| |||||||||||||||||||||||||||
|
  Dangers of Prenatal Testing April 13, 2008  Watch part one Watch part twoReporter: Jane Hansen An ultrasound screening test at 12 weeks is now considered a routine part of pregnancy, but an eminent British Radiologist claims the test, designed to pick up chromosomal abnormalities like Down syndrome, is highly inaccurate. In a controversial interview with the Sunday program, Dr Hylton Meire claims that the screening test known as a Nuchal Translucency, has a false positive rate of 95%, and that healthy babies are put at risk of miscarriage from more invasive tests to confirm or deny the original suspicion. He argues up to three healthy babies die for every one Downs syndrome baby that goes to term. Dr Meire’s claims are strongly opposed by Dr Andrew McLennan, a consultant obstetrician who introduced the nuchal tests to Australia in 1998. Dr Meire tells Sunday; “95% of positive tests don’t have a chromosomally abnormal baby” and he says, “patients probably aren’t aware of the fact they have a significant risk of losing a normal baby as a result of further tests.” At Tracey Young’s 12 week ultrasound screening test, she was told she had a one in 100 chance of having a baby with a chromosomal abnormality not ‘compatible with life’. Doctors recommended an amniocentesis, a procedure that involves a needle taking amniotic fluid from the pregnancy to confirm the diagnosis. There was a risk of miscarriage, but Tracey just wanted to make sure the baby was ok. ‘I did it for reassurance’ she recalls, and the result came back clear; the little boy was chromosomally perfect. Two weeks later she lost the baby. ‘All the amniotic fluid had leaked out and he died inutero’ Tracey is not alone. Deborah Thomas, editor of Women’s Weekly was 42 when she first fell pregnant and opted for a chorionic villus sampling, or cvs test where a sample of placenta is harvested via needle from the pregnancy. She painfully recounts how her amniotic sac developed a leak as a result and her baby boy was lost at 23 weeks. Dr Meire claims that the rate of false negatives for the Nuchal Translucency test is also high, with some 40% of Down syndrome babies not picked up by the test. At 25, Carrie Daniels was a young mum and not considered to be at high risk. Her NT screening test results concluded her baby had a less that one in 2000 chance of a chromosomal abnormality. Six months later, Harrison Daniels was born with Down syndrome. In another case 12 weeks into Harrison Catesby’s embryonic life, his parents were given the shocking news that he had a 1 in 8 chance of Down syndrome.However the original test was wrong. Harrison was born perfectly normal, but not before his parents opted for a cvs test, without thinking of the miscarriage rate. Is Ultrasound as useful as we think? The Donald, MacVicar and Brown lecture 2006 By Dr H B Meire INTRODUCTION I have been fortunate enough to spend the large majority of my professional career involved solely in developing and using the clinical applications of diagnostic ultrasound. My objective during the Donald, MacVicar and Brown lecture was to review briefly the extraordinary progress that has been achieved in improving the technical quality of ultrasound scans but, more importantly, to review the evidence that is available to prove the efficacy or otherwise of several well-established common uses of diagnostic ultrasound. The first ultrasound scans I had to work with in the early 1970s were bistable images with no grey scale whatever and few, if any, parenchymal echoes (figure 1). In the mid-1970s commercial grey scale image storage systems became available and images such as those in figure 2 were achieved and immediately enhanced our ability to detect parenchymal abnormalities. A B Figure 1 A. 1970 bistable longitudinal scan showing the liver and IVC. No parenchymal echoes are seen within the liver. Figure 1 B. 1970 bistable scan further to the right showing the right lobe of the liver and the right kidney. A B Figure 2 A. 1975 greyscale scan of the right lobe of liver and kidney. Parenchymal echoes are now seen. Figure 2 B. 1975 greyscale scan transversely through the upper abdomen showing the liver, pancreas, common bile duct and right kidney. Ascites is present. Within the field of obstetrics the improvements in both grey scale and beam width were dramatic during the 1970s, such that by 1980 high dynamic range and relatively high resolution grey scale images were being achieved (figure 3). Progress has continued to the present day and figure 4 shows a contemporary image of the fetal chest and abdomen with outstanding dynamic range and resolution. A B C Figure 3 A. 1970 bistable scan through a fetal abdomen. Figure 3 B. 1975 greyscale scan through a fetal abdomen. Tissue differentiation is now seen. Figure 3 C. 1980 scan through a fetal abdomen. The greyscale and spatial resolution have improved markedly. It is probably difficult for those of you who have come to ultrasound in recent years to appreciate the excitement experienced by those involved in the early days of clinical ultrasound where structures were becoming visible for the very first time. Not surprisingly the euphoria associated with this new-found imaging modality gave rise to many enthusiastic claims for the utility of ultrasound and, in the absence of the concept of ‘evidence-based medicine’ and clinical audit, there was a tendency for some of the applications of ultrasound to enter routine clinical use without having been scientifically validated. Figure 4. 2006 longitudinal scan through a fetal chest and abdomen. There has been further improvement in both the spatial and contrast resolution. In the late 1970s I was working for the Medical Research Council (MRC) and was aware that many obstetric units throughout the country, and abroad, were beginning to apply ultrasound scanning on a routine basis to all obstetric patients. It seemed reasonable to assume that the wealth of anatomical information obtained would be beneficial to the mother and fetus and would help obstetricians in planning the management of the pregnancies. This same outlook has persisted in many of the applications of ultrasound in obstetric management and I thought that it would be prudent therefore to take this opportunity to assess the evidence, or lack thereof, to confirm the utility of some of the well-established and more recent obstetric applications of ultrasound. In reviewing the utilisation of any new technology after its initial introduction to the market place, many authors have shown statistics that give rise to a graph such as that shown in figure 5. Not surprisingly, as soon as some new and revolutionary technology is developed, there is an initial steep rise in enthusiasm for the technology, frequently leading to overrated and unsubstantiated statements concerning its utility. After some time the limitations of the technology become apparent and there is a rapid decline in enthusiasm, usually with a degree of overshoot, and only with time does the true level of utility become evident, usually associated with a gentle rise to a plateau level of utilisation. My reasoning for introducing the reader to this enthusiasm curve will become evident later in the paper when I review the clinical evidence associated with the utility of a few of the applications of clinical ultrasound in obstetrics. Figure 5. The enthusiasm curve. THE 20 WEEK ANOMALY SCAN Whilst I was working with the MRC in the late 1970s I was concerned by the rapid utilisation of routine obstetric ultrasound scans, especially the routine anomaly scan. I was, nevertheless, convinced that the scan was immensely invaluable and the research that we undertook showed that, as equipment and expertise continued to improve, there was a very reasonable expectation that the utility of the technique would continue to advance. However, in about 1980 I approached the MRC with a request for support to set up a large multicentre study to evaluate critically the possible benefit of routine anomaly scans to a low risk obstetric population. I was informed that such a study could not be performed as it would be statistically, financially and ethically unfeasible. The statistical constraints arose from the relatively low incidence of pregnancy abnormalities, which would therefore require the study to recruit about 100,000 subjects and a matched number of un-scanned control subjects from the same institutions. All the ensuing pregnancies would have to be monitored until after birth to establish the true and false positive and negative rates. This would obviously be extremely time consuming and therefore expensive. Finally, the trial was considered ethically unacceptable because it was thought that it would not be acceptable to withhold the benefits of ultrasound from the un-scanned control group. It became clear that the hierarchy within the MRC were already, by that stage, convinced of the clinical utility of obstetric ultrasound. The prevalence of this view within the obstetric community was underlined by a paragraph within the 1984 report from the Royal College of Obstetricians and Gynaecologists (RCOG) working party on routine ultrasound examination in pregnancy1. Within this report they stated ‘We believe there are cogent reasons to expect benefit to all mothers and babies from a well-performed scan between 16 to 18 weeks of pregnancy but there is a need for a large, well-planned prospective study…’. In retrospect it is interesting to note that the authors of this text assumed that the benefit would be to ‘all mothers and babies’. It is also noteworthy that the RCOG authors recognised the absence of any valid statistical evidence to support their statement. 7 years later a study group set up by the RCOG revisited the subject2 and made the recommendation that all mothers should have a routine anomaly scan at 18-20 weeks. However, in the 1997 RCOG3 report, whilst referring to their previous publication, they stated ‘the evidence for these recommendations was tenuous’. The situation seems to have changed little by 2000 when Bricker et al4 in their NHS Health Technology Assessment concluded ‘Further research is needed to assess the effect of detection of fetal abnormalities on substantive outcome in terms of short and long term morbidity and mortality for both mother and child.’ It seems extraordinary that the large majority of obstetricians were prepared to subject their patients to an unproven technique, to accept the costs associated with the wholesale application of ultrasound, to accept that it was unproven and yet fail to undertake the necessary scientific studies to prove that the technique was as helpful as they believed. In their 1984 report1 the RCOG group calculated that ‘72% of mothers in England have a routine scan during the antenatal period,’ though I believe that most of these would have been merely to check for the presence of a live fetus and to confirm the duration of gestation. The 1997 RCOG publication3 identified the fact that during the study period prior to this publication 82% of NHS hospitals claimed to be offering a routine 20 week scan. It would seem to be true that in the early 1990s nearly 20% of obstetric departments did not offer routine ultrasound and I believe that is probably still the case 10 years later. Outside the UK other bodies were also assessing the utility of routine obstetric ultrasound scans and in 1998 the World Health Organisation5 stated ‘clear benefits of the routine use of ultrasound scanning [in pregnancy] have not been established.’ In 2002 the Institute for Clinical Systems Improvement in the USA6 stated ‘routine ultrasound in early pregnancy (before 24 weeks gestation) has not been shown to reduce perinatal mortality.’ In the same year the Canadian Guidelines Advisory Committee7 stated ‘serial ultrasound screening in the second and third trimesters does not have a statistically significant effect on perinatal illness, perinatal death or birth weights.’ Despite the more-or-less wholesale application of routine ultrasound it is interesting to note that even the leaders in obstetrics accept that there are circumstances under which routine scans can reasonably be withheld. For example in their 2000 publication8 the RCOG state ‘if a unit considers it cannot deliver scans to this minimum standard then the 20 week scan should be abandoned.’ This therefore presupposes that there will be some institutions in which either the equipment or expertise to undertake satisfactory scans are not available and that the patients at these institutions will therefore not receive routine obstetric ultrasound scanning. Similarly, in 2006, the UK National Screening Committee9 noted that ‘around half of the maternity units in Scotland routinely offer a routine fetal anomaly scan.’ The converse of this therefore indicates that half the obstetric departments in Scotland do not offer routine obstetric ultrasounds. This fact is not commented upon and there is therefore presumably a tacit acceptance that these departments are not in any way negligent by failing to offer this service. I was delighted to note that during the December 2006 BMUS Congress there was a debate on the motion ‘Routine ultrasound is a luxury the NHS can no longer afford.’ The chairman of the debate asked for a show of hands in advance of the debate to identify the percentage of the audience who agreed with this hypothesis. To my amazement 45% of the delegates agreed with the motion. Interestingly, after a lively debate, there was a small reduction in those prepared to support the motion. However, it was noteworthy that the audience had increased in size significantly during the period of the debate and thus the pre- and post-debate percentages cannot reliably be compared. In conclusion it is with regret that I have to note that there is still a lack of good evidence in favour of the wholesale application of routine anomaly scanning in pregnancy. One has to ask whether or not it is even now too late to undertake a scientifically valid study to answer this question once and for all. From a purely financial point of view such a study would almost certainly be worthwhile, since we are already investing millions of pounds per annum into the application of a technique that remains unproven. OBSTETRIC DOPPLER During my reviews of the scientific literature in preparation for the Donald MacVicar Brown lecture I was interested to note a number of studies recommending the routine application of obstetric Doppler studies of both the uterine arteries and umbilical arteries in all pregnancies. One of the earliest was the 1984 RCOG report1 where they stated ‘…measurement of fetal and uteroplacental blood flow which may prove to be one of the most important obstetrical applications of diagnostic ultrasound.’ Whilst one has sympathy with their enthusiasm for all things ultrasonic at that time there was no real scientific basis for the expressed views. The routine use of Doppler has not, in fact, eventuated and I note that the NHS Health Technology Assessment in 200010 stated ‘routine Doppler ultrasound in pregnancy has not been shown to be of benefit and may even increase the incidence of adverse outcome.’ Some 6 years later the Cochrane Review11 came to the same conclusion stating ‘Routine Doppler ultrasound in pregnancy does not have any health benefits for women or babies and may do some harm.’ It would seem, therefore, that at least this application of routine obstetric ultrasound has been critically evaluated and its unwarranted wholesale application has been averted. DETECTION OF ANEUPLOIDY Some years ago the concept of ultrasound ‘soft markers’ for aneuploidy, especially trisomy, was introduced. These are a disparate collection of ultrasound observations including, but not confined to, the items included in table 1. Use of these markers to identify cases of possible aneuploidy presupposes that any confirmed cases will warrant termination. As we will see below this policy leads to a great many interventions in order to achieve a modest reduction in aneuploidy live births. Table 1 Ultrasound ‘soft markers’ for aneuploidy. Nuchal thickening Echogenic bowel Mild ventriculomegaly Cardiac golf ball Choroid plexus cyst Single umbilical artery Enlarged cisterna magna Pyelectasis Clinodactyly Short femur Short humerus Absent nasal bone Brachycephaly Iliac angle Ear length Sandal gap It is well recognised that the risk of conceiving a fetus with trisomy rises steeply with increasing maternal age. However, in view of the fact that younger mothers produce the large majority of babies, the majority of trisomic babies are born to younger mothers. Any screening technique thus has to be applied to the entire obstetric population. The huge variation in the incidence of trisomy seen in the papers referred to below is almost certainly due to population selection, though this is not always explicitly explained by the authors. I felt it would be salutary to document some aspects of the clinical utilisation of a few of these aneuploidy markers and to see what lessons can be learnt from the resulting evidence. Before presenting the results of this study it is important for us to be aware of the natural history of aneuploidy in obstetrics. There is some evidence that 50% of all fertilised eggs suffer from aneuploidy12. Studies of first trimester spontaneous abortion material indicate aneuploidy rates between 50%13 and 72 %14. Surprisingly similar figures have been obtained in studies where the chromosome complement of pre-implantation IVF embryos have been studied with values of 40.9%15 to 81%16 aneuploidy being reported. It is likely that the high rate of aneuploidy in the latter study reflects the greater range of possible abnormalities searched for in this study. The combined incidences of trisomies 21 and 18 at 12 to 14 weeks of pregnancy is variously reported as ranging from 0.85% of 8,514 patients17 to 2.5% of 3,788 patients18. Despite this evidence the combined incidence of both at birth is 0.139%19. These data indicate that only one in six to one in 18 cases of trisomy 21 or 18 that are present in the first trimester actually survive until birth. Thus any test undertaken during the first trimester will diagnose large numbers of aneuploidy embryos or fetuses that would not, in fact, have reached term alive. It is against this background that we must judge the clinical utility of the various ultrasound markers. Choroid plexus cysts The identification of choroid plexus cysts (CPCs) was first reported in the mid 1980s20 and in 1989 Gabrielli et al21 published a paper reporting 4 cases of trisomy 18 in 65 fetuses with CPC’s. All these fetuses had other structural anomalies. Figure 6 shows results from selected papers21-29 on the topic over the ensuing years. The similarity between the curve shapes in figures 5 and 6 is noteworthy. Figure 6. The published aneuploidy rates for fetuses with choroid plexus cysts. It can be seen that in 1992 it had been claimed that the incidence of aneuploidy in CPC fetuses was 23%21 but, by 1995, Gonen25 reported no cases of aneuploidy in 108 cases of fetuses with isolated CPC. Since that date the majority of papers have indicated extremely low levels of aneuploidy in CPC fetuses and the mean incidence from the last five papers in figure 6 is 1 in 761, 0.13%. We can therefore now assume that the entire CPC saga was completely unwarranted and that there is no association between the presence of isolated choroid plexus cysts and the presence of aneuploidy. This hypothesis is supported by Cheng et al30 who, in 2006, stated ‘in pregnancies complicated by isolated CPCs fetal karyotyping is not indicated’. More importantly, in the same year Cristofalo31 stated ‘detection of CPC prenatally can invoke profound negative maternal emotional responses despite accurate provider counselling’. This reinforces my own clinical observations of a number of parents having had the enjoyment of their entirely normal pregnancies destroyed by the reported findings of choroid plexus cysts in an otherwise normal ultrasound examination. Indeed, I know of one patient in whom, despite repeated professional counselling, the mother remains convinced that her child, now aged nearly 10 years old, must still have an abnormal brain and that the CPC is responsible for every headache or other minor cerebral malaise that her unfortunate child suffers. It seems reasonable, therefore, to conclude that the identification of isolated choroid plexus cysts should not be reported either to the patient or the obstetrician. However, if additional ‘soft markers’ are seen to be associated with CPC’s referral for an expert fetal medicine opinion is warranted. Nuchal thickness The ultrasound detection of fluid within the soft tissues on the back of the fetal neck was first reported in 197532 and was probably a genuine cystic hygroma. By 1984 the detection of ‘oedema’ on the back of the neck was reported and its association with aneuploidy identified33. A year later Benacerraf et al34 reported ‘increased skin or soft tissue thickening at the back of the neck’ in five of 11 fetuses later shown to have Down’s syndrome and by 198735 she had increased her series to 9 cases of Down’s in 21 fetuses with ‘soft tissue thickening at the back of the fetal occiput’. In 1992 Nicolaides et al36 published a paper in which they reported a 35% association between nuchal thickening and aneuploidy. They stated that ‘fetal nuchal translucency greater than or equal to 3 mm is a useful first trimester marker for fetal chromosome abnormalities.’ It is important to note that their cohort of 827 fetuses were all drawn from a high risk group. As a teacher I have to emphasise that the use of the term ‘translucency’ in this paper is extremely unfortunate. The term has, of course, no meaning whatever within the context of ultrasound and I consider that it would be more helpful if we were to use the term ‘nuchal thickness’ in preference. As with choroid plexus cysts it is salutary to review the scientific papers reporting the incidence of aneuploidy with increased nuchal thickness (NT). The results are shown graphically in figure 7. In 1992 Nicolaides et al reported an incidence of 35%36 but by 1993, Nadel37 had found a 68.3% association between aneuploidy and nuchal thickening equal to or greater than 4mm. Figure 7. The published aneuploidy rates for fetuses with increased nuchal thickness. In 1995, in a study on a low risk population, Bewley et al38 reported the incidence of trisomy as 2.7% (2 of 75) of patients with an NT ≥3mm. Their cohort also included 3 trisomic fetuses with normal NT. In 1998 Nicolaides’ group39 reported on a multicentre study of 96,127 patients. 8,428 (8.77%) of the patients had a positive NT measurement and of these 521 (6.18%) were aneuploid. In a 2005 paper from the same group40, including serum biochemistry and NT, a positive result was obtained in 8.5% of 30,564 patients and 11.6% of positive subjects had an aneuploid fetus. The following year the same group reported aneuploidy in 19.2% of 11,315 NT positive patients41 and 10.9% of 17,446 patients42. It is not clear whether any of these patients were also included in the previous publications. Regrettably the virtual monopoly of NT measurement by the Fetal Medicine Foundation seems to have prevented independent workers from investigating the clinical utility of NT measurements. One of the major difficulties confronted by an obstetrician in counselling a patient with an ultrasound ‘soft marker’ for aneuploidy in early pregnancy derives from the fact that there is currently no non-invasive confirmatory test to determine the fetal chromosome complement. It is therefore necessary to undertake CVS or amniocentesis on all ultrasound positive patients if a definitive answer is to be obtained. Regrettably, CVS is not without its hazards with a post-procedure abortion rate of 2.1% of 1,914 patients reported in 200443. In 2006 a study of 9,886 subjects44 found an excess loss rate after CVS of 2.29% compared with the loss rate after amniocentesis. It is therefore likely that the CVS induced loss rate is of the order of 2%. In the 1998 paper from Nicolaides’ group39 they reported the findings of a multicentre trial in which 96,127 patients were recruited from 22 separate centres. Of these 8,428 patients (8.77%) were reported as being ultrasound positive for increased nuchal thickness. As all of these patients received a confirmatory CVS it is likely that about 170 fetuses were lost as a result of the procedure. Since the aneuploidy rate in the screen positive subjects in this paper was 6.1% about 10 of these lost fetuses would be aneuploid, leaving 160 normal fetuses lost due to the procedure. However, as we have seen above, the aneuploidy rate at 13 weeks is reported as between 0.85% and 2.5% whilst at term it is less than 0.2%, thus about 9 out of every ten aneuploid fetuses identified during the first trimester will later abort spontaneously. Therefore, in the above study, only about 50 of the 513 aneuploid fetuses would have gone to term. In order to diagnose these 50 aneuploid fetuses 160 normal fetuses will have been lost. One has to ask whether or not this is a price worth paying and whether, in this context, nuchal thickness measurement can be considered to be a clinically valid tool. I note that in their 1995 paper Bewley et al38 stated ‘there is, at present, insufficient data to warrant its [nuchal thickness] introduction for screening of the general population…’. More importantly in 2000 the NHS Health Technology Assessment Group decided to undertake a review of the published clinical papers on the utility of nuchal thickness measurements10. The reader may be aware that the NHS Health Technology Assessment Group evaluates the scientific and statistical validity of all papers before including them in their analyses. Regrettably they came to the conclusion that ‘none of the limited number of reports on these topics met our criteria for inclusion in systematic reviews. Until these data are available, the evidence does not support screening in the clinical service’. In the following year the Journal of the American Medical Association45 concluded ‘the overall sensitivity of this finding is too low for it to be a practical screening test for Down’s syndrome’. In the same year another authoritative American review46 concluded ‘Despite encouraging data and general enthusiasm for first trimester screening for fetal Down syndrome and other aneuploidies, a number of questions remain about its implementation in the United States.’ Scientific critique of nuchal thickness measurement I have four additional concerns regarding nuchal thickness measurement that warrant further discussion. Firstly, the software used by the Fetal Medicine Foundation to calculate the patients’ risk of aneuploidy has never been published nor subjected to independent scientific analysis. Such secrecy is, to my knowledge, unique within the medical scientific community. I have been privileged to evaluate the program on a confidential basis and my observations confirm that there is a need for this software to be subjected to unbiased scientific scrutiny. Secondly, utilisation of the risk-calculating software based on nuchal thickness measurements show that in certain clinical situations there is a profound change in apparent risk of aneuploidy for every 0.1 mm change in apparent nuchal thickness. This point was well illustrated by Chudleigh47 who showed that, using the FMF software, a 36 year old patient with an NT measurement of 2.2mm would have an aneuploidy risk of 1:275. However, if the NT was measured as 2.1mm the risk falls to 1:469 and if the NT measurement was 2.3 the risk rises to 1:153. All these NT values are well within the range of uncertainty and make it impossible to be certain whether the patient lies in the high or low risk group. The publishers of this software and I suspect the vast majority of users, seem unaware of the fact that ultrasound will not measure with a precision of 0.1 mm. There is clearly a supposition that because the calliper systems on the scanners give readings that vary by increments of 0.1 mm this represents the precision of the ultrasound measurement. In fact the limit of resolution along the length of the ultrasound beam is determined by the wavelength of the ultrasound. At 5 MHz this is 0.3 mm and at 3 MHz 0.5 mm. The theoretical limit of resolution is half the wavelength. The resolution is likely to be further compromised in modern ultrasound machines since many of these now default to ‘harmonic imaging’ mode that gives an apparent improvement in image quality. However, the user may be unaware that the transmitted frequency in harmonic mode is significantly reduced, usually by a factor of two, thus doubling the wavelength of the ultrasound and further reducing the precision of axial measurements. In clinical practice it is likely that the irreducible variability due to the ultrasound wavelength is of the order of ±0.3 to 0.5mm, at best. Clearly ultrasound machines can measure increments of less than 1mm but in order to display fractions of a millimetre the calliper systems have to read in increments of 0.1mm. This does not indicate the 0.1mm is the accuracy of the measurement. Thirdly, the few available studies of the accuracy and reproducibility of the measurements indicate errors approaching ± 1mm. In a French study48 of 36 investigators in 12 centres the authors concluded ‘…the method does not appear suitable for population screening because of the high variability in the results among investigators.’ In an American study49 of 20 experienced operators each examining the same five patients, the two standard deviation limit of the inter-observer variability was ± 1.12mm. Calculating an aneuploidy risk on data this variable would be meaningless. The problem was also addressed by the UCH group50 who found that repeating the measurement with a different operator caused 18.8% of all measurements to change their classification between normal and abnormal. They concluded that ‘The poor reproducibility of nuchal translucency measurements could diminish its usefulness as a screening test for Down’s syndrome.’ A limited study by the Nicolaides group51 found intra- and inter-observer variability of more than ± 0.5mm; they claim this to be ‘highly reproducible’. A phantom based study presented at the 2006 BMUS Congress by Evans et al52 found that the mean standard deviation of multiple measurements was 0.4mm indicating that the 95% confidence interval for a single measurement is ± 1.6mm. Fourthly, the technique recommended by Nicolaides for the measurement of nuchal thickness recommends the use of the ‘on-to-on’ end points. This strange terminology translates to measurement from the trailing edge of the first echo to the leading edge of the second echo. The measurements are thus dependent on the width of the echoes that are in turn related to equipment output and gain settings. Scientifically this is an unsatisfactory measurement technique and must lead to additional errors and variations of unknown magnitude, though Evans (personal communication) has suggested that the error from this source could also be in the range of ± 0.5mm. Soft markers Table 1 includes some of the range of structures and appearances which have been claimed to be ‘soft markers’ for an increased risk of aneuploidy. Experience indicates that none of these is reliable if it is the only marker present. However, if more than one marker is present, or if one is also associated with the presence of a major anomaly known to be associated with aneuploidy, the risk of fetal aneuploidy being present is very high and invasive testing is indicated. In 1997 an editorial in the British Medical Journal53 concluded ‘detecting them [soft markers] may do more harm than good.’ These authors also comment that the known rate of loss of aneuploid fetuses during pregnancy ‘…throws uncertainty over the precise value of the test.’ In addition they state that such testing ‘..may not be ethically acceptable when identifying these markers increases anxiety, usually unnecessarily and often without prior counselling. In 2001 in the Journal of the American Medical Association45 the following view was expressed; ‘using these markers as a basis for deciding to offer amniocentesis will result in more fetal losses than cases of Down’s syndrome detected …’. Also in 2001 Current Opinion in Obstetrics and Gynecology54 stated ‘the implications for pregnancy management when one of these so-called minor ultrasound markers is detected have been a matter of continuous controversy in the field of pre-natal diagnosis and yet the definitive answer on their clinical significance in the low risk population is still debated.’ More recently Susanne et al55 commented ‘A false positive test of fetal screening for Down syndrome by ultrasound examination may cause strong reactions of anxiety and even rejection of the pregnancy. The prevalence of such reactions and possible long-term effects needs further investigation.’ Finally some sound advise was offered in the North American Journal of Obstetrics and Gynecology in 200456 who stated ‘As obstetric ultrasound moves forward, particularly into the uncharted waters of clinical use of 3 and 4 dimensional ultrasound, one can expect a whole new crop of ultrasound findings with uncertain clinical significance. Clinicians are well advised to await well-designed studies to determine the clinical significance of these findings before altering clinical care.’ Conclusion Despite the use of ultrasound to investigate obstetric patients routinely for more than 30 years we have still not produced scientifically valid studies that confirm the utility of this technique in unselected populations. The use of nuchal thickness measurement for the assessment of aneuploidy risk must be seriously questioned since a systemic lack of scientific rigour was found in all the published papers on the subject and there are serious questions regarding the scientific validity of the measurement technique and presumed accuracy of the ultrasound measurements. It may well be that measurement of the nuchal thickness does have a valid role to play in a comprehensive integrated screening programme for aneuploidy risk but I would venture to suggest that it should not be included until such time as both the technique and associated software have been subjected to serious scientific scrutiny. REFERENCES 1. Report of the RCOG working party on routine examination in pregnancy. (1984) RCOG London. 2. Drife JO, Donnai D. Eds. (1991) Antenatal diagnosis of fetal abnormalities. RCOG London. 3. Ultrasound screening for fetal abnormalities. (1997) Report of the RCOG working party RCOG London. 4. Bricker L, Garcia J, Henderson J, et al (2000) Ultrasound screening in pregnancy: a systematic review of the clinical effectiveness, cost-effectiveness and women’s views. Health Technol Assess. 2000; 4(16) 5. Belizan J (1998) Ultrasound for fetal assessment in pregnancy: RHL commentary. WHO Reproductive Health Library. 6. Prenatal ultrasound as a screening test. (2002) Institute for Clinical Systems Improvement. 7. Prenatal ultrasound: Routine Screening (2002) Guidelines Advisory Committee Canadian Task Force on Preventive Health Care. 8. Routine Ultrasound Screening in Pregnancy. Protocol, Standards and Training. (2000) RCOG London 9. UK National Screening Committee’s Policy Positions, July 2006 10. Bricker L, Garcia J, Henderson J, et al (2000) Ultrasound screening in pregnancy: a systematic review of the clinical effectiveness, cost-effectiveness and women’s views. Health Technology Assessment www.hta.nhsweb.nhs.uk/execsumm/summ416.htm 11. Bricker L, Neilson JP. (2006) Routine Doppler ultrasound in pregnancy (Cochrane Review) Cochrane Library, Issue 3, 2006. Chichester. 12. Boue J, Bou A, Lazar P. (1975) Retrospective and prospective epidemiological studies of 1500 karyotyped spontaneous human abortions. Teratology 12(1):11-26 13. Eiben B, Bartels I, Bahr-Porsch S et al (1990) Cytogenetic analysis of 750 spontaneous abortions with the direct preparation method of chorionic villi and its implications for studying genetic causes of pregnancy wastage. Am J Hum Genet. 47(4):656-663 14. Fritz B, Hallermann C, Olert J et al (2001) Cytogenetic analysis of culture failures by comparative genomic hybridisation (CGH) – Re-evaluation of chromosome aberration rates in early spontaneous abortions. Eur J Hum Genet. 9(7):539-547 15. Kahraman S, Benkhalifa M, Donmez E et al (2004) The results of aneuploidy screening in 276 couples undergoing assisted reproductive techniques. Prenat Diagn. 24(4):307-311 16. Baart EB, van den Berg I, Martini E et al (2007) FISH analysis of 15 chromosomes in human day 4 and 5 preimplantation embryos: the added value of extended aneuploidy detection. Prenat Diagn. 27(1):55-63 17. Wapner R, Thom E, Simpson JL et al (2003) First-trimester screening for trisomies 21 and 18 N Engl J Med. 349(15)1405-1413 18. Cicero S, Longo D, Rembouskos G, Sacchini C, Nicolaides KH (2003) Absent nasal bone at 11-14 weeks of gestation and chromosomal defects. Ultrasound Obstet Gynecol. 22(1):31-35 19. Davies T (1994) Prenatal screening for Trisomy 18 should not be contemplated. (letter) Brit Med J. 308(6926)471 20. Chudleigh P, Pearce JM, Campbell S (1984) The prenatal diagnosis of transient cysts of the fetal choroid plexus. Prenat Diagn. 4(2):135-137 21. Gabrielli S, Reece EA, Pilu G et al (1989) The clinical significance of prenatally diagnosed choroid plexus cysts. Am J Obstet Gynecol. 160(5 Pt 1):1207-1210 22. Werner-von der Burg W, Hatzmann W, Kuchenbecker K (1992) Fetal plexus cysts. An indication for prenatal karyotyping. Geburtshilfe Frauenheilkd 52(11):674-680 23. Porto M, Murata Y, Warneke LA, Keegan KA Jr. (1993) Fetal choroid plexus cysts: an independent risk factor for chromosomal anomalies. J Clin Ultrasound. 21(2):103-108 24. Kupferminc MJ, Tamura RK, Sabbagha RE et al (1994) Isolated choroid plexus cyst(s): an indication for amniocentesis Am J Obstet Gynecol. 171(4):1068-1071 25. Gonen R, Dar H, Degani S (1995) The karyotyping of fetuses with anomalies detected by second trimester ultrasonography. Eur J Obstet Gynecol Reprod Biol. 58(2):153-155 26. Reinsch RC (1997) Choroid plexus cysts- association with Trisomy: prospective review of 16,059 patients. Am J Obstet Gynecol. 176(6):1381-1383 27. Geary M, Patel S, Lamont R (1997) Isolated choroid plexus cysts and association with fetal aneuploidy in an unselected population. Ultrasound Obstet Gynecol. 10(3):171-173 28. Leonardi MR, Wolf HM, Lanouette JM et al (1998) The apparently isolated choroid plexus cyst: importance of minor abnormalities in predicting the risk of aneuploidy. Fetal Diagn Ther. 13(1):49-52 29. Morcos CL, Platt LD, Carlson DE et al (1998) The isolated choroid plexus cyst. Obstet Gynecol. 92(2):232-236 30. Cheng PJ, Shaw SW, Soong YK (2006) Association of fetal choroid plexus cysts with Trisomy 18 in a population previously screened by nuchal translucency thickness measurement. J Soc Gynecol Investig. 13(4):280-284 31. Christofalo EA, Dipietro JA, Costigan KA et al (2006) Women’s responses to fetal choroid plexus cysts detected by prenatal ultrasound. J Pernatol. 26(4):215-223 32. Morgan CL, Haney A, Christakos A, Phillips J (1975) Antenatal detection of fetal structural defects with ultrasound. J Clin Ultrasound. 3(4):287-290 33. Redford DH, McNay MB, Ferguson-Smith ME, Jamieson ME (1984) Aneuploidy and cystic hygroma detectable by ultrasound. Prenat Diagn. 4(5):377-382 34. Benacerraf BR, Frigoletto FD, Laboda LA (1985) Sonographic diagnosis of Down syndrome in the second trimester. Am J Obstet Gynecol. 153(1):49-52 35. Benacerraf BR, Frigoletto FD Jr, Cramer DW (1987) Down Syndrome: sonographic sign for the diagnosis in the second-trimester fetus. Radiology. 163(3):811-813 36. Nicolaides KH, Azar G, Byrne D, Mansur C, Marks K (1992) Fetal nuchal translucency: ultrasound screening for chromosomal defects in first trimester of pregnancy. Brit Med J. 304(6831):867-869 37. Nadel A, Bromley B, Benacerraf BR (1993) Nuchal thickening or cystic hygroma in first- and early second-trimester fetuses: prognosis and outcome. Obstet Gynecol. 82(1):43-48 38. Bewley S, Roberts LJ, Mackinson AM, Rodeck CH (1995) First trimester fetal nuchal translucency: problems with screening the general population. 2. Br J Obstet Gynaecol. 102(5):386-388 39. Snijders RJ, Noble P, Sebire N, Souka A, Nicolaides KH (1998) UK Multicentre project on assessment of risk of Trisomy 21 by maternal age and fetal nuchal-translucency thickness at 10-14 weeks of gestation. Fetal Medicine Foundation First Trimester Screening Group. Lancet 352(9125):343-346 40. Avigidou K, Papageorghiou A, Bindra R, Spencer K, Nicolaides KH (2005) Prospective first-trimester screening for Trisomy 21 in 30,564 pregnancies. Am J Obstet Gynecol. 192(6):1761-1767 41. Kagan KO, Avgidou K, Molina FS, Gajewska K, Nicolaides KH (2006) Relation between increased fetal nuchal translucency thickness and chromosomal defects. Obstet Gynecol. 107(1):6-10 42. Chitty LS, Kagan KO, Molina FS, Waters JJ, Nicolaides KH (2006) Fetal nuchal translucency scan and early prenatal diagnosis of chromosomal abnormalities by rapid aneuploidy screening: observational study. Brit Med J. 332(7539):452-455 43. Phillip J, Silver RK, Wilson RD et al (2004) Late first-trimester invasive prenatal diagnosis: results of an international randomized trial. Obstet Gynecol. 103(6):1164-1173 44. Caughey AB, Hopkins LM, Norton ME (2006) Chorionic villus sampling compared with amniocentesis and the difference in the rate of pregnancy loss. Obstet Gynecol. 108(3 Pt 1):612-616 45. Smith-Bindman R, Hosmer W, Feldstein VA, Deeks JJ, Goldberg JD (2001) Second-trimester ultrasound to detect fetuses with Down syndrome: a meta-analysis. JAMA. 285(8):1044-1055 46. Souter VL, Nyberg DA (2001) Sonographic screening for fetal aneuploidy: first trimester. J Ultrasound Med. 20(7):775-790 47. Chudleigh P (2006) Standardising fetal measurements: an overview. Paper presented at annual congress of BMUS. 48. Grandjean H, Sarramon MF (1995) Sonographic measurement of nuchal skinfold thickness for detection of Down syndrome in second-trimester fetus: a multicentre prospective study. The AFDPHE Study Group. Obstet Gynecol. 85(1):103-106 49. Donnenfeld AE, Meister D, Allison J et al (1995) Interobserver variability of sonographically determined second-trimester nuchal skinfold thickness measurements. Ultrasound Obstet Gynecol. 5(2):119-122 50. Roberts LJ, Bewley S, Mackinson AM, Rodeck CH (1995) First trimester fetal nuchal translucency: problems with screening the general population. 1. Br. J Obstet Gynaecol. 102(5):381-385 51. Pandya PP, Altman DG, Brizot ML, Pettersen H, Nicolaides KH (1995) Repeatability of measurement of fetal nuchal translucency thickness. Ultrasound Obstet Gynecol. 5(5):334-337 52. Evans JA, Watson A, Moore AC, Brettle DS. (2006) How accurate are NT measurements? Paper presented at the 2006 BMUS annual congress. 53. Whittle M (1997) Ultrasonographic “soft markers” of fetal chromosomal defects. Brit Med J. 314(7085):918 54. Sepulveda W, Lopez-Tenorio J (2001) The value of minor ultrasound markers for fetal aneuploidy. Curr Opin Obstet Gynecol. 13(2):183-191 55. Susanne GO, Sissel S, Ulla W, Charlotta G, Sonja Ol (2006) Pregnant women’s responses to information about an increased risk of carrying a baby with Down syndrome. Birth. 33(1):64-73 56. Rochon M, Eddleman K (2004) Controversial ultrasound findings. Obstet Gynecol Clin North Am. 31(1):61-99 |
|
|||||||||||||||||||||||||
 |
