Cell free DNA screening is not a simple blood test

A recent article in the Boston Globe presents a disturbing picture of prenatal testing, in a report describing how “Oversold prenatal tests spur some to choose abortions.” The article goes on to discuss prenatal screening with cell free DNA, and to present a number of patient cases in which the screening test indicated that a patient was at high risk for aneuploidy when in fact, the fetus was normal. 

It is important for providers to remember that cell free DNA is a screening test, and does not have the diagnostic accuracy of amniocentesis.  By its very nature, a screening test does not tell with 100% certainty whether or not a fetus will be affected by a given disorder. Unfortunately, in part because of the high stakes in this very competitive market, the tests are being presented as having >99% accuracy, the same accuracy as is used to describe amniocentesis and CVS. 

While the detection rates for trisomy 21 are very high, and the false positive rate is very low, the test is less accurate and effective for detection of other disorders such as trisomy 18.  In fact, most of the patient stories in the article involved false positive and false negative tests for trisomy 18, not trisomy 21.

As providers, it is our responsibility to understand the characteristics of the tests that we order and interpret for our patients. Ideally, patients should meet with a genetic counselor prior to cfDNA testing. Providers should review the following key points with patients:

• cfDNA tests are screening tests, not diagnostic tests. 
• Abnormal results must be confirmed with diagnostic testing via CVS or amniocentesis before irreversible action, such as pregnancy termination, is undertaken.
• Although cfDNA tests have greater sensitivity and specificity than traditional serum screening, false positives and false negatives still occur.
• The likelihood that a patient with a positive cfDNA has an affected fetus – the positive predictive value – is lower if her background risk is low. For low risk women and for rare disorders, a positive test is more likely to be a false positive.
• cfDNA testing is therefore not recommended for low-risk women.
• Because the background risk for microdeletions is extremely low, a high-false positive rate is associated with cfDNA detected microdeletions.
• Genetic counseling services are an important part in providing information in the care for patients.  The SMFM recommends payers provide adequate reimbursement for these services to provide ideal care for patients.

Companies offering cfDNA should take steps to ensure that providers and patients interpret test results correctly. The Society suggests the following steps:

• Test results should be reported with a positive-predictive value or patient-specific risk, as is done with traditional serum screening.
• Given the risk of false positives with screening for rare disorders such as microdeletions, these tests should be offered as "opt-in," rather than "opt-out" options, ideally only after counseling by a genetic counselor.

Additional research is needed to determine how best to utilize this new technology:

• Post-marketing surveillance is needed to quantify patient and provider understand of cfDNA testing and the impact of this technology on families. 
• Studies should evaluate the impact of marketing materials on patient and provider understanding of these tests.
• Further research is needed to determine what counseling techniques and educational tools optimize patient understanding and minimize the morbidity of false positive and false negative results.

The Society for Maternal-Fetal Medicine has stated that all positive cell free DNA screening results require confirmatory diagnosis before an irreversible action, such as pregnancy termination, is undertaken.  Furthermore, SMFM has indicated that cell free DNA testing is a good screening test for high-risk women, but that the test is not recommended at present for low risk patients.  Any screening test will have different performance based on background risk, and in low risk patients the rates of aneuploidy are very low. Therefore, a positive test is usually more likely to be a false positive. 

This was demonstrated by a paper this year on cfDNA in low risk patients (Bianchi et al, N Engl J Med 2014;370:799-808) in which the positive predictive value of cfDNA was found to be less than 50% in low risk patients – that is, more than half of the positive cfDNA results were false positives.  Unfortunately, as reported in the Boston Globe article, that is often not well understood by providers, not explained to patients, and not clearly emphasized by the laboratories.  We would call upon the companies reporting these results to be clearer as to the appropriate interpretation of a positive test, ideally providing a positive predictive value as is done with traditional serum screening, in which a patient specific risk is presented. 

Cell free DNA is not “just a simple blood test” -- it is a genetic test with tremendous implications and consequences if misunderstood.