Ispd Position Statement
Aneuploidy screening: a position statement from a committee on behalf of the Board of the International Society for Prenatal Diagnosis, January 2011
Definitive prenatal diagnosis of Down syndrome and certain other fetal aneuploidies through chromosome analysis of amniocytes or chorionic villus samples (CVS) is an accepted part of prenatal care. But these procedures carry some degree of risk for miscarriage or other pregnancy complications (Tabor and Alfirevic, 2010). Therefore, in most developed countries it is now a routine practice to provide a woman's personal risk for aneuploidy (screening) and to offer definitive diagnosis through amniocentesis or CVS if the risk exceeds a fixed cut-off. However, in the United States it has been recommended that amniocentesis and CVS should be available to all women whether or not they have screening (American College of Obstetricians and Gynecologists, 2007a), although it is recognized that screening can be helpful to women before they decide whether to accept or reject amniocentesis or CVS (American College of Obstetricians and Gynecologists, 2007b).
Fetal aneuploidy risk can be evaluated on the basis of a combination of maternal age, prior family history, maternal serum biochemical tests and fetal ultrasound markers (Cuckle and Benn, 2010). Risk evaluation provides an opportunity to re-assure most women that their fetus is unlikely to be affected by a chromosomal disorder and also to reduce the number of unnecessary invasive procedures performed. Those women who are identified as being at high risk can receive genetic counseling, additional testing and appropriate follow-up obstetric care.
Because Down syndrome is the most common significant aneuploidy, prenatal screening has emphasized the detection of this disorder. However, it is recognized that many of the screening tests have a variable potential to detect other aneuploidies, some other genetic disorders, specific fetal anatomic abnormalities and pregnancy complications such as preeclampsia.
GOAL OF RISK EVALUATION
Every pregnant woman should have the opportunity to accept or reject amniocentesis or CVS on the basis of the best possible estimate of her personal risk for fetal aneuploidy. Programs involved with risk evaluation aim to provide timely and accurate individual patient-specific estimates of risk for the most common and clinically significant fetal aneuploidies.
COUNSELING AND THE PROVISION OF PRENATAL SCREENING
Aneuploidy risk assessment is a component of a broader set of prenatal clinical services that can include genetic counseling, chromosome analysis, midwifery and obstetrics.
Prior to receiving prenatal screening, women should be given information on the screening process and be provided with an opportunity to discuss this with a health professional before making a personal decision to accept or decline screening. Following the screening test, results should be explained in the context of the hazards and benefits of definitive diagnosis through amniocentesis and CVS.
Information must be provided through non-directive counseling. Each patient makes their own determination whether they wish to receive these services. Respect for ethical values, sensitivities and the decisions made by each patient are of key importance in the delivery of these services.
Prenatal aneuploidy risk assessment services often vary according to the healthcare systems that are present in different countries. Furthermore, service delivery may be modified to reflect individual women's clinical conditions such as infertility, co-existing risk for other genetic disorders or their moral and ethical values. Use of risk cut-offs in recommendations for diagnostic testing, sequential versus concomitant offers of screening and diagnostic testing, and other programatic differences exist. The Committee members had differing opinions on these standards of care and recognized that there are diverse approaches to these patient services that are compatible with beneficence to both individual women and to the populations served.
A range of maternal serum biochemical and fetal ultrasound markers have well-documented efficacy in distinguishing between affected and unaffected pregnancies. Each has validity within a specified time interval in pregnancy and should not be offered at earlier or later gestational ages. Combination of markers is valid, provided the correlation between them has been taken into consideration in the risk calculation.
Non-invasive approaches to screening and diagnosis through the analyses of fetal cells or nucleic acids in maternal circulation are under development. These tests have not yet been validated in clinical trials and are therefore not ready for clinical use.
MEASURING EFFICACY OF PROTOCOLS
The efficacy of risk assessment protocols has traditionally been based on the detection rate (DR, or sensitivity), false-positive rate (FPR) and positive predictive value (PPV), or odds of being affected given a positive result (OAPR). These population-based screening performance indices are of considerable value in comparing different protocols.
The relative efficacy of different protocols can be assessed by either fixing the FPR (between 1 and 5%) and comparing the DR, or fixing the DR (between 75 and 90%) and comparing the FPR. For a fixed risk cut-off, both the DR and FPR will vary between protocols. Statistical modeling using observational data is a reliable way of estimating the DR, FPR and OAPR of different screening protocols. Intervention studies can overestimate the screening performance, but may provide important information on the practicality of a specific protocol.
CHOICE OF PROTOCOL
The use of maternal age as a sole criterion for aneuploidy risk assessment is not justifiable.
Various first and second trimester approaches to aneuploidy screening as well as combinations of the two are listed in Table 1. Results are based on nuchal translucency (NT) at 12 weeks of gestational age. This is generally preferred over 11 weeks in order to facilitate optimal patient scheduling, because fetal anatomy is more clearly visualized and is better than 13 weeks because the screening performance is superior.
Table 1. Model-predicted Down syndrome detection rate for a 3% false-positive rate and positive predictive value for various screening protocols
|1a||PAPPA + freeβ (10), NT (12)||82||29|
|1b||PAPPA + hCG (10), NT (12)||80||29|
|1c||PAPPA + freeβ (12), NT (12)||80||29|
|1d||PAPPA + hCG (12), NT (12)||79||30|
|2a||AFP + freeβ + uE3 + InhA (15–19)||64||36|
|2b||AFP + hCG + uE3 + InhA (15–19)||60||39|
|3a||PAPPA + freeβ (10), NT (12), contingent AFP + freeβ + uE3 + InhA (15–19)||90||26|
|3b||PAPPA + hCG (10), NT (12), contingent AFP + hCG + uE3 + InhA (15–21)||88||27|
|3c||PAPPA + freeβ (10), NT (12), stepwise AFP + freeβ + uE3 + InhA (15–21)||92||25|
|3d||PAPPA + hCG (10), NT (12), stepwise AFP + hCG + uE3 + InhA (15–21)||91||26|
|4a||PAPPA (10), NT (12), AFP + freeβ + uE3 + InhA (15–19)||91||26|
|4b||PAPPA (10), NT (12), AFP + hCG + uE3 + InhA (15–19)||89||26|
|4c||PAPPA + freeβ (10), NT (12), AFP + freeβ + uE3 + InhA (15–19)||93||25|
|4d||PAPPA + hCG (10), NT (12), AFP + hCG + uE3 + InhA (15–19)||91||26|
|5a||PAPPA + freeβ (10), NT + NB (12)||91||26|
|5b||PAPPA + freeβ (10), NT (12), contingent NB||89||26|
|5c||PAPPA + freeβ (10), NT (12), contingent TCR||88||27|
|5d||PAPPA + freeβ (10), NT (12), contingent DV||88||27|
|6a||PAPPA + freeβ (10), NT (12), ANOMALY (18 + )||88||27|
|6b||PAPPA + hCG (10), NT (12), ANOMALY (18 + )||86||27|
|7a||ANOMALY (18 + )||56||41|
|7b||AFP + freeβ + uE3 + InhA (15–19), ANOMALY (18 + )||80||29|
|7c||AFP + freeβ + uE3 + InhA (15–19), contingent ANOMALY (18 + )||77||30|
|8a||PAPPA + freeβ (10), NT (12), AFP + freeβ + uE3 + InhA (15–19), ANOMALY (18 + )||96||25|
|8b||PAPPA + hCG (10), NT (12), AFP + hCG + uE3 + InhA (15–19), ANOMALY (18 + )||95||25|
A protocol based on first trimester measurement of NT for all women, with no additional tests, is insufficient. First trimester aneuploidy screening (the ‘combined’ test) is more advantageous than second trimester screening (the ‘quadruple’ test) not only because information is available earlier in pregnancy but also because the screening has greater efficacy (compare protocols 1a, b, c, d with 2a, b in Table 1). The quadruple test can be provided from 14 to 21 weeks but 15–19 weeks is preferred because 15–19 weeks is optimal for open neural tube screening using AFP.
Many women who receive a first trimester risk estimate that is intermediate between very high or moderately low risk may benefit from the provision of additional screening tests in the second trimester (‘contingent’ screening) and this can be associated with highly effective screening (protocols 3a, b). Additional testing for those with low first trimester risks (‘step-wise’ screening) can also be considered (protocols 3c, d), although this should not be needed for the majority of cases with very low first trimester risks (e.g. < 1 in 1500 at term). For both contingent and step-wise screening, it is essential that the second trimester risk estimation incorporates both the first and second trimester tests that have been performed. The provision of separate risk assessments based on first trimester markers alone and second trimester markers alone (‘independent’ screening) should not be carried out as it is associated with a significantly higher overall FPR and difficulties with second trimester counseling using two separate risk estimates. Protocols which include first and second trimester tests but only provide a risk figure after all screening tests are complete (‘integrated’ screening) are also associated with a high DR and low FPR but will restrict women's options in the first trimester (protocols 4a, b). When the same marker is tested in both trimesters (‘repeat measures’) there can be an additional benefit (protocols 4c, d).
The provision of additional first trimester sonographic markers can obviate the need for second trimester aneuploidy screening (Sonek and Nicolaides, 2010). The most widely used markers are absence of a fetal nasal bone (NB), tricuspid regurgitation (TCR) determined by pulse wave Doppler ultrasound and abnormal blood flow in the ductus venosus (DV). The routine use of these markers can substantially increase detection (protocol 5a), but good results are also obtained when this is done contingently at specialist centers (protocols 5b, c, d). The use of ultrasound needs to be consistent with fetal safety recommendations, i.e. with an ultrasound exposure that is as low as reasonably achievable (AIUM Practice Guideline, 2007).
Aneuploidy screening can also be improved by additional second trimester ultrasound markers (the second trimester ‘anomaly scan’ or ‘genetic sonogram’) provided the performance of the marker identification has been prospectively validated by the ultrasound center providing the service (Aagaard-Tillery et al., 2009). Findings with demonstrated utility include major malformations (MM), increased nuchal fold (NF) thickness, short femur or humerus length (FL or HL), echogenic intracardiac focus (EIF), pylectasis (P), echogenic bowel (EB) and ventriculomegaly (VM). NF, FL and HL should be expressed as continuous variables (e.g. with results expressed as MoMs) rather than categorical (i.e. on the basis of a value above or below a specific cut-off). Presence of EIF, P, EB and VM needs to be based on objective criteria. The anomaly scan should be performed at 18 weeks or later.
The anomaly scan can be used for patients who have received first trimester screening (protocols 6a, b), second trimester screening (protocols 7b,c) or both (protocols 8a,b). Although the second trimester anomaly scan can be used simply to modify the maternal age-specific aneuploidy risk alone, it is not a very effective screening test (protocol 7a). Using it to modify the risk following other aneuploidy screening can improve detection but when, as often happens, this is restricted to women with screen-positive results it can actually reduce detection.
Aneuploidy risks based on both NT and serum markers can be provided for twin pregnancies, despite poorer performance of the serum markers than in singletons. First trimester screening should take into consideration chorionicity; monochorionic twins are assumed to be monozygotic with an identical risk for each fetus, while dichorionic twins will be associated with separate risks for each fetus. First trimester serum markers require the use of gestation-specific and chorionicity-specific correction factors (Madsen et al., 2011). Second trimester screening with serum markers alone is considerably less accurate than that in singleton pregnancies. For triplets and higher multiplies, risks should be based on ultrasound markers alone. In the situation where there has been an early fetal loss (‘vanishing twin’), the serum markers may be un-interpretable (Spencer et al., 2010).
It is recognized that there is often limited availability of proficient sonographers. There may also be limitations in the availability of CVS, amniocentesis and counseling, restricted access to some of the serum tests or the computer programs used to calculate risks. Early pregnancy referral patterns and economic considerations are also likely to result in geographic differences in the protocols used. The choice of protocol must also take into consideration the need to screen for open neural tube defects either through second trimester AFP or second trimester ultrasound. No single combination of markers or screening cut-offs will therefore be appropriate for all situations.
SCREENING PROTOCOL RECOMMENDATIONS
On the basis of both observational studies and intervention projects the Committee recommends the following for women who wish to receive aneuploidy risk assessment:
- i.Ultrasound NT at 11–13 weeks combined with serum markers at 10–13 weeks. (Completed weeks e.g. 11–13 weeks = 11 weeks 0 days to 13 weeks 6 days).
- ii.Extending (i) to include other first trimester sonographic markers, provided performance has been prospectively validated by the ultrasound center where the screening is to be performed.
- iii.A ‘contingent’ test whereby women with borderline risks from (i) have (ii) at a specialist center and risk is subsequently modified.
- iv.Four maternal serum markers (quadruple test) at 15–19 weeks, for women who first attend after 13 weeks 6 days.
- v.Combining (i) and (iv) in either a stepwise or contingent protocol—provided that all screening test data are included in the final risk assessment. Integrated screening can be offered when CVS is not available.
- vi.Contingent second trimester ultrasound to modify risks for aneuploidy (sometimes referred to as the ‘anomaly scan’ or ‘genetic sonogram’) for women having (i), (iv) or (v). Performance must be prospectively validated by the ultrasound center where the screening is performed.
Laboratories providing maternal serum screening tests must participate in proficiency testing and monitor their performance through epidemiologic monitoring. Ultrasonographers must participate in an on-going audit of performance. Computer programs used in calculating risk should be checked for design accuracy. Comprehensive registries of aneuploidy should be encouraged, provided confidentiality of individual patient data can be assured. These registries can provide validation of the risks and also have considerable research value.
- i.Definitive diagnosis of Down syndrome and other fetal aneuploidies can only be achieved through amniocentesis or CVS.
- ii.The use of maternal age alone to assess fetal Down syndrome risk in pregnant women is insufficient.
- iii.A combination of ultrasound NT measurement and maternal serum markers in the first trimester should be available to women who want an early risk assessment.
- iv.A four marker serum test should be available to women who first attend for their prenatal care after 13 weeks 6 days of pregnancy.
- v.Protocols that combine first trimester and second trimester markers are valid.
- vi.Second trimester ultrasound can be a useful adjunct to other protocols.