Crandall, Robinson, and Grau
11. Burton BK. Outcome of pregnancy in patients with unexplained elevated or low levels of maternal serum alphafetoprotein. Obstet Gynecol 1988;72:709-13. 12. Salafia CM, Silberman L, Herrera NE, Mahoney MJ. Placental pathology at term associated with elevated midtrimester maternal serum a-fetoprotein concentration. AM J OBSTET GYNECOL 1988; 158: 1064-6.
September 1991 Am J Obstet Gynecol
13. Rodeck CH, Campbell S, Biswas S. Maternal plasma AFP in normal and complicated pregnancies. Br J Obstet Gynaecol 1976;83:24-32. 14. O'Brien WF, Stern light D, Torres C, Knuppel RA, Montenegro R. The value of early third trimester maternal serum alpha-fetoprotein determination. Prenat Diagn 1990; 10: 183-9.
Study of 156 cases of polyhydramnios and congenital malformations in a series of 118,265 consecutive births Claude G. Stoll, MD, Yves Alembik, MD, and Beatrice Dott, MD Strasbourg, France Polyhydramnios associated with congenital anomalies was studied over nine years in 118,265 consecutive pregnancies. The prevalence of this association was 1.32% (156 cases). A case-control study allowed the examination of genetic and environmental factors for the origin of polyhydramnios associated with congenital malformations. Diagnosis of polyhydramniOS associated with congenital malformations was performed prenatally in 41 % of the cases; 16% of the infants were stillborn. Fifty-five percent of the cases had more than one malformation, 13.4% of them had a chromosomal aberration, and 32% had multiple malformations that do not constitute a syndrome. There was an increase of consanguinity in the parents of our patients. The incidence of polyhydramniOS and congenital anomalies in first-degree relatives was 3.8%, and first-degree relatives had more malformations than the controls had (8.3% vs 3.2%). Our study demonstrated the low capacity of a general prenatal screening program because the diagnosis of malformations associated with polyhydramnios was made in only 41 % of the cases and only six of 21 chromosomal abnormalities were diagnosed prenatally. We recommend the use of fetal chromosome analysiS and careful ultrasonographic examination in every pregnancy complicated by polyhydramniOS. (AM J OBSTET GYNECOL 1991 ;165:586-90.)
Key words: Polyhydramnios, congenital anomalies, congenital malformations, prenatal diagnosis, chromosonal abnormalities The incidence of polyhydramnios vanes between 0.4% and 1.5%.2 The diagnosis of polyhydramnios, which was usually made clinically, was associated with an increased frequency of complications in both mother and fetus. In 358 clinically diagnosed cases of polyhydramnios Queenan and Gadow' found that 24.6% had diabetes mellitus, 11.5% had erythroblastosis fetalis, 8.4% had multiple births, 1.5% had acute polyhydramnios, and 20% had congenital anomalies; 34% From the Institut de Puericulture, Centre Hospitalo-Universitaire. Supported by grants from the Comite National des Registres (Institut National de la Sante et de la Recherche Medicale and Direction General de la Sante). Received for publication June 17, 1990; revised February 22, 1991; accepted March 1, 1991. Reprint requests: Claude Stoll, MD, Institut de Puericulture 23, Rue de la Porte de I'H8pital, Centre Hospitalo-Universitaire, 67091 Strasbourg Cedex, France. 6/1/29188
586
of the cases of polyhydramnios were idiopathic. Since that landmark report, the definition of polyhydramnios has shifted from a clinical to an ultrasonographic one. Here we consider only polyhydramnios associated with congenital malformations. Data on 156 cases of polyhydramnios occurring among 118,265 consecutive pregnancies delivered in Alsace in northeastern France from 1979 to 1987 are reported. A case-control study allowed the examination of genetic and environmental factors for the origin of polyhydramnios associated with congenital malformations. Complete ascertainment of pregnancies with polyhydramnios and congenital malformations in a defined geographic area provides an opportunity to study the epidemiologic association.
Material and methods The malformations in this study came from 118,265 consecutive births of known outcome, including 814
Polyhydramnios and congenital anomalies
Volume 165 Number 3
587
Table I. Prevalence of polyhydramnios with one fetal malformation and with more than one fetal malformation
Polyhydramnios plus one malformation Polyhydramnios with several malformations Chromosomal abnormalities Nonchromosomal recognized syndromes Multiple malformations without recognized syndromes and without chromosomal abnormalities
stillborn babies, registered in our registry of congenital malformations, which covers the "Department du BasRhin" in which Strasbourg is located. Our registry is described elsewhere. 3 Briefly, the newborns of 11 hospitals were examined for the period from Jan. 1,1979, to Dec. 31, 1987. No deliveries took place at home in the area under study. All newborn children were registered within the first nine days post partum, as were all fetuses with a minimum age of 20 weeks. New cases diagnosed up to 1 year after birth also were registered. When a suspected or confirmed case was located, the requested information was checked by the doctor in available records: prenatal consultation records, maternity files, neonatal unit files, autopsy reports, outpatient clinic files, and pediatric and surgery files. Diagnosis of polyhydramnios was made ultrasonographically. All scans were obtained with a real-time sector scanner with the use of a 3.5 to 5.0 or 7.5 MHz transducer of appropriate focal length. The scans were obtained in a routine fashion between 16 and 27 weeks of gestation. Polyhydramnios was diagnosed when more than one pocket of amniotic fluid >7 cm" was noted in a vertical dimension. More than 90% of the pregnant women of our region have antenatal ultrasonography. For each case more than 50 factors included in the registration forms were studied: parity and previous pregnancies, parental age, residence, education, ethnic origin, length, head circumference and weight at birth, inheritance, cytogenetics, occurrence in twins, environmental factors, seasonality, and pregnancy. For each case a control was studied. The control was a normal child of the same sex born after the case in the same maternity hospital. For seasonality studies the monthly occurrence of all normal births from 1979 to 1987 was chosen as a control rate. For sex ratio, controls were the normal newborn population. Statistical analysis was obtained by the Statistical Analysis System' procedure software package. Statistical comparison was by X2 test with Yates' correction for the characteristic under study in cases and in controls and by Student's t test for comparison of means.
70 86 21 15 50
44.9 55.0 13.4 9.6 32.0
Prevalence per 10,000 births
Sex ratio (male /female)
5.91 7.27 1.77 1.26 2.70
1.18 1.15 1.33 1.14 1.08
When it could not be assumed that vanances were equal, an adapted t test was used. Odds ratio values were analyzed by X2 test. Results
During the study period 156 cases of polyhydramnios were detected in a total of 118,265 consecutive births, representing an overall prevalence of 1.319 for polyhydramnios associated with congenital malformations detectable at birth and until 1 year of age (Table I). The numbers of cases by year were 15, 14,21,25,26, 14, 11, 12, and 18 from 1979 to 1987, respectively. One hundred twenty-one children (77.5%) were live births, 25 (16.0%) were stillbirths, and 10 (6.4%) were terminated pregnancies. Diagnosis of malformation was performed at birth in 68 cases (43.6%), during the first week of life in 16 newborns (10.2%), and during the first year of age in eight children (5.1 %). Diagnosis of an anomaly was performed prenatally in 64 of 156 cases (41.0%). There were 70 (44.9%) cases with only one malformation and 86 (55.0%) cases with more than one malformation. The malformations associated with polyhydramnios are listed in Table II. The expected number of each specific malformation was calculated from the frequency of that specific malformation in infants with multiple malformations without polyhydramnios who had been registered during the period from 1979 to 1987; chromosomal abnormalities were excluded. These other malformations allowed us to separate these cases into recognized syndromes (chromosomal and nonchromosomal) and nonrecognized syndromes (multiple malformations) (Table I). The sex ratios are shown in Table I. Overall sex ratio was 1.16 (84 males, 72 females). There were no differences between parents of fetuses with polyhydramnios with malformations and control parents with respect to the distribution of parity and previous pregnancies, previous unproductive pregnancies, previous stillbirths, parental age (mean maternal and paternal ages, 27.7 and 30.5, respectively), ethnic origin, residence, and education.
588
Stoll, Alembik, and Dott
September 1991 Am J Obstet Gynecol
Table II. Malformations in infants born after pregnancy complicated by polyhydramnios, with chromosomal abnormalities excluded and expected malformations by standardization for our registry of congenital anomalies Infants with multiple malformations and polyhydramnios (%)
No.
Congenital heart disease Ventricular septal defect Atrial septal defect Atrioventricular canal Ductus arteriosus Pulmonary stenosis Aortic stenosis Mitral stenosis Left heart hypoplasia Dextrocardia Coarctation of aorta Other Skeletal malformations Congenital hip dislocation Clubfoot Polydactyly Syndactyly Limb reduction defects Other Digestive system Esophageal atresia Intestinal atresia Anal atresia Common mesentery Other Renal anomalies Hydronephrosis Vesicoureteral reflux Megaureter Renal cyst Renal agenesis (unilateral) Urethral valve Other Cleft Cleft palate Cleft lip or palate Central nervous system Anencephaly Spina bifida Agenesis of corpus callosum Other cerebral agenesis Hydrocephalus Other Ocular Microphthalmia Anophthalmia Cataract Other Ear, face, and neck Facial dysmorphia H ypertelorism Other Diaphragmatic hernia Abdominal wall Genital Other TOTAL
60 IS 5 4 5 2 2 4 3 1 3 13 3 13 3 4 10 13 17 15 1 6 6 S 3 5 1 1 1 4 II
7 I
2 2 7 9 2 3
5.6 l.5 l.2 l.5 0.6
Expected infants with multiple malformations without polyhydramnios (%)
IS.7 4.6 2.5 0.07 l.6
l.2 0.9
46
45
0.9 4.0 0.9 1.2 3.1 4.0 5.3 4.6
23
IS 23
l.S l.S 2.5 0.9 1.5 0.3 0.3
3.4 2.2 0.3 0.6
14.3
4.0 0.3t O.St 2.4 7.1
5.6 7.1
2.2 2.8 10
0.9
3.1
I I
5 8 10 7
3.1 25
13 3 2 53
2.5 3.1 2.2
7.8
4.0 0.9 0.6 16.5
321
Numbers do not sum because some patients had malformations in more than one additional system. = 0.05. tfJ = 0.001.
*P
1.0 l.S* 1.6 l.S 2.4
1.3 1.2 1.3 1.0 l.S
2.6 1.6 0.5 l.S 0.4 1.6 2.7 0.9 5.3 1.2 l.7 1.7* 2.4 3.5
Volume 165 Number 3
At birth, the mean weight, length, and head circumference of malformed children with polyhydramnios were, respectively, 3003 gm (controls, 3297; p < 0.05), 47.1 em (controls, 49.4; p < 0.001), and 34.4 em (controls, 34.6; not significant) and the weight of the placenta was 520 gm (controls, 596; p < 0.05). Analysis after standardization for sex and gestational age «36, 36 to 38, and >38 weeks) gave the same concluSIOns. Factors associated with malformed fetuses with polyhydramnios Genetic factors CONSAI\GUINITY OF PARENTS. Consanguinity of parents was admitted in 5.1 % of the cases (controls, 1.1 %; p < 0.001). Inherited mendel ian conditions were excluded. An inherited mendelian condition was present in nine cases. Six infants had nonchromosomal syndromes. CYTOGENETICS. Karotypes were obtained in 80 cases (57.5%). Of these, 59 were normal; the others were as follows: six Down syndrome, 13 trisomy 18 syndrome, and two autosomal deletions (one 18q - and one 4p -). Five of the infants with Down syndrome had no visceral malformation, and one had duodenal atresia and ventricular septal defect. Eleven of the 13 children with trisomy 18 had cardiac defect (ventricular or atrial septal defect, patent ductus arteriosus, bicuspid aortic or pulmonary valves), intestinal malrotation, megaureter, and cleft palate. The two infants with autosomal deletions had facial dysmorphia only. Of the 25 cases in which karyotype was obtained prenatally, only six abnormalities were present (one Down syndrome and five trisomy 18 syndrome). Among our 156 index patients, four twins (2.5%) were observed (controls, 2.6%). Each twin had an unaffected twin sibling. One mother and five siblings had the same nonmendelian inherited anomaly as the proband (3.8%), 11 second-degree relatives had the same anomaly as the proband, and 13 first-degree relatives had other malformations (8.3%) (controls, 3.2%, p < 0.05). Environmental factors. Odds ratios for risk factors were evaluated. No significant difference was found between congenital malformations associated with polyhydramnios and controls with regard to the risk factors studied: diabetes, epilepsy, x-ray films, arterial hypertension, fever, "flu," medication, cigarettes, occupational exposure, and parental age (Table III). Comment
Polyhydramnios is common, occurring m 0.4% to 1.5% of pregnancies. I. 2 Congenital anomalies are more common than polyhydramnios.' Both are found to-
Polyhydramnios and congenital anomalies
589
Table III. Odds ratio and 95% confidence interval for risk factors compared with controls Risk factors
Diabetes Epilepsy X-ray films Arterial hypertension Fever Contraception "Flu" Cigarettes Drugs Occupational exposure Mother Father Maternal age Paternal age
95% Confidence interval
1.80 1.01 0.02 0.52 0.029 0.48 0.77 1.15 0.03
0.45-7.15 0.99-1.03 0.006-0.099 0.21-1.24 0.002-0.397 0.18-1.24 0.28-2.08 0.69-1.93 0.005-0.316
1.61 1.35 0.63 0.67
0.85-3.03 0.72-2.53 0.38-1.07 0.39-1.15
gether in as many as 20% of the cases of polyhydramnios. I They should therefore lend themselves to epidemiologic studies. However, there are no unequivocal epidemiologic data on congenital malformations associated with polyhydramnios, because both are difficult to ascertain accurately. Data on congenital malformations associated with polydramnios have been collected in different ways. I , 4, 6, 7 In our series >90% of pregnant women had ultrasonographic examination; therefore the overall incidence of congenital malformations associated with polyhydramnios (1.32%) is certainly not too far from the true incidence of this association. The comparison of our data seen in the tables with those of other series shows a lower incidence in the study of Hill et al. 6 (0.98%). Few series have evaluated an un selected obstetric population for ultrasonographically detectable polyhydramnios. 8 ,9 Ten of the 12 cases of Quinlan et al." had a congenital anomaly that led to death in all but two of these infants. Hill et al. 6 detected 102 cases of polyhydramnios among 10,214 patients studied; 10 of these 102 infants were malformed. In the review of pregnancies complicated by hydramnios undertaken by Landy et al. 7 99 cases were ascertained. Thirteen of them (three being twins or triplets) had associated malformations. In our series the most common congenital defects associated with polyhydramnios were congenital heart disease, musculo-skeletal malformations, and anomalies of the gastrointestinal tract. This is in agreement with what is known of the physiopathologic features of polyhydramnios. s, 10, 11 We also noted more diaphragmatic hernias and facial malformations than expected and fewer cases of spina bifida, unilateral renal agen-
590 Stoll, Alembik, and Dott
esis, renal cysts, and abdominal wall defects than expected. Prevalence of chromosomal abnormalities was 1.77% (9.6% of our cases). As this prevalence is not given in the other series it is difficult to know if this prevalence is high or low. In the French nationwide program for prenatal diagnosis of chromosomal abnormalities a new indication was added in 1986, the study of the fetal karyotype after the presence of "signes d'appel" ("calling signs" such as the discovery of a fetal malformation, a small-for-dates fetus, polyhydramnios, or oligohydramnios).12 Of 2563 fetal karyotypes studied so far, 299 were abnormal (3.7%), which is only about a third of what we have found in our series that took only polyhydramnios into account. Consanguinity of the parents of fetuses with congenital malformations associated with polyhydramnios is not mentioned in the other studies. In our study 5.1 % r of the parents were consanguineous. Six first-aegree relatives of patients with congenital malformations associated with polyhydramnios had the same nonmendelian inherited malformations. This last frequency is more than twice that of the controls. These conclusions have to be taken into account for genetic counseling. Maternal epilepsy, hypertension, and fever during pregnancy were no more common among the mothers of the fetuses with congenital malformations associated with polyhydramnios than among controls. During the pregnancy in question 13.1 % of the women had threatened abortions, which is significantly different from the controls (p < 0.01). Diabetes mellitus was more frequent in the mothers of the children with congenital malformations associated with polyhydramnios than in the controls (4.5% vs 2.1 %), but the difference was not statistically significant. In the study of the literature (see Hill et al. 6 ) polyhydramnios was associated with maternal diabetes mellitus in 14.7% to 26.3% of cases, in fetuses with or without congenital malformations. In our series polyhydramnios was taken into consideration only when it was associated with fetal malformations. Our study demonstrates that there is no statistically significant increase in diabetes mellitus among the mothers of infants having polyhydramnios associated with congenital malformations. In our series the diagnosis of congenital malformations was performed prenatally in 41.0% of the cases. In the series of Landy et al.,7 99 cases of pregnancy complicated by hydramnios were studied. Prenatal diagnosis of congenital malformations was performed in only seven cases out of 21 with congenital malformations associated with polyhydramnios. In our patients with chromosomal abnormalities a
September 1991 Am J Obstet Gynecol
congenital malformation was detected prenatally at ultrasonographic examination in only one fetus with trisomy 21 syndrome (duodenal atresia) out of six cases and in five with trisomy 18 syndrome out of 13 cases (28.5% of all cases). In the 15 other cases of chromosomal abnormality no anomaly except polyhydramnios was detected at routine ultrasonographic examination. Benacerraf et aI. 13 were able to detect 80% of fetuses with trisomy 18 at ultrasonographic examination. In our series seven cases of trisomy 18 could have been detected if level II and III ultrasonographic examinations had been performed, but seven fetuses with chromosomal anomalies (4.5%) (three Down syndrome, two trisomy 18, and two autosomal deletions) would not have had ultrasonographic detection because they had no malformations except mild facial dysmorphia and mental retardation. Therefore we recommend the use of level II or III ultrasonographic examination in every pregnancy complicated by polyhydramnios. l:l When no malformation at level III ultrasonographic examination is detected, fetal chromosome analysis has to be discussed. REFERENCES 1. QueenanJT, Gadow Ee. Polyhydramnios: chronic versus acute. AM J OBSTET GYNECOL 1970; 108:349-52. 2. Wallenburg HCS, Wladimoroff VW. The amniotic fluid. II. Polyhydramnios and oligohydramnios. J Perinat Med 1977;5:193-205. 3. Stoll C, Roth MP, Dott B, Bigel P. Etude des malformations congenitales dans Ie Nord de l'Alsace. Med Hyg 1984;42:505-12. 4. Sivit Cj, Hill MC, LarsenJW, Lande 1M. Second trimester polyhydramnios: evaluation with US. Radiology 1987: 165:467-9. 5. SAS/STAT. Guide for personal computers. Version 6 edition. Cary, North Carolina: SAS Institute, 1987:773-85. 6. Hill LM, Breckle R, Thomas ML, Fries JK. Polyhydramnios: ultrasonically detected prevalence and neonatal outcome. Obstet Gynecol 1987;69:21-5. 7. Landy HJ, Isoda NB, LarsenJW. Genetic implications of idiopathic hydramnios. AM J OB5TET GYNECOL 1987; 157: 114-7. 8. Zamah NM, Gillieson MS, Walters JH, et al. Sonographic detection of polyhydramnios: a 5-year experience. AM J OB5TET GYNECOL 1982;143:523-6. 9. Quinlan RW, Cruz AC, Martin M. Hydramnios ultrasound diagnosis and its impact on perinatal management and pregnancy outcome. AM J OB5TET GYNECOL 1983; 145:306-11. 10. Seeds AE. Current concepts of amniotic fluid volume. AM J OBSTET GYNECOL 1980;138:575-86. 11. Hobbins JC, Grannum PA, Berkowitz RL, et al. Ultrasound in the diagnosis of congenital anomalies. AM J OBSTET GYNECOL 1979;134:331-45. 12. Briard ML. Recherche d'anomalies chromosomiques sur signes d'appel. La Depeche 1988; 13:29-46. 13. Benacerraf B, Miller WA, Frigoletto FD. Sonographic detection of fetuses with trisomies 13 and 18; accuracy and limitations. AMJ OB51'£T GYNECOL 1988;158:404-9.
11. Burton BK. Outcome of pregnancy in patients with unexplained elevated or low levels of maternal serum alphafetoprotein. Obstet Gynecol 1988;72:709-13. 12. Salafia CM, Silberman L, Herrera NE, Mahoney MJ. Placental pathology at term associated with elevated midtrimester maternal serum a-fetoprotein concentration. AM J OBSTET GYNECOL 1988; 158: 1064-6.
September 1991 Am J Obstet Gynecol
13. Rodeck CH, Campbell S, Biswas S. Maternal plasma AFP in normal and complicated pregnancies. Br J Obstet Gynaecol 1976;83:24-32. 14. O'Brien WF, Stern light D, Torres C, Knuppel RA, Montenegro R. The value of early third trimester maternal serum alpha-fetoprotein determination. Prenat Diagn 1990; 10: 183-9.
Study of 156 cases of polyhydramnios and congenital malformations in a series of 118,265 consecutive births Claude G. Stoll, MD, Yves Alembik, MD, and Beatrice Dott, MD Strasbourg, France Polyhydramnios associated with congenital anomalies was studied over nine years in 118,265 consecutive pregnancies. The prevalence of this association was 1.32% (156 cases). A case-control study allowed the examination of genetic and environmental factors for the origin of polyhydramnios associated with congenital malformations. Diagnosis of polyhydramniOS associated with congenital malformations was performed prenatally in 41 % of the cases; 16% of the infants were stillborn. Fifty-five percent of the cases had more than one malformation, 13.4% of them had a chromosomal aberration, and 32% had multiple malformations that do not constitute a syndrome. There was an increase of consanguinity in the parents of our patients. The incidence of polyhydramniOS and congenital anomalies in first-degree relatives was 3.8%, and first-degree relatives had more malformations than the controls had (8.3% vs 3.2%). Our study demonstrated the low capacity of a general prenatal screening program because the diagnosis of malformations associated with polyhydramnios was made in only 41 % of the cases and only six of 21 chromosomal abnormalities were diagnosed prenatally. We recommend the use of fetal chromosome analysiS and careful ultrasonographic examination in every pregnancy complicated by polyhydramniOS. (AM J OBSTET GYNECOL 1991 ;165:586-90.)
Key words: Polyhydramnios, congenital anomalies, congenital malformations, prenatal diagnosis, chromosonal abnormalities The incidence of polyhydramnios vanes between 0.4% and 1.5%.2 The diagnosis of polyhydramnios, which was usually made clinically, was associated with an increased frequency of complications in both mother and fetus. In 358 clinically diagnosed cases of polyhydramnios Queenan and Gadow' found that 24.6% had diabetes mellitus, 11.5% had erythroblastosis fetalis, 8.4% had multiple births, 1.5% had acute polyhydramnios, and 20% had congenital anomalies; 34% From the Institut de Puericulture, Centre Hospitalo-Universitaire. Supported by grants from the Comite National des Registres (Institut National de la Sante et de la Recherche Medicale and Direction General de la Sante). Received for publication June 17, 1990; revised February 22, 1991; accepted March 1, 1991. Reprint requests: Claude Stoll, MD, Institut de Puericulture 23, Rue de la Porte de I'H8pital, Centre Hospitalo-Universitaire, 67091 Strasbourg Cedex, France. 6/1/29188
586
of the cases of polyhydramnios were idiopathic. Since that landmark report, the definition of polyhydramnios has shifted from a clinical to an ultrasonographic one. Here we consider only polyhydramnios associated with congenital malformations. Data on 156 cases of polyhydramnios occurring among 118,265 consecutive pregnancies delivered in Alsace in northeastern France from 1979 to 1987 are reported. A case-control study allowed the examination of genetic and environmental factors for the origin of polyhydramnios associated with congenital malformations. Complete ascertainment of pregnancies with polyhydramnios and congenital malformations in a defined geographic area provides an opportunity to study the epidemiologic association.
Material and methods The malformations in this study came from 118,265 consecutive births of known outcome, including 814
Polyhydramnios and congenital anomalies
Volume 165 Number 3
587
Table I. Prevalence of polyhydramnios with one fetal malformation and with more than one fetal malformation
Polyhydramnios plus one malformation Polyhydramnios with several malformations Chromosomal abnormalities Nonchromosomal recognized syndromes Multiple malformations without recognized syndromes and without chromosomal abnormalities
stillborn babies, registered in our registry of congenital malformations, which covers the "Department du BasRhin" in which Strasbourg is located. Our registry is described elsewhere. 3 Briefly, the newborns of 11 hospitals were examined for the period from Jan. 1,1979, to Dec. 31, 1987. No deliveries took place at home in the area under study. All newborn children were registered within the first nine days post partum, as were all fetuses with a minimum age of 20 weeks. New cases diagnosed up to 1 year after birth also were registered. When a suspected or confirmed case was located, the requested information was checked by the doctor in available records: prenatal consultation records, maternity files, neonatal unit files, autopsy reports, outpatient clinic files, and pediatric and surgery files. Diagnosis of polyhydramnios was made ultrasonographically. All scans were obtained with a real-time sector scanner with the use of a 3.5 to 5.0 or 7.5 MHz transducer of appropriate focal length. The scans were obtained in a routine fashion between 16 and 27 weeks of gestation. Polyhydramnios was diagnosed when more than one pocket of amniotic fluid >7 cm" was noted in a vertical dimension. More than 90% of the pregnant women of our region have antenatal ultrasonography. For each case more than 50 factors included in the registration forms were studied: parity and previous pregnancies, parental age, residence, education, ethnic origin, length, head circumference and weight at birth, inheritance, cytogenetics, occurrence in twins, environmental factors, seasonality, and pregnancy. For each case a control was studied. The control was a normal child of the same sex born after the case in the same maternity hospital. For seasonality studies the monthly occurrence of all normal births from 1979 to 1987 was chosen as a control rate. For sex ratio, controls were the normal newborn population. Statistical analysis was obtained by the Statistical Analysis System' procedure software package. Statistical comparison was by X2 test with Yates' correction for the characteristic under study in cases and in controls and by Student's t test for comparison of means.
70 86 21 15 50
44.9 55.0 13.4 9.6 32.0
Prevalence per 10,000 births
Sex ratio (male /female)
5.91 7.27 1.77 1.26 2.70
1.18 1.15 1.33 1.14 1.08
When it could not be assumed that vanances were equal, an adapted t test was used. Odds ratio values were analyzed by X2 test. Results
During the study period 156 cases of polyhydramnios were detected in a total of 118,265 consecutive births, representing an overall prevalence of 1.319 for polyhydramnios associated with congenital malformations detectable at birth and until 1 year of age (Table I). The numbers of cases by year were 15, 14,21,25,26, 14, 11, 12, and 18 from 1979 to 1987, respectively. One hundred twenty-one children (77.5%) were live births, 25 (16.0%) were stillbirths, and 10 (6.4%) were terminated pregnancies. Diagnosis of malformation was performed at birth in 68 cases (43.6%), during the first week of life in 16 newborns (10.2%), and during the first year of age in eight children (5.1 %). Diagnosis of an anomaly was performed prenatally in 64 of 156 cases (41.0%). There were 70 (44.9%) cases with only one malformation and 86 (55.0%) cases with more than one malformation. The malformations associated with polyhydramnios are listed in Table II. The expected number of each specific malformation was calculated from the frequency of that specific malformation in infants with multiple malformations without polyhydramnios who had been registered during the period from 1979 to 1987; chromosomal abnormalities were excluded. These other malformations allowed us to separate these cases into recognized syndromes (chromosomal and nonchromosomal) and nonrecognized syndromes (multiple malformations) (Table I). The sex ratios are shown in Table I. Overall sex ratio was 1.16 (84 males, 72 females). There were no differences between parents of fetuses with polyhydramnios with malformations and control parents with respect to the distribution of parity and previous pregnancies, previous unproductive pregnancies, previous stillbirths, parental age (mean maternal and paternal ages, 27.7 and 30.5, respectively), ethnic origin, residence, and education.
588
Stoll, Alembik, and Dott
September 1991 Am J Obstet Gynecol
Table II. Malformations in infants born after pregnancy complicated by polyhydramnios, with chromosomal abnormalities excluded and expected malformations by standardization for our registry of congenital anomalies Infants with multiple malformations and polyhydramnios (%)
No.
Congenital heart disease Ventricular septal defect Atrial septal defect Atrioventricular canal Ductus arteriosus Pulmonary stenosis Aortic stenosis Mitral stenosis Left heart hypoplasia Dextrocardia Coarctation of aorta Other Skeletal malformations Congenital hip dislocation Clubfoot Polydactyly Syndactyly Limb reduction defects Other Digestive system Esophageal atresia Intestinal atresia Anal atresia Common mesentery Other Renal anomalies Hydronephrosis Vesicoureteral reflux Megaureter Renal cyst Renal agenesis (unilateral) Urethral valve Other Cleft Cleft palate Cleft lip or palate Central nervous system Anencephaly Spina bifida Agenesis of corpus callosum Other cerebral agenesis Hydrocephalus Other Ocular Microphthalmia Anophthalmia Cataract Other Ear, face, and neck Facial dysmorphia H ypertelorism Other Diaphragmatic hernia Abdominal wall Genital Other TOTAL
60 IS 5 4 5 2 2 4 3 1 3 13 3 13 3 4 10 13 17 15 1 6 6 S 3 5 1 1 1 4 II
7 I
2 2 7 9 2 3
5.6 l.5 l.2 l.5 0.6
Expected infants with multiple malformations without polyhydramnios (%)
IS.7 4.6 2.5 0.07 l.6
l.2 0.9
46
45
0.9 4.0 0.9 1.2 3.1 4.0 5.3 4.6
23
IS 23
l.S l.S 2.5 0.9 1.5 0.3 0.3
3.4 2.2 0.3 0.6
14.3
4.0 0.3t O.St 2.4 7.1
5.6 7.1
2.2 2.8 10
0.9
3.1
I I
5 8 10 7
3.1 25
13 3 2 53
2.5 3.1 2.2
7.8
4.0 0.9 0.6 16.5
321
Numbers do not sum because some patients had malformations in more than one additional system. = 0.05. tfJ = 0.001.
*P
1.0 l.S* 1.6 l.S 2.4
1.3 1.2 1.3 1.0 l.S
2.6 1.6 0.5 l.S 0.4 1.6 2.7 0.9 5.3 1.2 l.7 1.7* 2.4 3.5
Volume 165 Number 3
At birth, the mean weight, length, and head circumference of malformed children with polyhydramnios were, respectively, 3003 gm (controls, 3297; p < 0.05), 47.1 em (controls, 49.4; p < 0.001), and 34.4 em (controls, 34.6; not significant) and the weight of the placenta was 520 gm (controls, 596; p < 0.05). Analysis after standardization for sex and gestational age «36, 36 to 38, and >38 weeks) gave the same concluSIOns. Factors associated with malformed fetuses with polyhydramnios Genetic factors CONSAI\GUINITY OF PARENTS. Consanguinity of parents was admitted in 5.1 % of the cases (controls, 1.1 %; p < 0.001). Inherited mendel ian conditions were excluded. An inherited mendelian condition was present in nine cases. Six infants had nonchromosomal syndromes. CYTOGENETICS. Karotypes were obtained in 80 cases (57.5%). Of these, 59 were normal; the others were as follows: six Down syndrome, 13 trisomy 18 syndrome, and two autosomal deletions (one 18q - and one 4p -). Five of the infants with Down syndrome had no visceral malformation, and one had duodenal atresia and ventricular septal defect. Eleven of the 13 children with trisomy 18 had cardiac defect (ventricular or atrial septal defect, patent ductus arteriosus, bicuspid aortic or pulmonary valves), intestinal malrotation, megaureter, and cleft palate. The two infants with autosomal deletions had facial dysmorphia only. Of the 25 cases in which karyotype was obtained prenatally, only six abnormalities were present (one Down syndrome and five trisomy 18 syndrome). Among our 156 index patients, four twins (2.5%) were observed (controls, 2.6%). Each twin had an unaffected twin sibling. One mother and five siblings had the same nonmendelian inherited anomaly as the proband (3.8%), 11 second-degree relatives had the same anomaly as the proband, and 13 first-degree relatives had other malformations (8.3%) (controls, 3.2%, p < 0.05). Environmental factors. Odds ratios for risk factors were evaluated. No significant difference was found between congenital malformations associated with polyhydramnios and controls with regard to the risk factors studied: diabetes, epilepsy, x-ray films, arterial hypertension, fever, "flu," medication, cigarettes, occupational exposure, and parental age (Table III). Comment
Polyhydramnios is common, occurring m 0.4% to 1.5% of pregnancies. I. 2 Congenital anomalies are more common than polyhydramnios.' Both are found to-
Polyhydramnios and congenital anomalies
589
Table III. Odds ratio and 95% confidence interval for risk factors compared with controls Risk factors
Diabetes Epilepsy X-ray films Arterial hypertension Fever Contraception "Flu" Cigarettes Drugs Occupational exposure Mother Father Maternal age Paternal age
95% Confidence interval
1.80 1.01 0.02 0.52 0.029 0.48 0.77 1.15 0.03
0.45-7.15 0.99-1.03 0.006-0.099 0.21-1.24 0.002-0.397 0.18-1.24 0.28-2.08 0.69-1.93 0.005-0.316
1.61 1.35 0.63 0.67
0.85-3.03 0.72-2.53 0.38-1.07 0.39-1.15
gether in as many as 20% of the cases of polyhydramnios. I They should therefore lend themselves to epidemiologic studies. However, there are no unequivocal epidemiologic data on congenital malformations associated with polyhydramnios, because both are difficult to ascertain accurately. Data on congenital malformations associated with polydramnios have been collected in different ways. I , 4, 6, 7 In our series >90% of pregnant women had ultrasonographic examination; therefore the overall incidence of congenital malformations associated with polyhydramnios (1.32%) is certainly not too far from the true incidence of this association. The comparison of our data seen in the tables with those of other series shows a lower incidence in the study of Hill et al. 6 (0.98%). Few series have evaluated an un selected obstetric population for ultrasonographically detectable polyhydramnios. 8 ,9 Ten of the 12 cases of Quinlan et al." had a congenital anomaly that led to death in all but two of these infants. Hill et al. 6 detected 102 cases of polyhydramnios among 10,214 patients studied; 10 of these 102 infants were malformed. In the review of pregnancies complicated by hydramnios undertaken by Landy et al. 7 99 cases were ascertained. Thirteen of them (three being twins or triplets) had associated malformations. In our series the most common congenital defects associated with polyhydramnios were congenital heart disease, musculo-skeletal malformations, and anomalies of the gastrointestinal tract. This is in agreement with what is known of the physiopathologic features of polyhydramnios. s, 10, 11 We also noted more diaphragmatic hernias and facial malformations than expected and fewer cases of spina bifida, unilateral renal agen-
590 Stoll, Alembik, and Dott
esis, renal cysts, and abdominal wall defects than expected. Prevalence of chromosomal abnormalities was 1.77% (9.6% of our cases). As this prevalence is not given in the other series it is difficult to know if this prevalence is high or low. In the French nationwide program for prenatal diagnosis of chromosomal abnormalities a new indication was added in 1986, the study of the fetal karyotype after the presence of "signes d'appel" ("calling signs" such as the discovery of a fetal malformation, a small-for-dates fetus, polyhydramnios, or oligohydramnios).12 Of 2563 fetal karyotypes studied so far, 299 were abnormal (3.7%), which is only about a third of what we have found in our series that took only polyhydramnios into account. Consanguinity of the parents of fetuses with congenital malformations associated with polyhydramnios is not mentioned in the other studies. In our study 5.1 % r of the parents were consanguineous. Six first-aegree relatives of patients with congenital malformations associated with polyhydramnios had the same nonmendelian inherited malformations. This last frequency is more than twice that of the controls. These conclusions have to be taken into account for genetic counseling. Maternal epilepsy, hypertension, and fever during pregnancy were no more common among the mothers of the fetuses with congenital malformations associated with polyhydramnios than among controls. During the pregnancy in question 13.1 % of the women had threatened abortions, which is significantly different from the controls (p < 0.01). Diabetes mellitus was more frequent in the mothers of the children with congenital malformations associated with polyhydramnios than in the controls (4.5% vs 2.1 %), but the difference was not statistically significant. In the study of the literature (see Hill et al. 6 ) polyhydramnios was associated with maternal diabetes mellitus in 14.7% to 26.3% of cases, in fetuses with or without congenital malformations. In our series polyhydramnios was taken into consideration only when it was associated with fetal malformations. Our study demonstrates that there is no statistically significant increase in diabetes mellitus among the mothers of infants having polyhydramnios associated with congenital malformations. In our series the diagnosis of congenital malformations was performed prenatally in 41.0% of the cases. In the series of Landy et al.,7 99 cases of pregnancy complicated by hydramnios were studied. Prenatal diagnosis of congenital malformations was performed in only seven cases out of 21 with congenital malformations associated with polyhydramnios. In our patients with chromosomal abnormalities a
September 1991 Am J Obstet Gynecol
congenital malformation was detected prenatally at ultrasonographic examination in only one fetus with trisomy 21 syndrome (duodenal atresia) out of six cases and in five with trisomy 18 syndrome out of 13 cases (28.5% of all cases). In the 15 other cases of chromosomal abnormality no anomaly except polyhydramnios was detected at routine ultrasonographic examination. Benacerraf et aI. 13 were able to detect 80% of fetuses with trisomy 18 at ultrasonographic examination. In our series seven cases of trisomy 18 could have been detected if level II and III ultrasonographic examinations had been performed, but seven fetuses with chromosomal anomalies (4.5%) (three Down syndrome, two trisomy 18, and two autosomal deletions) would not have had ultrasonographic detection because they had no malformations except mild facial dysmorphia and mental retardation. Therefore we recommend the use of level II or III ultrasonographic examination in every pregnancy complicated by polyhydramnios. l:l When no malformation at level III ultrasonographic examination is detected, fetal chromosome analysis has to be discussed. REFERENCES 1. QueenanJT, Gadow Ee. Polyhydramnios: chronic versus acute. AM J OBSTET GYNECOL 1970; 108:349-52. 2. Wallenburg HCS, Wladimoroff VW. The amniotic fluid. II. Polyhydramnios and oligohydramnios. J Perinat Med 1977;5:193-205. 3. Stoll C, Roth MP, Dott B, Bigel P. Etude des malformations congenitales dans Ie Nord de l'Alsace. Med Hyg 1984;42:505-12. 4. Sivit Cj, Hill MC, LarsenJW, Lande 1M. Second trimester polyhydramnios: evaluation with US. Radiology 1987: 165:467-9. 5. SAS/STAT. Guide for personal computers. Version 6 edition. Cary, North Carolina: SAS Institute, 1987:773-85. 6. Hill LM, Breckle R, Thomas ML, Fries JK. Polyhydramnios: ultrasonically detected prevalence and neonatal outcome. Obstet Gynecol 1987;69:21-5. 7. Landy HJ, Isoda NB, LarsenJW. Genetic implications of idiopathic hydramnios. AM J OB5TET GYNECOL 1987; 157: 114-7. 8. Zamah NM, Gillieson MS, Walters JH, et al. Sonographic detection of polyhydramnios: a 5-year experience. AM J OB5TET GYNECOL 1982;143:523-6. 9. Quinlan RW, Cruz AC, Martin M. Hydramnios ultrasound diagnosis and its impact on perinatal management and pregnancy outcome. AM J OB5TET GYNECOL 1983; 145:306-11. 10. Seeds AE. Current concepts of amniotic fluid volume. AM J OBSTET GYNECOL 1980;138:575-86. 11. Hobbins JC, Grannum PA, Berkowitz RL, et al. Ultrasound in the diagnosis of congenital anomalies. AM J OBSTET GYNECOL 1979;134:331-45. 12. Briard ML. Recherche d'anomalies chromosomiques sur signes d'appel. La Depeche 1988; 13:29-46. 13. Benacerraf B, Miller WA, Frigoletto FD. Sonographic detection of fetuses with trisomies 13 and 18; accuracy and limitations. AMJ OB51'£T GYNECOL 1988;158:404-9.
