Scott Topper

Panel testing: one tool, many uses

Scott Topper, Invitae Clinical Genomics

Clinical genetics is an incredibly fast-moving field. It’s propelled forward by improvements in technology, by the efforts of dedicated clinicians and scientists, and by the generosity of patients and families who consent to let the world learn from their experiences. Each day we better understand the genetic basis of certain diseases, the pathogenicity of particular genetic variants, and the full range of symptoms that can present in a clinical condition.

But while science, diagnostics, and medicine all move forward, they do so at different paces. The speed of these advancements can create profound opportunities and serious challenges for the practice of clinical genetics.

Within genetic diagnostics, panel testing—the testing of several genes at once—has been widely adopted as an effective tool in navigating between the urgencies of evidence-based clinical practice and the promise of better information to come. A panel test gives clinicians access to the highest-quality, most useful information available right now and puts them and their patients in the best position to act quickly when new information becomes available.

There is sometimes a lack of clarity about the multiple utilities of a genetic test, and this confusion can impact discussions on clinical utility, billing, patient education, and other subjects. Ultimately, this confusion stems from trying to use a single lens to understand something that has multiple facets.

Most diagnostic panels actually contain three classes of genes:

  • Genes that have been proven to cause or contribute to the disease in question
  • Genes that have been proven to cause a related disease
  • Genes that are suspected (but not yet proven) to cause disease

Among these classes of genes, there are differences in both the types of possible results and the most effective use of those results.

Class 1: Genes that have been proven to cause or contribute to the disease in question

For some genes, the question of causality has been clearly resolved: an abundance of evidence establishes that pathogenic variants in these genes can cause the disease seen in a particular individual or family. Integrating this certainty into clinical practice is straightforward and routine, and the results from this class of genes directly influence clinical decision making and follow-up.

For example, pathogenic variants in the BRCA1 gene can cause hereditary breast and ovarian cancer. Finding a pathogenic variant in the BRCA1 gene in a patient with hereditary breast cancer explains his or her condition, can be used to guide treatment decisions, and can be used to determine the risk for family members. Testing, management, and treatment guidelines are clear.

Class 2: Genes that cause disease but not necessarily the disease in question

A gene can be clearly established as causative of a particular disease, but the full range of presentations for that disease may not yet be known. In some cases, anecdotal evidence suggests that a particular condition may be caused by a gene that is primarily associated with a different condition.

For example, a patient with a family history of breast cancer might be tested using a panel that includes the PMS2 gene. The PMS2 gene is well established as a cause of Lynch syndrome, which is most commonly characterized by significant increases in colorectal and endometrial cancer risk. However, there exist suggestions that individuals with disease-causing variants in PMS2 can also present with isolated breast cancer. This connection has not yet been conclusively proven, but there is widespread suspicion that breast cancer may be part of the PMS2 presentation spectrum, and there is an expectation that additional cases may support this.

The difference between Class 1 and Class 2 genes comes down to context and the presentation of the patient. In the context of a panel used to diagnose a history of isolated breast cancer, PMS2 is a Class 2 gene. But in the context of a panel used to diagnose a history of Lynch syndrome, PMS2 is a Class 1 gene.

Within Class 2, positive findings are possible and can mean one of two things: A pathogenic finding could be the explanation for the patient’s condition. On the other hand, there are individuals and families who are affected by more than one condition. A pathogenic finding in a Class 2 gene could be irrelevant to the patient’s current condition but could instead be an actionable, incidental finding indicating that a second disease may be present or imminent. Either way, pathogenic findings in these genes can be used to guide management, monitoring, and family testing.

Class 3: Genes that are suspected (but not yet proven) to cause disease

There can be a long lag between the first suggestions that a gene may cause disease and definitive proof one way or the other. In many cases, additional research eventually conclusively proves that the gene causes disease, but in some cases additional research determines that the original assertions were incorrect.

Until research proves that genes in Class 3 either cause or do not cause disease, testing these genes cannot yield definitive positive results. However, there are still two very good clinical reasons to test patients for genes that are in this class.

The first is related to the efficient use of clinical resources. A variant discovered in a Class 3 gene may in fact be a disease-causing variant—it just may be impossible to know with certainty at the time the test is performed. However, having tested for this variant puts the clinician in a position of being able to provide additional guidance as soon as definitive evidence becomes available proving that the variant either does or does not cause disease. At Invitae, the cost of the test is the same regardless of the size of the panel, so this information can be made available without needing to dedicate additional resources. We also make a commitment to our clients to keep them informed as new evidence becomes available so they can stay up to date without needing to continuously monitor the genetic literature.

The second reason is a shared commitment to improving our understanding of clinical genetics. By testing additional people for genes in Class 3, we can support efforts to resolve these important questions. With engaged families, we support collaborations with clinicians who aim to publish useful case reports, and we support collaborations with studies and organizations looking to address larger questions. We are strong supporters of the Free the DataClinGen, and ClinVar efforts, and we believe that data is more powerful when shared—while always recognizing that patients control their own data.

Variants of uncertain significance in Class 1, 2, and 3 genes

Testing can, in some cases, reveal a variant of uncertain significance (VUS): a variant whose relation to disease has not yet been proven or disproven. At a particular point in time, there may be insufficient evidence available to resolve the question with certainty. VUSs are possible in Class 1, 2, and 3 genes, yet VUSs in each of these three categories mean something slightly different, and different follow-up becomes appropriate.

Even in well-studied genes and diseases, the scientific community’s understanding of the genetics of disease can be incomplete. When VUSs are found in Class 1 genes, family studies and segregation analysis may provide information that can be useful for a more certain classification (called VUS resolution). In most cases, resolving the question of pathogenicity of a particular variant requires data from multiple unrelated families. VUS tracking in a particular family contributes to this effort and puts the clinician in a position of being able to provide additional guidance as soon as new information becomes available.

Resolving the question of pathogenicity for a particular variant is more complicated within Class 2 genes. VUS resolution depends on identifying a correlation between the presence of the variant and the presence of the associated condition. When it is not certain that the condition is truly caused by the gene, this correlation cannot be made reliably. For this reason, variant tracking of VUSs in families for Class 2 genes is generally not helpful for pursuing a more definitive classification.

In Class 3 genes (sometimes called genes of uncertain significance), positive findings are, by definition, not possible. VUSs (or benign variants) are the only possible results. Familial testing for VUSs in Class 3 genes is not recommended, as the fundamental scientific question of the gene’s relationship to disease must be answered first. Testing Class 3 genes today allows clinicians to provide additional guidance as soon as the gene and the variant are better understood.

Flexible testing today

At Invitae, we recognize that the decision about which genes to test belongs to patients and their clinicians and that different practices prompt different approaches. We offer flexibility by allowing clinicians to build custom panels and to test only the genes that they are comfortable with or that are relevant to a particular patient. We also offer clinical consultation services for clinicians who wish to discuss cases with genetic-counseling experts and we support a re-requisition for the analysis of additional genes related to the same clinical indication, at no additional charge. This allows clinicians and patients to move through different kinds of information in the way that makes the most sense for their practice.

Looking ahead

One day, when the scientific community’s understanding of the genetics of disease is complete, we will have just one class of genes: those that cause or increase the risk of disease. Testing more patients and sharing the results brings us closer to that complete understanding. Until then, though, we think it’s extremely important to understand the different reasons to test for each gene and to have the right expectations about what the results could mean for a specific patient today. We hope that this understanding will ensure that patients and families in the doctor’s office today are tested for the genes (and for the potential results) they are comfortable with—no more, no less.