Incidental Variants Are Critical for Genomics

Abstract

The topic of incidental variants detected through exome and genome sequencing is controversial, both in clinical practice and in research. The arguments for and against the deliberate analysis and return of incidental variants focus on issues of clinical validity, clinical utility, autonomy, clinical and research infrastructure and costs, and, in the research arena, therapeutic misconception. These topics are briefly reviewed and an argument is made that these variants are the future of genomic medicine. As a field, we should take full advantage of all opportunities to study these variants by searching them out, returning them to patients and research participants, and studying their utility for predictive medicine.

Main Text

This essay provides a rationale for aggressively pursuing both the research and clinical utilization of incidental genomic findings, because these variants are the future of clinical genomics. I make this plea for a host of reasons, recognizing the many challenges, arguing nonetheless that seizing this opportunity is essential for the progress of the field of research in clinical genomics, justifiable as proper medical care in the clinical context and necessary to maintain public support of the scientific enterprise. It is a challenge to the claim that such variants are “a threat to genomic medicine”¹.

Moving forward with incidental variants means that as researchers and as clinicians, we should maximize our opportunities to identify secondary variants, both in the research and in the clinical arenas. In the clinical arena, we should return those variants to patients when they meet reasonable standards for proof of causality and can significantly improve the medical care of our patients. In the research arena, we should study incidental variants to learn what they can tell us about the full spectrum of genotypes and phenotypes. Because this research improves our knowledge of incidental variants, they can be moved onto, or perhaps in some cases off of, the lists of genes and variants known to be medically useful.

At the same time, it is important to acknowledge the limitations and challenges that we currently face in both the clinical and research realms. These issues have led some to suggest that we should avoid acquiring or redact incidental variants from exome and genome results, withhold the variants that we do find from research subjects, or not include reporting of such variants from clinically indicated genome and exome testing results. A key limitation regarding incidental variants is that our ability to interpret these variants is inadequate.^1–5 Specifically, we do not have adequate data on the full spectrum of genotype-phenotype correlations⁶ and we cannot interpret the overwhelming majority of such variants. A second important issue is that, in the research realm, the analysis and return of such variants can blur the distinction of the roles of researcher versus clinician and increase the risk of therapeutic misconception.^7–9 A third limitation is logistic: in neither the research nor clinical domains do we have adequate infrastructure—physical (e.g., informatics), procedural (e.g., informed IRBs and consent processes),^10–14 or human (e.g., researchers able to identify variants and clinicians able and willing to act on such results)^4,15–17—to analyze and use incidental variants to improve the medical care of research participants or patients. These limitations and concerns are real and valid but represent only the “glass half empty” side of the equation.

The “glass half full” side of the equation is more compelling. Notwithstanding all of the challenges described above, incidental variants with important health implications can be identified^18–20 and consensus on the propriety of returning results is attainable for carefully delimited sets of genes and variants.^21,22 Also, the American College of Medical Genetics has recently recommended return of specific types of incidental variants for 57 genes and 24 conditions,²³ although there is much debate and concern about this recommendation. It is also clear that in spite of the important ethical concerns that have been raised, the ethical consensus is shifting toward returning incidental findings.^22,24–27 In addition, there is a social-political issue to address. Most genomic scientists depend entirely or primarily on the generosity and support of the public through their funding of research. There is also the altruism and generosity of the research participants^28,29, and the general public has high expectations of what they want researchers to do on behalf of participants.^30,31 The views and desires of research participants drive the effort to translate basic genomics into clinical genomics—we cannot continue to promise that the health benefits of the Genome Project³² are forever around the next corner. Incidental variants, beyond the primary variants we seek in clinical care and research, are an important additional opportunity to improve health (Box 1). The public and the body politic are likely to react negatively to the scientific enterprise if subjects are harmed by policies that preclude or discourage the return of incidental findings on the basis that it is too expensive, too difficult, too time consuming, requires too much paperwork, or is not legally obligatory to identify and return results to research participants with life-threatening predispositions to diseases.^1,3,33–36 We cannot simultaneously argue that genomics can deliver major health advances and at the same time ignore incidental mutations for severe diseases in the participants that we study. I predict that ignoring such variants would lead to a serious backlash in public opinion regarding genomic research, with major repercussions for the field.

Box 1. A Case Study in the Value of Returning Incidental Findings.

A 47-year-old man was enrolled in the ClinSeq® study because of a personal and family history of early onset myocardial infarction in the absence of hyperlipidemia. Whole-genome sequencing was performed to identify the cause of this disorder, which, to date, has not been successful. During a general review of his whole-genome sequence data, it was apparent that the average read-depth coverage of a segment of chromosome 17p was reduced by about 50%. This region corresponded to the classical deletion seen in patients with hereditary nerve and pressure palsies (HNPP; MIM 162500), and the deletion was confirmed by CLIA-validated testing. The result was returned to the participant with medical and genetic counseling, whereupon it was learned that he had been misdiagnosed with herniated lumbar discs with radiculopathy and had been recommended to undergo surgery for this diagnosis. The surgery was declined and the patient has undergone physical and occupational therapy for HNPP with good results.

If the glass is indeed half full, we must find a way forward toward an approach to healthcare that is based as much on prevention and prophylaxis as it is on symptomatic diagnosis and treatment and overcome the limitations and challenges described above. One approach would be to organize large, new cohorts of healthy patients to perform prospective studies to explore the ability to predict and avert disease. Regrettably, the United States is somewhat behind in these endeavors as compared to other countries such as Canada and the European Union. A major impediment to such studies is the enormous costs of organizing such studies and the long timelines necessary for recruitment, as well as data generation and subsequent analysis. A second, less expensive, and faster way forward may be to leverage currently funded clinical genomics research studies (and perhaps clinical genomic testing as well) to pilot approaches for predictive medicine. Investigators have described thousands of individuals who are being sequenced for a myriad of indications; intellectual disability, malformations, cancer, heart disease, undiagnosed diseases, etc. Although we celebrate the successes of many of these endeavors to understand the genetic basis of these disorders, something is being missed. These exome and genome studies are rightly focused on the primary disease that led to the sequencing, but the sequence data from all of these studies could provide a nearly free set of genomic data upon which predictive medicine exploratory research could be performed. For every child with intellectual disability enrolled in a trio sequencing study, there are four parental alleles that may harbor a mutation in a gene associated with cancer susceptibility,²⁰ cardiomyopathy, malignant hyperthermia, or other potentially occult and serious medical disorders. For every thousand such trios, there are likely to be dozens of individuals with a rare, apparently pathogenic variant in a gene that predisposes them to the future development of a significant medical condition. I suggest that research participants with incidental variants in these studies should be aggregated and studied for that secondary disorder as a part of the primary protocol, or they should be enrolled in an incidental variants protocol where the penetrance, clinical utility of reporting of the variant, and long-term outcome could be assessed and followed. It would be critical to design such studies in a manner that allowed integration of data across the multiple genes and disorders to test the hypothesis that the detection of a set of such variants improved the overall health and well-being of these incidental cohorts. The power of genomics is that it provides a broad view of the landscape of heritability – it is wasteful to discard variants in all genes other than the primary causative variant.

In spite of the challenges and expenses of addressing incidental variants, I believe that they provide us with an opportunity to build the foundation of preventive, individualized medicine.³⁷ I also believe that research participants, the public, and their elected representatives (our funders) expect us to do this and that ongoing support of clinical genomics is dependent upon us taking up this challenge and moving forward. When a field is faced with a daunting challenge, there are always a myriad of reasons not to go forward, and there will always be skeptics. But we should not ignore the challenge of incidental findings and risk missing the opportunity of all that they will teach us; the time is now.

Web Resources

The URL for data presented herein is as follows:

• Online Mendelian Inheritance in Man (OMIM), http://www.omim.org/