
Instead of looking at a population of treated patients and trying to identify molecular or genomic features that can distinguish responders from non-responders, GeneCentric develops classifiers a priori, based on differences in expression of cancer-related genes and pathways, immune signals, and other features.
Rather than taking drug response as the distinguishing feature and then working backward, the company defines a set of universal biological subtypes, which can then be studied in the context of drug response.
In the bladder cancer study presented at ASCO, for example, the researchers first used TCGA data from about 400 patients to assign four gene expression subtypes. They then used GeneCentric's methodology to narrow the subtype signature down from an initial 2,700 genes to 60 genes and validated it in two additional data sets.
With this in hand, the company could begin to look for insights with potential utility in the pharma space, and ultimately for companion diagnostics and patient selection in the clinic.
One insight, the authors wrote, is that there were clear differences between the four subtypes in the expression of genes that are now promising therapeutic targets in bladder cancer, such as FGFR3 and ERBB2. The bladder subtypes also showed variability in immune profiles that could inform the response to immunotherapy.
The company has already made headway in making a case for its approach with several pharma companies, announcing a partnership with Bristol-Myers Squibb in January 2017 to explore whether its platform can identify translational biomarkers for BMS's immunotherapy Opdivo (nivolumab).
More recently, the firm said it is also collaborating with G1 Therapeutics to evaluate the potential of its CSP to identify lung cancer responders to G1's oral CDK4/6 inhibitor G1T38.
Both of those projects focus on lung cancer, which is also where GeneCentric developed and commercialized its first clinical test, HistoPlus Lung Cancer, now licensed to and offered by LabCorp.
The company hasn't announced any partnerships focused on PARP inhibitors yet, but it did share data at the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics last year on the potential of its lung cancer subtyping assay to predict patient susceptibility to PARP inhibition.
Researchers analyzed the expression of homologous recombination genes and previously published BRCAness signatures, concluding that there appear to be biologic differences between the squamous cell carcinoma subtypes defined by the company's classifier, which could affect susceptibility to DNA damage response inhibitors.
Lai-Goldman said the firm has also developed a subtyping system for head and neck cancers, though she did not detail any applications of that in the pharma space. On its website, GeneCentric also lists renal, colorectal, and prostate cancers as targets that are "coming soon."
At the time it developed the LabCorp HistoPlus lung cancer test, GeneCentric was anticipating that its commercial path would be focused on in-licensing IP related to various subtyping signatures, building evidence for specific classifiers, and then licensing those to companies that would commercialize specific tests.
But it has since changed gears, building more in-house capabilities and focusing on integrating its technology into pharmaceutical development.
"We have really shifted our focus in the last few years … we have built in-house capabilities, and have understood some of the gaps in markers in drug development, and so we have shifted gears to serving the pharmaceutical community from being a strict diagnostic company — we even changed name last year to GeneCentric Therapeutics from GeneCentric Diagnostics," Lai-Goldman said.
That does not mean that the company no longer aims to develop clinical tests, though, she explained. "The shift is not in the ultimate mission, but in the means to get there, [because] gaining clinical utility for companion diagnostics is [hard for] a diagnostics company on their own," she said.
"So we felt we had to shift gears to integrate ourselves into the drug development process, so we are getting clinical utility data through the drug trial with the ultimate hope that [our classifier] will be required in the end as a companion test."