Three years on, what have we learned about non-invasive prenatal screening?

Sequenom's updated logo

Sequenom’s updated logo

Three years ago, a new prenatal test for Down syndrome entered the market offering earlier, safer, and more accurate results. What have we learned about this new technology since then?

It’s non-invasive prenatal screening, not diagnostic.

On October 17, 2011, Sequenom, a San Diego biotech company, launched MaterniT21, a prenatal test for Down syndrome and other conditions with an extra chromosome. Competitors have since entered the market offering their own brand: Ariosa’s Harmony, Verinata’s verifi, and Natera’s Panorama. These tests promise very high detection rates and very low false negatives based only on a test of the mother’s blood.

When the testing was being developed, it was thought it would be diagnostic. When it came to market, however, it still had false positives and false negatives, rendering it not a diagnostic test, but a screening test. To emphasize this point, the American College of Medical Genetics & Genomics (ACMG) labeled the testing “Non-Invasive Prenatal Screening” or “NIPS.”

No matter how the test results are returned or told by a medical provider, a MaterniT21, verifi, Harmony, or Panorama test is never positive.

Tests placental DNA, not fetal DNA.

At the time of the launch, NIPS was described as testing cell free fetal DNA (cffDNA). It turns out the testing is of DNA from cells in the developing placenta, not the fetus. There is a phenomenon known as “placental mosaicism” where cells in the placenta can have a third copy of the 21st Chromosome, the most common cause for Down syndrome, but the fetus does not have Down syndrome. This is one of the reasons these tests will have false positives and false negatives, because they are not testing DNA from the fetus.

Tests’ accuracy is limited to high-risk populations.

The study Sequenom relied on to take its test to market was of a population with an incidence rate as high as 1-in-8, meaning one out of every eight pregnancies was positive for Down syndrome. This high of a rate does not happen in the natural population. While the chance for Down syndrome increases with the age of the mother, even a 45-year-old mom’s chance of having a child with Down syndrome is 1-in-35. The other testing companies have also reported their accuracy levels in populations already designated as “high-risk,” usually women 35 and older or those who already had a prenatal test showing an increased chance for Down syndrome. As a result, the guidelines published by the National Society of Genetic Counselors (NSGC), the International Society for Prenatal Diagnosis, and the American Congress of Obstetricians and Gynecologists and the Society for Maternal-Fetal Medicine, recommend NIPS be offered only to women who are already considered “high-risk.” The stated accuracy of the new tests are not reaching their advertised level in the commercial setting.

Sequenom’s technology is not patentable.

Soon after each test entered the market, lawsuits were filed against each other for claims of patent infringements. In a dispute between Sequenom and Ariosa, it was ruled that Sequenom’s technology is not patentable because it was not innovative and used conventional methods. More on that court decision can be found here.

Insurance coverage is limited to high-risk moms.

Due to the professional recommendations, while NIPS is increasingly being covered by more and more insurance plans, that coverage remains limited to tests performed on mothers considered “high-risk.” As shown in the charts below, this is because the clinical studies have yet to show that NIPS provides reliable results in low-risk moms.

When a “positive” means a 50/50 (or lower) chance.

The professional recommendations are supported by NIPS having a much lower predictive value in low-risk populations.

Positive predictive value (PPV) means how likely a “positive” test result actually means the tested-for condition is present. PPV is the number of true positives divided by the total number of positive tests returned, both true and false positives. When the new test is performed in populations considered low-risk, a positive result can have a PPV of less than 50%.

For example, a 30 year old mother has about a 1-in-1,000 chance of having a child with Down syndrome. At that incidence rate, the chart below shows the number of true and false positives the new tests would report at a detection rate of 99.5% and a specificity rate (the rate of ruling out unaffected pregnancies) of 99.9% for 100,000 30-year old pregnant women, 100 of which would actually be positive for Down syndrome:

Down syndrome pregnancy

Unaffected pregnancy

Test Positive

99.5 (True positive)

100 (False positive)

Test Negative

0.5 (False negative)

99,800 (True Negative)

Therefore, a 30-year old woman receiving a “positive” test result could be in the 99.5 who are actually carrying a child with Down syndrome or the 100 who were not. This gives her a 50/50 chance. With the chance for having a child with Down syndrome tied to a mother’s age, it is more likely that a woman under 30 receiving a “positive” NIPS result is not actually expecting a child with Down syndrome. (More on this analysis here).

Imperceptible reassurance for younger women.

The overwhelming majority of pregnant women are not carrying a child with Down syndrome. Therefore, almost all women considering prenatal testing are not doing so because they expect to receive a positive result. Rather, they are seeking reassurance through a negative result. This is what is promised by NIPS’ low false negative rate. But, for younger mothers, Robert Resta, a genetic counselor and former editor of the Journal of Genetic Counseling, considered that level of reassurance “not exactly a great scientific achievement.”

This is because younger mothers already have close to, if not the same, chance of not having a child with Down syndrome as the test’s ability to screen out those pregnancies that are not affected.

The NIPS labs quote a specificity rate, the ability to rule out those unaffected pregnancies, of 99.9%. But, as the table for Negative Predictive Value (NPV) below shows, mothers under 30 are already at that 99.9% rate based solely on their age:

Figure 3 NPV maternal age

NPV maternal age

Therefore, the high accuracy rate is largely imperceptible in low-risk populations, particularly mothers under the age of 30.

NIPS is not the holy grail of Prenatal Testing.

In 2013, the former chair of ACOG’s ethics committee recounted her recent experience of being an older mom—over 40—and expecting her fourth child. She reported how her obstetrician exulted with the new testing, “this is the Holy Grail. It’s what we’ve been waiting for all these years.” This reaction has been common to the new prenatal testing, but it is incorrect, and not even shared by the testing companies themselves.

The notion of a “holy grail” of prenatal testing is a test that can diagnose the fetus without exposing the pregnancy to the risk of miscarriage. Current diagnostic tests—a first-trimester chorionic villus sampling or the more commonly performed second-trimester amniocentesis—both have a chance, however slight, of procedure-related miscarriage. This is due to a needle being inserted into the womb. However, these tests remain the only way still to receive a diagnosis.

Every professional organization’s statement on NIPS makes clear this point: that diagnostic testing is necessary to confirm results and women receiving a NIPS test should be offered diagnostic testing. Therefore, because NIPS is still a screening test, and not a diagnostic test, it is not the holy grail of prenatal testing.

But, one of the predictions when Sequenom launched has proven true.

Dramatically fewer invasive diagnostic procedures.

Due to its high rates, it was predicted that mothers would choose to rely on their NIPS results and forego exposing their pregnancies to the risk of miscarriage from diagnostic testing. This is, in fact, what is happening.

Mothers receiving a negative NIPS result are choosing to avoid the needle of diagnostic testing. Those receiving a positive NIPS result who intend to continue their pregnancy similarly are relying on the NIPS result to prepare for having a child with Down syndrome rather than risk miscarriage to know for certain. The testing companies have found that the rate of diagnostic procedures has plummeted following the introduction of NIPS.

The need for educational materials for patients and providers persists.

When Sequenom launched its test, the study it relied upon recognized then that “educational materials for both patients and providers” were needed. Since 2008, in fact, federal law has recognized the need for patients to understand the accuracy of prenatal tests, to receive written materials about the tested-for condition, and to be referred to parent support organizations. In the summer before Sequenom launched, the NSGC and the American Academy of Pediatrics each published guidelines recognizing that patients receiving a positive prenatal test result for Down syndrome should also receive accurate written materials about Down syndrome and the chance to contact fellow parents. Since launching, all of the NIPS laboratories have emphasized their role in providing information to patients.

But, none of them have yet to provide information consistent with federal law and professional guidelines when they deliver a test result.

Sequenom has had a role in the disunity of information to be provided, while Ariosa and Natera are providing the Lettercase booklet Understanding a Down Syndrome Diagnosis to their provider networks. Whether the patient receives written materials or referral to support organizations is dependent upon their health care provider. But, when ACOG surveyed its own members after its guidelines changed to offering all women prenatal testing, less than 3-in-10 said they provided educational materials to their patients.

In 2013, a mom received her test result over the phone. She sat in her car in the parking lot crying after hanging up. She wanted more information, but, as she said,

Nobody offered anything. Nobody told me about any resources.

The federal law has gone unfunded and unimplemented by the Department of Health & Human Services. Tired of waiting, Massachusetts, in 2012, and then Kentucky in 2013 passed a law requiring health care providers deliver with each test result for Down syndrome educational materials from the National Center for Prenatal & Postnatal Down Syndrome Resources and referral to the patient’s local parent support organization. In 2014, four other states passed their own versions of the law. The materials from the National Center are the only ones that have been both approved by the NSGC and recognized by the ACMG as resources that patients should receive with a test result.

The NIPS market was estimated at $1.6 billion in 2013. Verinata was acquired by Illumnia for $350 million. Increasingly the cost of testing is being covered by insurance and government health programs. Yet, none of the testing manufacturers are providing the recommended information to accompany their tests and insurance companies are not covering the costs for health care providers to provide this information.

What we’ve learned about NIPS.

Three years on, much has been revealed about the new prenatal test for Down syndrome. While not the holy grail of prenatal testing, it is a screening test, of placental DNA, effective in high-risk populations, that is resulting in fewer invasive diagnostic procedures, and is increasingly being covered by insurance. But the need identified when NIPS launched in 2011, and by a federal law passed in 2008, for patients to receive educational materials about Down syndrome from their health care providers and from the testing companies remains, despite these resources being available.

Comments

  1. Mary Davenport says:

    The new cell free DNA tests are a huge advance over predecessors. Thousands of fetal lives have been saved by the avoidance of amnios and chorionic villus samples.

  2. When Premaitha Health launch their test in January 2015, is their accuracy not higher still? It does not look like the test will be available in the US to begin with

  3. I’m not sure you appreciate the difference that this accuracy has compared to quad marker screening.

    Even if it is only 50/50 for low risk moms that is a huge improvement over the 1:200 (.005%) that is considered a “positive” screen using the quad marker test. Due to existing low risk there is no value to an amnio with only a .005% chance of abnormality, so the decline in amnios should be expected. While this reduces the detection of abnormalities coincidentally picked up during an unnecessary amnio for a woman with greater than 1:200 risk, the solution is to continue to improve genetic screening methodology and not to encourage more unnecessary amnios.

    ALL positive NIPS results should result in further diagnostic testing, and 1st trimester ultrasound should continue to be offered. But it is shameful that in an age when highly accurate tests are available women are still being told “you screened positive for T21” due to a risk slightly over 1:200. Quad marker screening needs to go away, and frankly I don’t think this website is helping.

    • I do appreciate the difference in accuracy between cfDNA and conventional screening for the limited number of conditions cfDNA screens for, and I have made the bioethical case that the principle of nonmalefecence can justify offering cfDNA prior to any diagnostic test for the reason you note of cfDNA screening’s ability to be even more accurate at ruling out an affected pregnancy. Quad screening however detects many more conditions and is dramatically more affordable than cfDNA screening and conventional screening of low risk mothers remains the most appropriate because cfDNA screening has yet to show the same levels of accuracy in the general population as in the high risk population. The number of comments from expectant mothers on many posts on this website expressing appreciation for providing helpful information to them which they did not receive from their health care provider or testing laboratories is the most effective counter to your final remark.

      • I undertook IONA test based on my quad results of 1-65 for Downs (Edwards/T18 was 1 – 2000) my IONA tests results returned as one in a million for Downs however, over 75% for Edwards T18. I opted for amino based on the high percentage for a fatal trisomy and my results have returned as ‘normal’
        How can 3 tests all come back with differing results?
        As you can imagine I am ecstatic to learn my baby is normal but the worry and despair I faced waiting for all these test results was horrendous.
        I feel that NIPS is not advertised correctly and women believe that if they receive a high NIPS then the baby is destined to have that particular disorder.

        • Thank you for sharing your experience, thought I wish it had not had so much anxiety. Screening tests have false positives and false negatives. Your quad result was a false positive for Down syndrome, your IONA was a false positive for Trisomy 18. This is why only diagnostic tests can tell you for certain about your child’s genetics. You’re right: tests like IONA are marketed as providing reassurance, but your experience demonstrates how that promise isn’t always kept. I wish you a healthy pregnancy and delivery.

Trackbacks

  1. […] Update: Since NIPT launched two years ago, much as been learned about what it is testing for, the level of its accuracy, with the laboratories emphasizing it remains a screening test requiring diagnostic testing to confirm NIPT results. These lessons learned since NIPT’s launch are collected in this post here. […]

  2. […] Update: For more lessons learned about NIPT (including why it’s NIPS) click here. […]

  3. […] NIPS first launched on October 17, 2011, Sequenom was the first in the market with its MaterniT21. Then in 2012 came Verinata’s […]

  4. […] are the most accurate option. Still, because these are screening tests, and not diagnostic tests, they can include false results —some babies with Down syndrome are not identified and others without Down syndrome will be […]

Leave a Reply

%d bloggers like this: