Germline variants in POT1 have been implicated in predisposition to melanoma, sarcoma and glioma. However, these tumor associations have been derived from very small studies, or those with cohorts highly ascertained for specific cancers. In our recent study in Genetics and Medicine regarding POT1 tumor predisposition syndrome, Ambry set out to determine the prevalence of cancer types in individuals with POT1 pathogenic variants (PVs) undergoing multigene panel testing (MGPT) for a broad variety of cancer indications.
Ambry’s study contributes to efforts to characterize genes through a gene-disease validity scoring process. Broadly speaking, gene-disease validity characterization is the process of evaluating the strength of evidence that pathogenic variants in a certain gene cause a higher risk of an associated disease. Many genes can initially appear to have a correlation to a patient’s symptoms, but without data to confirm their gene-disease validity, they are referred to as “limited evidence genes.”
Evidence linking a gene and a disease is assigned points based on the quality and quantity of the evidence, and the points correspond to five categories that designate our confidence in the association (definitive, strong, moderate, limited, and disputed). When gene-disease associations are legitimate, evidentiary data tends to be generated in a timely manner. Cancer predisposition genes that remain in the “limited evidence” category for around three years tend to remain limited or are downgraded or recategorized as disputed.
Characterization of a gene-disease association is the first step in developing a gene’s clinical utility. For patients to benefit from these classifications, it is crucial to understand how a positive result should affect medical management. Once a gene is newly characterized, appropriate recommendations and management guidelines must be established. In addition to gene characterization, accurate interpretation of specific variants in the gene determines the appropriateness of recommended guidelines; therefore, robust classification efforts from laboratory partners plays an important role in medical management.
With next-generation sequencing, we are able to access huge amounts of genetic data, but without interpretation, the data is just numbers. Especially in the context of common phenotypes such as cancer, initial studies that show enrichment, meaning a possible gene/disease link, don’t necessarily translate to clinical validity. More data is needed to establish reproducibility, define increased risk, and eventually recommend medical management. An example of this concept is if we were to see an enrichment of variants in a gene in a breast cancer population that aren’t seen in the control population. The next question is, is the link and its association with breast cancer risk real?
Once gene-disease validity is established, it is still a dynamic process that evolves according to online literature review, not just regarding GDV for the gene, but phenotype-specific associations. These genes get incorporated into published guidelines. Laboratories must determine how guidelines feed into test design to ensure the test best serves clinicians and providers.
Initial observations of gene-disease associations are typically made in a “phenotype-first” approach, when variants are identified in a disease-specific cohort. Sometimes, once a clinical lab starts amassing data and testing people with all phenotypes, the same associations are not observed in a “genotype-first” setting, and we start to see these alterations in people who weren’t part of the original ascertainment.
Papers like this, which evaluate a more ancestrally and clinical diverse cohort, are important to establish the gene-disease validity.
POT1 and Tumor Predisposition
Every day, members of Ambry’s Gene Team work to better characterize genes using gene-disease validity. POT1-associated tumor predisposition (POT1-TBD) has an interesting phenotype and biological mechanism and was a fascinating subject for further research, because the only prior literature studied small cohorts; there wasn’t a lot of data on the full phenotype of individuals with POT1 mutations. When the gene first emerged, there were some case control studies, especially in the melanoma setting.
We wanted to see how this plays out in people who were ascertained for hereditary cancer, but not specifically hereditary melanoma, and whether that prevalence carries through when a different population is tested. Looking at our cohort, we realized we might have hundreds of individuals to study, and because our cohort demonstrated different tumor indications, we could ascertain the phenotype from a larger group. We found that a predisposition to melanoma did hold true in our cohort.
This paper analyzed around 220 individuals with POT1 mutations, compared to individuals with multi-gene panel testing that had no identified mutations. None of the individuals with POT1 were ascertained from a melanoma-specific panel. While still enriched for cancers in general, pan-cancer testing cohorts remove one level of bias, since the results aren’t from a specifically melanoma-enriched group. We calculated a statistically significant odds ratio, showing an enrichment of melanoma. This study showed a sevenfold increase in the presence of melanoma in the positive cohort vs the wild-type negative cohort. (An interesting aside: because our “wild-type” cohort also includes patients who were tested at Ambry, we have to keep in mind when we design our studies that our wild type often shows more cancer than actual wild type, so there is likely a greater than seven-fold increase in melanoma compared to the general population.)
Being able to show this association in another cohort lends more credibility to the association with melanoma, but we also found some other enrichments due to the large sample size we had access to. Another tumor that had been reported in association with POT1 mutations was sarcoma, specifically angio-sarcoma. It’s a rare tumor so only a few families show that possible association. We had a good number of individuals that reported sarcoma, and we had enough power to determine a statistically and clinically significant odds ratio for sarcoma in general, which broadened the possible phenotypic spectrum.
Ambry is dedicated to further study, until all human disease is understood. For a deeper dive into these topics, watch the full discussion between Carrie Horton, Jennifer Herrera-Muller, and Marcy Richardson.
