CURRENT DEVELOPMENTS
The present status of prenatal detection of neural tube defects SHAKUNTALA CHESTER New
York,
CHAUBE, A.
New
SWINYARD,
PH.D. M.D.,
PH.D.
York
In experimentally induced myelocele in rats, efforts to find neural cells in amniotic @id (AF) were unsuccessful. Creatine phosphokinase (CPK) and aldolase concentrations studied in serum of 118 and cerebrospinal fluid (CSF) in 9 patients with myelomeningocele showed serum CPK to be significantly elevated and more responsive to additional muscle injury than aldolase, but both enzymes appeared in lower concentrations in patients w&h myelomeningocele than those with infantile atrophy or cerebral palsy. In CSF, CPK, and aldolase concentrations averaged 4.2 I.U. and 2.7 S.L.U. per milliliter, respectively. Significant CPK elevation (p < 0.001) was also found in AF from myeloschitic fetuses and maternal rat serum. Although these findings suggest that increased CPK concentration is an indicator of myelocele in rats, the technique is impractical for prenatal detection of human neural tube defects (NTD) because muscle denervation in the human fetus occurs too late in gestation. This does not, however, preclude the value of CPK for detecting onset of paraparesis. In all myeloschitic human fetuses, the CSF communicates directly with AF for at least 3 to 4 weeks. This implies that CSF is probably the principal source of increased alphafetoprotein concentration encountered in AF of all pregnancies with NTD. When biological variables are recognized, it is evident that increased concentration of amniotic fluid alpha fetoprotein is a reliable indicator of fetuses with open myelocele and/or anencephalus.
P A R E N T s who have been delivered of one child with myelomeningocele (spina bifida) are at risk (3 to 5 per cent) to deliver a similar defective chi1d.l Those of us concerned with rehabilitation of children with this birth defect are constantly
confronted with the problem of counselling these “at risk” parents and feel keenly the urgency of finding a reliable marker for prenatal detection of myelocele. This problem and need also includes anencephaly and other varieties of serious developmental defects of the neural tube which have their onset near the end of the first month of gestation. At the present time more than 50 different pathologic states can be detected prenatally. An excellent comprehensive review of prenatal detection of disease recently appeared in this JOURNALS but neural tube malformation could only be given brief mention. The objective of this paper is to present clinical and experimental data related to prenatal detection
Supported by grants from the Sam and Rose Mitchell Memorial Fund and the Irwin Strasburger Memorial Medical Foundation, New York, New York, and by the Social and Rehabilitation Service, Department of Health, Education, and Welfare, under the designation of New York Uniuersity as a Rehabilitation Research and Training Center. Reprint requests: Dr. Shakuntala 400 E. 34th St., New York, New
Chaube, York 10016.
429
430
Chaube
and
February 1, 1173 Am. J. Obstct. Gynecol.
Swinyard
of neural tube defect, to review recent advances in this field, and to present embryogenetic data of importance in interpreting the significance of changes in prenatal indicators of neural tube defects. Possible cellular indicators Our initial approach to this problem was based on the concept that a specific neural cell such as an ependymal cell might become separated from the neural plate and be found in the amniotic fluid (AF) of pregnancies resulting in neonatal myelocele. We therefore produced this birth defect in fetal rats by injecting pregnant rats with 50 mg. per kilogram of Trypan blue (1 per cent solution in normal saline) subcutaneously on days 7 and 8 of gestation. Control animals of similar gestational age were injected with equal volumes of normal saline. All animals were autopsied on day 20 of gestation and the AF from individual fetuses was collected and examined for the presence of ependymal cells. We were unable to find any ependymal or other unusual cells free in the AF. Ependymal cells appear to be more tightly fixed to the myelocele than are fetal surface epithelial cells which are found in the AF of all pregnancies. However, Sutherland and Brock” have recently described the presence of fetal macrophages in AF of 2 anencephalic pregnancies in which the diagnosis was based upon raised alphafetoprotein (AFP) levels. They suggested that presence of these cells along with elevated AFP in AF cultures might be of importance in the early detection of severe central nervous system (CNS) anomalies. Their suggestion was confirmed by Nelson, Ruttiman, and Brock” who found similar fetal macrophages in AF of a pregnancy in which the fetal skull, brain, and face were defective, and the AFP level was raised. Change enzymes
in
concentration
of
certain
muscle
The second area of our concern was the possibility that enzymes normally found in highest concentration in striated muscles (glutamic oxaloacetic transaminase [GOT], aldolase, creatine phosphokinase [CPK] ) might be found in the AF during pregnancies with fetuses having myelomeningocele. The logic of this approach was based upon numerous studies which indicate that when striated muscles fragment during dystrophic process5 or become denervated,6 GOT, aldolase, and CPK effuse from the muscles, and the serum concentration of these enzymes increases. The serum concentration of ‘these enzymes has well-known diagnostic,r prognostic, and
geneti? implications in the progressive muscular dystrophies, atrophies, and myocardial infarction. We elected to study CPK and aldolase since there are data indicating that these enzyrnes are also elevated in the cerebrospinal fluid (CSF) of newborn infants and infants with meningoencephalitis,!’ rnyelomeningocele,‘” and hydrocepha1us.l” Serum samples were obtained from 70 patients, and CSF from 9 patients, with myelomeningocele. Th e patients from whom the CSF was obtained required ventricular tap because of hydrocephalus. Creatine phosphokinase and aldolase were also quantitated in the maternal serum and AF of fetal rats which had experimentally produced neural tube defects. Enzyme
quantitation
technique
CPK activity was determined by the method of Nuttal and Wedin’l based on the reaction, adenoCPK sine-5-triphosphate (ATP) + creatine -> adenosine-5-diphosphate (ADP) + creatine phosphate, and the enzyme activity is expressed in International Units (I.U./ml.), i.e., pmoles of pyruvate (equivalent to creatine phosphate) produced per liter of serum per minute at 37’ C.* Aldolase activity was determined by the method of Sibley and Lehningerr? based on the reaction, aldolase fructose- 1-6,diphosphate (FDP) -> dihydroxyacetone phosphate + glyceraldehyde-3-phosphate. The activity is expressed as Sibley-Lehninger units (S.L.U./ml. ) , i.e., the amount of enzyme required to split 1 mm.” of FDP per hour at 37’ C.* Normal values for serum CPK are numerous but the methods used to quantitate the enzyme are varied. Several investigators have used the forward reactionl:3-l!J and some have added cysteine to the reaction mixture with the impression that a higher and more accurate value is obtained by this modification.l”-l7 Our quantitations were done without addition of cysteine. I* Available normal serum aldolase and CPK (without cysteine) activity reported in children are shown in Table I. The normative data show CPK concentration in the newborn infant to be about 10 times higher than the adult. It is remarkably increased over the first 24 hours of life, followed by a decrease thereafter until it reaches approximate adult level by the fifteenth month of postnatal life. Normal serum aldolase activity has a similar circadian pattern; howSt.
“Dade Louis,
Enzyme Missouri.
Assay
Kit,
Sigma
Chemical
Company,
Volume Number
Prenatal
121 3
Table I. Normal
CPK
and aldolase
activity
(IU/ml.) CPK
1
“N”:
i
20
39.4 +_ 4.151
19
10
1 day
12
69.02
7.35
19
30
20.3 2
5.9
19
21
4.1 - 11.4) 3.1 - 12)
14
months
8
Years : o-1
22 (0
l-3
32 (0
1
No.
Birth
l-12
of neural
tube
defects
431
in children
No.
Age
detection
S!i%?~l.
“N”of:
lO.Sf (7.9 - 14.2) 15.3 (8.1 - 25.0) 12.4 (7.9 - 19.0)
16 16 16
14 16
1-5 3-6
46
1-13 6-10
47 43
10-15
47
15-20
9
O-20
3.9 (0 - 15.1) 11.1 f. 1.29 2.6 (0 - 10.7) 2.9 (0 - 13.2) 3.6 (0 - 6.2)
14 14
23
(3.48:2
19 14
39
6.7
16
14
31
6.7
16
14
36
5.5
16
12
12
1 (0
> 20
48 (0
*Mean
+ standard
tMean; $Mean
figures in parentheses not given.
3.0 - 13.2)
9.8)
14
121
18 0.5)
5.5
16
deviation. are the range.
ever, its concentration in children of similar age groups is much less than CPK. The results of our study on serum CPK activity in 65 (33 male, 32 female) and serum aldolase activity in 53 (26 male, 27 female) patients with myelomeningocele, and normal values for these enzymes obtained by other investigators for children of similar agesI I67 18, I9 are shown in Figs. 1 and 2. About 88 per cent of the children had CPK concentrations which were above normal, but it was not elevated in one newborn infant and a 5yZ-week-old infant. Mean CPK value (and range) for age groups 1 to 3, 4 to 10, 11 to 20, and > 20 were 14.8 (6.7 to 36.5), 10.9 (4.0 to 24.5). 11.0 (3.75 to 37.75), and 8.9 (1.5 to 19.0) I.U. per milliliter, respectively. These values are from 3 to 4vZ times higher than the normal values reported for children of similar ages. Aldolase activity, on the other hand, was not affected in any age group. In another study, CPK and adolase activity was quantitated in 33 patients with myelomeningocele ranging in ages from 5 to 20 years, and was compared with the values on patients with infantile atrophy and three clinical types of cerebral palsy
Table II. CPK and aldolase activity in patients with myelomeningocele, infantile atrophy, and cerebral palsy, ranging in ages from 5 to 20 years
I NO. I Normal Myelomeningocele” Amyotonia congenita* Spastic quadriparetic* Athetoid’ Spastic hemiparetic* * = t =
99 36
p < 0.001. Mean + standard
CPK (I.U./ml.) 3.814 (0 - 13.2) 10.8 +_ 7.9t
1 No. I
Aldolase (S.L.U./ ml.)
53
5.31s
36
4.7 2 2.3
8
16.22
3.2
8
7.0 + 1.7
6 6
17.1 + 23.8+
7.6 8.5
6 6
8.1 + 2.5 8.5 + 2.0
6
19.7 f 10.7
6
12.8f
6.1
deviation.
(Table II). The concentration of these enzymes was 2 to 4 times higher in the latter four groups. This was not surprising since the severity and extent of muscle weakness is much greater in the latter conditions. Previous studies have shown that CSF contains only the brain isoenzyme of CPK.20 The concentration of this enzyme increases in a number of neuro-
432
Chaube
and Swinyard
Am.
February J. Obstet.
1, 1Yi.T Gynecol.
”
iii-------
MONTHS Fig. 1. Serum
YEARS AGE CPK activity in patients with myelomeningocele.
logical diseases, especially progressive hydrocephalus and symptomatic epilepsy.21, 22 The clinical significance of CSF aldolase has been discussed elsewhere.23 We examined the CSF of 9 children with myelomeningocele (aged 4 days to 4 years) (Fig. 3) and were unable to detect any CPK or aldolase activity in 2 patients (a 30-day-old and a 2yz-yearold child). In the remaining 7 children (aged 3 months to 4 years), mean and range of CPK and aldolase activity was 4.2 (1.7 to 10.7) I.U. per milhliter, and 2.7 (1.0 to 2.3) S.L.U. per milliliter, respectively. Normal CPK and aldolase values for CSF of children are practically non-existent and those for adults vary greatly.20-22~ 24 Aronsor+ used 33 patients (aged 13 months to 9 years) as controls when studying aldolase activity in patients with amaurotic familial idiocy and Niemann-Pick disease. All of his so-called controls, however, had either cerebral palsy, CNS anomalies, obstructive hydrocephalus, or craniopharyngioma. The marked responsiveness of serum CPK and aldolase to muscle trauma is well known and also occurs in children with spina bifida who already have an increased enzyme concentration. ‘The change in serum concentration incident to a surgical
procedure is illustrated in Fig. 4 which shows the mean concentration in 5 patients with myelomeningocele before and after ureteroileostomy. In our studies of maternal serum and amniotic fluid from rats with experimentally induced myelomeningocele, the concentration of CPK was significantly elevated (p < 0.001) near the end of gestation (autopsy day 20) (Table III). The marked elevation of CPK in amniotic fluid of rats with paralysis incident to experimentally induced myelocele (autopsied on gestation day 20)) and the significant elevation of this enzyme in serum of patients with myelomeningocele suggested that concentration of this enzyme might be a reliable marker for neural tube defect. We previously presented electromyographic evidence of denervation in paretic limbs of infants and children with myelocele.25 However, detailed maternal accounts of fetal movement in pregnancies which resulted in myelocele indicated that among 100 mothers, three fourths of them noted a marked change in the intensity of fetal movement during the sixth to eighth month of pregnancy as compared with other pregnancies resulting in normal neonates. This circumstantial evidence suggested that the paresis in fetuses
Volume Number
121 3
Prenatal
detection
of
neural
tube
defects
433
A Male
A Female
,.,-NORMAL (Ret161
MON:HS
YE,&AGE Fig.
2. Serum
aldolase
activity
with myelocele occurs near the beginning of the third trimester. The foregoing data led us to study the development of the sciatic nerve in human fetuses with lumbosacral myelocele. Serial sections of a number of myeloschitic human fetusesZG indicated that neuroblasts develop within the myelocele and their axons reach the anlage of their respective lower limb muscles (Fig. 5). It is therefore apparent that the muscle denervation which accounts for elevated serum CPK levels in children with myelocele occurs too late in pregnancy to make practical use of elevated CPK levels in amniotic fluid as a marker for prenatal detection of human myelocele. The relatively infrequent occurrence of arthrogryposis in neonates with myelocele supports these general conclusions. It is quite possible, however, that such studies in the third trimester would indicate more precisely when the paresis occurs and enable us to learn the etiologic factors related to the muscle weakness which confronts the infants and children. Increase
in proteins
in amniotic
fluid
During the past two years, attention has been directed toward the possibility of using changes in AF concentration of certain proteins as possible indicators of a fetus with myelomeningocele. In 1944, Pederson”’ described a globulin in calf serum which was specific for the fetus, and 12 years later Bergstrand and CzarZs identified this protein in human fetal serum and referred to it as AFP. Its presence has since been confirmed in both fen@ and adult3” human serum. A protein which is immunologically identical to human AFP has also been found in patients with hepatocellular cancer and embryonal teratoma31 and in the adult and fetal sera of several animal species.31-34
in patients
with
myelomeningocele.
. : MALE
0 mdCPK . =MALE 0 : FEMALEWOLASE
0
I
vu
36
4
'
48
lo\
AGE IN MONTHS Fig. from
3. CPK and hydrocephalic
aldolase patients
activity with
in cerebrospinal myelomeningocele.
fluid
Pertinent data with reference to AFP concentrations in maternal and fetal sera and AF in normal human pregnancies and those with neural tube defect are shown in Fig. 6. 35-51 During normal pregnancy AFP is detectable in fetal serum at 6 weeks of gestation (66.5 pg per milliliter) ; it increases in concentration more than fortyfold by the twelfth to fourteenth week (2,800 ,ug per milliliter) and diminishes thereafter (18 pg per milliliter at thirtyfifth week) .35 A similar pattern is seen in maternal serum.36 If one relates AFP concentrations of the maternal to fetal serum at similar gestational stages, the ratios at the midpoint of the first, second, and third trimesters are approximately 1: 1,466, 1: 7,700, and 1: 366, respectively. By applying a sensitive radioimmunoassay technique, SJ Brock, Bolton, and Monaghan”l were able to detect anencephaly in which the first indication of abnormality was increased
434
Chaube
and
Swinyard Am.
Table III. and Trypan fetuses
CPK activity blue-induced
in serum and AF of normal myelomeningocele in rat
Compound
Treated (Trypan blue) (Pregnant rats were given 50 mg. per kilogram of trypan blue subcutaneously on days 7 and 8, and autopsied on day 20 of gestation) Control (saline)
-------SERUM --SERUM
16 14
I
Maternal serum CPK (I.lJ./ml.) 128.6f
5.1* (5)t
106.6 + 10.6
(6) *Mean * standard deviation. tFigures in parentheses indicate analyzed.
number
I
\ *.._
21.82 11.4 (10)
of
CPK ALDOLASE
A--t
Amniotic fluid CPK (I.U./ml.)
9.8+ (11)
February 1, 1975 J. Obstet. Gynecol.
fi'cDtGE
NORMAL
---.__* -'.,
3.2
samples
concentration of maternal serum AFP in the sixteenth week of gestation. This is an important step since theoretically it raises the possibility of screening all mothers to detect those whose serum AFP is elevated and provide them with amniocentesis and AF study for confirmation of a possible neural tube defect. In AF, AFP is present at the 6th week of gestation (1.5 pg per milliliter) 57 and although its concentration is much less (e.g., at twelfth to fifteenth weeks, its concentration is one-fiftieth to one-onehundredth that of fetal serum), its circadian pattern is essentially similar to that seen for fetal semm 37-39, 41 In 1972, Brock and Sutcliffess reported that AFP concentration was higher in AF in pregnancies resulting in myeloceles or in a more serious neural tube defect-anencephaly. Their observation was a retrospective one in which the study was made on AF in which the outcome of the pregnancy was already known. The publication provoked a small flood of prospective studies. in man38l 40-JQ which have been published during the last 18 months. Despite variation in the actual amounts of AFP present in the AF reported by different authors, they all demonstrate that AFP concentration between the fourteenth to twenty-second week of gestation is significantly higher in pregnancies resulting in neural tube defects than in normal pregnancies of equivalent gestational ages. However, there appears to be no significant difference in fetal serum AFP between normal fetuses and those with neural tube defects. This relationship suggests that there might be a corresponding increase in CSF AFP since virtually all reports indicate a much higher AFP concentration
4. Serum CPK and aldolase levels in patients with myelomeningocele before and after iliopsoas-transfer operation; P = Preoperative; 0 = time of operation.
in AF in those neural tube defects in which there is a continuing communication between the AF and CSF compartments. We have had an opportunity to study many human embryos with myeloceles” and the following embryogenetic considerations are pertinent to the problem. When a lumbosacral myelocele develops about the thirtieth day of gestation, there is always direct communication between the CSF and AF compartments. The choroid plexus invaginates the roof of the ventricle about the 36th day and its secretory function begins the thirty-seventh to fortieth day. The dura does not encircle the neural tube until the ninth to the tenth week and does not achieve a significant degree of compactness until the twelfth week. Therefore, when the myelocele occurs, there is undoubtedly free communication between the AF and the CSF in all neural tube myeloceles for at least 3 to 4 weeks. Meningeal closure and/or skin covering of the lesion when seen at birth would only mean that duration of communication was relatively more limited. Nothing is known about the rate of AFP metabolism in the fetus, placenta, or mother, and the exchange in the AF between these three sites. Thus the relatively few instances where normal AFP concentrations were recorded in Al? from pregnancies with neural tube defects clearly illustrates one of the difficulties in using this marker late in pregnancy (Fig. 6) .38-39 The specificity of AFP as a marker is further reduced by evidence that it is elevated in maternal serum and AF in instances of intrauterine fetal *Dr. University
H.
Nishimura, Professor of Anatomy Medical School, made the specimens
in Kyoto available.
Volume Number
121 3
Prenatal
detection
of neural
tube
defects
435
Fig. 5. A photomicrograph of a cross section through a human embryo with a lumbosacral myelocele. The boxed area is enlarged to show the sciatic nerve (arrow) innervating the developing muscles (C.-R. length, 17.9 mm.; C. A., 50 days).
death,Jz fetal anomalies,3Fp 42* 53 occurrence of prior abortions4 in first pregnancies as compared to subsequent ones,55 and in twin versus single conceptions5ti, 67 (Fig. 7). All these possibilities must thus be considered before conclusions can be drawn regarding the significance of elevated serum levels of AFP as a possible indicator of fetal neural tube anomalies
The
in utero.
lack of AFP
specificity
led some investiga-
tors
to search
for
a more
specific
protein
be found in high concentrations in the the aid of immunoelectrophoresis and against normal human CSF, Clausen strated the presence of beta trace protein CSF. Clausen’s findings were confirmed investigators.“9-“1 Its mean concentration healthy persons is 2.6 t 0.8 mg. per 100 trace protein has also been demonstrated
that would CSF. With antiserum demonin human by other in CSF of ml.“l Beta in human
436
Chaube
and
Swinyard Anl.
February I, lY7.i J. Obstet. Gynccol.
0,135), +,1361. D .137), I J381, .,(39) ( r,(40), nJ41). c,(42), 9,(43), +,t441, o .(45), 0,1461, e-147 ard 49). a,(48), l ,151)
NORMAL FETAL SERUM 1351
-WMAl.
AMNIOTIC FLUID (47ond491
.” --NORMAL
.p
AMNIOTIC FLUID (361
F NOWAL CMNKlTlC FLUID (391
IJr
NORMAL AMNICTIC FLUID (371 NORMAL MATERNAL SERUM (36)
01
’ 5
’ IO
: I5
t m
: 25
: 30
I 35
! 40
1 45
GES’TA?XWALAGF /IV WEiiKS Fig. 6. Alpha-fetoprotein (AFP) in maternal and fetal sera and.amniotic fluid of normal pregnancies and those with neural tube defects. N, normal; A, anencephalus; S, spina bifida (myelomeningocele) ; T, neural tube defect. When these scripts appear alone, they represent values for amniotic fluid. When the scripts are followed by the letter M or F, the values represent those of maternal serum (NM, AM, SM) or fetal serum (NF, AF). Normal and elevated AFP values taken from the same reference are designated by the same symbol.
Volume 121 Number 3
Prenatal
-WIN
NORMAL-
GESTAUONAL AGE IN WEEKS
Fig. 7. Concentration of a-fetoprotein in maternal serum at different gestational ages in normal single, twin, primiparous, and multiparous pregnancies. (Based on data of Ishigura and Nishimura.sjv 56. 57)
pleural and ascitic fluids,59 urine,59p 62 and serum.61 It has been reported absent in AF from normal human6? and rate” pregnancies but present in pregnancies with anencephalye3 and experimentally produced myelomeningocele.G4 We believe that additional studies of this type are urgently needed. Further observations on the concentration of CPK and aldolase in AF might enable us to learn more about abnormal fetal morphogenesis especially with regard to the time of onset of prenatal paraparesis while evaluation of AFP and beta trace protein by the recently developed counter-
detection
of neural
tube
defects
437
electrophoretic (rocket) techniques which improve their accuracy and sensitivity add to their importance as diagnostic techniques for prenatal detection of myelomeningocele and anencephaly. Preoccupation with cellular and/or biochemical indicators of neural tube defects found in AF obtained by amniocentesis should not cause us to overlook the assistance available from roentgenography,6” amniography,66 fetography,6’ ultrasound fetoscopy,68 or direct visualization of the fetus by fiberoptic microscopy.69 Proper selection of techniques now available makes it possible to detect prenatally a high percentage of the fetuses with anencephaly and myelomeningocele. Addendum Since this manuscript was submitted, additional evidence supporting the validity of using AFP levels in maternal serum70, T1 and AF72-75 have appeared. Ward and StewartT6 have pointed out that, since AFP concentration in fetal blood is in the milligrams per milliliter range and in micrograms per milliliter levels in amniotic fluid, hemorrhage of only 1 ml. of fetal blood into amniotic fluid will increase the AFP level in AF above the norm at 14 weeks. They presented two such cases and recommended that possibility of this occurrence be detected by screening AF for fetal hemoglobin by the Klechauer technique. An editorial” has emphasized the need for refinement of technique, additional data from normal “high-risk” mothers, accurate detection of ideal gestational age for detection, and additional data on plural gestation. In a retrospective study, Macri and associates78 confirmed the reliability of beta trace protein (BTP) as a marker in 15 cases of neural tube defect; however, Olsson and colleagues79 emphasized variability of BTP concentration and recommended use of AFP as a marker. The technical skills of Misses Patricia Joan M. Koechel are acknowledged.
A. Muck
and
REFERENCES
1. Carter, C. O., David, P. A., and Laurence, K. M.: J. Med. Genet. 5: 81, 1968. 2. Burton, B. K., Gerbie, A. B., and Nadler, H. L.: AM. J. 0asTa.r. GYNECOL. 118: 718, 1974. 3. Sutherland, G. R., and Brock, D. J. H.: Lancet 2: 1098, 1973. 4. Nelson, M. M., Ruttiman, M. T., and Brock, D. J. H.: Lancet 1: 504, 1974. 5. Schapira, G., and Dreyfus, J.-C.: Biochemistry of progressive muscular dystrophy, in Bourne, G. H., and Golarz, M. N., editors: Muscular Dystrophy in Man and Animals, New York, 1963, Hafner Publishing Company, p. 48.
6. Kouffen, H., and Consbrugh, U.: Nervenarzt. 41: 599, 1970. 7. Heyck, H., and Laudahn, G.: Muscle serum enzymes in muscular dystrophy and neurogenic muscular atrophy, A comparative study, in Milhorat, A. T., editor: Exploratory Concepts in Muscular Dystrophy and Related Disorders, Amsterdam, 1965, Excerpta Medica Foundation, p. 232. 8. Milhorat, A. T., Shafiq, S. A., and Goldstone, L.: Ann. N. Y. Acad. Sci. 138: 246, 1966. 9. Lui, W. Y.: Dev. Med. Child Neurol. 14: 467, 1972. 10. Drummond, M. B., and Belton, N. R.: Arch, Dis. Child. 47: 672, 1972.
438
11. 12. 13. 14.
15. 16. 17.
Chaube
and
Febluary
Swinyard
Nuttall, F. Q., and Wedin, D. S.: J. Lab. Clin. Med. 68: 324, 1966. Sibley, J. A., and Lehninger, A. L.: J. Biol. Chem. 177: 859, 1949. Pearce, J. M. S., Pennington, R. J., and Walton, J. N.: J. Neurol. Neurosurg. Psychiat. 27: 1, 1964. Okinaka, S., Sugita, H., Momoi, H., Toyokura, Y., Watanabe, T., Ebashi, F., and Ebashi, S.: J. Lab. Clin. Med. 64: 299. 1964. Bodensteiner, J. B.: and Zellweger, H.: J. Lab. Clin. Med. 77: 853, 1971. Zellweger, H., and Hanson, J. Med. _ W.: Excerpta 186: 53, 1969. Okinaka, S., Kumagai, M., Ebashi, S., Sugita, H., Momoi. * , Y.: Arch. Neural.’ 4:H.. 64, Tovokura. 1961. ’ Y.. ’ and Fuiie.
Am. J.
44. 45. 46. 47. 48. 49. 50. 51. 52.
18. 19.
20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 3 1. 32. 33. 34. 35. 36. 37. 38.
39. 40. 4.1. 42. 43.
Bray, G. M., and Ferrendelli, J. A.: Neurology 18: 480, 1968. Garrido Malo, M. A., de la Torre, C. T., Deutsch, S., and Falcone, A.: Bull. Sot. Chim. Bidl. (Paris) 52: 1467. 1970. Sherwin,’ A. L., Norris, J. W., and Bulche, J. A.: Neurology 19: 993, 1969. Herschkowitz, N., and Cumings, J. N.: J. Neurol. Neurosurg. Psychiat. 27: 247, 1964. Katz, R. M., and Liebman, W.: Am. J. Dis. Child. 120: 543, 1970. Aronson, S. M., Saifer, A., Perle, G., and Volk, B. W.: Proc. Sot. Exp. Biol. Med. 97: 33 1, 1958. Lisak, R. P., and Craig, F. A: J. A. M. A. 199: 160, 1967. Chantraine, A., Lloyd, K., and Swinyard, C. A.: Dev. Med. Child Neurol. 6: 7, 1964. Swinyard, C. A., Chaube, S., and Nishimura, H.: Pediatr. Ann. 2: 31, 1974. Pederson, K. 0.: Nature 154: 578, 1944. Bergstrand, C. G., and Czar, B.: Stand. J. Clin. Lab. Invest. 8: 174, 1956. Tatarinov, Y. S.: Fed. Proc. (Transl. Suppl.) 24: T916, 1965. Ruoslahti, E., and Seppall, M.: Nature 235: 161, 1972. Tatarinov, Y. S.: Vop. med. Khim. 11: 20, 1965. Oakes, D. A., Shuster, J., and Gold, P.: Cancer Res. 32: 2753, 1972. Pihko, H., and Ruoslahti, E.: Int. J. Cancer 12: 354, 1973. Gitlin, D., and Boesman, M.: Comp. Biochem. Physiol. 21: 327, 1967. Gitlin, D., and Boesman, M.: J. Ciln. Invest. 45: 1826, 1966. Seppali, M., and Ruoslahti, E.: AM. J. OBSTET. GYNECOL. 112: 208, 1972. Sepplla, M., and Ruoslahti, E.: AM. J. OBSTET. GYNECOL. 114: 595, 1973. Allen, L. D., Ferguson-Smith, M. A., Donald, I., Sweet, E. M., and Gibson, A. A. M.: Lancet 2: 522, 1973. Brock, D. J. H., and Sutcliffe, R. G.: Lancet 2: 197, 1972. Brock, D. J. H., and Scrimgeour, J. B.: Lancet 2: 1252, 1973. Field, B., Mitchell, G., Garrett, W., and Kerr, C.: Lancet 2: 798, 1973. Nevin, N. C., Nesbitt, S., and Thompson, W.: Lancet 1: 1382, 1973. Lorber, J., Stewart, C. R., and Milford Ward, A.: Lancet 2: 1187, 1973.
53.
54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79.
Obstet.
1 1975 Gynecol.
Sepplll, M., and Ruoslahti, E.: Lancet 1: 155, 1973. Seller, M. J., Coltart, T. M., Campbell, S., and Singer, J. D.: Lancet 2: 73, 1973. Leek, A. E., Ruoss, C. F., Kitau, M. J., and Chard, T.: Lancet 2: 385, 1973. Randle, G. H., and Cumberbatch, R. N.: J. Obstet. Gynaecol. Br. Commonw. 80: 1054, 1973. Harris, R., Jennison, R. F., Laurence, K. M., Ruoslahti, E., and Seppala, M.: Lancet 1: 429, 1974. Seller, M. J., Singer, J. D., Coltart, T. M., and Campbell, S.: Lancet 1: 428, 1974. Ruoslahti, E., and SeppllH, M.: Int. J. Cancer 8: 374, 1971. Brock, D. J. H., Bolton, A. E., and Monaghan, J. M.: Lancet 2: 923, 1973. Seppall, M., and Ruoslahti, E.: AM. J. OBSTET. GYNECOL. 115: 48, 1973. Laurence, K. M., Turnbill, A. C., Harris, R., Jennison, R. F., Ruoslahti, E., and SeppHlL, M.: Lancet 2: 860, 1973. SeppalP, M., and Ruoslahti, E.: Br. Med. J. 4: 769, 1972. Ishiguro, T.: Lancet 1: 1509, 1973. Ishiguro, T.: Lancet 2: 1214, 1973. Ishiguro, T., and Nishimura, T.: AM. J. OBSTET. GYNECOL. 116: 27, 1973. Clausen, J.: Proc. Sot. Exp. Biol. Med. 107: 170, 1961. Hochwald, G. M., and Thorbecke, G. J.: Proc. Sot. Exp. Biol. Med. 109: 91, 1962. Pepe, A. J., and Hochwald, G. M.: Proc. Sot. Exp. Biol. Med. 126: 630, 1967. Link, H.: Acta Neural. Stand. 43 (Suppl. 28): 7, 1967. Ericsson, J., Link, H., and Zettervall, 0.: Neurology 19: 606, 1969. Macri, J. N., Joshi, M. S., Weiss, R. R., and Evans, M. I.: Lancet 1: 14, 1974. Macri, J. N., Joshi, M. S., and Evans, M. I.: Nature (New Biol.) 246: 89, 1973. Russell, G. G. B.: J. Obstet. Gynaecol. Br. Commonw. 76: 345, 1969. Queenan, J. T., and Gadow, E. C.: Obstet. Gynecol. 35: 648, 1970. Aqiiero, O., and Zighelboim, T.: Surg. Gynecol. Obstet. 130: 649, 1970. Thompson, H. E.: Obstet. Gynecol. Surv. 23: 903, 1968. Takaaki, M., Chie, M., and Fujiko, Y.: J. Jap. Obstet. Gynecol. 15: 87, 1968. Wald, N. J., Brock, D. J. H., and Bonnar, J.: Lancet 1: 765, 1974. Brock,.D. J. H., Bolton, A. E., and Scrimgeour, J. B.: Lancet 1: 767, 1974. Milunsky, A., Alpert, E., and Charles, D.: Obstet. Gynecol. 43: 592, 1974. Seller, M. J., Greasy, M. R., and Alberman, E. D.: Br. Med. J. 2: 524, 1974. Emery, A. E. H., Brock, D. J. H., and Eccleston, D.: J. Obstet. Gynaecol. Br. Commonw. 81: 512, 1974. Medical News: J. A. M. A. 229: 1706, 1974. Ward, A. M., and Stewart, C. R.: Lancet 2: 345, 1974. Editorial: Lancet 1: 907, 1974” Macri, J. N., Weiss, R. R., and Joshi, M. S.: Lancet 1: 1109, 1974. Olsson, J. E., Sherman, M. S., and Kjessler, B.: Lancet 2: 347, 1974.
The present status of prenatal detection of neural tube defects SHAKUNTALA CHESTER New
York,
CHAUBE, A.
New
SWINYARD,
PH.D. M.D.,
PH.D.
York
In experimentally induced myelocele in rats, efforts to find neural cells in amniotic @id (AF) were unsuccessful. Creatine phosphokinase (CPK) and aldolase concentrations studied in serum of 118 and cerebrospinal fluid (CSF) in 9 patients with myelomeningocele showed serum CPK to be significantly elevated and more responsive to additional muscle injury than aldolase, but both enzymes appeared in lower concentrations in patients w&h myelomeningocele than those with infantile atrophy or cerebral palsy. In CSF, CPK, and aldolase concentrations averaged 4.2 I.U. and 2.7 S.L.U. per milliliter, respectively. Significant CPK elevation (p < 0.001) was also found in AF from myeloschitic fetuses and maternal rat serum. Although these findings suggest that increased CPK concentration is an indicator of myelocele in rats, the technique is impractical for prenatal detection of human neural tube defects (NTD) because muscle denervation in the human fetus occurs too late in gestation. This does not, however, preclude the value of CPK for detecting onset of paraparesis. In all myeloschitic human fetuses, the CSF communicates directly with AF for at least 3 to 4 weeks. This implies that CSF is probably the principal source of increased alphafetoprotein concentration encountered in AF of all pregnancies with NTD. When biological variables are recognized, it is evident that increased concentration of amniotic fluid alpha fetoprotein is a reliable indicator of fetuses with open myelocele and/or anencephalus.
P A R E N T s who have been delivered of one child with myelomeningocele (spina bifida) are at risk (3 to 5 per cent) to deliver a similar defective chi1d.l Those of us concerned with rehabilitation of children with this birth defect are constantly
confronted with the problem of counselling these “at risk” parents and feel keenly the urgency of finding a reliable marker for prenatal detection of myelocele. This problem and need also includes anencephaly and other varieties of serious developmental defects of the neural tube which have their onset near the end of the first month of gestation. At the present time more than 50 different pathologic states can be detected prenatally. An excellent comprehensive review of prenatal detection of disease recently appeared in this JOURNALS but neural tube malformation could only be given brief mention. The objective of this paper is to present clinical and experimental data related to prenatal detection
Supported by grants from the Sam and Rose Mitchell Memorial Fund and the Irwin Strasburger Memorial Medical Foundation, New York, New York, and by the Social and Rehabilitation Service, Department of Health, Education, and Welfare, under the designation of New York Uniuersity as a Rehabilitation Research and Training Center. Reprint requests: Dr. Shakuntala 400 E. 34th St., New York, New
Chaube, York 10016.
429
430
Chaube
and
February 1, 1173 Am. J. Obstct. Gynecol.
Swinyard
of neural tube defect, to review recent advances in this field, and to present embryogenetic data of importance in interpreting the significance of changes in prenatal indicators of neural tube defects. Possible cellular indicators Our initial approach to this problem was based on the concept that a specific neural cell such as an ependymal cell might become separated from the neural plate and be found in the amniotic fluid (AF) of pregnancies resulting in neonatal myelocele. We therefore produced this birth defect in fetal rats by injecting pregnant rats with 50 mg. per kilogram of Trypan blue (1 per cent solution in normal saline) subcutaneously on days 7 and 8 of gestation. Control animals of similar gestational age were injected with equal volumes of normal saline. All animals were autopsied on day 20 of gestation and the AF from individual fetuses was collected and examined for the presence of ependymal cells. We were unable to find any ependymal or other unusual cells free in the AF. Ependymal cells appear to be more tightly fixed to the myelocele than are fetal surface epithelial cells which are found in the AF of all pregnancies. However, Sutherland and Brock” have recently described the presence of fetal macrophages in AF of 2 anencephalic pregnancies in which the diagnosis was based upon raised alphafetoprotein (AFP) levels. They suggested that presence of these cells along with elevated AFP in AF cultures might be of importance in the early detection of severe central nervous system (CNS) anomalies. Their suggestion was confirmed by Nelson, Ruttiman, and Brock” who found similar fetal macrophages in AF of a pregnancy in which the fetal skull, brain, and face were defective, and the AFP level was raised. Change enzymes
in
concentration
of
certain
muscle
The second area of our concern was the possibility that enzymes normally found in highest concentration in striated muscles (glutamic oxaloacetic transaminase [GOT], aldolase, creatine phosphokinase [CPK] ) might be found in the AF during pregnancies with fetuses having myelomeningocele. The logic of this approach was based upon numerous studies which indicate that when striated muscles fragment during dystrophic process5 or become denervated,6 GOT, aldolase, and CPK effuse from the muscles, and the serum concentration of these enzymes increases. The serum concentration of ‘these enzymes has well-known diagnostic,r prognostic, and
geneti? implications in the progressive muscular dystrophies, atrophies, and myocardial infarction. We elected to study CPK and aldolase since there are data indicating that these enzyrnes are also elevated in the cerebrospinal fluid (CSF) of newborn infants and infants with meningoencephalitis,!’ rnyelomeningocele,‘” and hydrocepha1us.l” Serum samples were obtained from 70 patients, and CSF from 9 patients, with myelomeningocele. Th e patients from whom the CSF was obtained required ventricular tap because of hydrocephalus. Creatine phosphokinase and aldolase were also quantitated in the maternal serum and AF of fetal rats which had experimentally produced neural tube defects. Enzyme
quantitation
technique
CPK activity was determined by the method of Nuttal and Wedin’l based on the reaction, adenoCPK sine-5-triphosphate (ATP) + creatine -> adenosine-5-diphosphate (ADP) + creatine phosphate, and the enzyme activity is expressed in International Units (I.U./ml.), i.e., pmoles of pyruvate (equivalent to creatine phosphate) produced per liter of serum per minute at 37’ C.* Aldolase activity was determined by the method of Sibley and Lehningerr? based on the reaction, aldolase fructose- 1-6,diphosphate (FDP) -> dihydroxyacetone phosphate + glyceraldehyde-3-phosphate. The activity is expressed as Sibley-Lehninger units (S.L.U./ml. ) , i.e., the amount of enzyme required to split 1 mm.” of FDP per hour at 37’ C.* Normal values for serum CPK are numerous but the methods used to quantitate the enzyme are varied. Several investigators have used the forward reactionl:3-l!J and some have added cysteine to the reaction mixture with the impression that a higher and more accurate value is obtained by this modification.l”-l7 Our quantitations were done without addition of cysteine. I* Available normal serum aldolase and CPK (without cysteine) activity reported in children are shown in Table I. The normative data show CPK concentration in the newborn infant to be about 10 times higher than the adult. It is remarkably increased over the first 24 hours of life, followed by a decrease thereafter until it reaches approximate adult level by the fifteenth month of postnatal life. Normal serum aldolase activity has a similar circadian pattern; howSt.
“Dade Louis,
Enzyme Missouri.
Assay
Kit,
Sigma
Chemical
Company,
Volume Number
Prenatal
121 3
Table I. Normal
CPK
and aldolase
activity
(IU/ml.) CPK
1
“N”:
i
20
39.4 +_ 4.151
19
10
1 day
12
69.02
7.35
19
30
20.3 2
5.9
19
21
4.1 - 11.4) 3.1 - 12)
14
months
8
Years : o-1
22 (0
l-3
32 (0
1
No.
Birth
l-12
of neural
tube
defects
431
in children
No.
Age
detection
S!i%?~l.
“N”of:
lO.Sf (7.9 - 14.2) 15.3 (8.1 - 25.0) 12.4 (7.9 - 19.0)
16 16 16
14 16
1-5 3-6
46
1-13 6-10
47 43
10-15
47
15-20
9
O-20
3.9 (0 - 15.1) 11.1 f. 1.29 2.6 (0 - 10.7) 2.9 (0 - 13.2) 3.6 (0 - 6.2)
14 14
23
(3.48:2
19 14
39
6.7
16
14
31
6.7
16
14
36
5.5
16
12
12
1 (0
> 20
48 (0
*Mean
+ standard
tMean; $Mean
figures in parentheses not given.
3.0 - 13.2)
9.8)
14
121
18 0.5)
5.5
16
deviation. are the range.
ever, its concentration in children of similar age groups is much less than CPK. The results of our study on serum CPK activity in 65 (33 male, 32 female) and serum aldolase activity in 53 (26 male, 27 female) patients with myelomeningocele, and normal values for these enzymes obtained by other investigators for children of similar agesI I67 18, I9 are shown in Figs. 1 and 2. About 88 per cent of the children had CPK concentrations which were above normal, but it was not elevated in one newborn infant and a 5yZ-week-old infant. Mean CPK value (and range) for age groups 1 to 3, 4 to 10, 11 to 20, and > 20 were 14.8 (6.7 to 36.5), 10.9 (4.0 to 24.5). 11.0 (3.75 to 37.75), and 8.9 (1.5 to 19.0) I.U. per milliliter, respectively. These values are from 3 to 4vZ times higher than the normal values reported for children of similar ages. Aldolase activity, on the other hand, was not affected in any age group. In another study, CPK and adolase activity was quantitated in 33 patients with myelomeningocele ranging in ages from 5 to 20 years, and was compared with the values on patients with infantile atrophy and three clinical types of cerebral palsy
Table II. CPK and aldolase activity in patients with myelomeningocele, infantile atrophy, and cerebral palsy, ranging in ages from 5 to 20 years
I NO. I Normal Myelomeningocele” Amyotonia congenita* Spastic quadriparetic* Athetoid’ Spastic hemiparetic* * = t =
99 36
p < 0.001. Mean + standard
CPK (I.U./ml.) 3.814 (0 - 13.2) 10.8 +_ 7.9t
1 No. I
Aldolase (S.L.U./ ml.)
53
5.31s
36
4.7 2 2.3
8
16.22
3.2
8
7.0 + 1.7
6 6
17.1 + 23.8+
7.6 8.5
6 6
8.1 + 2.5 8.5 + 2.0
6
19.7 f 10.7
6
12.8f
6.1
deviation.
(Table II). The concentration of these enzymes was 2 to 4 times higher in the latter four groups. This was not surprising since the severity and extent of muscle weakness is much greater in the latter conditions. Previous studies have shown that CSF contains only the brain isoenzyme of CPK.20 The concentration of this enzyme increases in a number of neuro-
432
Chaube
and Swinyard
Am.
February J. Obstet.
1, 1Yi.T Gynecol.
”
iii-------
MONTHS Fig. 1. Serum
YEARS AGE CPK activity in patients with myelomeningocele.
logical diseases, especially progressive hydrocephalus and symptomatic epilepsy.21, 22 The clinical significance of CSF aldolase has been discussed elsewhere.23 We examined the CSF of 9 children with myelomeningocele (aged 4 days to 4 years) (Fig. 3) and were unable to detect any CPK or aldolase activity in 2 patients (a 30-day-old and a 2yz-yearold child). In the remaining 7 children (aged 3 months to 4 years), mean and range of CPK and aldolase activity was 4.2 (1.7 to 10.7) I.U. per milhliter, and 2.7 (1.0 to 2.3) S.L.U. per milliliter, respectively. Normal CPK and aldolase values for CSF of children are practically non-existent and those for adults vary greatly.20-22~ 24 Aronsor+ used 33 patients (aged 13 months to 9 years) as controls when studying aldolase activity in patients with amaurotic familial idiocy and Niemann-Pick disease. All of his so-called controls, however, had either cerebral palsy, CNS anomalies, obstructive hydrocephalus, or craniopharyngioma. The marked responsiveness of serum CPK and aldolase to muscle trauma is well known and also occurs in children with spina bifida who already have an increased enzyme concentration. ‘The change in serum concentration incident to a surgical
procedure is illustrated in Fig. 4 which shows the mean concentration in 5 patients with myelomeningocele before and after ureteroileostomy. In our studies of maternal serum and amniotic fluid from rats with experimentally induced myelomeningocele, the concentration of CPK was significantly elevated (p < 0.001) near the end of gestation (autopsy day 20) (Table III). The marked elevation of CPK in amniotic fluid of rats with paralysis incident to experimentally induced myelocele (autopsied on gestation day 20)) and the significant elevation of this enzyme in serum of patients with myelomeningocele suggested that concentration of this enzyme might be a reliable marker for neural tube defect. We previously presented electromyographic evidence of denervation in paretic limbs of infants and children with myelocele.25 However, detailed maternal accounts of fetal movement in pregnancies which resulted in myelocele indicated that among 100 mothers, three fourths of them noted a marked change in the intensity of fetal movement during the sixth to eighth month of pregnancy as compared with other pregnancies resulting in normal neonates. This circumstantial evidence suggested that the paresis in fetuses
Volume Number
121 3
Prenatal
detection
of
neural
tube
defects
433
A Male
A Female
,.,-NORMAL (Ret161
MON:HS
YE,&AGE Fig.
2. Serum
aldolase
activity
with myelocele occurs near the beginning of the third trimester. The foregoing data led us to study the development of the sciatic nerve in human fetuses with lumbosacral myelocele. Serial sections of a number of myeloschitic human fetusesZG indicated that neuroblasts develop within the myelocele and their axons reach the anlage of their respective lower limb muscles (Fig. 5). It is therefore apparent that the muscle denervation which accounts for elevated serum CPK levels in children with myelocele occurs too late in pregnancy to make practical use of elevated CPK levels in amniotic fluid as a marker for prenatal detection of human myelocele. The relatively infrequent occurrence of arthrogryposis in neonates with myelocele supports these general conclusions. It is quite possible, however, that such studies in the third trimester would indicate more precisely when the paresis occurs and enable us to learn the etiologic factors related to the muscle weakness which confronts the infants and children. Increase
in proteins
in amniotic
fluid
During the past two years, attention has been directed toward the possibility of using changes in AF concentration of certain proteins as possible indicators of a fetus with myelomeningocele. In 1944, Pederson”’ described a globulin in calf serum which was specific for the fetus, and 12 years later Bergstrand and CzarZs identified this protein in human fetal serum and referred to it as AFP. Its presence has since been confirmed in both fen@ and adult3” human serum. A protein which is immunologically identical to human AFP has also been found in patients with hepatocellular cancer and embryonal teratoma31 and in the adult and fetal sera of several animal species.31-34
in patients
with
myelomeningocele.
. : MALE
0 mdCPK . =MALE 0 : FEMALEWOLASE
0
I
vu
36
4
'
48
lo\
AGE IN MONTHS Fig. from
3. CPK and hydrocephalic
aldolase patients
activity with
in cerebrospinal myelomeningocele.
fluid
Pertinent data with reference to AFP concentrations in maternal and fetal sera and AF in normal human pregnancies and those with neural tube defect are shown in Fig. 6. 35-51 During normal pregnancy AFP is detectable in fetal serum at 6 weeks of gestation (66.5 pg per milliliter) ; it increases in concentration more than fortyfold by the twelfth to fourteenth week (2,800 ,ug per milliliter) and diminishes thereafter (18 pg per milliliter at thirtyfifth week) .35 A similar pattern is seen in maternal serum.36 If one relates AFP concentrations of the maternal to fetal serum at similar gestational stages, the ratios at the midpoint of the first, second, and third trimesters are approximately 1: 1,466, 1: 7,700, and 1: 366, respectively. By applying a sensitive radioimmunoassay technique, SJ Brock, Bolton, and Monaghan”l were able to detect anencephaly in which the first indication of abnormality was increased
434
Chaube
and
Swinyard Am.
Table III. and Trypan fetuses
CPK activity blue-induced
in serum and AF of normal myelomeningocele in rat
Compound
Treated (Trypan blue) (Pregnant rats were given 50 mg. per kilogram of trypan blue subcutaneously on days 7 and 8, and autopsied on day 20 of gestation) Control (saline)
-------SERUM --SERUM
16 14
I
Maternal serum CPK (I.lJ./ml.) 128.6f
5.1* (5)t
106.6 + 10.6
(6) *Mean * standard deviation. tFigures in parentheses indicate analyzed.
number
I
\ *.._
21.82 11.4 (10)
of
CPK ALDOLASE
A--t
Amniotic fluid CPK (I.U./ml.)
9.8+ (11)
February 1, 1975 J. Obstet. Gynecol.
fi'cDtGE
NORMAL
---.__* -'.,
3.2
samples
concentration of maternal serum AFP in the sixteenth week of gestation. This is an important step since theoretically it raises the possibility of screening all mothers to detect those whose serum AFP is elevated and provide them with amniocentesis and AF study for confirmation of a possible neural tube defect. In AF, AFP is present at the 6th week of gestation (1.5 pg per milliliter) 57 and although its concentration is much less (e.g., at twelfth to fifteenth weeks, its concentration is one-fiftieth to one-onehundredth that of fetal serum), its circadian pattern is essentially similar to that seen for fetal semm 37-39, 41 In 1972, Brock and Sutcliffess reported that AFP concentration was higher in AF in pregnancies resulting in myeloceles or in a more serious neural tube defect-anencephaly. Their observation was a retrospective one in which the study was made on AF in which the outcome of the pregnancy was already known. The publication provoked a small flood of prospective studies. in man38l 40-JQ which have been published during the last 18 months. Despite variation in the actual amounts of AFP present in the AF reported by different authors, they all demonstrate that AFP concentration between the fourteenth to twenty-second week of gestation is significantly higher in pregnancies resulting in neural tube defects than in normal pregnancies of equivalent gestational ages. However, there appears to be no significant difference in fetal serum AFP between normal fetuses and those with neural tube defects. This relationship suggests that there might be a corresponding increase in CSF AFP since virtually all reports indicate a much higher AFP concentration
4. Serum CPK and aldolase levels in patients with myelomeningocele before and after iliopsoas-transfer operation; P = Preoperative; 0 = time of operation.
in AF in those neural tube defects in which there is a continuing communication between the AF and CSF compartments. We have had an opportunity to study many human embryos with myeloceles” and the following embryogenetic considerations are pertinent to the problem. When a lumbosacral myelocele develops about the thirtieth day of gestation, there is always direct communication between the CSF and AF compartments. The choroid plexus invaginates the roof of the ventricle about the 36th day and its secretory function begins the thirty-seventh to fortieth day. The dura does not encircle the neural tube until the ninth to the tenth week and does not achieve a significant degree of compactness until the twelfth week. Therefore, when the myelocele occurs, there is undoubtedly free communication between the AF and the CSF in all neural tube myeloceles for at least 3 to 4 weeks. Meningeal closure and/or skin covering of the lesion when seen at birth would only mean that duration of communication was relatively more limited. Nothing is known about the rate of AFP metabolism in the fetus, placenta, or mother, and the exchange in the AF between these three sites. Thus the relatively few instances where normal AFP concentrations were recorded in Al? from pregnancies with neural tube defects clearly illustrates one of the difficulties in using this marker late in pregnancy (Fig. 6) .38-39 The specificity of AFP as a marker is further reduced by evidence that it is elevated in maternal serum and AF in instances of intrauterine fetal *Dr. University
H.
Nishimura, Professor of Anatomy Medical School, made the specimens
in Kyoto available.
Volume Number
121 3
Prenatal
detection
of neural
tube
defects
435
Fig. 5. A photomicrograph of a cross section through a human embryo with a lumbosacral myelocele. The boxed area is enlarged to show the sciatic nerve (arrow) innervating the developing muscles (C.-R. length, 17.9 mm.; C. A., 50 days).
death,Jz fetal anomalies,3Fp 42* 53 occurrence of prior abortions4 in first pregnancies as compared to subsequent ones,55 and in twin versus single conceptions5ti, 67 (Fig. 7). All these possibilities must thus be considered before conclusions can be drawn regarding the significance of elevated serum levels of AFP as a possible indicator of fetal neural tube anomalies
The
in utero.
lack of AFP
specificity
led some investiga-
tors
to search
for
a more
specific
protein
be found in high concentrations in the the aid of immunoelectrophoresis and against normal human CSF, Clausen strated the presence of beta trace protein CSF. Clausen’s findings were confirmed investigators.“9-“1 Its mean concentration healthy persons is 2.6 t 0.8 mg. per 100 trace protein has also been demonstrated
that would CSF. With antiserum demonin human by other in CSF of ml.“l Beta in human
436
Chaube
and
Swinyard Anl.
February I, lY7.i J. Obstet. Gynccol.
0,135), +,1361. D .137), I J381, .,(39) ( r,(40), nJ41). c,(42), 9,(43), +,t441, o .(45), 0,1461, e-147 ard 49). a,(48), l ,151)
NORMAL FETAL SERUM 1351
-WMAl.
AMNIOTIC FLUID (47ond491
.” --NORMAL
.p
AMNIOTIC FLUID (361
F NOWAL CMNKlTlC FLUID (391
IJr
NORMAL AMNICTIC FLUID (371 NORMAL MATERNAL SERUM (36)
01
’ 5
’ IO
: I5
t m
: 25
: 30
I 35
! 40
1 45
GES’TA?XWALAGF /IV WEiiKS Fig. 6. Alpha-fetoprotein (AFP) in maternal and fetal sera and.amniotic fluid of normal pregnancies and those with neural tube defects. N, normal; A, anencephalus; S, spina bifida (myelomeningocele) ; T, neural tube defect. When these scripts appear alone, they represent values for amniotic fluid. When the scripts are followed by the letter M or F, the values represent those of maternal serum (NM, AM, SM) or fetal serum (NF, AF). Normal and elevated AFP values taken from the same reference are designated by the same symbol.
Volume 121 Number 3
Prenatal
-WIN
NORMAL-
GESTAUONAL AGE IN WEEKS
Fig. 7. Concentration of a-fetoprotein in maternal serum at different gestational ages in normal single, twin, primiparous, and multiparous pregnancies. (Based on data of Ishigura and Nishimura.sjv 56. 57)
pleural and ascitic fluids,59 urine,59p 62 and serum.61 It has been reported absent in AF from normal human6? and rate” pregnancies but present in pregnancies with anencephalye3 and experimentally produced myelomeningocele.G4 We believe that additional studies of this type are urgently needed. Further observations on the concentration of CPK and aldolase in AF might enable us to learn more about abnormal fetal morphogenesis especially with regard to the time of onset of prenatal paraparesis while evaluation of AFP and beta trace protein by the recently developed counter-
detection
of neural
tube
defects
437
electrophoretic (rocket) techniques which improve their accuracy and sensitivity add to their importance as diagnostic techniques for prenatal detection of myelomeningocele and anencephaly. Preoccupation with cellular and/or biochemical indicators of neural tube defects found in AF obtained by amniocentesis should not cause us to overlook the assistance available from roentgenography,6” amniography,66 fetography,6’ ultrasound fetoscopy,68 or direct visualization of the fetus by fiberoptic microscopy.69 Proper selection of techniques now available makes it possible to detect prenatally a high percentage of the fetuses with anencephaly and myelomeningocele. Addendum Since this manuscript was submitted, additional evidence supporting the validity of using AFP levels in maternal serum70, T1 and AF72-75 have appeared. Ward and StewartT6 have pointed out that, since AFP concentration in fetal blood is in the milligrams per milliliter range and in micrograms per milliliter levels in amniotic fluid, hemorrhage of only 1 ml. of fetal blood into amniotic fluid will increase the AFP level in AF above the norm at 14 weeks. They presented two such cases and recommended that possibility of this occurrence be detected by screening AF for fetal hemoglobin by the Klechauer technique. An editorial” has emphasized the need for refinement of technique, additional data from normal “high-risk” mothers, accurate detection of ideal gestational age for detection, and additional data on plural gestation. In a retrospective study, Macri and associates78 confirmed the reliability of beta trace protein (BTP) as a marker in 15 cases of neural tube defect; however, Olsson and colleagues79 emphasized variability of BTP concentration and recommended use of AFP as a marker. The technical skills of Misses Patricia Joan M. Koechel are acknowledged.
A. Muck
and
REFERENCES
1. Carter, C. O., David, P. A., and Laurence, K. M.: J. Med. Genet. 5: 81, 1968. 2. Burton, B. K., Gerbie, A. B., and Nadler, H. L.: AM. J. 0asTa.r. GYNECOL. 118: 718, 1974. 3. Sutherland, G. R., and Brock, D. J. H.: Lancet 2: 1098, 1973. 4. Nelson, M. M., Ruttiman, M. T., and Brock, D. J. H.: Lancet 1: 504, 1974. 5. Schapira, G., and Dreyfus, J.-C.: Biochemistry of progressive muscular dystrophy, in Bourne, G. H., and Golarz, M. N., editors: Muscular Dystrophy in Man and Animals, New York, 1963, Hafner Publishing Company, p. 48.
6. Kouffen, H., and Consbrugh, U.: Nervenarzt. 41: 599, 1970. 7. Heyck, H., and Laudahn, G.: Muscle serum enzymes in muscular dystrophy and neurogenic muscular atrophy, A comparative study, in Milhorat, A. T., editor: Exploratory Concepts in Muscular Dystrophy and Related Disorders, Amsterdam, 1965, Excerpta Medica Foundation, p. 232. 8. Milhorat, A. T., Shafiq, S. A., and Goldstone, L.: Ann. N. Y. Acad. Sci. 138: 246, 1966. 9. Lui, W. Y.: Dev. Med. Child Neurol. 14: 467, 1972. 10. Drummond, M. B., and Belton, N. R.: Arch, Dis. Child. 47: 672, 1972.
438
11. 12. 13. 14.
15. 16. 17.
Chaube
and
Febluary
Swinyard
Nuttall, F. Q., and Wedin, D. S.: J. Lab. Clin. Med. 68: 324, 1966. Sibley, J. A., and Lehninger, A. L.: J. Biol. Chem. 177: 859, 1949. Pearce, J. M. S., Pennington, R. J., and Walton, J. N.: J. Neurol. Neurosurg. Psychiat. 27: 1, 1964. Okinaka, S., Sugita, H., Momoi, H., Toyokura, Y., Watanabe, T., Ebashi, F., and Ebashi, S.: J. Lab. Clin. Med. 64: 299. 1964. Bodensteiner, J. B.: and Zellweger, H.: J. Lab. Clin. Med. 77: 853, 1971. Zellweger, H., and Hanson, J. Med. _ W.: Excerpta 186: 53, 1969. Okinaka, S., Kumagai, M., Ebashi, S., Sugita, H., Momoi. * , Y.: Arch. Neural.’ 4:H.. 64, Tovokura. 1961. ’ Y.. ’ and Fuiie.
Am. J.
44. 45. 46. 47. 48. 49. 50. 51. 52.
18. 19.
20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 3 1. 32. 33. 34. 35. 36. 37. 38.
39. 40. 4.1. 42. 43.
Bray, G. M., and Ferrendelli, J. A.: Neurology 18: 480, 1968. Garrido Malo, M. A., de la Torre, C. T., Deutsch, S., and Falcone, A.: Bull. Sot. Chim. Bidl. (Paris) 52: 1467. 1970. Sherwin,’ A. L., Norris, J. W., and Bulche, J. A.: Neurology 19: 993, 1969. Herschkowitz, N., and Cumings, J. N.: J. Neurol. Neurosurg. Psychiat. 27: 247, 1964. Katz, R. M., and Liebman, W.: Am. J. Dis. Child. 120: 543, 1970. Aronson, S. M., Saifer, A., Perle, G., and Volk, B. W.: Proc. Sot. Exp. Biol. Med. 97: 33 1, 1958. Lisak, R. P., and Craig, F. A: J. A. M. A. 199: 160, 1967. Chantraine, A., Lloyd, K., and Swinyard, C. A.: Dev. Med. Child Neurol. 6: 7, 1964. Swinyard, C. A., Chaube, S., and Nishimura, H.: Pediatr. Ann. 2: 31, 1974. Pederson, K. 0.: Nature 154: 578, 1944. Bergstrand, C. G., and Czar, B.: Stand. J. Clin. Lab. Invest. 8: 174, 1956. Tatarinov, Y. S.: Fed. Proc. (Transl. Suppl.) 24: T916, 1965. Ruoslahti, E., and Seppall, M.: Nature 235: 161, 1972. Tatarinov, Y. S.: Vop. med. Khim. 11: 20, 1965. Oakes, D. A., Shuster, J., and Gold, P.: Cancer Res. 32: 2753, 1972. Pihko, H., and Ruoslahti, E.: Int. J. Cancer 12: 354, 1973. Gitlin, D., and Boesman, M.: Comp. Biochem. Physiol. 21: 327, 1967. Gitlin, D., and Boesman, M.: J. Ciln. Invest. 45: 1826, 1966. Seppali, M., and Ruoslahti, E.: AM. J. OBSTET. GYNECOL. 112: 208, 1972. Sepplla, M., and Ruoslahti, E.: AM. J. OBSTET. GYNECOL. 114: 595, 1973. Allen, L. D., Ferguson-Smith, M. A., Donald, I., Sweet, E. M., and Gibson, A. A. M.: Lancet 2: 522, 1973. Brock, D. J. H., and Sutcliffe, R. G.: Lancet 2: 197, 1972. Brock, D. J. H., and Scrimgeour, J. B.: Lancet 2: 1252, 1973. Field, B., Mitchell, G., Garrett, W., and Kerr, C.: Lancet 2: 798, 1973. Nevin, N. C., Nesbitt, S., and Thompson, W.: Lancet 1: 1382, 1973. Lorber, J., Stewart, C. R., and Milford Ward, A.: Lancet 2: 1187, 1973.
53.
54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79.
Obstet.
1 1975 Gynecol.
Sepplll, M., and Ruoslahti, E.: Lancet 1: 155, 1973. Seller, M. J., Coltart, T. M., Campbell, S., and Singer, J. D.: Lancet 2: 73, 1973. Leek, A. E., Ruoss, C. F., Kitau, M. J., and Chard, T.: Lancet 2: 385, 1973. Randle, G. H., and Cumberbatch, R. N.: J. Obstet. Gynaecol. Br. Commonw. 80: 1054, 1973. Harris, R., Jennison, R. F., Laurence, K. M., Ruoslahti, E., and Seppala, M.: Lancet 1: 429, 1974. Seller, M. J., Singer, J. D., Coltart, T. M., and Campbell, S.: Lancet 1: 428, 1974. Ruoslahti, E., and SeppllH, M.: Int. J. Cancer 8: 374, 1971. Brock, D. J. H., Bolton, A. E., and Monaghan, J. M.: Lancet 2: 923, 1973. Seppall, M., and Ruoslahti, E.: AM. J. OBSTET. GYNECOL. 115: 48, 1973. Laurence, K. M., Turnbill, A. C., Harris, R., Jennison, R. F., Ruoslahti, E., and SeppHlL, M.: Lancet 2: 860, 1973. SeppalP, M., and Ruoslahti, E.: Br. Med. J. 4: 769, 1972. Ishiguro, T.: Lancet 1: 1509, 1973. Ishiguro, T.: Lancet 2: 1214, 1973. Ishiguro, T., and Nishimura, T.: AM. J. OBSTET. GYNECOL. 116: 27, 1973. Clausen, J.: Proc. Sot. Exp. Biol. Med. 107: 170, 1961. Hochwald, G. M., and Thorbecke, G. J.: Proc. Sot. Exp. Biol. Med. 109: 91, 1962. Pepe, A. J., and Hochwald, G. M.: Proc. Sot. Exp. Biol. Med. 126: 630, 1967. Link, H.: Acta Neural. Stand. 43 (Suppl. 28): 7, 1967. Ericsson, J., Link, H., and Zettervall, 0.: Neurology 19: 606, 1969. Macri, J. N., Joshi, M. S., Weiss, R. R., and Evans, M. I.: Lancet 1: 14, 1974. Macri, J. N., Joshi, M. S., and Evans, M. I.: Nature (New Biol.) 246: 89, 1973. Russell, G. G. B.: J. Obstet. Gynaecol. Br. Commonw. 76: 345, 1969. Queenan, J. T., and Gadow, E. C.: Obstet. Gynecol. 35: 648, 1970. Aqiiero, O., and Zighelboim, T.: Surg. Gynecol. Obstet. 130: 649, 1970. Thompson, H. E.: Obstet. Gynecol. Surv. 23: 903, 1968. Takaaki, M., Chie, M., and Fujiko, Y.: J. Jap. Obstet. Gynecol. 15: 87, 1968. Wald, N. J., Brock, D. J. H., and Bonnar, J.: Lancet 1: 765, 1974. Brock,.D. J. H., Bolton, A. E., and Scrimgeour, J. B.: Lancet 1: 767, 1974. Milunsky, A., Alpert, E., and Charles, D.: Obstet. Gynecol. 43: 592, 1974. Seller, M. J., Greasy, M. R., and Alberman, E. D.: Br. Med. J. 2: 524, 1974. Emery, A. E. H., Brock, D. J. H., and Eccleston, D.: J. Obstet. Gynaecol. Br. Commonw. 81: 512, 1974. Medical News: J. A. M. A. 229: 1706, 1974. Ward, A. M., and Stewart, C. R.: Lancet 2: 345, 1974. Editorial: Lancet 1: 907, 1974” Macri, J. N., Weiss, R. R., and Joshi, M. S.: Lancet 1: 1109, 1974. Olsson, J. E., Sherman, M. S., and Kjessler, B.: Lancet 2: 347, 1974.
