Conventional aneuploidy screening remains “most appropriate” choice for general population


OpinionThe American Congress of Obstetricians and Gynecologists (ACOG) have updated their guidance on cell-free DNA (cfDNA) screening tests for fetal aneuploidy. In it, they state that any patient (i.e. women at high-risk OR low-risk for having an affected pregnancy) may choose cfDNA testing but they caution that conventional screening tests are more appropriate. This document replaces an earlier opinion, published in 2012, which clearly stated that cfDNA screening tests should not be offered to the general obstetrical population because they are considered to be at low-risk.

So ACOG went from recommending that cfDNA testing not be performed on low-risk women to say that they may choose cfDNA testing. Why the subtle change? Well, as ACOG correctly notes, the landscape of cfDNA is changing rapidly. New studies are published frequently and those that have examined the performance of cfDNA tests in  low-risk women have reported that the test performs just as well in them as it does in high-risk women.

However, they make an important point about a metric that doesn't get the attention it deserves. The positive predictive value (PPV). See here for background. Because the prevalence of fetal aneuploidy in low-risk women is lower than it is in high-risk women, a "positive" or "abnormal" test result in low-risk women is more likely to be a false-positive result. For example, a positive result in a 25-year-old woman gives a 33% chance that the fetus is affected but that chance increases to 87% in a high-risk woman.

The report also calls out the "no result" problem. cfDNA tests fail to produce a result in 1-8% of samples tested, usually due to a low amount of fetal DNA in the blood sample. It's becoming clear that women with samples that fail to produce a result are at increased risk of having an affected fetus. According to ACOG, these women she be offered diagnostic testing such as fetal karyotyping using amniotic fluid obtained by amniocentesis.

Other notable points contained within the updated guidance include:

  • Caution about not routinely performing microdeletion screening (offered by some labs) because it has not been fully validated in clinical studies.
  • Clearly indicating that a negative or normal result does not rule out the possibility of an affected fetus.
  • Providing genetic counseling to patients about test limitations and that decisions such as pregnancy termination should not be based on these screening tests.
  • A reminder that cfDNA tests do not screen for neural tube or ventral wall defects

This certainly won't be the final say that ACOG has on cfDNA aneuploidy screening tests. Indeed, they state that "It will be critical to remain abreast of this rapidly changing technology to provide patients with the most effective, accurate, and cost-conscious methods for aneuploidy screening."

Finally, an hCG blood test that can be performed at the point-of-care


Testing urine samples for the presence or absence of hCG is commonly performed in hospitals and clinics for the rapid assessment of a woman's pregnancy status. This topic has been discussed several times in this blog (see here and here).

Urine hCG tests are hugely popular because they can be performed near the patient and they are granted waived status by the Clinical Laboratory Improvement Amendments.

However, urine is not a suitable sample type for pregnancy assessment for many reasons, such as:

  • Urine hCG tests are prone to false-negative results.
  • hCG becomes detectable in the urine after it appears in the blood, so urine tests are not as sensitive as blood hCG tests.
  • Urine tests provide a positive/negative results whereas blood tests provide a quantitative concentration.

Because of these limitations, in particular the risk of false negative results, I've often said that urine hCG testing is inappropriate in healthcare delivery settings. Blood tests for hCG are much more reliable but they take longer to produce results because of the time required for sample transport and processing by a centralized laboratory.

Finger StickAt long last, a rapid, quantitative blood test for hCG is finally available in the US from Abbott Point of Care, Inc. on their i-STAT instrument. Naturally, people will want to know how the test performs and we recently published a paper on the analytical performance of the test.

We evaluated the test using whole blood and plasma (the sample types that were cleared for use) as well as serum. Overall, the test works quite well in all sample types and is suitable for use in healthcare settings. It provides the reliability of laboratory blood-based hCG testing but with the convenience of point-of-care testing.

There are two important limitations to note:

  1. The analytical measuring range of the i-STAT hCG test is limited to 5-2,000 IU/L. While this is similar to the measuring ranges of many laboratory hCG tests, the instruments used in labs can automatically dilute and re-test samples that have a high hCG concentration, something that the i-STAT is unable to do. When tested with the i-STAT, a sample with an hCG concentration greater than 2,000 IU/L will be reported as such. While this type of result indicates that hCG is present in the blood, not having an absolute concentration will be insufficiently informative in some clinical situations.
  2. The test is susceptible to the high-dose hook effect at an hCG concentration somewhere between 400,000 and 600,000 IU/L. This means that when a sample with an extremely high hCG concentration is tested, the result can be falsely decreased. However, even though the result is falsely decreased, it is still very likely to be interpreted as "positive" and the risk of a false-negative result is extremely remote.

This type of test is long overdue in the US. The use of urine hCG testing to determine a woman's pregnancy status is fraught with difficulties and is known to cause harm to patients. Despite their problems, urine hCG testing won't be going away any time soon but the availability of a test that performs hCG blood tests close to the patient is a step in the right direction.

Momentum grows for use of cell-free DNA Down syndrome screening tests in all pregnant women


Low risk

The use of cell-free DNA (cfDNA) testing to screen for fetal aneuploidies has been the topic of several posts on this blog. Large clinical studies that have evaluated the performance of cfDNA tests have all arrived at the same conclusion: cfDNA testing is superior to traditional biochemical screening tests for the detection of trisomy 21 (Down syndrome) and other trisomies. However, most of these studies have tested women who are considered to be at high risk (e.g. over 35 years of age or who have had an abnormal biochemical screening test) of having an affected fetus. Fewer studies have evaluated test performance in women considered to be at low risk. Because of limited data in low-risk women, the majority of professional societies recommend restricting the use of cfDNA screening tests to only high-risk women. 

This is certainly going to change, and sooner rather than later.

The New England Journal of Medicine recently published a very large, well-designed study that compared the performance of a cfDNA screening test to a biochemical screening test (the first trimester combined test) in an unselected population of almost 15,841 women.

The results were rather unsurprising. There were 38 pregnancies affected by Down syndrome. All 38 (100%) were identified by cfDNA testing but only 30 (79%) were identified by biochemical testing. While that was a significant difference in the detection rate there was a greater significant difference in the false-positive rates. There were 854 false-positive results from biochemical screening and only 9 from cfDNA screening. These numbers translate into a false-positive rate of 5.4% and 0.06% for biochemical and cfDNA screening, respectively.

As the proportion of true positive results divided by the number of all positive results, the positive predictive value answers the question: "What is the probability of an affected fetus given a positive result?” In this study, these predictive values were 3.4% for biochemical screening and 80.9% for cfDNA screening. Clearly, cfDNA offers a huge improvement.

I must stress (as I’ve done several times before) that cfDNA tests are screening tests. The better performance of cfDNA tests has, unfortunately, created the perception that cfDNA tests produce conclusive results and, as such, are diagnostic tests. This could not be further from the truth. Just as with a positive biochemical screening test, a positive result from cfDNA testing should be followed by invasive diagnostic testing. Consider, for example, that the positive predictive value of the cfDNA test that was reported in this study for the 14,947 low-risk women was 50%. That’s a coin toss! Without a doubt it is vastly better than biochemical screening but no woman should make a decision to terminate her pregnancy based on cfDNA testing alone.

So is cfDNA testing an appropriate Down syndrome screening strategy for low-risk women? Yes, it is. It’s just a matter of time before professional societies recognize that fact. Indeed, the International Society for Prenatal Diagnosis did just that in their new position statement

Stay tuned…

False negative pregnancy tests: the story continues


This post was written by Robert D. Nerenz, PhD, an assistant professor at the University of Kentucky, in Lexington.

Neg pregnancy testIn previous posts, we have discussed false negative pregnancy test results caused by hCG beta core fragment (hCGβcf), the predominant form of hCG found in urine after six weeks of pregnancy. False negative pregnancy tests can result in undesirable outcomes (including loss of pregnancy) if inappropriate treatment is given. In a recent study evaluating the performance of devices used in a hospital setting, 9 of 11 devices were found to be susceptible to false negative results when used to test urine solutions containing hCGβcf concentrations observed in normal pregnancy.

After our study was published, we were frequently asked whether these devices actually performed poorly in clinical practice or if the results we observed only occurred in a controlled laboratory environment. To address this question, we searched the database of medical device malfunctions reported to the FDA (Manufacturer and User Facility Device Experience, or MAUDE) to find reports of false negative pregnancy test results in a clinical setting. Using the search term “MAUDE pregnancy false negative” we found 707 reports between 2000 and 2014 that described false negative urine hCG results in women shown to be pregnant at the time of testing. 91 different POC hCG devices were described from 14 manufacturers, including 10 of the 11 devices evaluated in our initial screening study.

Based on the description in the MAUDE report, we subdivided the false negatives by potential cause. In 433 reports the cause of the false negative result was unknown.  132 were likely because the hCG concentration was so low that it was below the limit of detection for the test device. Of these 132, nine were associated with adverse events. We also found 142 reports that were likely due to hCGbcf hook effect and ten (or 7%) of those were associated with adverse events, including delayed prenatal care, delayed treatment of ectopic pregnancy, performance of inappropriate imaging studies and even surgery leading to loss of pregnancy. In addition to the reports documented on the FDA website, it is virtually certain that many more false negative results have occurred and have gone unreported to the FDA.

The fact that false negative pregnancy test results occur relatively frequently in clinical practice was an important finding because it highlighted the limitations of currently available devices and emphasized that this problem is not limited to one or two devices. Rather, the fact that the vast majority of pregnancy test devices performed poorly in our study and were reported to generate false negative results in clinical practice indicates that this is a much larger problem.

To decrease the occurrence of false negative pregnancy test results, contributions from multiple different groups will be required. First, the FDA should insist that device manufacturers market devices that generate positive results in all pregnant women, including those with high urine concentrations of hCGβcf. Second, clinicians at large hospitals should request that pregnancy testing be performed on serum using a quantitative assay, especially in patients with abdominal pain, vaginal bleeding or other symptoms that strongly suggest the patient might be pregnant. Quantitative serum assays can generate results in less than an hour, can detect lower concentrations of hCG than point-of-care test cartridges and are not affected by hCGβcf because hCGβcf is not present in serum. Lastly, laboratorians should work to decrease the time required to generate test results in order to make quantitative testing more appealing to clinicians. At institutions where urine point-of-care testing must be performed, laboratorians should evaluate all available options and select the device that provides an optimal combination of sensitivity and lack of susceptibility to interference caused by elevated concentrations of hCGβcf.

Unfortunately, currently available pregnancy test devices present a risk to patients. It is our hope that a coordinated effort from the FDA, manufacturers, clinicians and laboratorians will eliminate that risk.

Confusion over NIPT invites catastrophe


Timing is everything. A week after I wrote about false-positive NIPT results, the Boston Globe published an article titled "Oversold prenatal tests spur some to choose abortions" written by Beth Daley of the New England Center for Investigative Reporting. The article describes non-invasive prenatal testing (NIPT) using relatively new cell free DNA tests with a focus on women who have experienced receiving incorrect results.

The article focuses on one woman who had Sequenom's MaterniT21 PLUS test that indicated her fetus had trisomy 18 or Edwards syndrome. She initially considered immediately terminating the pregnancy and her doctor helped her locate a physician who could perform the procedure the next day. Only hours later did her doctor caution her to consider diagnostic testing which confirmed the fetus did not have trisomy 18. Additional cases in the article tell of one woman who experienced a false-positive result (confirmed by diagnostic testing) but so trusted the results of the DNA test that she aborted her pregnancy anyway and a woman who experience the trauma of a false-negative result.

Stories like this indicate a clear lack of understanding regarding the limitations of NIPT and demonstrate that physicians and consumers don't always appreciate the fact that these are screening tests. In a post on its blog about the Globe article, the Society for Maternal Fetal Medicine emphasizes just that by stating "It is important for providers to remember that cell free DNA is a screening test, and does not have the diagnostic accuracy of amniocentesis." They also point out that doctors who order DNA-based screening tests need to understand the test characteristics and they emphasize the role of genetic counseling for women who undergo screening for aneuploidy. The Society's statement was the focus of a follow-up piece by the New England Center for Investigative Reporting.

Whether aneuploidy screening is performed using DNA-based tests or by traditional biochemical screening, it is a screening test. Neither are diagnostic tests. Abnormal results from any screening test must be followed up by diagnostic testing to confirm (or not) the results of the screening test. To be misinformed on this basic fact of laboratory medicine is to flirt with disaster.

The ugly stepsister: false positive NIPT test results


Positive Negative

© Stuart Miles – Fotolia.com

NIPT (non-invasive prenatal testing) continues to get lots of attention lately. Indeed, we've written about it extensively on this blog. None of this is suprising because NIPT is a new technology that is continually evolving. Two years ago, I wrote about NIPT here and provided information showing it's excellent diagnostic sensitivity and specificity. To be clear: these tests are more accurate than traditional biochemical screening for detecting fetal aneuploides but they are still screening tests, meaning that positive (or abnormal) test results must be confirmed with diagnostic testing.

As is commonplace, with time comes experience and the lens of scruitiny has recently been focused on the positive predicitive value (PPV) of NIPT. What's a PPV? It's the proportion of true positive results divided by the number of all positive results. For NIPT testing, it answers the question: "What is the probability that a positive result means that the fetus is affected?" It is very important to stress that the PPV of any test is not intrinsic to the test. The PPV is also dependent on the prevalence of the condition in the tested population. If the condition is very rare in the tested population, then the PPV will likely be low, meaning that a positive result is more likely to be a false positive. The opposite is also true (positive test results are more likely to be "true" when the condition is highly prevalent).

NIPT is done to screen for fetal aneuploidies (extra copies of specific chromosomes) such as Down syndrome (trisomy 21), Edwards syndrome (trisomy 18), and Patau syndrome (trisomy 13). The prevelance of each of these disorders is influenced by the woman's age. As examples, the prevalence of each in a 35-year-old woman with a fetus at 10 weeks’ gestational age is 1:185, 1:470, and 1:1,500, respectively. As you might expect, the less prevalent a condition is, the more likley a positive result will be falsely positive.

This has been demonstrated for NIPT. A study published earlier this year evaluated the concordance of NIPT and cytogenetic results among cases with positive or negative NIPT results. The study examined test results from 109 consecutive specimens that were either prenatally and/or postnatally studied by fluorescence in situ hybridization, karyotyping, and/or oligo–single-nucleotide polymorphism microarray (as the definitive, or diagnostic, test). NIPT testing was performed with the Panorama (Natera, San Carlos, CA), Harmony (Ariosa Diagnostics, San Jose, CA), MaterniT21 (Sequenom, San Diego, CA), or Verifi (Illumina, Redwood City, CA).

The PPV for T21 was highest at 93% followed by a 64% PPV for T18.  The PPV for T13 was only 44%. Given the prevalence of each of these conditions, these data aren't all that surprising but they are still rather alarming. Why? Because several studies have claimed NIPT tests are >99% specific (e.g. ~1% false-positive rate). As the authors of the study described here state: "To an average clinician, the claim that a test is >99% specific leads him or her to expect that the false-positive rate will be <1%."

As I stated this above and in several other posts on this blog (but is worth emphasizing again): NIPT is a screening test, not a diagnostic test and it cannot be considered a replacement for diagnostic testing.

Improved Qualitative Pregnancy Devices


In the past, we have blogged about false negative urine qualitative hCG tests in both point-of-care (POC) hospital devices and over-the-counter (OTC) devices due to the presence of high concentrations of hCGbcf. We feel this represents a real problem for patients and clinicians trying to diagnose pregnancy and could results in harm to mother and/or fetus.

It is our understanding that the FDA is requiring device manufacturers to address this problem in any new devices going through the FDA approval process. However, it is up to manufacturers if they want to voluntarily change their existing devices. Hence we have urged manufacturers to modify their devices to eliminate false negatives due to hCGbcf.

Recently, we were made aware of two manufacturers that had apparently modified their qualitative pregnancy devices: The Cen-Med Elite Plus One-Step Pregnancy Test (a hospital POC device) and the First Response Early Result OTC device (an over-the-counter device). In order to evaluate these modifications we compared the old and new devices using the screening test we have developed previously.   Our results demonstrated that indeed, the new version of each device perform better than the previous version.  Both original devices demonstrated significantly diminished signal when 500 pmol/L hCG was tested in the presence of 500,000 pmol/L hCGbcf. However the modified devices gave faint or clear positive signals in the presence of the same hCG concentrations (see figure). Figure for Blog
It is clear that improvement of qualitative urine hCG devices is possible and we encourage all manufacturers to design devices that are not inhibited by hCGbcf. 

The hCG discriminatory zone in ectopic pregnancy: does assay harmonization matter?


Ectopic_Pregnancy_DiagramEctopic pregnancy is a potentially life threatening complication of pregnancy. It occurs when the embryo implants outside the uterine cavity, most commonly in the fallopian tube. Common symptoms include nausea, abdominal pain, and slight vaginal bleeding. Approximately 2% of all pregnancies are ectopic pregnancies and they require emergency treatment to either prevent rupture of the fallopian tube or treat a woman with a ruptured ectopic pregnancy.

Diagnosis of an ectopic pregnancy requires the exclusion of a normal, intrauterine pregnancy. This is often facilitated by the use of transvaginal ultrasound (TVUS) to visualize the location of the yolk sac or embryo, which should be able to be seen by the end of the 6th week of gestation. However, such precise dating is often not available for patients that seek medical attention with symptoms of ectopic pregnancy and so hCG testing is often used as a surrogate marker for gestational age. The serum hCG result is interpreted against the “hCG discriminatory zone.” This is the hCG concentration that, once exceeded, means that the embyro should be large enough for TVUS to always detect it. However, it is inappropriate to diagnose a patient with an ectopic pregnancy even if no embryo is observed in the uterus when the hCG concentration has exceeded the discriminatory zone. 

The hCG discriminatory zone is commonly described as an hCG concentration between 1,500 and either 2,000, 2,500 or 3,000 IU/L. Guidelines from the American College of Obstetricians and Gynecologists define it as 1,500 to 2,000 IU/L.

The management of patients with a suspected ectopic pregnancy would be easier if hCG assays were harmonized, meaning that, regardless of the assay used to measure the hCG concentration, the results would be equivalent. In other words, it would be ideal if a sample tested for hCG on one manufacturer’s test platform was equivalent to the result obtained from another manufacturer’s test platform. Unfortunately, hCG assays are not harmonized, yet many clinicians are unaware of this limitation.

Along with colleagues, I recently published the results of a study that evaluted if the lack of hCG test harmonization would affect the interpretation of an hCG result with regards to the hCG discriminatory zone. The findings confirmed that there was, indeed, a lack of harmonization between 7 commonly used hCG assays. However, when we calculated what the expected hCG discriminatory zone should be for the different assays, all of them, with one exception, would have been within 9% of the frequently cited hCG discriminatory zone cutoffs of 1,500 to 3,500 IU/L. That amount of variation is very reasonable and is not cause for alarm. The single exception was an hCG assay for which a discriminatory zone of 2,000 to 4,900 IU/L would be estimated to be equivalent. We concluded that, despite significant differences in hCG concentrations across different hCG tests, an hCG result within a discriminatory zone of 1,500–3,500 IU/L could be used without regard to a specific assay for all but one commonly used hCG test.

Watch a short video of me explaining this study!

The dawn of obsolescence for fetal lung maturity tests


Just over 3 years ago, I blogged about the declining use of fetal lung maturity (FLM) tests and questioned their clinical usefulness. I wanted to revisit this same issue again because a recent report has validated my sentiments. Actually, there have been two recent reports on this topic but the authors on that second report include myself and my co-blogger, Dr. Ann Gronowski, so in the interest of impartiality, I won't discuss our article here.

Historically, FLM tests have been used to determine the maturity of fetal lungs for a wide variety of indications (see table below) with the intent to deliver the infant if lung maturity was demonstrated.

FLM test indications

The authors of the report observed an incredible amount of disparity between doctors at different centers in regards to the use of FLM tests for different indications. Further, they speculated (accurately, I might add) that this lack of consistency suggested that as obstetricians became more comfortable with performing amniocentesis, more amniocenteses were performed which, in turn, expanded the reasons for FLM testing! They point out that this could be a major reason why there was an increase in late preterm deliveries in the United States. It was a classic self-fullfilling prophesy: more amniocentesis for FLM testing lead to more non-term deliveries when fetal lung maturity was demonstrated.

The authors then point to studies that have clearly demonstrated that infants delivered before 39 weeks of gestation are at increased risk of adverse outcomes. Obviously, this argues against such practice and elective delivery before 39 weeks is strongly discouraged by major medical associations. Indeed, it's the number one recommendation of the American College of Obstetricians and Gynecologists in the Choosing Wisely campaign. In light of compelling evidence, the authors state that "the only remaining indications (for FLM testing) are restricted to those patients in whom accurate fetal dating is absent."

I could not agree more with these authors! If you want to hear more, you can watch a short video of me discussing this topic here.

FLM testing is being used with much less frequency than it has in the past and that's a good thing. Healthcare reform (in the US) is focusing its lens on quality healthcare practices and laboratory tests that don't contribute in a positive way should not be performed. I would argue that, with rare exceptions, FLM testing is not clinically useful. I have no doubt that if they were to vanish overnight the medical community would barely take notice. I look forward to that day.

Screening for Down syndrome in the United States


This year marks the 30th anniversary of the use of laboratory tests to screen pregnancies for Down syndrome. The tests have evolved over the last three decades and as have their ability to detect affected fetuses. Currently there are several different lab tests available to screen for Down syndrome during pregnancy, including the recently developed cell-free DNA tests (commonly referred to as non-invasive prenatal testing, or NIPT).

How has the landscape of Down syndrome screening evolved over the last few decades? That was the question considered by investigators of a recent report that sought to determine the number of women screened for Down syndrome in the United States in 2011 and 2012, along with the type of test they received.

The results of this report were based on surveys completed by 168 laboratories that offered Down syndrome screening tests in 2011 and/or 2012 and are rather interesting:

  • Of the 4.13 million pregnancies that occur each year in the United States, approximately 72% are screened for Down syndrome.
  • The most common screening test is the quadruple test (50%) followed by the first-trimester combined test (18%), and the sequential test (14%). The other types of tests (e.g. triple test, full integrated test, and serum integrated test) were less commonly performed.
  • The 6 largest laboratories (those that screened more than 100,000 pregnancies each year) performed 61% of all screening tests.
  • The 32 smallest laboratories (those that screened fewer than 1,500 pregnancies each year) performed only 1% of all screening tests.
  • Between 2011 and 2012, the use of the first trimester and integrated testing increased by 27% and 6% and the use of the quad test decreased by 1.2%.
  • As shown in the figure below, there was an inverse relationship between the percent of laboratories that offered testing only in the second trimester (e.g. triple or quadruple tests) and the number of pregnancies screened each year.

Labs offering only 2nd trimester tests

That last bullet point is an important one because nearly 70% of women have their first prenatal visit in the first trimester and it is recommended that integrated screening be offered at this visit. However, because the number of laboratories offering only second trimester testing is much greater than the number of labs that offer all types of screening tests, this recommendation may not be easily adopted.