LETTERS TO THE EDITOR 521

at 10 weeks followed by ultrasound NT scanning at11 weeks for those with high serum-based risks. Themodeling method is described in Wright et al. (2004);the maternal age distribution is for England the Walesfrom 1996–1998, gestational age is based on ultrasoundand serum markers are corrected for maternal weight.The mean log (MoM) of NT at 11 weeks’ gestation wastaken from Wald et al. (2004).

Figure 1 shows the receiver-operating characteristiccurves when NT measurement is provided to the 10,20 and 30% of women with the highest serum-basedrisks. For comparison, curves are also shown when NTis not offered (0%) and when all women receive ascan (100%). More than half of the detection advantageof adding NT to serum markers can be obtained byscanning just 10% of women and there is little loss ofdetection when 30% are scanned. For example, at a 3%false-positive rate, selecting 10% for NT increases thedetection rate from 66 to 77%, whereas with NT for allwomen, it would increase to 84%. If NT were providedonly to the 30% of women with the highest serum-basedrisks, the detection rate would be 83%.

We have previously shown that sequential screeningin the first and second trimester is more efficient using

60

65

75

70

80

85

90

1 1.5 2 2.5 3 3.5 4 4.5 5

False Positive Rate (%)

Det

ectio

n R

ate

(%)

100%30%20%

10%

0%

Figure 1—Receiver-operating characteristic curves for contingentscreening with PAPP-A and free β-hCG at 10 weeks according tothe proportion of women receiving NT measurement at 11 weeks

a contingent approach than non-disclosure screening(Wright et al., 2004). The efficiency would not be greatlyimpaired if the first stage of the sequence was alsocontingent, as described above. For example, considera protocol involving serum PAPP-A and free β-hCG at10 weeks and NT at 11 weeks, with serum AFP, uE3and free β-hCG in the second-trimester contingent onthe first-trimester risk. Fixing the overall detection rateat 85%, the first-trimester detection rate at 60% and theproportion of tests completed in the first trimester at80%, the false-positive rate is 0.9%. If the policy weremodified so that NT was contingent on the serum-basedrisk so as to complete 70% of tests at 10 weeks and10% at 11 weeks, the false-positive rate only increasesto 1.1%.

First-trimester and sequential screening policies havehigher detection rates than second-trimester screening.However, the limited availability of NT scanning facili-ties may restrict the establishment of early pregnancyscreening in some localities. In these circumstances,contingent screening within the first trimester possiblyfollowed by sequential contingent screening is a goodalternative.

Dave Wright1, Ian Bradbury2, Peter Benn3,Howard Cuckle4 and Karen Ritchie51Department of Mathematics and Statistics, University ofPlymouth, UK2School of Biomedical Sciences, University of Ulster,Coleraine, UK3Department of Genetics and Developmental Biology,University of Connecticut, USA4Reproductive Epidemiology, School of Medicine, Universityof Leeds, UK5NHS Quality Improvement, Scotland, UKDOI: 10.1002/pd.1147

REFERENCES

Wald N, Rodeck C, Rudnicka A, Hackshaw A. 2004. Nuchaltranslucency and gestational age. Prenat Diagn 24(2): 150–151.

Wright D, Bradbury I, Benn P, Cuckle H, Ritchie K. 2004. Contin-gent screening for Down syndrome is an efficient alternative to non-disclosure sequential screening. Prenat Diagn 24(10): 762–766.

Prenatal imaging of limb-body wall complex by magnetic resonance imaging

A 28-year-old, gravida 4, para 1 woman was referred at15 weeks’ gestation to evaluate fetal structural abnor-malities. Her husband was 35 years old. The couplehad a 10-year-old healthy daughter. There was no fam-ily history of congenital malformations. The womanand her husband were non-consanguineous and healthy.She was not on any medication and did not have any

recent infection or diabetes mellitus during this preg-nancy. Initial two-dimensional (2D) ultrasound revealedabdominoschisis and megacystis. Subsequent investiga-tion by three-dimensional (3D) ultrasound showed anattachment of the fetus to the placenta and a defi-ciency of the right lower limb. Ultrafast magnetic reso-nance imaging (MRI) scans further depicted a detailed

Copyright 2005 John Wiley & Sons, Ltd. Prenat Diagn 2005; 25: 516–528.

522 LETTERS TO THE EDITOR

(a)

(b)

Figure 1—Fetal MRI scans in the sagittal section at 15 weeks’gestation show (a) extracorporeal urinary bladder (B), liver (L),and intestines (I), deficiency of a lower limb, abdominoplacentalattachment, and (b) scoliosis. LL, lower limb; PL, placenta; SP, spine

anatomic relationship between the fetus and the placenta,fetoplacental attachment, fetal scoliosis with extracorpo-real urinary bladder, intestines, and liver, and absenceof a normal umbilical cord (Figure 1). A diagnosis oflimb-body wall complex (LBWC) was established. TheMRI images were used to demonstrate the detailed imag-ing findings to the parents. The parents opted to ter-minate the pregnancy. The postmortem findings wereconsistent with LBWC with right lower limb deficiency,umbilical cord dysgenesis, urinary bladder distension,ambiguous genitalia, and anal atresia. The cranioface

was normal. Cytogenetic analysis of the fetus revealeda 46,XX karyotype.

The case presented here shared many characteris-tics with the phenotype of LBWC without craniofacial(CF) defects. Cases with LBWC without CF defectsusually present major anomalies of the lower limbs,abnormal genitalia, anal atresia, urinary tract malfor-mations, an abdominoplacental attachment, persistenceof the extra-embryonic coelom, and umbilical cordabnormalities (Russo et al., 1993; Russo and Vecchione,1996). In contrast, cases of LBWC with CF defectsmay show severe anomalies of the upper limbs, con-strictive amniotic bands, and a cranioplacental attach-ment. Russo et al. (1993) suggested that LBWC with-out CF defects is related to a defective lateral andcaudal folding process of the embryonic disk, whileLBWC with CF defects is caused by an early vas-cular disruption. Prenatal diagnosis of LBWC withoutCF defects should include a differential diagnosis ofamniotic band syndrome, omphalocele, gastroschisis,pentalogy of Cantrell, thoracoabdominal syndrome, andomphalocele-exstrophy-imperforate anus-spinal defects(OEIS) complex.

In utero detection of LBWC by 2D and 3D ultrasoundhas been well described (Chen et al., 2000, 2001, 2002).To the best of our knowledge, the present case is thefirst report of prenatal imaging of LBWC by MRI. Thepresent case shows that MRI is able to enhance the mor-phological evaluation of the complex anomalies of fetallimbs, spine, cranioface, internal organs, umbilical cord,and the fetoplacental attachment associated with LBWC.We suggest that MRI is a powerful adjunct to ultrasoundby providing more comprehensive images of congenitalanomalies, which are very helpful in parental counsellingand perinatal decision-making for pregnancies associ-ated with fetal abdominal wall defects.

Chih-Ping Chen1,2,3,4, Sho-Jen Cheng5, Yi-Hui Lin1

and Wayseen Wang21Department of Obstetrics and Gynecology, MackayMemorial Hospital, Taipei, Taiwan, Republic of China2Department of Medical Research, Mackay MemorialHospital, Taipei, Taiwan, Republic of China3School of Nursing, National Yang-Ming University, Taipei,Taiwan, Republic of China4College of Chinese Medicine, China Medical University,Taichung, Taiwan, Republic of China5Department of Radiology, Mackay Memorial Hospital,Taipei, Taiwan, Republic of ChinaDOI: 10.1002/pd.838

REFERENCES

Chen C-P. 2001. Prenatal diagnosis of limb-body wall complex withcraniofacial defects, amniotic bands, adhesions and upper limbdeficiency. Prenat Diagn 21: 418–419.

Chen C-P, Shih J-C, Chan Y-J. 2000. Prenatal diagnosis of limb-bodywall complex using two- and three-dimensional ultrasound. PrenatDiagn 20: 1020.

Chen C-P, Tzen C-Y, Chang T-Y, Yeh L-F, Wang W. 2002. Prenataldiagnosis of acrania associated with facial defects, amniotic bands

Copyright 2005 John Wiley & Sons, Ltd. Prenat Diagn 2005; 25: 516–528.

LETTERS TO THE EDITOR 523

and limb-body wall complex. Ultrasound Obstet Gynecol 20:94–95.

Russo R, D’Armiento M, Angrisani P, Vecchione R. 1993. Limbbody wall complex: a critical review and a nosological proposal.Am J Med Genet 47: 893–900.

Russo R, Vecchione R. 1996. Limb body wall complex: craniofacialdefects as a distinctive factor. Birth Defects 30: 157–164.

Contribution of ultrafast magnetic resonance imaging in prenatal diagnosis ofsonographically undetected cerebral tuberous sclerosis associated with cardiacrhabdomyomas

A 28-year-old, gravida 5, para 2, woman had an uncom-plicated pregnancy until 34 weeks’ gestation whenintrauterine growth restriction (IUGR) and fetal car-diac tumors were detected on routine ultrasound. Shehad experienced two abortions. Her first child died ofprematurity and hydrocephalus. Her second child wasnormal. This was her fifth pregnancy. She and her hus-band were healthy and did not have any family his-tory of tuberous sclerosis complex or cardiac tumors.Detailed sonographic examination of the fetal heartat 36 weeks’ gestation demonstrated two solid tumorsabout 0.5 cm in diameter arising from the interventric-ular septum and two tumors about 0.5 cm in diameterin the left ventricle. A diagnosis of cardiac rhabdomy-omas was made. Detailed sonographic examination ofthe fetal brain was normal. Ultrafast magnetic reso-nance imaging (MRI) performed at 36 weeks’ gesta-tion revealed small subependymal tubers (Figure 1) andcardiac rhabdomyomas (Figure 2). Subsequent weekly

Figure 1—Prenatal cerebral axial T2-weighted MRI scan at 36 weeks’gestation shows subependymal tubers (arrows)

Figure 2—Prenatal MRI scan with double inversion recovery and darkblood sequence of the fetal heart in a four-chamber view at 36 weeks’gestation shows a mass (arrowheads) in the left ventricle (LV) and asecond mass (arrowheads) arising from the interventricular septum(IVS) and protruding towards the right ventricle (RV)

follow-ups showed IUGR, but no signs of cardiac com-promise. At 39 weeks’ gestation, a 2262-g female infantwas delivered uneventfully. The karyotype was 46,XX.Neonatal echocardiography demonstrated two intracar-diac tumors in the left ventricles with the largest onebeing 0.389 cm in diameter and two from the inter-ventricular septum with the largest one being 0.215 cmin diameter. The infant’s ventricular outlets were notobstructed and the cardiac functions remained withinnormal limits. Brain ultrasound and renal ultrasoundwere unremarkable. MRI scans of the brain, however,revealed small subependymal tubers (Figure 3).

We describe the first reported case on prenatal MRIdemonstration of concomitant fetal cardiac rhabdomy-omas and cerebral tuberous sclerosis. This case high-lights the contribution of ultrafast MRI in the perinatal

Copyright 2005 John Wiley & Sons, Ltd. Prenat Diagn 2005; 25: 516–528.