The False-negative Phenotype

(1)

The False-negative Phenotype

John D. Lantos, MD

School of Medicine, University of Missouri-Kansas City and Children’s Mercy Bioethics Center, Kansas City, Missouri

Dr Lantos wrote, reviewed, and revised the manuscript and approved the final manuscript as submitted. DOI: https:// doi. org/ 10. 1542/ peds. 2018- 1099G

Accepted for publication Jul 3, 2018

Address correspondence to John D. Lantos, MD, Children’s Mercy Bioethics Center, 2401 Gillham Rd, Kansas City, MO 64108. E-mail: [email protected] PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

FINANCIAL DISCLOSURE: The author has indicated he has no financial relationships relevant to this article to disclose.

FUNDING: Supported by the National Institutes of Health Eunice Kennedy Shriver National Institute of Child Health & Human Development and the National Human Genome Research Institute under awards U19HD077627, U19HD077632, U19HD077671, U19HD077693. Funded by the National Institutes of Health (NIH).

POTENTIAL CONFLICT OF INTEREST: The author has indicated he has no potential conflicts of interest to disclose.

NIH

Ethical controversies may arise when genome sequencing reveals a genetic variant that is thought to be pathogenic, but the patient has no symptoms. This could be due to variable penetrance or expressivity. It could also result from a misclassification of the gene as pathogenic. In this article, I analyze 2 possibilities when such a situation occurs. The first is straightforward. We could conclude that the sequencing results should be considered a “false-positive” test result. The second is a bit more counterintuitive. In some cases, we could consider the test result to be a true-positive but in way that has not yet led to phenotypic findings. Somewhat playfully, we imagine that, in such cases, we could consider the patient’s phenotype to be falsely negative. Sometimes, as odd as it seems, we act is if that is what we believe.

(2)

I am a general pediatrician and bioethicist, not a geneticist. Over the past 5 years, I had the privilege of working with a number of geneticists as part of the Newborn Sequencing in Genomic Medicine and Public Health (NSIGHT) consortium. In this role, I came to understand the power and the subtlety of genomic sequencing as a diagnostic tool. I also became aware of some of the ethical controversies that arise because many genomic test results are difficult to interpret, and the implications of those ambiguous results are difficult to convey to patients. In some cases, a genetic variant that has been thought likely to be pathogenic is found in a person who has no symptoms or phenotypic findings. There are many reasons why this could happen. Genes can have variable expressivity1

and incomplete penetrance.2_The

previous studies may have been in nonrepresentative or skewed populations.3_{It may be difficult to}

do or interpret functional testing; in a recent consensus report, it was noted, “not all functional studies are effective in predicting an impact on a gene or protein function.” 4

Sometimes, as a result of these ambiguities, cases arise in which a patient has a genetic variant that is classified as very likely to be pathogenic, but the person is asymptomatic with no phenotypic findings of the predicted pathology. At a recent conference on genomic sequencing in newborns, experts were discussing such a puzzling case. It involved a previously healthy 10-year-old who collapsed while playing soccer. She was successfully resuscitated and found to have hypertrophic cardiomyopathy (HCM). Genetic testing revealed a gene variant that had been previously reported in several unrelated patients with HCM (MYBPC3 NM_000256.3:c.1484G>A, p.Arg495Gln). This variation is seen in low frequency in the Broad Institute’s ExAC database, raising

the likelihood that it is, in fact, pathogenic. Furthermore, in silico prediction models supported a deleterious role for this variant.5

The girl’s parents and siblings were advised to have clinical evaluations and genetic testing. Her 16-year-old brother, a high school basketball player, was found to carry the same genetic variant. A comprehensive cardiology evaluation revealed no sign of HCM or other cardiac pathology. The cardiologists then had a dilemma. What should the 16-year-old and his parents be t16-year-old? Does he need periodic screening for HCM? If so, how often and for how long? How certain do we need to be that he will not develop disease before we recommend no further testing? Some of the complexities of interpreting genome sequencing results are illustrated in this case. Professional societies recommend a careful and complex process to interpret the many variants that are inevitably found in any individual’s genomic sequence. Richards et al4

write, “Our understanding of the clinical significance of any given sequence variant falls along a gradient, ranging from those in which the variant is almost certainly pathogenic for a disorder to those that are almost certainly benign.” To classify genetic variants along this spectrum, interpreters must take into consideration population data on the frequency of the allele in question, computational and predictive data (“in silico models”) in which a deleterious effect from the variant is suggested, functional studies, family history, and studies revealing that people with the variant have the disease and those without the variant do not. It is a laborious process that can be considered as much an art as a science. The art requires value judgments about the risks of “calling” a variant as pathogenic. There are risks to both false-positives and false-negatives.

There are clearly situations in which the classification of a genetic variant changes over time as a result of new information. For example, a variant in the TCAP3 gene (TCAP NM_003673.3: c.37_39delGAG, p.Glu13del) had been reported to be associated with HCM. However, further research revealed that this variant was also common in asymptomatic control populations, and it was reclassified from “likely pathogenic” to “likely benign.” 6

We are in the early days of clinical genome sequencing, so such reclassifications are likely to be common. Many will be corrected over time as we gather more data. We will also learn more about the many factors that can lead to different phenotypic findings among people with the same genetic variants, factors such as penetrance, expressivity, and epigenetic factors. With our current imperfect state of knowledge, it may be hard to know whether a finding of a pathogenic variant in an asymptomatic patient ought to be considered a false-positive test result or whether, instead, such a test result should be considered a warning flag indicating a higher-than-average probability that the person will develop disease in the future. Some have referred to people in this situation as “patients-in-waiting.” 7_{I would like to suggest,}

in a manner only partially tongue-in-cheek, that such a situation could also be characterized as a “false-negative phenotype.” With either label, the implications are that the patient will be treated as if they were at risk for developing disease in the future. This may lead to increased anxiety, extra diagnostic testing, and, in some cases, even treatment of disease that may or may not ever occur.

Ackerman et al8_{reported a case in}

which the possibility of inappropriate treatment is illustrated.8_Doctors

tested a first-degree relative of a patient who died of sudden cardiac death. The relative had a genetic finding that was interpreted as likely

(3)

pathogenic for long QT syndrome (LQTS). The man had no signs or symptoms of LQTS at the time of the molecular diagnosis. Nevertheless, on the basis of that genomic result, the doctor recommended, and the patient received, an implantable defibrillator. The authors criticized the decision and warned that, “The mere presence of a rare variant in a bona fide LQTS-susceptibility gene should not compel a pathogenic, probably deleterious variant rendering.” We do not know how common such situations are, but we do know that interpretations of the likelihood that a particular variant will be classified as

pathogenic are constantly changing.9

Such findings will always create uncertainty among both doctors and patients.

KRABBE SCREENING AND THE IMPLICATIONS OF FALSE-NEGATIVE PHENOTYPES

Although no geneticist would actually believe in the theoretical idea of a false-negative phenotype, doctors sometimes behave as if this phenomenon exists. A good example is in the interpretation of newborn screening for Krabbe disease (KD).10_{In 2006, the state of}

New York mandated the screening of all newborns for KD. Over 2 million infants have now been screened. The initial screening is metabolic testing for low levels of galactocerebrosidase. Infants with low levels of galactocerebrosidase enzyme undergo genomic testing to reduce the number of false-positives that result from testing enzyme levels alone.11_{Since 2006, when the}

program began, 51 infants had both low levels of enzyme activity and genetic variants that were classified as likely pathogenic. Of those, only 5 have developed clinical signs of disease. The other 90% have remained asymptomatic.7

The New York screening program had to decide how to counsel parents

whose children were diagnosed as being at high risk for developing KD. The parents were advised to bring their children in for frequent neurologic evaluations, with the goal of diagnosing disease as early as possible to improve outcomes after stem cell transplantation. Some of these “high-risk” children are now 10 years old. Most remain asymptomatic. Wasserstein et al10

write of these children, “They remain at risk for later-onset phenotypes, which have variable age of onset and a broad phenotypic spectrum.” There is currently no way to know how many, if any, of the currently asymptomatic children with low galactocerebrosidase enzyme levels and pathogenic genetic variants will develop disease in the future. The only way to answer that question will be to manage them for many years. In the meantime, however, such children and their families will be treated as if they have, or are at high risk of developing, a life-threatening disease and that they need to be carefully monitored.

Such testing leads to questions about how to think about the concept of “molecular diagnoses” in apparently healthy people. Given a genomic variant that would be classified as likely pathologic in an apparently healthy person, there are 2 possibilities. The tests could be wrong. Or the people could, in fact, have disease but have not yet developed phenotypic manifestations of illness.

DO WE JUST NEED MORE DATA?

One view of interpretive ambiguities is that they arise because we are so early in the era of genomic medicine. The hope is that, as we learn more about genomic variation, we will be able to give more precise interpretations of each genomic variant. That will only happen, of course, if the resources are available to do the long-term follow-up

studies that quantify differences in expressivity, penetrance, and epigenetic modification. Given the ubiquity of genomic variation, however, it is unlikely that such long-term follow-up studies will be done for even a small fraction of known (or yet-to-be-discovered) genomic variants.

The hope that uncertainty will decrease in the future is not an evidence-based or realistic hope. Instead, it seems likely that, for all but a small fraction of genomic variations, our interpretive

uncertainties will not decrease over time but will, instead, likely increase. We will see more and more cases in which a genomic variant that is accurately identified as likely to be pathologic does not correlate with phenotypic findings. In such situations, we will have to decide whether to call the test result a false-positive or the phenotype a false-negative. The decision will have implications for our assessment of the risks, costs, and benefits of genomic testing. If the test results are seen as false-positives, then genomic sequencing will seem far less useful than if they are seen and reported as accurate molecular diagnoses with false-negative phenotypes. The proper response to a false-negative test result is to reassure patients and families that they are not at risk. The proper response to a false-negative phenotype is to tell people that they are at risk, and, depending on the situation, recommend some forms of enhanced surveillance. Such enhanced surveillance will be expensive, worrisome, and, as in Ackerman et al’s8_{case report,}

(4)

ABBREVIATIONS

HCM: hypertrophic cardiomyopathy KD: Krabbe disease LQTS: long QT syndrome

REFERENCES

1. D’Amours G, Lopes F, Gauthier J, et al. Refining the phenotype associated with biallelic DNAJC21 mutations. Clin Genet. 2018;94(2):252–258

2. Kammenga JE. The background puzzle: how identical mutations in the same gene lead to different disease symptoms. FEBS J. 2017;284(20):3362–3373

3. Chong JX, Buckingham KJ, Jhangiani SN, et al; Centers for Mendelian Genomics. The genetic basis of mendelian phenotypes: discoveries, challenges, and opportunities.

Am J Hum Genet. 2015;97(2): 199–215

4. Richards S, Aziz N, Bale S, et al; ACMG Laboratory Quality Assurance Committee. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405–424

5. van Velzen HG, Schinkel AFL, Oldenburg RA, et al. Clinical characteristics and long-term outcome of hypertrophic cardiomyopathy in individuals with a MYBPC3 (myosin-binding protein C) founder mutation. Circ Cardiovasc Genet. 2017;10(4):e001660

6. Bos JM, Poley RN, Ny M, et al. Genotype-phenotype relationships involving hypertrophic cardiomyopathy-associated mutations in titin, muscle LIM protein, and telethonin. Mol Genet Metab. 2006;88(1):78–85

7. Timmermans S, Buchbinder M. Patients-in-waiting: living between sickness and health in the

genomics era. J Health Soc Behav. 2010;51(4):408–423

8. Ackerman JP, Bartos DC, Kapplinger JD, Tester DJ, Delisle BP, Ackerman MJ. The promise and peril of precision medicine: phenotyping still matters most [published online ahead of print October 8, 2016]. Mayo Clin Proc. doi: 10. 1016/ j. mayocp. 2016. 08. 008

9. McNally EM, George AL Jr. New approaches to establish genetic causality. Trends Cardiovasc Med. 2015;25(7):646–652

10. Wasserstein MP, Andriola M, Arnold G, et al. Clinical outcomes of children with abnormal newborn screening results for Krabbe disease in New York State. Genet Med. 2016;18(12):1235–1243

11. Orsini JJ, Kay DM, Saavedra-Matiz CA, et al; New York State Krabbe Disease Consortium. Newborn screening for Krabbe disease in New York State: the first eight years’ experience. Genet Med. 2016;18(3):239–248

(5)

DOI: 10.1542/peds.2018-1099G

2019;143;S33

Pediatrics

John D. Lantos

The False-negative Phenotype

Services

Updated Information &

http://pediatrics.aappublications.org/content/143/Supplement_1/S33

including high resolution figures, can be found at:

References

BIBL

http://pediatrics.aappublications.org/content/143/Supplement_1/S33#

This article cites 11 articles, 1 of which you can access for free at:

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtml

in its entirety can be found online at:

Information about reproducing this article in parts (figures, tables) or

Reprints

http://www.aappublications.org/site/misc/reprints.xhtml

(6)

DOI: 10.1542/peds.2018-1099G

2019;143;S33

Pediatrics

John D. Lantos

The False-negative Phenotype

http://pediatrics.aappublications.org/content/143/Supplement_1/S33

located on the World Wide Web at:

The online version of this article, along with updated information and services, is

the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2019 has been published continuously since 1948. Pediatrics is owned, published, and trademarked by Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it

at Viet Nam:AAP Sponsored on August 28, 2020 www.aappublications.org/news