Short course in genetics and screening
Every pregnancy is at risk of chromosomal abnormalities, which are generally more common than most professionals and pregnant women appreciate, about 1 in 150 live births. With many early losses occurring due to aneuploidy, the prevalence is higher the earlier in pregnancy one looks. All ob/gyns are familiar with trisomy 21, which is an aneuploidy (abnormal number of chromosomes). However, many other types of chromosomal abnormalities (eg microdeletions and microduplications) are responsible for the difference between the frequency of trisomy 21 (1 in 800 live births) and the 1 in 150 number cited above.
While anomalies, maternal age, and a history of a prior affected pregnancy all increase risk of chromosomal abnormalities, nothing except a diagnostic test (ie, chorionic villus sampling or amniocentesis) can eliminate risk, at least to the limits of current technology.* Even then, I tell patients we can say that there is no chromosomal abnormality to the limits of current technology. (Affordable, rapid whole exome sequencing may eventually represent the ultimate test for this in the future).
Practice Bulletin #163 (PB) contains a short course in genetics and screening, with a great deal of important information for ob/gyns. A few basic points made in the PB always bear repeating.1
1. Screening works in low-risk populations, where the detection rate (sensitivity) has to be balanced with avoiding false alarms (population false-positive rate). The false-positive rate is often confused with the positive predictive value (PPV). How often have we all heard a patient say she doesn’t want a screening test “because it’s almost always wrong”? That is a function of the low PPV of most of our tests. A screening test that shows a 1/100 risk of trisomy 21 essentially is “incorrect” 99% of the time.
2. When you perform multiple screening tests, unless they are designed to function together (eg, the quad test, where each of the 4 blood tests has been studied and found to be independent of the others), you risk increasing the number of false-positive results with an uncertain improvement in the actual detection rate for the abnormality.
3. Every pregnancy is at risk of chromosomal abnormalities, and “[a]neuploidy screening or diagnostic testing should be discussed and offered to all women…”
This PB contains a thorough review of the many screening tests available, 9 by my count, with the advantages and disadvantages of each in a very helpful table. The limitations of ultrasound are also succinctly outlined: among others, low detection rate for trisomy 21 (likely 50% to 60% as a screening test), lack of standardization, and variable sonographer and sinologist experience.
Cell-free DNA (cfDNA) has been extensively marketed to professionals and patients alike. Despite some claims, the PB points out that they all have similar detection and false-positive rates, although there are limited data on direct comparisons in large patient samples. Each technique available has advantages and disadvantages. Perhaps the most important point in the PB on cfDNA is the reminder that patients with low fetal fractions leading to “no reportable result” have a heightened risk of fetal aneuploidy. Adding these patients to reported calculations would REDUCE the sensitivity of cfDNA tests. Conversely, considering these patients as screen-positive will decrease specificity and increase screen-positive rates.
Risks of cfDNA testing:
1. False reassurance of normality, because despite being very sensitive to trisomy 21, the test is only 80% in detecting all chromosomal abnormalities2; and
2. Given the rarity of microdeletion syndromes like diGeorge (del22q11), and trisomy 13, it is uncertain what positive or negative screening test results mean. In fact, the PB recommends against screening for microdeletion disorders at present.
The PB contains a table that has recommendations for management of ultrasonographic markers for aneuploidy, some of which do not work well in the real world. Specifically, while it is useful in research protocols to define specific levels of risk as screen-positive or -negative, life is messier than that. One patient may consider 1 in 149 an acceptable risk and desire no further testing, while another may want invasive testing for a risk of 1 in 151. I would argue that the patient deserves to know as much as she can absorb about the risk number, not simply be told “you are screen-negative.”
Not addressed, and beyond the scope of a PB on screening, are other clinical dilemmas we now face with fetal anomaly detection. We increasingly face the scenario of women with very high risks for aneuploidy or microdeletion disorders, for example fetal tetralogy of Fallot, who decline diagnostic testing. cfDNA’s imperfect rate for detection of trisomy 21 is less of an issue, because > 99% of cases will be detected. The detection rate for del22q11, however, is only about 85%, so there will be far more false-positive and false-negative tests than with diagnostic testing.
As usual, the Summary of Recommendations and Conclusions contains the key bullet points for patient care, and there are too many to cite here specifically.
*Some minor microdeletions and epigenetic abnormalities are still not detected, and until economical and rapid whole exome sequencing is readily available, this is the current state of the art.