Skip to main content
Premium Trial:

Request an Annual Quote

Akoya Demonstrates Reproducibility of Protein Imaging System as Part of Clinical Push

Premium

NEW YORK – Akoya Biosciences and its collaborators have completed a reproducibility study of the company's Vectra Polaris protein imaging system as the firm works to move the platform into clinical applications.

Detailed in a paper published this month in the Journal for ImmunoTherapy of Cancer (JITC), the study looked at the reproducibility of a multiplexed immunofluorescence panel targeting six proteins for characterizing the PD-1/PD-L1 axis and, potentially, patient response to immunotherapy.

The researchers analyzed the reproducibility of the assay both within individual sites and across six research centers, finding that it was generally within the ranges required for clinical work.

"The error range of these multiplexed fluorescence assays from lab to lab is in the same range [as common clinical assays]," said David Rimm, professor of pathology at Yale University School of Medicine and an author on the study. "So I think you can argue that this has been proved successful in that it is a test that could work in routine pathology labs."

He noted that while the staining was done at each of the six participating sites, the image analysis was done centrally by Akoya. He suggested that the company would likely aim in future work to demonstrate that this analysis could reproducibly be done across multiple sites, as well.

Such a demonstration is a key step for Akoya as it seeks to move the system into clinical trials and, ultimately, develop it for companion diagnostics use.

"This is a scientific endorsement of the robustness of the system to serve that [clinical] market," said Akoya CEO Brian McKelligon. He noted that in addition to demonstrating the system's reproducibility, the study showed it offered the throughput required for clinical applications, with the platform able to run 30 slides at a time with total assay times running around 12 to 13 hours.

Marlborough, Massachusetts-based Akoya offers both a high-multiplexing research-focused tissue imaging system, called CODEX (Co-Detection by Indexing), and its Phenoptics platform (which includes the Vectra Polaris system), a lower-multiplexing, higher-throughput platform that is targeted to translational and clinical work.

Akoya acquired the Phenoptics platform from PerkinElmer in 2018. Since then, it has placed 132 CODEX systems and 455 Phenoptics systems.

Rimm noted that while labs like his have been investigating quantitative fluorescence assays like those offered by Akoya in a research setting for a decade or more, the technology has made little to no headway in clinical pathology where chromogenic immunohistochemistry, or IHC, is still the mainstay for protein analysis.

In theory, a multiplexed, quantitative fluorescence platform like the Phenoptics system could provide more fine-grained biomarker measurements, as well as more reproducible, automated readings. However, Rimm said, there has been little work done to date on analytical validation of the technology.

"There are lots and lots of publications with quantitative fluorescence, but very few inter-lab validations suggesting that it is ready to go into the clinic," he said. "That's the new thing here."

Rimm said that adoption of quantitative fluorescence among pathologists has also been held back by a perceived lack of need for the technology. Many companion diagnostics were developed and validated as either binary or semi-quantitative tests, he noted, which has entrenched this technology.

While an automated platform like the Phenoptics system may be capable of determining relatively fine quantitative differences, pathologists are less able to do so.

"We're pretty good if you just ask us to tell positive or negative," Rimm said. "But we can't reproducibly score on a 255-point scale like a machine. In fact, our reproducibility is pretty awful even with a four-point scale. So there's sort of a conflict of interest in moving to quantitative testing in that [it] requires tools that we are not familiar with, and so we'll only go down that road as a field kicking and screaming."

The emergence of cancer immunotherapies, however, has provided a potential impetus for change.

"We've had such a poor ability to select patients for immunotherapies that it is likely some of the big pharma companies are dipping their toes into the quantitative water," he said. "I don't know that this is happening, but I suspect it is, that they are trying to do quantitative work with quantitative fluorescence."

He added that this shift likely influenced Akoya's decision to organize and fund the JITC study.

"I suspect that they are trying to prepare the field for new quantitative testing that they are already doing with pharma, because the [existing] binary tests for PD-L1 pretty much fail," he said.

"This paper is really the foundation of our dialog with all biopharma," said McKelligon. "Step number one is to prove the analytical robustness and the throughput. And now, having done so, a lot of our work is around the clinical evidence that shows that this is highly predictive and that it can really impact patient care."

He highlighted recent work with Johns Hopkins dermatology professor Janis Taube, also first author on the JITC study, that identified a six-marker panel that was predictive of immunotherapy responses and outcomes in melanoma patients.

Additionally, last year, the company inked a collaboration agreement with the University of California, San Francisco Helen Diller Family Comprehensive Cancer Center to use the Phenoptics platform to develop biomarkers to help select the most effective neoadjuvant and adjuvant immunotherapies for patients with early-stage breast cancer.

In June, Akoya announced an agreement with AstraZeneca to use the Phenoptics platform for spatial biomarker analysis in immuno-oncology drug development, clinical trials, and biomarker discovery. The company also announced this month the rollout of a new service offering called Advanced Biopharma Solutions, which targets biopharma firms looking to implement the company's imaging tools in their translational pathology and immunotherapy work.

"Now it's really about getting the clinical evidence proven so that we can continue to expand our business with biopharma, with leading academic medical centers, doing translational work to ultimately get to a clinical solution and impact patient care," McKelligon said.

Rimm noted that in addition to immunotherapy, new anti-HER2 drugs that appear to work in patient populations scored as HER2-negative by traditional IHC could also be a driver of quantitative fluorescence in pathology. (Some mass spectrometry researchers are also tackling this question.)

He said that his lab is working to develop quantitative fluorescence that it hopes to begin using this fall to assess HER2 status in a research setting.

Rimm said that while Akoya is probably the current leader in terms of moving multiplexed quantitative fluorescence into the clinic, several other companies are also targeting the space, including Budapest-based 3DHistech and Seattle-based RareCyte as well as Zeiss and Leica (a subsidiary of Danaher). He added that Roche and Agilent, two of the largest providers of IHC services, are also likely working on quantitative fluorescence technologies, though they haven't made these efforts public yet.

One potential concern regarding Akoya's system is the expertise required to operate it. "Our experience was that it was hard to use, and it takes a significant amount of training," he said. "We're a little concerned about bringing it into our CLIA lab because of how hard it is to use. Will we have to hire special higher-level techs if we bring that instrument into our lab? It's not clear how that is going to work out yet."

Rimm said that other firms he sees moving into the space are offering platforms that are simpler and less expensive, which could prove an advantage.

"The Akoya platform is terrific because it's really sophisticated and it can do anything, but it's expensive and hard to use," he said. "That's the usual technology trade-off. [A platform] has less capability but it is cheaper and easier to use, or it has more capability but it is harder to use and more expensive. It's hard to predict how it is all going to play out."