Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
True Value of Scanning
- James S Smeltzer (4 March 2000)
-
Down's syndrome screening policy
- Howard Cuckle (7 March 2000)
-
Down's screening - A large independent trial may be useful
- Tim Reynolds (10 March 2000)
-
Down's syndrome screening in Southampton - what is the evidence?
- Bent Norgaard-Pedersen (13 March 2000)
-
Re: Down's syndrome screening policy
- David T Howe (14 March 2000)
-
Screening for Down's Syndrome
- P A Boyd (21 March 2000)
-
Combined modality approach -- the way forward for Down's syndrome screening ?
- Bidyut Kumar (28 March 2000)
|
|
|||
|
James S Smeltzer, Perinatologist Wellstar Health System
Send response to journal:
|
David Howe and colleagues are to be commended for their excellent population-based study. They confirm my personal experience with a captive population in which antenatal sonography has been at least 95% sensitive for the detection of Down's Syndrome. An additional benefit is the ability of a quality sonographic examination to find other major chromosomal and life-thereatening major anomalies, most of which are not amenable to serum marker screening. In addition, the sonogram provides a very real bonding experience for the family, and the potential for motivation to other health promoting behaviors, such as smoking cessation, avoidance of alcohol and proper nutrition. One limitation is, however, that this modality is very operator dependent. These data indicate what CAN be done with a good sonographic examination, not necessarily what will be done in a particular clinical setting, which depends on the individual sonographer, the sonologist, and the focus on quality of the practice within which the sonographer and sonologist work - as well as a minor dependence on equipment. The authors have found a way to measure this benefit appropriately. But the ingredients in this stew are hard to reproduce without considerable effort and a fanatical attention to quality. It is clear from these data and others that the conjunction of early nuchal lucency determination and a quality targeted mid-pregnancy examination can lead to more effective and cost-effective case-finding of genetically and anomaly-affected pregnancy than rampant serum screening and amniocentesis. Dr. Howe and the sonographers are to be commended for both turning this potential into a reality and documenting it. Perhaps the three optimal quality markers of any prenatal diagnosis program should be the overall sensitivity and specificity of prenatal diagnoses of major anomalies, and the percent of all prenatal chromosome studies that are abnormal. James S Smeltzer, MD |
|||
|
|
|||
|
Howard Cuckle, Professor Reproductive Epidemiology, 26 Clarendon Road, University of Leeds, Leeds LS2 9NZ
Send response to journal:
|
Howe and colleagues assert that the detection rate in Southampton using maternal age and second trimester ultrasound anomaly scanning is considerably higher than found elsewhere with serum screening.1 That is incorrect. Twenty-one large prospective intervention studies of multi-marker second trimester serum screening have been published.2 In total 353,000 women were screened including 514 with Down's syndrome. Protocols varied: 13 used three markers, eight two; 16 screened everyone, five only young women; there were 13 different cut-offs. The combined results yielded a 67% detection rate and 4.5% false- positive rate compared with 68% and 6.6% respectively in Southampton. Intervention studies are positively biased because of the detection of non -viable true-positives. Assuming 25% non-viability the rates become 60% and 61% respectively. Statistical modelling is a reliable tool in formulating Down's syndrome screening policy. This clearly shows that maternal age alone is ineffective: for England and Wales in 1993-7 only 30% of affected pregnancies are predicted in the oldest 5% of women. And that the observed second trimester serum screening results are in line with predictions.2 However, there are no sufficiently validated models of second trimester anomaly scanning. In contrast modelling is possible for first trimester nuchal translucency scanning. This predicts a 73% detection rate for a 5% false-positive rate and the observed rate of 78% for 8%, after allowing for viability bias, in the Fetal Medicine Foundation study is in line with predictions.3 Combining nuchal translucency with first trimester serum markers predicts 88% detection for 5% false-positives.4 The Southampton results are good but anomaly scanning studies are not always reproducible. Until confirmatory results are reported planners would best be guided by modelling. This favours serum screening, ideally in the first trimester and combined with nuchal translucency scanning. 1. Howe D, Gornall R, Wellesley D, Boyle T, Barber J. Six year survey of screening for Down's syndrome by maternal age and mid-trimester ultrasound scans. Brit Med J 2000;320:606-610. 2. Cuckle H, Murray J. Screening for neurological disorders. In: Fetal and Neonatal Neurology and Neurosurgery. (Eds. MI Levene, MJ Whittle, FA Chervenak, J Punt, MJ Bennett). Churchill Livingstone, London, UK. 2000. 3. Snijders RJM, Noble P, Sebire N, Souka A, Nicolaides KH. UK multicentre project on assessment of risk of trisomy 21 by maternal age and fetal nuchal-translucency thickness at 10-14 weeks of gestation. Lancet 1998;352:343-346. 4. Cuckle HS, van Lith JMM. Appropriate biochemical parameters in first trimester screening for Down's syndrome. Prenat Diagn 1999;19(6):505 -12. |
|||
|
|
|||
|
Tim Reynolds, Professor of Chemical Pathology Queen's Hospital, Burton-on-Trent
Send response to journal:
|
Many years ago, screening for Down's syndrome in pregnancy was only available by amniocentesis which was performed if a pregnant women was older than a certain age threshold. Everything changed in 1988 when Wald et al published their triple screen method (1). Since then we have seen increasingly Byzantine screening protocols proposed to wring out the last possible percentage of detection. Recently, straining ethics to the limit, it has even been suggested that screening by ultrasound and biochemical testing should be carried out in the first trimester, the results should be withheld to be later combined with second trimester results (2). All this despite a lack of any evidence that adding extra analytes provides any statistically significant increase in detection rates (3). It is therefore refreshing that Howe et al have challenged the accepted orthodoxy and shown that detection rates similar to those achievable by serum screening can be achieved by ultrasound screening alone (4). Howe's report indicates that it would be ethical to carry out a randomised trial comparing ultrasound with serum screening since ultrasound screening alone should not disadvantage any patient. It would be essential to examine a very large number pregnancies (i.e. at least 250,000 pregnancies / group) with 50% of patients in each participating site receiving the results of their serum screen, whilst the remainder only receive ultrasound [but with serum samples collected] so that true differences in detection rates could be identified. A trial of that size is essential because the confidence limits about detection rates are so large (3). Furthermore, it would be essential that anyone organising such a trial should be seen to be beyond reproach: i.e. to have no vested interest in proving that either form of screening should prevail by virtue of patent ownership, etc (3). This has the potential to save the NHS large sums of money - a proper examination of the comparative efficiency of the two methods should be carried out soon! REFERENCES 1. Wald NJ, Cuckle HS, Densem JW, Nanchahal K, Royston P, Chard T, Haddow JE, Knight GJ, Palomaki GE, Canick JA. Maternal serum screening for Down's syndrome in early pregnancy. BMJ 1988;297:883-7 2. Wald NJ, Watt HC, Hackshaw AK. Integrated screening for Down’s syndrome based on tests performed during the first and second trimesters. NEJM 1999:341;461-7 3. Reynolds TM. Down's Syndrome Screening: A controversial test, with more controversy to come! J Clin Path 2000; In press 4. Howe DT, Gornall R, Wellesley D, Boyle T, Barber J. Six year survey of screening for Down's syndrome by maternal age and mid-trimester scans. BMJ 2000; 320: 606-10 |
|||
|
|
|||
|
Bent Norgaard-Pedersen, Professor, head of Department Statens Serum Institut, Department of Clinical Biochemistry
Send response to journal:
|
Howe and colleagues(1) suggest that screening for Down’s syndrome (DS) by age and mid-trimester ultrasound scans may perform as well as conventional screening using serum markers and age. This conclusion is wrong, as they report a detection rate (DR) of 68% for a false positive rate (FPR) of 6.6% and a detection time as late as 24 weeks. This is clearly inferior to the cumulative results of prospective serum screening where a similar DR can be obtained with a FPR of only 4.5%(2), and the confirmatory amniocentesis can be performed prior to week 18. In accordance with this, a recent serum screening program for DS covering one county in Denmark(3), and comprising 17023 low-risk pregnancies, showed a DR of 68% and a FPR of 4.1%. Additional scanning reduced that figure to 3.0%, corresponding to 30 amniocentesises per DS case compared with 105 invasive tests per DS case in the age indication group (>35 years) in the same county. The screening procedures in Southampton are based upon maternal age >35 years, 19 weeks anomaly scan, and apparently also alfafoetoprotein (AFP) screening for spina bifida. The existence of serum AFP results will certainly improve the performance of anomaly scans. It is difficult to understand how the age distribution can be estimated from AFP laboratory records as most women >35 years usually have amniocentesis without any prior AFP determination? With only 10% of pregnant women above 35 years of age, how is it possible that 2/3 of DS cases are found in this group? These figures raise a suspicion of selection bias since only 1/4 of DS cases are expected in the older group, according to data from UK. Using the prospective data from a serum screening program in Cambridge(4) and the age distribution from Southampton1 one should expect a DR of 84% with a FPR of < 5%. Furthermore, a recent similar prospective Scandinavian Multicenter study(5) on ultrasound anomaly scanning in 18-20 week, found a 10 times lower detection rate for DS, namely 6.3%, among women <35 years of age. In conclusion, the Southampton data cannot support the claim that it is necessary to conduct a controlled comparison of maternal age and anomaly scan versus conventional serum screening in second trimester. On the contrary, serum screening is highly evidence-based and resources were better spent on controlled trials of first trimester serum and ultrasound screening. 1. Howe D, Gornall R, Wellesley D, Boyle T, Barber J. Six year survey of screening for Down's syndrome by maternal age and mid-trimester ultrasound scans. Br Med J 2000; 320: 606-610. 2. Cuckle H. Established markers in second trimester maternal serum screening. Early Hum Dev 1996; 47 (Suppl): 27-29. 3. Christiansen M, Petersen PL, Permin M, Larsen LA, Nørgaard-Pedersen B. Maternal serum screening for congenital malformation and Down syndrome in Sønderjyllands County. Eight years of experience. Ugeskr Laeg 1999; 161: 6928-6934. 4. Goodburn SF, Yates JRW, Raggatt PR, Carr C, Ferguson-Smith ME, Kershaw AJ et al. Second trimester maternal serum screening using alpha- fetoprotein, human chorionic gonadotrophin, and unconjugated oestriol: experience of a regional programme. Prenatal Diagnosis 1994; 14: 391-402. 5. Jørgensen FS et al. Multiscan - a Scandinavian multicenter second trimester obstetric ultrasound and serum screening study. Acta Obstet Gynecol Scand 1999; 78: 501-510. |
|||
|
|
|||
|
David T Howe, Consultant in FetoMaternal Medicine Princess Anne Hospital, Southampton
Send response to journal:
|
Editor, We thank Professor Cuckle for his interest in our study(1), but are disappointed that he has misrepresented(2) what we concluded. We did not state that using maternal age and ultrasound as a screening policy for Down's was better than serum screening. Instead we said that our detection rate was much higher than the predictions made for screening by maternal age in the demonstration projects of serum screening to which we referred. There are considerable public health implications in selecting the best screening policy for Down's syndrome because of the costs involved. We also have a responsibility as a profession not to cause unhappiness and confusion to pregnant women in our care by giving them inconsistent advice about the best method of screening. We believe the great strength of our study is that it describes what was actually achieved in detecting affected pregnancies in our entire local population, over a prolonged period. The statistical modelling he advocates predicts we should have detected only 30% of affected pregnancies, instead of the 68% actually achieved, since it makes no allowance for the additional effect of routine ultrasound. In addition, there are two sources of bias in the demonstration projects of serum screening to which Professor Cuckle refers, which tend to overestimate its effectiveness and may reduce the accuracy of statistical modelling. Firstly, many demonstration projects include only the screened population(3) whereas uptake of screening, where reported, is around 75%(4). Babies with Down's syndrome born to women in the local population who did not have serum screening are thus not counted. Secondly, women who were deemed "screen positive" but who declined amniocentesis are counted as having had successful antenatal detection of Down's, when the diagnosis was really not made until after delivery(4). If we applied the same logic to our own study, we would count women who were offered but declined amniocentesis as successfully detected and claim an antenatal pick-up rate of 81% (43/53). This is clearly not what really happened. We leave it to others, particularly those involved in health care planning, to judge whether such statistical modelling can be relied on as "a reliable tool in formulating ..policy". Our advice remains that a higher standard of evidence, based on properly controlled trials, should be demanded before any new screening methods are widely introduced. David Howe
Diana Wellesley
Tracy Boyle 1. Howe DT, Gornall R, Wellesley D, Boyle T, Barber J. Six year survey of screening for Down's syndrome by maternal age and mid-trimester ultrasound scans. BMJ 2000;320:606-10. 2. Cuckle H. Down's syndrome screening policy (electronic response). eBMJ 2000 3. Haddow JE, Palomaki GE, Knight GJ, Williams J, Pulkkinen A, Canick JA, Saller, D N, Bowers, G B. Prenatal screening for Down's syndrome with use of maternal serum markers. New England Journal of Medicine 1992;327:588-593. 4. Goodburn SF, Yates JRW, Raggatt PR, Carr C, Ferguson-Smith ME, Kershaw AJ, et al. Second trimester maternal serum screening using alpha- fetoprotein, human chorionic gonadotrophin, and unconjugated oestriol: experience of a regional programme. Prenatal Diagnosis 1994;14:391-402. |
|||
|
|
|||
|
P A Boyd
Send response to journal:
|
Editor- We were interested in the findings of Howe et al1 regarding screening for Downs syndrome and their suggestion that the benefits of mid trimester serum screening have been overrated. The Women's Centre at the John Radcliffe Hospital in Oxford has a similar population in terms of number of deliveries (33,920 local women for the years 1993-98 inclusive), percentage of women having prenatal karyotyping (7%) and screening policy for Downs syndrome (amniocentesis offered to women aged >35 at expected date of delivery; no serum or nuchal screening on the NHS although some organise this privately). However, there is a higher maternal age distribution (16% age over 35 years compared with 10% in Southampton). Using data from the Oxford Congenital Anomaly Register, Oxford Genetics Laboratory and Oxford maternity data system we have performed a similar analysis on the 78 ongoing pregnancies of local women that resulted in a Down's syndrome fetus/infant delivered between the years 1993 and 1998 inclusive and present the findings in the Table.
____________________________________________________________
Year Primary reason for karyotyping
in cases prenatally diagnosed
before 24 weeks gestation
_____________________________
Maternal
age Serum
screen Suspicious
scan Diagnosed Total+ (%
postnatally* prenatally
detected)
__________________________________________________________
1993 2 0 0 3 5 (40%)
1994 4 1 2 8 15 (47%)
1995 3 1 4 6 14 (57%)
1996 3 2 2 4 11 (64%)
1997 2 2 4 4(2) 12 (67%)
1998 1 3 7 10(3) 21 (57%)
Total 15 9 19(1) 35 78 (55%)
__________________________________________________________
* Or at 24 weeks' gestation.
+ With Down's syndrome.
(1) 10 / 19 women were aged < 35 years.
(2) one had a positive serum screen, was age > 35
and declined amniocentesis; one had nuchal thickening
+ round head and declined amniocentesis.
(3) two aged > 35 late bookers with soft markers
diagnosed >24 weeks.
Of the 78 cases of Down's syndrome 43 (55%) were detected prenatally. The prenatal detection rate in mothers aged under 35 years was 41%.Of the 35 cases not prenatally diagnosed before 24 weeks eight were born to mothers aged over 35 years who would have been eligible for amniocentesis. In Oxford there has been an increase in the prenatal detection of Down's syndrome in the six years analysed. This increase is mostly due to detection because of a suspicion at anomaly scan. However the detection rate in Oxford is lower than that reported in Southampton in spite of a higher maternal age distribution. Whilst we agree that there are advantages to using the anomaly scan as a screening tool for Down's syndrome (more economical, single intervention, increased diagnosis of other chromosome abnormalities) there are also disadvantages. For example, lack of awareness by most women that the anomaly scan may lead to the detection of Down's syndrome and relatively late diagnosis (in this study 12/19 cases diagnosed because of a suspicion on scan the gestation at this scan was >=20 weeks). We feel that the introduction of biochemical screening would have some advantages; first earlier diagnosis, second (with appropriate pre test counselling) more informed consent and third, higher detection rate for a lower false positive rate (and hence fewer normal babies lost by miscarriage). We agree with Howe et al that it is important to obtain evidence of the effectiveness of new screening methods such as ultrasound in combination with serum screening2 and compare them with current practice before their widespread introduction. PA Boyd AJM Crocker M Jefferies PF Chamberlain 1. Howe DT, Gornall R, Wellesly D, Boyle T, Barber J Six year survey of screening for Down's syndrome by maternal age and mid-trimester ultrasound scans BMJ 2000;320:606-10 2. Wald NJ, Watt HC, Hackshaw AK Integrated screening for Down's syndrome based on tests performed during the first and second trimesters N Engl J Med 1999;341:461-7 |
|||
|
|
|||
|
Bidyut Kumar
Send response to journal:
|
Editor- The article by Howe et al1 was interesting and thought provoking. We would like to make a few points lest readers get the impression that Down's syndrome (DS) screening by maternal age and 20 week ultrasound scan (USS) is universally better than antenatal serum screen (ANSS) recommended by the Royal College of Obstetricians and Gynaecologists(RCOG). The age related of incidence of DS dictates that efficacy of screening for this disorder will be influenced by age distribution of the population being screened and quite simply that applies to both ANSS as well as screening by USS. In women aged >/= 35 years in this study (about 3100 in number), assuming amniocentesis uptake rate of 75%, Howe et al performed 2325 amniocenteses to detect 22 DS cases giving a positive predictive value of 0.9% in this group of women. It is most likely that these figures can be bettered by prior use of ANSS to select patients for amniocentesis and this in turn will reduce the amniocentesis pregnancy loss rate in women of 35 and over. Referring to Table 4 in Howe's article, 5 cases should have been excluded from the number of DS detected. Women with previous history of DS (3 cases) do not need screening and first trimester nuchal translucency (NT) measurement (1 case) and one case of ANSS are precluded by the stated objective of this review. Therefore the actual detection rate is 42% (6/14) in <35 years, 73% (25/34) in >/= 35 years and 64% (31/48) overall. In this context one should remember that the RCOG report (July 1993) also mentioned that routine US pregnancy dating increased detection rate with ANSS from 58% to 67%. Screening by routine 20 week US scan is controversial because there is no consensus on risk estimate of DS in presence of one or more scan detectable abnormalities including 'soft markers'. Moreover, while scaning, one should be prepared to visualise and explain findings totally unrelated to DS. Obviously proper counselling is of utmost importance to avoid undue worries and that can be extremely time-consuming as well as confusing. Notwithstanding these problems of US scaning, I think the way forward is screening by a combined modality approach. Roberts et al2 achieved a detection rate of 80% with an invasive procedure rate of 4.8% among 36410 women with a median age of 27 years with the use of (i) second trimester ANSS (ii) offering karyotyping for maternal age over 39 or prior high risk and (iii) fetal US scan at 18-22 weeks. Others incorporating NT measurement with ANSS achieved detection rates of 100% (5% false positive)3 and 85% (0.9% false positive)4 while Sullivan et al showed that ANSS is an useful adjunct to targeted ultrasonography in the evaluation of fetuses with choroids plexus cysts5. Bidyut Kumar.
Geeta Kumar References 1) Howe DT, Gornall R, Wellesley D, Boyle T, Barber J. Six year survey of screening for Down's syndrome by maternal age and mid-trimester ultrasound scans. BMJ 2000;7235:606-610 (4 March) 2) Roberts D, Walkinshaw SA, McCormack MJ, Ellis J.Prenatal detection of trisomy 21: combined experience of two British hospitals. Prenatal Diagnosis 2000; 20: 17-22. 3) Benattar C, Audibert F, Taieb J, Ville Y et al. Fetal Diagnosis and Therapy 1999;14:112-117. 4) Wald NJ, Watt HC, Hackshaw AK. N Engl J Med 1999;341(7):461- 467.(12 August) 5) Sullivan A, Giudice T, Vavelidis F, Thiagarajah S. Am J Obstet Gynecol 1999; 181: 260-265. |
|||