Emerging Polygenic Risk Score Tests Demonstrate Potential for Improving Cancer Screening

Polygenic risk score testing (PRS) is one of the most promising fields of genetic testing. PRS tests use algorithms to analyze multiple genetic risk factors, often in combination with family history, age, and other personal medical information, to generate a score estimating an individual’s risk of developing, redeveloping, or having certain diseases, such as diabetes, dementia, and cancer. Although PRS technology and adoption still have a way to go, it is progressing quickly, especially in the cancer and oncology space. This year has already seen some key developments in PRS breast, prostate, and colorectal cancer testing. Here is a quick briefing:

Enhanced PRS Model Demonstrates Potential to Improve Breast Cancer Detection

Breast cancer detection has been a focal point for development and validation of PRS tests. Many of these PRS tests are performed in conjunction with the Tyrer-Cuzick (T-C) model that uses a wide array of genetic, family history, age, and other personal risk factors to calculate the likelihood that a person will develop breast cancer within the next 10 years and over their lifetime.¹ However, some lab companies have developed products based on their own proprietary models to rival or supplement T-C. Among the latter is a cross-ancestry, integrated risk score (caIRS) test developed by personal genomics firm MyOme that is used in combination with T-C.

A retrospective validation study that MyOme performed with Natera concludes that the caIRS model was more effective than standard T-C alone in predicting breast cancer. Published in JCO Precision Oncology, the study evaluates the association between the caIRS and breast cancer risk in retrospective data from two validation cohorts with longitudinal follow-up from 130,000 women.² The researchers then compared the caIRS associations with breast cancer to five-year and lifetime breast cancer predictions generated by T-C alone.

The study found that caIRS outperformed T-C used by itself among all populations in both cohorts, with Black and Hispanic women seeing the biggest gains. For example:

• Positive predictive value of the caIRS showed a twofold increase over T-C alone in identifying high-risk Black women in cohort 1
• Use of the caIRS improved the odds ratio per standard deviation—i.e., risk associated with study participants in terms of change in risk per standard deviation—among Hispanic women in cohort 1 from 1.31 to 1.88
• The caIRS reclassified the estimated five-year and lifetime risk for 8.8 percent of cohort 1 and 6.7 percent of cohort 2, as compared to the T-C model alone

MyOme is planning to launch an integrated breast cancer PRS lab-developed test based on the caIRS sometime in 2023, according to an article on the company’s site.³

PRS Shows Promise for Improving Colorectal Cancer Screening

Typical colorectal cancer screening begins when individuals reach a certain age and may involve stool tests, flexible sigmoidoscopy, colonoscopy, and/or computed tomography (CT) colonography.⁴ However, screening does not consider the subject’s individual genetic factors. At the recent annual conference of the European Society of Human Genetics (ESHG), a group of scientists from the Institute for Molecular Medicine Finland (FIMM) presented the results of a study demonstrating the potential of PRS to improve screening.

Using data from FinnGen, a collection of health and genome data from more than 400,000 Finnish individuals,⁵ the researchers calculated a genome-wide PRS for colorectal cancer. “A challenge of many prior PRS studies is that they have been performed in smaller datasets that are not representative of the general population, but in this study we used epidemiological and statistical approaches to calibrate our estimates with that population,” explained FIMM physician Max Tamlander, MD, who presented the study results on behalf of the research team, in an ESHG press release.⁶

The researchers found that PRS could be useful in assessing future colorectal cancer risk after a colonoscopy and identifying people who might benefit from more frequent surveillance rather than basing such screening on just age or family history. “Our findings are well in line with other studies on PRSs in breast cancer, another common cancer with organized population-level screening,” Tamlander noted in the press release.⁶

Combining PRS with PSA Testing May Improve Prostate Cancer Screening

The lack of reliability of the prostate-specific antigen (PSA) as a biomarker undermines the effectiveness of PSA blood tests currently relied on as the principal means of screening for prostate cancer risk. Thus, while high PSA levels are associated with prostate cancer, the cancer is often low grade or indicative of other conditions that pose no real threat to the patient. This often leads to anxiety and costly, invasive interventions that are not medically necessary.⁷

However, recent studies indicate that use of PRS in combination with PSA testing can improve prostate cancer diagnosis and obviate the need for medically unnecessary biopsy and treatment. The most recent example is a new study published in Nature Medicine on June 1.⁸ Based on analyses of the genomes of nearly 96,000 men without prostate cancer, researchers from Stanford University and their colleagues identified 128 specific sites in the genome that may affect PSA levels. The genetic “noise” generated by these sites accounts for between 30 percent to 40 percent of PSA level, according to the team. Use of a PRS could predict nearly 10 percent of PSA level variations. The researchers also found that the score was more effective among men of European ancestry than those of East Asian or African ancestry.

After applying the score to a group of men with and without prostate cancer, as confirmed by biopsy, the team concluded that use of the PSA test in combination with the PRS could have spared 30 percent of those men from having to undergo a biopsy. The PRS-adjusted PSA level testing was especially effective in detecting more aggressive forms of prostate cancer in men of European ancestry.

While expressing their determination to develop a score that would benefit men of all ancestries, the researchers noted in a press release that “even a small improvement in screening could save lives, given that one in nine men in the United States will be diagnosed with prostate cancer and one in 40 will die from it.”⁹

So What?

PRS testing remains a work in progress. The industry got a sobering reminder of how much work they still have to do to win over providers and payers in early June when Medicare Administrative Contractor Novitas Solutions published a local coverage determination imposing strict new coverage limits on 13 cancer genetic tests, including products from Abbott, Castle Biosciences, and Interpace Biosciences.¹⁰ Others have also raised concerns about PRS testing. In a recent statement highlighting issues with PRS and the lack of clinical guidelines for its use, the American College of Medical Genetics and Genomics recently advocated against clinical use of PRS testing “unless the provider and patient have a clear understanding of the limitations of the testing and applicability to the specific patient, including how the results will be used to guide evidence-based clinical care.”¹¹ However, despite the need for caution and more evidence to support its use, PRS testing continues to hold great potential for improving screening for a variety of diseases.

Cancer Genetic Tests Subject to New Novitas Medicare Coverage Limits

References:

1. https://ibis-risk-calculator.magview.com/
2. https://ascopubs.org/doi/full/10.1200/PO.22.00447
3. https://myome.com/news/myome-natera-show-cross-ancestry-polygenic-risk-score-improves-breast-cancer-prediction/
4. https://www.cdc.gov/cancer/colorectal/basic_info/screening/tests.htm
5. https://www.finngen.fi/en
6. https://www.eshg.org/news-home/for-media/2023
7. https://www.g2intelligence.com/prostate-cancer-national-guidelines-have-not-stopped-overuse-of-low-value-psa-screening-in-older-males/
8. https://www.nature.com/articles/s41591-023-02277-9
9. https://med.stanford.edu/news/all-news/2023/05/prostate-cancer-personalized-psa.html
10. https://www.cms.gov/medicare-coverage-database/view/lcd.aspx?lcdId=39365&ver=80
11. https://www.g2intelligence.com/acmg-statements-highlight-ongoing-concerns-around-polygenic-risk-scores/