Seven years ago, Northeastern graduate Susan Dina Ghiassian was the first employee of Scipher Medicine Corp., a precision medicine company co-founded by Northeastern professor Albert-László Barabási.

The Waltham-based company where Ghiassian is senior director of data science and network medicine now has 100 employees and recently received Medicare approval for a groundbreaking blood test, PrismRA, that helps target best treatments for patients with rheumatoid arthritis.

“This is the first clinically validated rheumatoid arthritis test that predicts treatment response,” says Ghiassian, who got her master’s degree and her doctorate in physics from Northeastern.

“Patients will get a chance to be put on the right therapy from day one,” she says.

Getting Medicare approval for PrismRA involved “multi-step teamwork,” says Barabási, who remains on Scipher’s board and is Robert Gray Dodge professor of network science and distinguished professor of physics at Northeastern.

Barabási highlighted Ghiassian’s role, saying his former Ph.D. student “was the one who defined the DNA of the company, bringing the tools and the vision that the company is pursuing now.”

“This Medicare approval is really a first step towards a better future when no patient will take a drug or treatment without first doing diagnostics to decide whether the drug will work for them or not,” Barabási says.

“Treatment decisions should not be a trial and error, and sadly, for many diseases, it still is. Treatment should be tailored to the individuals. Scipher has shown this is possible,” he says.

Ghiassian talked with Northeastern Global News about how machine learning models and big data sets are connecting patients with the right medicines right away — and how network-based platforms initially developed in Barabási lab in Northeastern and later matured and commercialized at Scipher are playing an important role in this development.

Early effective treatment is key

About 1.3 million Americans suffer from rheumatoid arthritis, an autoimmune disorder affecting the joints and other body systems, according to the National Institutes of Health.

The first line of biologic treatment is usually an anti-inflammatory drug from the tumor necrosis factor inhibitor (TNFi) family, Ghiassian says. Infliximab and adalimumab are two of several TNFis approved by the FDA.

But the success rate after six months of TNFI treatment for RA is only 30%, she says.

“Most physicians take a trial-and-error approach,” Ghiassian says, typically taking RA patients off ineffective TNFI treatments only after months or even years have passed.

“That means there are 70% of these patients who are losing time” that could be spent managing the disease, which is a particular problem because progressive RA can cause irreversible bone damage, Ghiassian says.

Researchers at Scipher realized they had a powerful platform for exploring and addressing unmet needs in health care, she says.

Road trip, then network science

“We literally took a road trip” and asked physicians and payers what predictive tests they desire to have access to, Ghiassian says.

“One of the major unmet needs is predicting responders and non-responders to TNFis, before patients are put on treatment,” she says. TNFis also are used to treat other immune-mediated diseases such as inflammatory bowel disease and psoriasis.

“We realized the importance of a model that could predict what RA patients are likely to benefit from their treatments from day one. Such models not only benefit patients but also reduce the wasted cost spent for ineffective therapy by the payers.”

Scipher researchers started working with the few publicly available datasets several years ago. They eventually got access to baseline blood data and TNFi response rates for hundreds of RA patients.

Using that data, they used machine-learning algorithms to predict which patients would benefit from TNFis using genetic material from biomarkers in their blood, Ghiassian says.

She says they also validated the results by recruiting about 700 RA patients in a prospective trial to study the clinical utility of the test. The results showed that patients who took the PrismRA biomarker test to guide therapy selection were three times more likely to enter remission than patients who had not, Ghiassian says.

Medicare approval

Medicare approved coverage of the PrismRA test for certain patients with RA in October.  

“That was a major hurdle,” Ghiassan says.

She says Medicare decisions around coverage are typically followed by other insurers. “This is really huge for Scipher,” she says.

Physicians writing to the Centers for Medicare and Medicaid Services urging approval of PrismRA said they needed precision medicine to guide treatments for rheumatoid arthritis and other autoimmune disorders the way precision medicine is helping target oncology treatments for certain cancer patients.

One physician wrote that joint damage caused by rheumatoid arthritis generally “occurs within the first two years of diagnosis. Up to 70% of patients that are inadequately treated become disabled within 10 years.”

“There are a lot of side effects that come with the biologic treatments for RA, which tends to occur in older women,” Ghiassian says.

To be tested by PrismRA, patients have their blood drawn in their doctor’s office, Ghiassian says.

She says an RNA sequencing analysis is performed and a likelihood of non-response is predicted by Scipher’s commercialized PrismRA tests. The results are reported back to the physician, who can put the patient on an alternative therapy, if the patient is likely to not respond to TNFis.

Precision medicine for other autoimmune diseases

Now that a biologic test has been developed for patient response to RA treatment, the next step is to develop similar blood marker tests for other autoimmune diseases, Ghiassian says.

Thanks to the patient recruitment and data collection of RA patients, “we have access to whole blood RNA-seq data from over 20,000 patients today. That is a very large molecular database.”

“The data captures more than just RA, and should be viewed as human molecular data. A lot of these RA patients have comorbid diseases,” Ghiassian says.

Scipher, which was also co-founded by Dr. Joe Loscalzo of Harvard Medical School, continues its connection with Northeastern by sponsoring post-doctoral students through a research collaboration agreement, she says.

Northeastern developed the technology and platform to study disease through the lens of human interactome, which are networks that comprise physical interactions among cellular components, Ghiassian says, adding Barabási is “one of the pioneers” in this field.

Cynthia McCormick Hibbert is a Northeastern Global News reporter. Email her at c.hibbert@northeastern.edu or contact her on X/Twitter @HibbertCynthia.