Understanding Genetic Variant Classification and Reclassification
Each human body contains over a trillion cells, including specialized cells like skin, brain, and heart cells. Within each cell resides our genetic material, DNA, which functions as the body's instructional manual. Each instruction in this manual is termed a gene, and each gene codes for a specific protein. To maintain organization, DNA is tightly wound into structures called chromosomes.
Genetic testing involves laboratories searching for changes, or genetic variants, in an individual's DNA. While most DNA sequences are consistent across individuals, everyone possesses variants. Some variants, known as pathogenic variants, can impact health, while others, termed benign variants, do not. Variants of uncertain significance (VUS), also referred to as unclear or unknown significance, lack sufficient information to determine if they are disease-causing (pathogenic) or benign.
The Variant Classification Process
Upon identifying a genetic variant, genetic testing laboratories initiate a classification process. This involves understanding the gene in which the variant was found, followed by assessing the variant's impact on the protein it codes for and its potential influence on health conditions. Unfortunately, many genetic test results yield VUS classifications, highlighting significant knowledge gaps in genetics and health. It is estimated that the genetic basis for to genetic conditions remains unidentified. Furthermore, between \% and \% of patients undergoing genetic testing in recent years have received a VUS result, underscoring the critical need for further research.
Gene-Disease Relationship Validity
Before classifying a specific variant, it is essential to determine if there is sufficient evidence that variants in that particular gene can cause a specific disease. This involves a rigorous process undertaken by curators or researchers. Researchers collect evidence from publicly published studies about the gene, including genetic evidence from individuals with variants in the gene and any observed health effects, as well as experimental studies on the gene's function. The gathered genetic and experimental evidence is then scored to determine the strength of the gene-disease relationship.
This strength is categorized into: Definitive or Strong, which is assigned when a gene's role in a disease has been consistently demonstrated in research and clinical settings over time; or Moderate, Limited, or No Known Disease Relationship, assigned as lesser evidence is available. Variants found in genes with limited, disputed, or refuted classifications are often categorized as VUS because the gene's function and its impact on health are not yet well understood.
These gene-disease validity classifications are crucial for guiding genetic testing and variant interpretation. Laboratories utilize this information to decide which genes to include on testing panels or to report in genome/exome sequencing, and clinicians also rely on this information for interpreting patient test results.
ClinGen and Gene Curation Expert Panels (GCEPs)
To standardize and build an authoritative resource for gene and variant clinical relevance, the Clinical Genome Resource Project (ClinGen), an NIH-funded initiative, was created. ClinGen aims to define which genes and variants are associated with disease and how genetic information can inform clinical care. GenomeConnect, a patient registry, contributes to this effort. ClinGen establishes expert panels, known as Gene Curation Expert Panels (GCEPs), which systematically review genes and publish gene-disease validity scores (definitive, strong, moderate, limited, no known). These GCEPs curate genes across numerous disease areas, providing a consistent framework for researchers, laboratories, and clinicians.
Specific Variant Classification (ACMG/AMP Guidelines)
Once a gene has been confidently associated with a health condition (moderate, strong, or definitive gene-disease relationship), the specific variant within that gene can be evaluated. Variant classification assesses the impact of a specific variant on health and disease. In , the American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP) published guidelines for classifying sequence variants. These guidelines outline criteria for deeming a variant pathogenic (disease-causing), benign (not disease-causing), or uncertain (VUS). Updated guidelines are currently in progress.
The guidelines utilize various pieces of evidence to determine a variant's classification, evaluating questions such as: Has the variant been studied in research or a laboratory setting? Scientists may examine symptoms in animal models (e.g., mice) or the variant's impact on cell growth. Understanding the predicted impact of the variant on the protein is critical. Synonymous (Silent) Variants are DNA sequence changes that do not alter the amino acid sequence, making them unlikely to impact protein function and often deemed benign (e.g., C changing to T, but both TCC and TCT code for Serine), suggesting the variant is unlikely to impact health. Nonsense Variants result in a premature stop codon, leading to a shortened protein. This shortening can significantly alter protein function, making such variants more likely to be disease-causing (e.g., G changing to T, resulting in a stop codon and a truncated protein).
Published literature and public databases are also critical, specifically regarding Case Examples: Have other individuals with the same variant and similar symptoms been reported in the literature? Sufficient case examples with similar symptoms may lead to a pathogenic classification. Population Data: Has the variant been observed in individuals without health conditions? A variant seen in \% or more of the general population is less likely to be the sole cause of a rare health condition. For instance, if a variant found in an individual with epilepsy, intellectual delays, and a heart defect is also reported in healthy individuals, it's unlikely to be the sole cause of these health issues.
Individual and family history considerations include: Type of Testing: Was the testing comprehensive (multiple genes) or targeted (single gene)? Other Variants: Were other variants found? Family History: Does anyone else in the family have symptoms or the variant? If a variant is new or unique (de novo) to the individual being tested, not present in biological parents who reported no health conditions, it is more likely to be disease-causing.
Variant Curation Expert Panels (VCEPs)
Similar to GCEPs, ClinGen also establishes Variant Curation Expert Panels (VCEPs) that provide gene-specific guidance for variant classification. While ACMG/AMP guidelines are general, VCEPs "flesh out" these guidelines with more detailed, context-specific information. For example, for a gene associated with a metabolic condition, a VCEP might specify additional blood work (e.g., enzyme levels) to assess protein function. A change in enzyme levels (e.g., lower than typical) in an individual with a specific genetic change would suggest the variant impacts gene/protein function, making it more likely disease-causing.
VCEPs follow these steps: First, they develop specifications by providing additional, gene-specific guidelines for evaluating variants. Second, they pilot and publish these specifications, testing them in a pilot program and, once finalized, publishing them for use by other labs, researchers, and clinicians. Third, after establishing the framework, VCEPs begin variant classification, often focusing on "tricky" variants with conflicting classifications from different laboratories or variants consistently classified as VUS (uncertain). All VCEP classifications are made publicly available on the ClinGen website and in ClinVar, a public database of variants.
The Importance of Data Sharing
Data sharing is crucial for improving the understanding of genes and variants and their relationship to health. By pooling resources and engaging in data sharing, individuals can provide invaluable case-level data that expert panels use for both gene-disease validity and variant classifications.
GenomeConnect's Role
Participants in ClinGen's GenomeConnect patient registry consent to de-identified data sharing and recontact. They upload genetic test results and provide health history via surveys. This de-identified genomic and health information is then shared with ClinVar.
ClinVar
As a publicly available database, ClinVar is widely used by laboratories, researchers, and clinicians during variant curation. GenomeConnect submits the original laboratory's classification along with additional clinical details such as the individual's age at testing, clinical features, and inheritance information (allele origin, de novo status). This case-level data provides crucial context that goes beyond the aggregate submissions often provided by laboratories, which may lack specific health or inheritance details.
Benefits of Data Sharing
Benefits of data sharing include: Population Data, where more individuals undergoing genetic testing contribute to a larger population dataset, helping determine how frequently or infrequently a variant is seen in the general population. Variant Comparison allows researchers to see if individuals with similar variants (e.g., nonsense variants in a specific gene region) exhibit similar symptoms, indicating that area's importance. Finally, Comprehensive Details from patient data sharing provide access to unpublished cases and critical case-level details (symptoms, inheritance, other variants found) that inform variant classification.
Variant Reclassification
Genetic variant classifications are not static; they can change as new information becomes available and scientific understanding evolves. The goal is often to reclassify VUS into more definitive categories (benign or pathogenic). Our understanding of already classified benign or pathogenic variants can also shift over time. A study by a cancer genetics laboratory found that over a -year period, \% of variants were reclassified, leading to approximately updated reports.
Responsibility for Recontact
While clinical laboratories and healthcare providers are expected to reassess genomic variants and recontact patients, there is no clear legal duty established. However, there is a consensus that recontacting patients about updates is beneficial. Significant barriers exist, including limited resources and challenges in tracking patients (e.g., patient relocation, change of providers).
GenomeConnect's Facilitation of Updates
GenomeConnect offers a service to help facilitate genetic report updates. If data sharing reveals possible updates, participants who have consented receive email notifications. They are encouraged to contact their healthcare provider to request an updated report. GenomeConnect emphasizes that it doesn't identify all variants and advises participants to routinely follow up with their provider (e.g., annually or biennially). To date, GenomeConnect has shared over report updates. Classification changes have occurred across all categories including: VUS to pathogenic or likely benign, pathogenic to uncertain (rare), and benign to likely benign. Individuals with questions about specific test results or who seek updated information are advised to contact the ordering doctor's office or a local genetics provider. The National Society of Genetic Counselors (NSGC) offers a search feature to find local genetic counselors.
Q&A Highlights
Accessing ClinGen and ClinVar Information
Information on ClinGen and ClinVar is publicly available. Users can visit the ClinGen website (clinicalgenomeresource.org) and use the search bar to look up specific genes. Previous webinars are available on the ClinGen playlist on how to track gene-disease relationship changes and specific variants in ClinVar for updates and alerts.
GenomeConnect Submissions to ClinVar
If GenomeConnect has submitted information for a specific variant, it will be indicated at the bottom of the variant's page in ClinVar. GenomeConnect submits "phenotyping only," meaning it reports the classification provided by the original laboratory and does not provide its own classification. This is sometimes reflected as "not provided" for GenomeConnect's own classification status to avoid conflicting with aggregate variant classifications. Additional clinical details from GenomeConnect submissions (e.g., lab name, report date, original classification) can be found by expanding the submission details on ClinVar. GenomeConnect aims to implement a better notification process for participants when their data has been shared with ClinVar.
The Role of AI in Variant Classification
Artificial intelligence (AI) is anticipated to be highly useful in variant classification, primarily by: automating tasks like checking population datasets to quickly determine variant commonality, assisting in identifying and pulling relevant publications with case-level data or experimental studies, and using AI-based "in silico" models that have long been used to predict a variant's impact on protein function and potential damage. However, AI is not expected to replace expert panels. These panels remain crucial for tackling complex and ambiguous classifications, requiring human experience, teamwork, and in-depth analysis. While AI can gather and present information more efficiently, the final layer of expert review is essential.
Downgrades from Pathogenic to VUS or Likely Benign
Downgrades of variants from pathogenic or likely pathogenic to VUS or likely benign are extremely rare. This rarity stems from the conservative and robust approach taken in initial classifications, which requires substantial evidence to categorize a variant as pathogenic or likely pathogenic. Such classifications are typically based on significant data, often upheld over extended periods. In the rare instances a downgrade occurs, individuals are advised to consult their healthcare provider and genetic counselor to discuss the implications and next steps, considering their personal and family history.