NEW YORK – New data highlighted this week at the San Antonio Breast Cancer Symposium has added evidence supporting the value of using polygenic risk scores (PRS) in assessing women's risk of developing breast cancer and confirmed that, when used in the clinic, PRS tests do lead to changes in patient screening. PRS scores also, according to data presented at the meeting, can refine cancer risk in individuals with benign breast disease.
While the breast cancer prevention community seems to be well sold on the notion that PRS can improve upon risk predictions, compared to clinical risk factors alone, studies of the impact this may have on patients' eventual outcomes are still pending.
To try to fill that gap, Myriad Genetics, which markets a combined test that includes both hereditary cancer mutation testing and PRS, conducted a study examining the impact of what it calls a combined risk score (CRS) — a score that couples its approximately 150-gene PRS with the Tyrer-Cuzick risk calculator, which factors in age, family history, and other clinical factors — on the rate of certain interventions. These include earlier mammography initiation and MRI imaging.
Guidelines generally recommend more intense screening for individuals who have a 20 percent or more increase in cancer risk, said Katie Johansen Taylor, Myriad's VP of clinical product research and partnerships, who present the results at the meeting on Wednesday.
PRS testing is not specifically recommended by the National Comprehensive Cancer Network as a tool to guide clinical care, but Myriad wanted to explore whether patients who do get a high-risk result with its test see the same increase in interventions as individuals who only read as high risk by the Tyrer-Cuzick algorithm.
The study used healthcare claims data from Optum Labs Data Warehouse for a total of 8,600 patients 18 or older, excluding those with past history of breast cancer and other potentially confounding historical health events.
Investigators divided the cohort into four subgroups based on predicted lifetime risk of breast cancer by either the Myriad CRS or Tyrer-Cuzick alone. The first group included all patients who were considered high risk by both measures. The second had average risk by both measures, and the final two had discrepant results, with one test indicating high risk and the other average risk.
About half the cohort had dual average risk, a little more than a quarter had dual high-risk results. The rest, about 18 percent, had conflicting results.
Investigators reported that among the CRS-high risk cases, Tyrer-Cuzick called about 26 percent of cases as average risk. Meanwhile, for Tyrer-Cuzick-high risk patients, only about 13 percent had an average-risk CRS result.
Cases with dual high-risk results had a significant increase in MRI screening, compared to the two subgroups with conflicting results. Screening mammography, though, didn't vary significantly across the groups, nor did the use of genetic counseling.
An audience member expressed surprise that the rate of MRI was so much higher for patients with high-risk CRS and Tyrer-Cuzick results when the guidelines don't directly recommend intensified screening based on PRS.
Taylor said that one of the goals of the study was to try to see if clinicians are taking PRS results seriously, and acting upon them, as clinical demand can be one factor in influencing future guidelines.
The current study did not look at cancer incidence and patient outcomes, but that could be something that the company could use claims data to help examine, Taylor said.
It's difficult to try to correlate test results with clinical outcomes, she added, "We might be waiting decades for a person with a high risk of breast cancer to actually develop breast cancer so we need to come up with some creative ways to measure and power that study."
Alongside exploring clinical impact, Myriad is also working to validate an updated version of its test targeting a larger panel of both cancer-risk and ancestry-related variants.
The test's performance across populations is a feature the company strongly emphasizes, and it was the focus of another presentation, on Thursday that reported results from a validation study of the CRS using the UK Biobank.
Myriad's UK Biobank study featured an updated version of its assay, which the company expanded to include 59 ancestry markers and about 300 breast cancer risk SNPs. The cohort included 200,000 women, a small subset of whom — about 12,000 — were non-White. Investigators compared CRS results to breast cancer incidence at a median of 11 months followup.
Comparing CRS testing with the use of Tyrer-Cuzick clinical risk assessment alone, the team saw that CRS did appear to outperform standalone risk testing. The calls Myriad's test made that Tyrer-Cuzick missed did show an increased incidence of breast cancer.
Fergus Couch, a researcher at the Mayo Clinic, called the test's performance stability across the White and non-White groups "really nice."
In a separate study, Mayo Clinic researchers also sought to demonstrate whether there is added value in using PRS to assess risk in patients with benign breast disease. This study was also an effort to replicate results from a prior collaborative study that showed that a PRS score consisting of 261 genes could identify patients with benign findings that were at a higher versus lower risk of developing malignancy.
Investigators tested banked samples from almost 4,000 women with either non-proliferative, proliferative without atypia, or atypical hyperplasia.
"Think of these as progressions events as you go along through the spectrum of benign breast disease," said Couch. The study compared cases — those who later developed breast cancer — with controls who did not.
In each of the three subgroups PRS was higher in those who developed cancer. Surprisingly, there was no significant difference between the three subtypes.
"That was a little unexpected," Couch said. "We would have thought maybe that the non-proliferative group, which is a very, very early form of benign breast disease, would not necessarily have read out that way."
Overall, he said, the data indicate that PRS can be combined with the obvious risk factor of having benign breast disease to better measure risk.
Couch said that while Myriad has been making strides in providing clinical PRS testing for breast cancer in the US, other countries are only now starting to embrace this type of approach.
"We're seeing it start to take hold, but there's still a number of implementation studies that are being done around the world … that are still trying to show the clinical efficacy," he said.