Ultrasound Obstet Gynecol 2001; 18: 432 – 436
The accuracy of antenatal ultrasound in the detection of facial clefts in a low-risk screening population
Blackwell Science Ltd
C. CASH, P. SET and N. COLEMAN* Departments of Radiology and *Histopathology, Addenbrooke’s NHS Trust and the University of Cambridge, Cambridge, UK
K E Y WO R DS: Antenatal ultrasound, Facial clefts, Screening
ABSTRACT Objectives To evaluate the accuracy of prenatal ultrasound in the detection of facial clefts in a low-risk screening population and to report on the outcome of these pregnancies. Design We retrospectively reviewed antenatal ultrasound records from the obstetric ultrasound department, postnatal records from the regional craniofacial unit and autopsy reports of fetuses over 16 weeks’ gestational age from the regional pathology department over a 5-year period (1993– 97). Cross-referencing between the three data sets identified all cases of facial clefts. Results Out of 23 577 live and still births, 30 had facial clefts; four were excluded from the study. Of the remaining 26 cases, 10 had associated major anomalies. There were 19 live births and seven terminations. Six of the seven terminations had other major abnormalities. Our detection rate for cleft lip and palate was 93% and the detection rate for isolated cleft palate was 22%. Isolated cleft lip was detected in 67% of cases. The overall detection rate for facial clefts was 65%. Conclusion From our results and a review of the literature it is clear that before standards can be set we need to define which facial clefts are sonographically demonstrable. Our data provide information for effective counseling and setting of standards for clinical practice.
INTRODUCTION Despite much controversy over the value of routine ultrasound screening for fetal abnormalities1,2 ultrasound is increasingly becoming part of routine antenatal care in Europe and the UK. There is a wealth of literature regarding the major abnormalities such as cardiac and craniospinal defects which have permitted standards for detection rates to be set. Although much has been written about antenatal diagnosis of facial clefts the reported detection rates are extremely variable, ranging from 0 to 91%3,4. In the current climate of clinical governance and evidence-based medicine, published
standards are needed in antenatal ultrasound. With the increasingly accessible medical information available to the public through Internet usage, questions on accuracy are now being raised. We evaluated the accuracy of antenatal ultrasound in the detection of facial clefts in a low-risk screening population at the Rosie Maternity Unit at Addenbrooke’s Hospital in Cambridge.
SUBJECTS AND METHODS From January 1993 to December 1997, antenatal records from the obstetric ultrasound department, postnatal records from the craniofacial department and autopsy reports on any termination or intrauterine death occurring after 16 weeks of gestation from the department of pathology were retrospectively reviewed. At our hospital, which supervises approximately 5000 births per annum, women attending for routine antenatal care are offered two ultrasound examinations, at 8 –12 weeks and at 18–20 weeks of gestation. The 20-week examination consists of a 30-min transabdominal study and includes a systematic structural survey of the fetus. Coronal views of the lips, tangential views of the maxillary alveolus and a profile view of the face form part of the survey (Figures 1 and 2). Repeat examinations are carried out in all situations in which the initial examination was incomplete due to either poor fetal position or large maternal body mass index. Thus the fetal lips and alveolus are imaged in all cases, with the exception of a negligibly small number of cases of women who have an extremely large body mass index. All examinations are completed before 24 weeks’ gestation. All anomaly ultrasound examinations are performed by trained radiographers with a diploma of medical ultrasound or a postgraduate diploma in ultrasound, and with recourse to consultant staff whenever there is any potential problem. In the case of an abnormal finding, patients are counseled by the fetal medicine consultant and all isolated facial clefts are also referred antenatally to the plastic surgeon. After confirmation of an abnormality an entry is made into an audit
Correspondence: Dr P. Set, Department of Diagnostic Radiology, Addenbrooke’s NHS Trust, Hills Road, Cambridge CB2 2QQ, UK (e-mail:
[email protected]) Received 10-1-00, Revised 15-5-01, Accepted 6-6-01
432
ORIGINAL PAPER
Antenatal ultrasound and facial clefts
Cash et al.
Figure 1 Ultrasound image (a) and corresponding line diagram (b) of the fetal maxillary alveolus seen on the tangential view. Shaded area represents the maxillary alveolus.
Figure 2 Ultrasound image (a) and corresponding line diagram (b) of the entire fetal maxillary alveolus seen on the transaxial view. Shaded area represents the maxillary alveolus.
file. This file has been kept since 1992 and records all fetal abnormalities detected. Each case is meticulously followedup postnatally. During the 5-year study period all examinations were performed on a Toshiba 140 or 250 (Toshiba, Tokyo, Japan) using the 5-MHz and /or the 3.5-MHz probes. We retrospectively reviewed all the antenatal records of facial clefts recorded in the audit file from January 1993 to December 1997. Our Craniofacial unit at Addenbrooke’s Hospital is the regional center for cleft repair for East Anglia, UK. During the study period all cases of cleft lip and palate were referred to this unit for counseling and, in all instances of ongoing pregnancy, for postnatal care. We reviewed the craniofacial surgical computerized database for all postnatal facial clefts treated during the 5-year period. From this cohort all Cambridge patients’ notes were reviewed. Cross reference between the obstetric ultrasound department, the craniofacial unit and the pathology department data sets was performed.
From the craniofacial surgical data for the population studied there were 21 live births with a cleft defect. During this period seven pregnancies with affected fetuses were terminated as a result of the antenatal ultrasound findings. In only one case was the decision to terminate made on the basis of a cleft defect alone: the parents were teenagers living in difficult social circumstances. In the other six cases the cleft defect was one of multiple abnormalities. In two late missed miscarriages, diagnosed at the anomaly ultrasound examination, subsequent autopsies revealed cleft defects. The overall incidence of facial clefts during this period was therefore 30 cases in 23 577 live or still births (0.127%). Review of the antenatal records of these 30 cases revealed that only 26 of these cleft defects could have been diagnosed at 18–20 weeks; four cases were excluded from further analysis for the following reasons: 1 In one of the live births the baby had a small cleft of the soft palate and Beckwith–Wiedemann syndrome. The cleft palate was not detected antenatally, nor was it detected clinically until the child was 3 years old. As it is highly unlikely that this anomaly could have been detected antenatally it has not been included in this study. 2 The mother of one baby born with a unilateral cleft lip and palate presented late and ultrasound was not performed until 29 weeks. At that stage abnormal limbs and cardiac anomalies were detected, but the lips and palate were not identified. At a second examination, performed by a consultant
RESULTS During the study period there were 23 577 live and still births. A total of 27 708 anomaly ultrasound examinations were performed. The number of anomaly studies exceeds the total number of births because of the inclusion of repeat examinations in the computer data.
Ultrasound in Obstetrics and Gynecology
433
Antenatal ultrasound and facial clefts
Cash et al.
obstetrician to confirm the anomalies already detected, the lips and palate were still not identified. This case has not been included in the study because the patient was examined beyond 24 weeks’ gestation. 3 The two missed miscarriages diagnosed at the anomaly ultrasound examination were excluded because it was clear early on during the ultrasound examination that there had been fetal demise and the examination was therefore terminated. Seventeen of the remaining 26 cases with cleft defects were detected by antenatal ultrasound (65%); these comprised 2 /3 (67%) cases of isolated cleft lip, 13/14 (93%) cases of cleft lip and palate, and 2 /9 (22%) cases of isolated cleft palate. In 12 of these 17 the diagnosis was completely accurate (70.5%). In the remaining five cases part of the defect was not detected, for example a left cleft lip and palate was detected but an incomplete right lip was missed. In two of these five cases, the face was not clearly demonstrated although a defect was suspected, but because the anomaly scan had detected other severe defects resulting in the decision to terminate the pregnancy, repeat examinations to study the face more clearly were not necessary. In nine cases the cleft defects were not detected antenatally (35%). There were no false-positive diagnoses.
Table 1 correlates the ultrasound reports with the postnatal or autopsy findings in each of the 26 cases. We achieved a sensitivity of 65% and a specificity of 100% for all facial clefts although the detection rate for cleft lip with or without palate (15/17 cases) was 88%. Given the total number of anomaly ultrasound examinations performed per live or still birth, we have shown an overall negative predictive value of 99.9% (nine false-negative and 23 551 true-negative diagnoses) and a positive predictive value of 100% (no false positive and 17 true positive diagnoses).
DISCUSSION Coupland and Coupland reported an incidence of 1.82 per 1000 live births of cleft lip and palate in the Trent region between 1973 and 19825. This is comparable to our incidence rate of 1.3 per 1000 live or still births. Before evaluating the accuracy of antenatal ultrasound in the detection of facial clefts, it is important to define which facial clefts are sonographically demonstrable. A review of the literature demonstrates that the cleft lip with or without cleft palate group constitutes the category which is detectable by antenatal ultrasound (Figures 3 and 4). An isolated cleft palate without other fetal abnormalities has not been
Table 1 Reported integrity of lips and alveolus at the 18 – 20-week anomaly ultrasound scan, pregnancy outcome and final clinical and pathological diagnosis in the 26 cases of facial cleft Pregnancy outcome
Final clinical and pathological diagnosis
Normal Normal Normal Normal Normal Normal Normal Unilateral cleft lip and alveolus Unilateral cleft lip and alveolus Unilateral cleft lip and alveolus Unilateral cleft lip and alveolus Unilateral cleft lip and alveolus Unilateral cleft lip Incomplete unilateral cleft lip Bilateral cleft lip and alveolus Bilateral cleft lip and alveolus Unilateral cleft alveolus Left cleft lip and alveolus Unilateral cleft alveolus Normal (but multiple other anomalies) Abnormal facial profile (and multiple anomalies) Midline facial defect (and multiple anomalies) Normal (but multiple anomalies)
Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Live birth Termination
Incomplete cleft soft palate in Twin 2 of a multiple pregnancy Unilateral cleft lip Unilateral complete cleft lip and palate Cleft palate Cleft palate, Pierre–Robin syndrome Cleft palate (mainly soft) Cleft soft palate Unilateral cleft lip and palate Unilateral cleft lip and palate; Turner syndrome Unilateral cleft lip and palate Unilateral cleft lip and palate Unilateral cleft lip and palate Unilateral cleft lip Incomplete unilateral cleft lip Bilateral cleft lip and palate plus double aortic arch Bilateral cleft lip and palate Unilateral cleft lip and palate; Opitz G syndrome Left cleft lip and palate; incomplete right cleft lip Unilateral incomplete cleft lip; cleft submucous soft palate Midline cleft palate; acrocephalosyndactyly; probable Aperts syndrome
Termination
Midline cleft palate; renal hypoplasia; coarctation of aorta; karyotype 46,XX
Termination
Midline cleft palate; multiple dysmorphic features consistent with 4p-syndrome
Termination
Bilateral cleft lip and alveolus (and multiple anomalies) Bilateral cleft lip and alveolus (and multiple anomalies) Bilateral cleft lip and alveolus
Termination Termination
Midline cleft palate; complete situs inversus; hydrocephalus; small VSD; karyotype 46,XY Bilateral cleft lip and palate; transposition of great vessels associated with pulmonary artery hypoplasia; VSD; karyotype 46,XX Bilateral cleft lip and palate plus severe anomalies; Roberts syndrome
Termination
Bilateral cleft lip and palate
Case
Ultrasound scan results
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
VSD, ventricular septal defect.
434
Ultrasound in Obstetrics and Gynecology
Antenatal ultrasound and facial clefts detected in studies reported in the English literature4,6–9. Turner and Twining9 suggest that isolated cleft palates are difficult to demonstrate because of acoustic shadowing from the facial bones. As an isolated cleft palate is not usually associated with an alveolar cleft; the diagnosis relies solely on visualizing the midline cleft through the palate, which is of the same echogenicity as the tongue. We propose that the difficulty in demonstrating isolated cleft palates may be due in
Figure 3 Coronal ultrasound image demonstrating unilateral cleft lip (arrow).
Cash et al. most instances to obliteration of the cleft defect by the tongue. A review of the literature suggests that isolated cleft palates may not be sonographically demonstrable without the use of power Doppler imaging. We therefore divided the study group into three subsets to allow better comparison of results between studies (Table 2). Our detection rate for cleft lip with or without palate was 88% (15/17 cases). An understanding of embryology10 permits a more accurate description and therefore diagnosis. In retrospect the cleft lip in cases 17 and 19 should have been predicted given that a cleft alveolus almost always coexists with a cleft lip. The high detection rate achieved in our study may be explained by a number of factors. Firstly, the time allocated for each ultrasound study was twice as long as that quoted in an earlier British series11 and the policy to reschedule the examination in cases of incomplete review may also have contributed to a better detection rate. Secondly, in our study, three views of the face were routinely obtained, while only the profile view was used by Shirley and colleagues11, and Chitty et al. did not routinely study the face until the second half of their study period12. Thirdly, although this study was performed in a teaching hospital setting with some consultant input, trained sonographers conducted the primary screening process. Our emphasis on active audit and continuing education for the sonographers probably contributes to the high detection rate. Fourthly, operating as a single center without antenatal tertiary referrals, our data did not suffer the inevitable machine and operator variability that affects multicenter pooled studies as demonstrated by Stoll et al.13. Recent studies will have benefited from more modern ultrasound technology which will no doubt enhance detection rates. Hafner and colleagues8, using modern technology to screen a low-risk population, achieved detection rates of 72%. Finally, the population studied may also affect the detection rate. In our series 31% of the study group had other abnormalities compared to none in the group of Shirley et al.11 and 49% in that of Benacerraf et al.6. For cleft lip with or without cleft palate, our detection rate of 88% is comparable to a rate of 90% for Benacerraf and colleagues. This study has recognizable limitations. Its retrospective nature does not permit a detailed analysis of the sequence of events, such as the initial number of false positives suspected by sonographers but not confirmed by consultant staff. Similarly, it is not possible in a retrospective study to demonstrate how many of the facial clefts were detected because of a detailed examination following detection of other abnormalities. In conclusion, our data suggest that studies of prenatal ultrasound diagnosis in a low-risk population are significant in providing information required for effective counseling in addition to producing data for setting national standards.
ACKNOWLEDGMENTS
Figure 4 Tangential ultrasound image demonstrating unilateral cleft maxillary alveolus (arrow).
Ultrasound in Obstetrics and Gynecology
We thank Sylvia Bishop, superintendent radiographer in the Rosie ultrasound department, and her staff, Mr G Hackett, consultant obstetrician, under whose care were some of these patients, and Mr P Hall, consultant maxillofacial surgeon.
435
Antenatal ultrasound and facial clefts
REFERENCES 1 Chitty LS. Ultrasound screening for fetal abnormalities. Prenat Diagn 1995; 15: 1241 – 57 2 Dubbins PA. Screening for chromosomal abnormality. Semin Ultrasound CT MR 1998; 19: 310 – 7 3 Levi S, Hyjazi Y, Schaaps JP, Defoort P, Coulon R, Buekens P. Sensitivity and specificity of routine antenatal screening for congenital anomalies by ultrasound: the Belgian multicentric study. Ultrasound Obstet Gynecol 1991; 1: 102 – 10 4 Bronshtein M, Blumenfeld I, Kohn J, Blumenfeld Z. Detection of cleft lip by early second trimester transvaginal sonography. Obstet Gynecol 1994; 84: 73 – 6 5 Coupland MA, Coupland AI. Seasonality, incidence and sex distribution of cleft lip and palate births in Trent region (1973– 82). Cleft Pal J 1988; 25: 33 – 7 6 Benacerraf BR, Mulliken JB. Fetal cleft lip and palate: sonographic diagnosis and postnatal outcome. Plast Reconstr Surg 1993; 92: 1045 – 51
436
Cash et al. 7 Boyd PA, Chamberlain P, Hicks NR. Six year experience of prenatal diagnosis in an unselected population in Oxford, UK. Lancet 1998; 352: 1577–81 8 Hafner E, Sterniste W, Scholler J, Schuchter K, Philipp K. Prenatal diagnosis of facial malformations. Prenat Diagn 1997; 17: 51–8 9 Turner GM, Twining P. The facial profile in the diagnosis of fetal abnormalities. Clin Radiol 1993; 47: 389–95 10 Sadler TW. Langman’s Medical Embryology, 7th edn. Baltimore: Williams & Wilkins, 1995: 331–40 11 Shirley IM, Bottomley F, Robinson VP. Routine radiographer screening for fetal abnormalities by ultrasound in an unselected low risk population. Br J Radiol 1992; 65: 564–9 12 Chitty LS, Hunt GH, Moore J, Lobb MO. Effectiveness of routine ultrasonography in detecting fetal structural abnormalities in a low risk population. BMJ 1991; 303: 1165–9 13 Stoll C, Dott B, Alembik Y, Roth MP. Evaluation of routine prenatal diagnosis by a registry of congenital anomalies. Prenat Diagn 1995; 15: 791–800
Ultrasound in Obstetrics and Gynecology
