Skip to content

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.


Prenatal Genetic Testing: When is it “Toxic Knowledge”?

Can prenatal genetic test results destroy the joy of pregnancy? (Credit Jane Ades, NHGRI)

When I was a fetus, I was left alone. My amniotic fluid and chorionic villi were untouched, and I arrived apparently normal, except for a mark on the part of my anatomy where a bullet “directly bit”  Forrest Gump.

Perhaps that’s too much information. But TMI is precisely the problem when prenatal genetic testing detects a DNA sequence variant, but we don’t know exactly what it means. The fuzzy line between “unusual or uncertain” and “abnormal” in test results can cause great anxiety. This is especially so when “abnormal” before birth doesn’t affect health after, thanks to a characteristic of genes termed incomplete penetrance: not everyone with a genotype develops the corresponding phenotype.

A recent article in Genetics in Medicine examines the distress of a sample of women receiving “abnormal” fetal test results. Lead investigator Barbara Bernhardt, MS, a genetic counselor and co-director of the Penn Center for the Integration of Genetic Healthcare Technologies, reports that some of the participants called the test results “toxic knowledge.”

Tales From Genetic Counseling
I’ve been providing genetic counseling at an ob/gyn practice since 1984. Back then, before the Internet educated so many health care consumers, I’d routinely explain amniocentesis, the prenatal test that had become a rite of passage for women of “advanced maternal age” – 35+ – and those with family histories of detectable syndromes.

Amnio brought good news for about 98% of patients. Others learned of a chromosome abnormality – extra, missing, swapped or inverted — that signaled a known condition. And a few patients received less meaningful information.

One fetus had an unusual inverted Y chromosome. Fortunately, so did the very normal dad. A female fetus had an extra X chromosome. Despite my assurances that the only “symptom” was tallness, the patient, seeking genetic perfection, ended the pregnancy.

Fetal chromosome checks continued through the 1990s. Finding the most common problems became faster, and more women chose chorionic villus sampling, done earlier in pregnancy. And the roster of single gene tests grew.

Then a new type of test came along that tracked repeats of DNA sequences. These “copy number variants” – CNVs — range from 1,000 to 3 million DNA bases. They’re smaller versions of the 5-million-base or larger deletions and duplications that chromosome stains and fluorescent markers reveal in traditional chromosome charts, called karyotypes.

DNA Repeats Are Information Too
Different numbers of repeats at a particular genomic address can track with specific medical conditions. Chromosomal microarray (CMA) technology uses DNA probes on a chip to count an individual’s repeats.

CMA to analyze CNVs – genetics has the worst abbreviations — edged into pediatrics in 2005, to support diagnoses of autism, developmental delay, and intellectual disability. Prenatal applications began a year later. But there’s a world of difference ,in what to do with genetic information, between prenatal and postnatal. Prenatal testing reveals genotypes before parents can see phenotypes (the results); postnatal testing attempts to assign genotypes that might explain phenotypes.

When CNV analysis provides concrete information, that’s good. It can clear up mysteries. For example, The American College of Obstetrics and Gynecology advisory from 2009 recommends a CMA test if ultrasound reveals a fetal anomaly, but the karyotype is normal. CMA is also used to clarify abnormal maternal serum marker levels and for patients who are unusually anxious about the pregnancy.

Like amnio results, most CMA test results are normal. The Genetics in Medicine paper uses the term “of known or probable clinical significance” for the few cases that have suspicious repeats, using specific criteria listed below. But some CMA test results use the term “variant of uncertain significance,” or “VOUS,” if the repeat number isn’t attached to a particular diagnosis.

One Mother’s Story
My student Sherrie shares her experience with finding a VOUS in her son’s genome.

“From birth, he was different from the other babies, but most of my concerns had been dismissed by the pediatricians who had been trained to looked for ‘horses, not zebras.’ Even as his unique cluster of physical, developmental and behavioral quirks sent up more red flags than a mini-golf course, my husband continued to be reassured by the doctors’ claims of our son’s ‘normalness.’”

Sherrie consulted a geneticist, who suggested CMA testing. It revealed a small DNA sequence repeat, like a genetic stutter, not reported in the medical literature.

“Knowing that our son has a VOUS is very unlikely to help him. Our pediatricians are now much more likely to attribute all of his quirks to his genetic differences.” Sherrie resents the stigmatization and the genetic determinism. “Maybe he’s bouncing off the walls because you kept us waiting in a tiny exam room for 90 minutes, not because he has a small genetic microduplication!” The couple is keeping their son’s genetic uniqueness private.

Sherrie is glad she learned of the duplication after she’d fallen in love with her son. “I cringe at the thought of receiving this news before I had a chance to hold him in my arms and gaze into his impossibly big brown eyes.”

Finding Copy Number Variants in a Fetus – A Different Story
The parents-to-be in the Genetics in Medicine study faced the word “abnormal” because the investigators weeded out “benign” DNA variants. But for some of the participants, uncertainty persisted. “For most of the women, the abnormal microarray results left them initially shocked, anxious, confused, and overwhelmed,” the study concluded.

Just weeks before the paper appeared, Diana Bianchi, M.D., Executive Director of the Mother Infant Research Institute at Tufts University School of Medicine wrote, presciently, in Nature Medicine, “With chromosomal microarray studies, the detection of CNVs of unknown clinical meaning or variants that have known effects but incomplete penetrance raises parental anxiety, as well as the possibility of termination of a clinically unaffected fetus.”

The 23 women in the Genetics in Medicine study were part of a large, multicenter prospective trial that is analyzing CNV detection using chorionic villus sampling or amniocentesis, among 4,450 women seeking prenatal testing for the usual reasons. The researchers screened DNA from fetal cells and DNA in blood samples from both parents for duplications or deletions associated with more than 80 syndromes, as well as more mysterious variants.

A committee of genetic counselors, doctors, and geneticists classified each extra or missing bit of DNA. It could:

• identify a specific condition (such as 22q11.2 deletion syndrome).
• nott identify a known condition; it is “likely benign”
• not identify a known condition, but be ”potentially clinically relevant.”

The physicians informed couples only for a specific condition or potentially clinically relevant results.

The researchers asked several questions to term a finding “potentially clinically relevant”:

1. Is a duplication or deletion also in a parent, like the boy with the inverted Y? Then it’s less worrisome. If the variant arose anew, or “de novo” in the fetus, and it isn’t in the medical literature, other questions arise.

2. Is the CNV in a known gene?

3. How many DNA bases are involved? Size matters. Bigger is worse.

“If a deletion or duplication is small, inherited, and contains no genes, it’s likely to be benign. But a de novo deletion or duplication that contains one or more functional genes and is a particular size is potentially clinically significant,” Barbara Bernhardt explains.

The researchers interviewed by phone the 23 women given abnormal results, 6 months after the births. Because they’d received normal karyotypes a week before the abnormal CMA results, the women were shocked, even though they’d agreed to be called if something turned up.

For 12 of the couples, one parent had the CNV, so all seemed ok. Of the 11 de novo cases, 7 were known conditions and 4 potentially clinically significant. Altogether, 7 couples chose to end the pregnancy, 6 of them with de novo CNVs.

The Anxiety of Uncertain Genetic Test Results
The researchers wouldn’t reveal the pregnancy outcomes – critical information, I think, for evaluating whether the worry was worth it – but the participants’ comments are telling.
• Women perceived CMA testing as “an offer too good to pass up.” It added no further risk to the invasive testing they were already getting.

• After normal chromosome tests, the women were “blindsided by the results” of the CMA.

• Parents-to-be believed that if they searched long enough, they’d find a name, description, and prognosis for a CNV.
• Some women were very disturbed when health care practitioners urged abortion.

• Many women regarded the CMA test results as “toxic knowledge” that had ruined the experience of pregnancy.


Half the women who continued their pregnancies still worry that their children have something wrong with them that hasn’t shown up yet. And the parents told the variant came from them are not exactly thrilled.

A similar situation arising from newborn screening is dubbed “patients in waiting,” when an abnormal metabolite level never blossoms into the anticipated disease. It’s a little like when Michael J. Fox’s character in Back to the Future sees his criminal uncle as a toddler behind the bars of his playpen and says, “Get used to it.” I wonder if we’ll get used to all this genetic information.

TMI Revisited
Time may take care of the TMI aspect of genetic testing in general, and in CMA testing in particular.  The International Standards for Cytogenomic Arrays Consortium is building a public database from clinical testing laboratories that “will help to classify CNVs of uncertain clinical significance and to translate this knowledge for the improvement of patient care.” But because we can already sequence a fetal genome from DNA in a woman’s circulation, the microarray-based test may be obsolete before it can become more meaningful.

Meanwhile, patients taking genetic tests must deal with sometimes fuzzy results. “Physicians and patients alike crave certainty,” wrote Pamela Hartzband, M.D. and Jerome Groopman, M.D. in a recent New England Journal of Medicine editorial. They were referring to the PSA marker test for prostate cancer, but the same can be said for genetic testing.

My student Sherrie sums up the situation of overreliance on genetic testing:

“Our son is not a microduplication. He’s not a variant of uncertain significance. He’s a little boy. I have to come up with the right words to share the information with him some day. Perhaps by then, the databases that should have been set up long before these tests were offered to the public will be filled with so many examples of genetic differences that it will be a non-issue. Perhaps I will be able to say to him, ‘Son, a long time ago the doctors did a blood test, and at first they were worried. But now they realize that we all have little differences in our DNA. That’s what makes us special.’”

  1. Thanks for an illuminating post about uncertainty, toxic information, and perhaps excess enthusiasm about prenatal testing. Wondering if sooner or later, this will fit in with the analysis in so much of medicine today warning of the hazards of overdiagnosis and overtreatment…but Americans – and maybe many elsewhere don’t like living with uncertainty.

    Best wishes on your new blog! It’s great to see you here.

  2. I turned up this article during a literature review and greatly appreciate your thoughtful analysis. I checked and see no other results published (yet). The study raises so many questions and how useful are these tests in improving the health of newborns and comfort to parents? How many women or their newborns with “normal” genetic tests experienced other type problems (e.g., PIH, gestational diabetes, or congenital heart defects). The article raises more question than answers.

    Thank you for your blog. It has become my “go to” for rational and compassionate information.

Leave a Reply

Your email address will not be published. Required fields are marked *

Add your ORCID here. (e.g. 0000-0002-7299-680X)

Back to top