Skip to main content
Premium Trial:

Request an Annual Quote

Theralink Aims to Improve Predictive Power of Immunotherapy Selection Assay With Vanderbilt Patent


This article has been updated from a version posted April 11, to correct company details and the type of input material used in Theralink's assay.

NEW YORK – Theralink said last week that it has acquired the license for a patent to measure levels of the MHC-II protein in patients tissue for immunotherapeutic purposes from the lab of Justin Balko at Vanderbilt University.

Specifically, the patent covers the measurement of MHC-II levels in tissue from a core needle biopsy for use in selecting the appropriate immunotherapy for a given patient.

Theralink has completed a technical validation of the biomarker for its proteomic breast cancer assay and has included MHC-II as one of that assay's 32 analytes for all patients.

Emanuel Petricoin, chair of Theralink's Scientific Advisory Board, contends that new and better biomarkers of therapeutic response are needed and that MHC-II shows considerable promise as one.

"The evidence that markers like PD-L1 can predict response to immunotherapeutics [such as] pembrolizumab and others is actually extremely weak," he said.

In contrast, he added, "the actual MHC-II protein that's expressed in the tumor cell epithelium … provides tremendous predictive potential in many solid tumors like breast cancer, melanoma, thyroid, [and] other malignancies."

Jarret Glasscock, CEO of Cofactor Genomics, whose OncoPrism assay predicts immune therapy response via RNA-seq, agreed that better biomarkers are needed.

With respect to the work being done with MHC-II pathway analysis, he said, a correlation between MHC-II-related gene set enrichment and immune cell infiltration has been observed.

"I think to a certain extent [MHC-II] is another proxy for immune cell infiltration," he said. "And that proxy is proving better, potentially better, at least in some studies, than, say, [tumor mutational burden]."

"I think there's been a lot of frustration in the industry with mutational burden as a proxy for immune infiltration," he added, explaining that it has proven a less sensitive biomarker than might be hoped for.

Another advantage of measuring MHC-II, he explained, is that the protein affects whole pathways related to immunotherapy response. These pathway signatures, he said, correlate with immune cell infiltrates and therefore also with predicting response.

Petricoin said that the company has "good levels of evidence" of MHC-II's predictive utility in breast cancer, as well as preclinical evidence to support its use in melanoma.

The Golden, Colorado-located company is now planning further studies to assess MHC-II's utility as a biomarker in a wider set of tumor types.

"We don't have clinical work in sarcoma, or cholangiocarcinoma, or every [other] tumor type that's out there," Petricoin said. "We don't know how MHC-II will behave as a predictive marker as immunotherapeutics make their way in to pan-tumor indications."

Petricoin commented that his company's Tumor Biomarker Platform is currently the only commercial CLIA-certified assay to include both phosphorylated and unphosphorylated proteins.

This flexibility adds value to the platform, he explained, because a protein's phosphorylation state affects both its function and whether or not a given therapy might effectively target it.

"Doctors making treatment decisions," he said, "[are] interested in all of these protein-drug targets because … you have a growing number of FDA-approved therapies for things like triple-negative breast cancer."

Cofactor's Glasscock agreed that state matters in assessing biomarker utility.

"You can't just look at [an analyte's] presence or absence," he said. "Ideally, you would look at the state, and that includes things like phosphorylation."

Last year, Cofactor published a study in Nature Scientific Reports showing that measuring the state of T-cell subtypes — whether they were activated or exhausted — outperformed both PD-L1 testing and TMB in predicting therapeutic response.

Glasscock suggested that two key factors differentiating RNA-based and protein panels is that one can get more information from less tissue, which can make a difference in clinical settings, and that transcriptomic assays can handle many more analytes than can protein panels.

Petricoin noted that Theralink's Tumor Biomarker Platform only requires a few sections of FFPE tissue from a resection, or core needle biopsy, and that in general, protein assays avoid the extrapolation of data intrinsic to transcriptomic assays. 

"Gene expression is an indicator of protein expression, but you cannot make assertions of protein abundance merely based on gene expression," he said. "The only way to effectively measure protein abundance is to measure it directly."

Petricoin said that the previous tool of choice in proteomics assays — immunohistochemistry — had been "plagued by variability in antigen retrieval, staining quality, and pathology interpretation," but that the field has been shifting toward significantly more quantitative calibrated proteomics assays.

"Our assay was originally developed to work within the clinical setting using the same tissue inputs as other genomics or transcriptomics test. This has likely allowed us to be an early entrant in the clinical proteomics space," he said.

For now, the Theralink breast cancer assay is offered as a lab-developed test, although Petricoin said that the company was open to seeking full regulatory approval should the opportunity present itself, such as in working with a biopharma on a companion diagnostic involving specific biomarkers.

Theralink is currently working with physicians to collect real-word evidence that the biomarkers in its assay impact clinical outcomes and is communicating with Vanderbilt's Balko lab about further licensing opportunities.