Ask a GC: Genetic Counseling for Polygenic Risk Assessment (PRA), Chapter 2

Series Introduction: Let’s be real - as a clinician, you have a lot of information to keep up with. Whether in primary care or a specialty field, the basic knowledge maintenance, continuing education requirements, and overwhelming barrage of new or updated innovations, companies, tests, methodologies, guidelines, etc. that clinicians have to keep up with can leave many feeling like there is barely any time to actually take care of patients. 

This can be especially true in regards to the ever-evolving field of Genomics. When our team at GenomicMD needs guidance wading through the chaos, we lean on our expert Genetic Counselors (GCs)-professionals with Master's degrees in Human Genetics & Genetic Counseling, which make them uniquely qualified to serve as translators for the alphabet soup that is genetic testing. This knowledge and experience has been so invaluable to the GMD team that it felt wrong to keep it all to ourselves, so 'Ask a GC' was created to 'share the wealth' with the clinicians (and patients) we serve. The next time you need a general update when it comes to Genomics (and particularly polygenic testing) let our GC team help guide you through the chaos and get to the heart of what you need to know, so you can get back out there where you want to be-with your patients!

Please note: Though GenomicMD's Ask a GC series is written under the guidance of genetics professionals and may refer to recent medical recommendations, it is not intended to be used as personalized medical advice. Patients seeking medical evaluation and clinicians seeking assistance with the development of healthcare plans should seek referral to an appropriate specialty provider.

Chapter 2 - The History of Polygenic Risk Scores

Hello, and welcome to the latest edition of Ask a GC! This series, and more specifically, our next few clinically-focused posts, will likely be most helpful for genetic counselors and other clinicians looking to hone their polygenic testing knowledge. However, please feel free to read on if you are interested in learning more about Polygenic Risk Scores (PRS), no matter your experience, background, or goals! 

PRS play an important role in personalized risk assessments and population health screening. However, there is a scarcity of published guidelines on how to implement them in a clinical setting. This series is not intended to serve as an official guide, though we do hope that it will help get you up to speed on this type of testing. Previously in Chapter 1, we reviewed some definitions, differentiated PRS from other types of genetic testing, summarized how they are created, and discussed clinical applications. Let’s take a moment now to go over PRS history, noting examples of the various types of diseases in which they can be useful for stratifying risk. 

Expanding on Chapter 1 - A Review of GWAS

We discussed in Chapter 1 that PRS are made possible by Genome-Wide Association Studies (GWAS). These studies allow for two things: the identification of dozens to millions of single nucleotide polymorphisms (SNPs), and the quantification of said SNPs’ effects on one’s risk for developing certain conditions. As a reminder, one's PRS are the summation of the quantified risk associated with the SNPs they are found to be carrying in their DNA. The PRS developed by GWAS are single values specific to an individual that represent that person’s risk of developing a condition relative to the population average (this is where the term "relative risk (RR)", a popular format in which to present PRS, comes from). For example, a PRS for Patient XYZ may show that they have twice the average risk for developing type 2 diabetes due to their unique SNP profile.

The first large-scale GWAS was completed in 2007 by the Wellcome Trust Case Control Consortium (WTCCC). The consortium brought together over 50 research groups from the United Kingdom that were involved in various aspects of human disease genetics research. The main experiment studied 2,000 cases and 3,000 controls for 7 diseases: 

1. Bipolar disorder (BD)
2. Coronary artery disease (CAD)
3. Crohn’s disease (CD)
4. Hypertension (HT)
5. Rheumatoid arthritis (RA)
6. Type 1 diabetes (T1D)
7. Type 2 diabetes (T2D) 

The study aimed to learn more about the genetic contributions to these diseases, as well as address quality control, design, and analytical methods important to GWAS. 

All-in-all, and after subtracting those that did not meet quality control measures, the study identified close to 400,000 SNPs that had a population prevalence of ≥1% (i.e., the SNPs were found in ≥1% of the study population). 

—>This 1% threshold is often used to help prioritize the efforts of a GWAS and prevent spurious disease associations from being (incorrectly) identified. As a small aside, these SNPs are typically called "common", while those found in fewer than 1% of the population are called "rare".

The identified SNPs were then used in the association analysis portion of the GWAS, which determined whether they were associated with an increased risk for disease (and, if so, to what degree). The study replicated the findings of previous research, which had identified SNPs in four regions of the genome for CD, four regions for T1D, and two regions for T2D, as well as many more significant disease associations. 

Since that time, thousands more GWAS have been performed, leading to advances in the ability to identify SNPs associated with common, complex diseases. Next, let’s look at some examples to give you a general idea of the areas in which this exciting type of genomic analysis is improving our understanding of, and approach to, disease.

Schizophrenia

The first use of a PRS model was in 2009 by the International Schizophrenia Consortium using GWAS from over 3,000 European individuals with schizophrenia and a similar number of controls. Their study supported that genetic contribution to schizophrenia risk is not caused by a single gene or mutation, but rather by the combined effects of multiple genes. They also found that polygenic schizophrenia risk is substantially shared with polygenic bipolar disease risk. 

Depression

• A 2013 study by the Cross-Disorder Group of the Psychiatric Genomics Consortium examined the relationship between five psychiatric disorders based on GWAS data from the Psychiatric Genomics Consortium, finding:
  • High correlation between schizophrenia (SCZ) and bipolar disorder (BPD),
  • Moderate correlation between SCZ and major depressive disorder (MDD), and ADHD and MDD, and
  • Low correlation between SCZ and autism spectrum disorders.
• In 2018, Wray et al. identified 44 risk variants associated with major depression by GWAS, citing that approximately one quarter of the heritability for MDD is due to common genetic variants. They sum up the potential benefits of better defining polygenic contributors to MDD risk by saying, “Subsequent empirical studies may show that it is possible to stratify MDD cases at first presentation to identify individuals at high risk for recurrence, poor outcome, poor treatment response, or who might subsequently develop a psychiatric disorder requiring alternative pharmacotherapy (e.g., schizophrenia or bipolar disorder). This could form a cornerstone of precision medicine in psychiatry.”

Alzheimer’s Disease (AD)

• In 2017, Desikan et al. used genotype data from 17,000 cases and over 37,000 controls from the International Genomics of Alzheimer’s Project to identify AD-associated SNPs. They created what they call polygenic hazard scores (PHS) for each participant, which quantify individual differences in age-specific genetic risk for AD. This is important in modifying risk for the disease beyond the more commonly-known APOE gene variants. 

Breast Cancer

• In 2007, Hunter et al. conducted a GWAS of breast cancer by genotyping over 528,000 SNPs in 1,145 women and 1,142 controls. They identified four SNPs in the FGFR2 gene that were highly associated with breast cancer, which was later confirmed by additional studies. 
• In 2013, Michailidou et al. completed a meta-analysis of 9 GWAS totalling over 10,000 breast cancer cases and over 12,000 controls. They selected close to 30,000 SNPs for further genotyping. Those SNPs were then analyzed in over 45,000 cases and 41,000 controls from 41 studies in the Breast Cancer Association Consortium (BCAC). The study identified SNPs at 41 new breast cancer susceptibility loci (physical sites or locations within a gene or other DNA segment of interest), with additional analysis suggesting the presence of more than 1,000 additional loci involved in breast cancer susceptibility.

Cardiovascular Disease 

• In 2015, Mega et al. examined the association of a 27-SNP PRS with coronary artery disease (CAD) using data from samples of ~48,000 individuals enrolled in several CAD studies (JUPITER, ASCOT, CARE, and PROVE IT-TIMI 22). Their study found that the PRS correctly identified those individuals at increased risk for CAD events. Interestingly, the study also found that individuals with high-risk scores on the PRS derived greater relative and absolute CAD risk reduction with statin therapy than those with lower-risk scores. 
• Of note, the American Heart Association recently (2022) published a scientific statement on their stance regarding the use of PRS for cardiovascular disease:  “Relatively recent advances in population genetics have uncovered the polygenic basis of…cardiovascular conditions…These observations point to the possibility of using genetic profiling to inform clinical practice in significantly larger groups of individuals than for whom monogenic cardiovascular variants are considered… Such PRSs may be appropriately considered in select scenarios, given the current evidence base.”

Colorectal Cancer (CRC)

• In 2012, Jo et al. completed a small study of 187 cases and 976 controls (all of Korean ancestry) analyzing the effects of conventional CRC risk factors (such as family history) in combination with PRS. They found that PRS improved the prediction of CRC when used together with age and family history risk factors.
• In 2013, Dunlop et al. combined the effects of age, gender, and family history with a 10-susceptibility loci genotype and found that the genotype data provided additional information that has the potential to stratify a population into CRC risk categories. This could help with targeted prevention and surveillance.
• In 2015, Hsu et al. used data collected from more than 12,000 participants in 6 studies to develop colon cancer risk prediction models. By adding a 27-SNP PRS to the typical risk calculation method using family history alone, they were able to increase statistically significant discriminatory accuracy between CRC cases and controls. They assert that these findings could lead to more individually tailor screening recommendations. More specifically, they argue that individuals at very low risk could require fewer colonoscopies in their lifetime, helping to reduce unnecessary procedures. Conversely, those with very high risk may need to start scans at earlier ages to reduce overall risk of developing CRC. 

Diabetes

• As early as 1994, Davies et al. were suggesting a polygenic inheritance mechanism for type 1 diabetes after identifying several chromosomal regions with evidence of disease linkage using semi-automated fluorescence-based technology and linkage analysis. 
• In 2020, Vujkovic et al. analyzed data from over 228,000 cases and over 1 million controls of multi-ethnic populations and reported on 568 genetic associations with type 2 diabetes. They found that a high PRS was strongly associated with increased risk of T2D-related retinopathy and modestly associated with chronic kidney disease, peripheral artery disease, and neuropathy. They assert that their findings may help identify therapeutic targets for T2D.
• In 2022, both Ge et al. and Prive et al. published studies working to better represent diabetes polygenic scores in populations other than the historically studied Europeans, among other goals. 
• Just last month, Billings et al. published a study demonstrating how PRS can help differentiate between type 1 and 2 diabetes in patients with atypical presentations of the disease that would otherwise lead to uncertainty of specific diagnosis.

Multi-Disease Studies

• In 2018, Khera et al. reported they developed and validated PRS for five diseases using an initial validation dataset of over 120,000 UK Biobank Phase 1 participants and then an assessment in an independent testing set of over 288,000 participants of the UK Biobank Phase 2 data set. They found that 8.0%, 6.1%, 3.5%, 3.2% and 1.5% of the population studied had greater than three-fold increased risk for coronary artery disease (CAD), atrial fibrillation, type 2 diabetes, inflammatory bowel disease, and breast cancer, respectively. 

Additional Diseases/Conditions with Established PRS Utility

• Lung cancer-associated variants
• Prediction of continuous traits in the population like BMI
• PRSs of left ventricular cardiovascular magnetic resonance phenotypes are predictive of heart failure events independently of clinical risk factors in the UK Biobank
• Rheumatoid arthritis
• Multiple Sclerosis
• Early-onset inflammatory bowel disease
• Ulcerative colitis
• Adiposity and fat distribution

Discussion 

Polygenic risk assessment is slowly but surely becoming a more prevalent method of accurately and successfully identifying a person’s individual risks for certain common multifactorial diseases. Its use is becoming more widespread in various clinical areas, examples of which have been touched upon in this blog post. Studies have shown that PRS are capable of improving healthcare outcomes both in individuals and in the healthcare system as a whole (stay tuned for more specifics on this in upcoming blog posts!). 

While becoming familiar with a new type of risk analysis may take some time and a little practice, it will be worth it. Personalized genomic medicine is becoming more and more common across our ever-evolving healthcare landscape. Whether you actively pursue ordering PRS or not, chances are high that you’ll see these sorts of reports trickle into your practice sooner or later. Familiarizing yourself with their history, methodology, and clinical utilization now will ensure that you’re ahead of the game when that day comes! 

Our next post in this series will dive into how PRS performance can be evaluated, how these scores can be communicated in a healthcare setting, and limitations to be aware of regarding this type of risk analysis.

Disclaimer 

This blog post is provided by GenomicMD’s certified genetic counselors solely to serve as a helpful resource and tool for genetic counselors and other clinicians. GenomicMD’s blog posts do not (and are not intended to) dictate an exclusive course of management, nor guarantee a particular outcome. 

Blog Glossary

• Genome-Wide Association Studies (GWAS): Large, population-based studies that scan the genomes of hundreds to thousands of people looking for patterns and other signs of disease associations in the form of SNPs. 
  
  
• Mendelian/monogenic genetic testing: Tests that analyze for mutations/variants in consecutive sequences of nucleotides within a gene, and in that gene a mutation impacts the gene’s function with sufficient magnitude to result in significant risk for disease.
  
  
• Personalized medicine: Healthcare plans/methods that are customized to each individual patient.
  
  
• Pharmacogenomic testing: Analyzes how a person’s genetic code affects their metabolism of certain medications.
  
  
• Polygenic Risk Assessment (PRA): The analysis for the presence of many genetic variants (SNPs) across the genome which, combined together, impact a person’s risk for developing certain common conditions or traits. The results of a PRA are presented as a single or group of PRS (typically one per disease).
  
  
• Polygenic Risk Score (PRS): A single value that represents a person’s relative risk for developing a condition compared to the risk for that condition in an average individual within the population. Note that, depending on the context, PRS can also mean polygenic risk scores (plural).
  
  
• Relative Risk: The likelihood the person will develop a given disease compared with (or relative to) the average individual. Some testing methodologies will compare results to a more specific subset of people, like those of a particular ancestry or background.
  
  
• Single-Nucleotide Polymorphism (SNP): The most common type of genetic variation among people. This type of genetic variation takes place at a single nucleotide placement in our DNA.
  
  
• Traditional medicine: The method of caring for patients using similar standards no matter the age/sex/race/etc. of the patient (with obvious exceptions, like pediatric vs. adult, and sex-specific screens/treatments).

Referenced Articles

• American Heart Association: O'Sullivan et al. Polygenic Risk Scores for Cardiovascular Disease: A Scientific Statement From the American Heart Association. Circulation. 2022 Aug 23;146(8):e93-e118. doi: 10.1161/CIR.0000000000001077. Epub 2022 Jul 18. PMID: 35862132; PMCID: PMC9847481.
  
  
• ASCOT: Moser M. The ASCOT trial. J Clin Hypertens (Greenwich). 2005 Dec;7(12):748-50. doi: 10.1111/j.1524-6175.2005.05298.x. PMID: 16330898; PMCID: PMC8109301.
  
  
• CARE: Sacks et al. The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events Trial investigators. N Engl J Med. 1996 Oct 3;335(14):1001-9. doi: 10.1056/NEJM199610033351401. PMID: 8801446.
  
  
• Chatterjee N, Shi J, García-Closas M. Developing and evaluating polygenic risk prediction models for stratified disease prevention. Nat Rev Genet. 2016 Jul;17(7):392-406. doi: 10.1038/nrg.2016.27. Epub 2016 May 3. PMID: 27140283; PMCID: PMC6021129.
  
  
• Cross-Disorder Group of the Psychiatric Genomics Consortium; International Inflammatory Bowel Disease Genetics Consortium (IIBDGC). Genetic relationship between five psychiatric disorders estimated from genome-wide SNPs. Nat Genet. 2013 Sep;45(9):984-94. doi: 10.1038/ng.2711. Epub 2013 Aug 11. PMID: 23933821; PMCID: PMC3800159.
  
  
• Davies et al. A genome-wide search for human type 1 diabetes susceptibility genes. Nature. 1994 Sep 8;371(6493):130-6. doi: 10.1038/371130a0. PMID: 8072542.
  
  
• Desikan et al. Genetic assessment of age-associated Alzheimer disease risk: Development and validation of a polygenic hazard score. PLoS Med. 2017 Mar 21;14(3):e1002258. doi: 10.1371/journal.pmed.1002258. Erratum in: PLoS Med. 2017 Mar 28;14 (3):e1002289. PMID: 28323831; PMCID: PMC5360219.
  
  
• Dunlop et al. Cumulative impact of common genetic variants and other risk factors on colorectal cancer risk in 42,103 individuals. Gut. 2013 Jun;62(6):871-81. doi: 10.1136/gutjnl-2011-300537. Epub 2012 Apr 5. PMID: 22490517; PMCID: PMC5105590.
  
  
• Ge et al. Development and validation of a trans-ancestry polygenic risk score for type 2 diabetes in diverse populations. Genome Med. 2022 Jun 29;14(1):70. doi: 10.1186/s13073-022-01074-2. PMID: 35765100; PMCID: PMC9241245.
  
  
• Hsu L et al. Colorectal Transdisciplinary (CORECT) Study; Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO). A model to determine colorectal cancer risk using common genetic susceptibility loci. Gastroenterology. 2015 Jun;148(7):1330-9.e14. doi: 10.1053/j.gastro.2015.02.010. Epub 2015 Feb 13. PMID: 25683114; PMCID: PMC4446193.
  
  
• Hunter et al. A genome-wide association study identifies alleles in FGFR2 associated with risk of sporadic postmenopausal breast cancer. Nat Genet. 2007 Jul;39(7):870-4. doi: 10.1038/ng2075. Epub 2007 May 27. PMID: 17529973; PMCID: PMC3493132.
  
  
• International Schizophrenia Consortium; Purcell SM, Wray NR, Stone JL, Visscher PM, O'Donovan MC, Sullivan PF, Sklar P. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature. 2009 Aug 6;460(7256):748-52. doi: 10.1038/nature08185. Epub 2009 Jul 1. PMID: 19571811; PMCID: PMC3912837.
  
  
• Jo J, Nam CM, Sull JW, Yun JE, Kim SY, Lee SJ, Kim YN, Park EJ, Kimm H, Jee SH. Prediction of Colorectal Cancer Risk Using a Genetic Risk Score: The Korean Cancer Prevention Study-II (KCPS-II). Genomics Inform. 2012 Sep;10(3):175-83. doi: 10.5808/GI.2012.10.3.175. Epub 2012 Sep 28. PMID: 23166528; PMCID: PMC3492653.
  
  
• JUPITER Study Group. Rosuvastatin to prevent vascular events in men and women with elevated C-reactive protein. N Engl J Med. 2008 Nov 20;359(21):2195-207. doi: 10.1056/NEJMoa0807646. Epub 2008 Nov 9. PMID: 18997196.
  
  
• Khera et al. Genome-wide polygenic scores for common diseases identify individuals with risk equivalent to monogenic mutations. Nat Genet. 2018 Sep;50(9):1219-1224. doi: 10.1038/s41588-018-0183-z. Epub 2018 Aug 13. PMID: 30104762; PMCID: PMC6128408.
  
  
• Liana K Billings et al. Utility of Polygenic Scores for Differentiating Diabetes Diagnosis Among Patients With Atypical Phenotypes of Diabetes, The Journal of Clinical Endocrinology & Metabolism, 2023;, dgad456, https://doi.org/10.1210/clinem/dgad456
  
  
• Mega et al. Genetic risk, coronary heart disease events, and the clinical benefit of statin therapy: an analysis of primary and secondary prevention trials. Lancet. 2015 Jun 6;385(9984):2264-2271. doi: 10.1016/S0140-6736(14)61730-X. Epub 2015 Mar 4. PMID: 25748612; PMCID: PMC4608367.
  
  
• Michailidou et al. Large-scale genotyping identifies 41 new loci associated with breast cancer risk. Nat Genet. 2013 Apr;45(4):353-61, 361e1-2. doi: 10.1038/ng.2563. PMID: 23535729; PMCID: PMC3771688.
  
  
• Privé F, Aschard H, Carmi S, Folkersen L, Hoggart C, O'Reilly PF, Vilhjálmsson BJ. Portability of 245 polygenic scores when derived from the UK Biobank and applied to 9 ancestry groups from the same cohort. Am J Hum Genet. 2022 Jan 6;109(1):12-23. doi: 10.1016/j.ajhg.2021.11.008. Erratum in: Am J Hum Genet. 2022 Feb 3;109(2):373. PMID: 34995502; PMCID: PMC8764121.
  
  
• PROVE IT-TIMI 22: Cannon et al. Pravastatin or Atorvastatin Evaluation and Infection Therapy-Thrombolysis in Myocardial Infarction 22 Investigators. Intensive versus moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med. 2004 Apr 8;350(15):1495-504. doi: 10.1056/NEJMoa040583. Epub 2004 Mar 8. Erratum in: N Engl J Med. 2006 Feb 16;354(7):778. PMID: 15007110.
  
  
• The Wellcome Trust Case Control Consortium. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 447, 661–678 (2007). https://doi.org/10.1038/nature05911
  
  
• Uffelmann, E., Huang, Q.Q., Munung, N.S. et al. Genome-wide association studies. Nat Rev Methods Primers 1, 59 (2021). https://doi.org/10.1038/s43586-021-00056-9
  
  
• Vujkovic et al. Discovery of 318 new risk loci for type 2 diabetes and related vascular outcomes among 1.4 million participants in a multi-ancestry meta-analysis. Nat Genet. 2020 Jul;52(7):680-691. doi: 10.1038/s41588-020-0637-y. Epub 2020 Jun 15. PMID: 32541925; PMCID: PMC7343592.
  
  
• Watanabe, K., Stringer, S., Frei, O. et al. A global overview of pleiotropy and genetic architecture in complex traits. Nat Genet 51, 1339–1348 (2019). https://doi.org/10.1038/s41588-019-0481-0
  
  
• Wray et al. Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium. Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depression. Nat Genet. 2018 May;50(5):668-681. doi: 10.1038/s41588-018-0090-3. Epub 2018 Apr 26. PMID: 29700475; PMCID: PMC5934326.