1012 LEVELS OF ALPHA-FETOPROTEIN IN MATERNAL BLOOD AS A SCREENING TEST FOR FETAL NEURAL-TUBE DEFECT
many patients with proven allergy and respiratory asthma as well as patients with allergic gastroenteropathy have serum-IgE levels within normal limits.15 16 The low serum-IgA levels in some of our patients are a little surprising since similar measurements in ulcerative colitis from the same laboratory have shown normal or raised values of serum-IgA." However, there were normal or possibly increased quantities of IgA present in the rectal-biopsy specimens of proctitis patients. Raised serum-IgM levels have also been reported in patients who have had colonic resection because of ulcerative colitis, suggesting that IgM binding occurs in the diseased
P. C. LEIGHTON M. J. KITAU
T. CHARD
Departments of Obstetrics, Gynœcology and Reproductive Physiology, St Bartholomew’s Hospital Medical College and London Hospital Medical College, London EC1A 7BE
Summary
bowel. 18 Our findings of cells with an
increased number of IgE-producof eosinophils in the rectal mucosa, and the clinical improvement in many of the patients treated with D.S.C.G., suggest that an immediate hypersensitivity reaction is important in the pathogenesis of this condition. These features are similar to those which have been described in children with cow’s milk allergy.16 These children had vomiting, diarrhoea, and chronic ill health, and the small-intestinal biopsy specimens contained large quantities of IgE cells.9 10 Allergic reactions involving the respiratory tract in asthma and also rhinitis have similar features. In all of these conditions it is proposed that antigen reacts with a cell-bound antibody, mainly IgE, which causes a breakdown of mast cells and the release of pharmacologically active substances which damage the local tissue. D.S.c.G. has been shown to prevent this immediate-type reaginic reaction as well as Arthus reactions, by inhibiting the breakdown of mast cells which follows antigen/antibody interaction. 19 It is probable that antigens, at present unidentified, stimulate the production of IgE in the rectal mucosa of susceptible individuals and that the subsequent damage to mast cells in the rectal wall is responsible for the inflammatory process. Such antigens may be present in the faeces, since ulcerative procto-colitis often originates in the distal colon or rectum at the site of maximum fsecal stasis. The demonstration of large quantities of IgE in the rectal mucosa, with a clinical response to D.S.C.G., provides strong evidence to support the view that an immediate hypersensitivity reaction is important in the pathogensis of proctitis and possibly other related conditions. an
ing
Y. B. GORDON A. E. LEEK*
excess
A range has been established for
mater-
nal plasma-alpha-fetoprotein (A.F.P.) levels in normal pregnancies (930 women). A.F.P. levels between 10 and 40 weeks gestation were examined in 51 pregnancies associated with fetal neural-tube defect. In 96% of cases (20 anencephalus, 6 spina bifida) examined between 16 and 26 weeks of gestation A.F.P. levels were above the 95th centile of the normal range. It is suggested that measurement of A.F.P. in maternal blood should become a screening test in all pregnancies, and a scheme for the further investigation of patients with abnormal results is described. Introduction
ALPHA-FETOPROTEIN (A.F.P.) is the major circulating protein of the early human fetus. It is also found in amniotic fluid, reaching peak levels at 12-14 weeks,’I and, at lower levels, in maternal blood reaching a peak 32 weeks.23 The finding that A.F.P. levels are often raised in maternal blood in association with a neuraltube defect of the fetus (table i) is an important advance at
TABLE I-LEVELS OF A.F.P. IN MATERNAL BLOOD IN PREGNANCIES ASSOCIATED WITH NEURAL-TUBE DEFECTS: CASES WITH BOTH ANNENCEPHALUS AND SPINA BIFIDA ARE CLASSED AS ANNENCEPHALUS
REFERENCES 1. 2.
Anderson, A. F. R. Am. J. dig. Dis. 1942, 9, 91. Wright, R., Truelove, S. C. Br med. J 1965, ii, 138. 3. Wright, R., Truelove, S. C. ibid. p. 142. 4. Pepys, J. Hypersensitivity Diseases of the Lungs due Dusts.
to
Fungi
and
Organic
Basle, 1969.
5. Altounyan, R. E. C. Acta allerg. 1967, 22, 487. 6. Cox, J. S. G. Nature, 1967, 216, 1328. 7. Stanworth, D. R. Proc. R. Soc. Med. 1969, 62, 971. 8. Heatley, R. V., Calcraft, B. J., Rhodes, J., Owen, E, Evans, B. K Gut, 1975, 16, 559. 9. Shiner, M., Ballard, J., Smith, M. E. Lancet, 1975, i, 136 10. Kilby, A., Walker-Smith, J. A., Wood, C B. S. ibid. p. 531.
*No normal range given; figures based on ranges of Brock et al. tNature of neural-tube defect not specified, classified under anencephalus, Cutoff point for normal range not specified; figures are estimates from data given tRalsed level in anencephalus was associated with spontaneous abortion.
11.
in obstetric
Skinner, J. M., Whitehead, R. J.
12. Lloyd, G., Fox,
H.
clin. Path.
1974, 27, 643.
Pathological Society of Great
Britain and
Ireland, 1974,
abstr. 19. 13.
Gelzayd,
E.
A., Sumner, C. K., Kirsner, J. P. Gastroenterology, 1968, 54,
334. 14. Gelzayd, E. A., Sumner, C. K., Fitch, F. W., Kirsner, J. P. ibid. p. 341. 15. Johansson, S. G. O. Lancet, 1967, ii, 951. 16. Freier, S., Berger, H. ibid. 1973, i, 913. 17. Bolton, P. M., James, S. L., Newcombe, R G, Whitehead, R H., Hughes,
since it presents the possibility of a screening programme leading to the early diagnosis and termination of these abnormal pregnancies. We have established a normal range for A.F.P. concentrations in maternal blood and present the results in 51 cases of neural-tube defect.
practice
L. E. Gut, 1974, 15, 213. 18. Gelernt, I. M., Present, D. H., Janowitz, H. D. ibid. 1972, 13, 21. 19. Pepys, J., Hargreave, F. E., Chan, M., McCarthy, D S. Lancet, 1968, 134.
ii,
*Present address: Institute of Cellular vain,
Brussels, Belgium.
Pathology, Catholic
Universe of Ion-
1013 Patients and Methods
TAR! F 11-MATERNAI PLASMA-A.F.P. I FVPI S IN NORMAl I’RH;NANC)
blood-samples were collected from 1322 pregnant part of a prospective study in the City of London and Hackney, Winchester, Southampton, and Wycombe districts. In the first district samples were collected at every visit throughout pregnancy; in the other three districts one sample was collected at the booking clinic. 930 pregnancies (3143 samples) could be defined as normal-i.e., the period of gestation was certain, there was no evidence of a threatened abortion, and the patient delivered a single live-born fetus without congenital abnormality. In the last trimester of pregnancy only3 patients without obstetric complications as previously defined 4400
women as
included. women in whom the fetus had a neural-tube defect were examined, and 96 blood-samples were assayed for A.F.P. The group included 8 patients who had had a previous infant with a neural-tube defect. In 18 cases (13 anencephalus, 5 spina bifida) termination was carried out. All cases were first identified by a raised maternal-blood A.F.P. A.F.P. was estimated by radioimmunoassay.3 The procedure is simple, rapid, and has a between-assay variation of 6%. The were
51
targeted to yield optimal precision in 40-640g/l plasma; the detection threshold is 10
dose-response curve the range
was
:. Results The levels of A.F.P. in maternal blood in normal pregnancy increase to a peak at 32 weeks’ gestation and subsequently fall to term (table u, fig. 1). 45 normal women had levels above the 95th centile between weeks 10 and 26 and provided a second blood-sample within 4 weeks; only 17 (38%) had a raised A.F.P. the second time. Of the 34 cases of anencephalus in which maternal blood was examined, 26 had A.F.P. levels greater than the 95th centile at some time during the pregnancy (fig. 1). In all 8 with normal levels the blood-samples were collected before the 16th week or after the 26th week. In the 3 pa-
I
I
I
I
*704 of these observations at 20 weeks of pregnancy .3 ject of a previous publication
or more
-
have been the sub-
tients in whom the pregnancy continued into the third trimester the levels were transiently raised between the 16th and 28th weeks of pregnancy and subsequently returned to the normal range (fig. 2); in 1 case A.p.p. levels before the 16th week were normal and subsequently rose. There were 12 cases of open spina bifida, 6 associated with maternal A.F.P. levels greater than the 95th centile at some time during the pregnancy (fig. 1). In 3 of the remaining cases samples were collected once at 10 weeks’ gestation, and in 2 more cases blood was examined only in late pregnancy; in the 1 case with a normal level of A.F.P. at 16 weeks the exact nature of the defect was not clearly recorded at the time of spontaneous abortion at 22 weeks, and it could have been a closed lesion. None of the 5 cases with a closed neural-tube defect had maternal blood levels of A.F.P. above the 95th centile. Thus, the detection-rate for neural-tube defects on the basis of maternal blood A.F.P. between weeks 16 and 26 was 97% for anencephalus and 92% for open spina bifida (table in). Discussion
Fig. 1-Median lines).
and 95th centile of maternal
plasma-A.F.P. throughout
normal pregnancy
(solid
Single or serial levels of maternal A.F.P. are shown for patients with a fetus having either anencephalus (8) or open spina bifida (o). Serial levels in individual cases are joined by interrupted lines.
This study confirms and extends previous findings on the use of maternal-blood
A.F.P.
determinations in the
early diagnosis of neural-tube defects of the fetus (table i). The number of
’
1014
further line of potential hazardous investigation. False-negative results may occur for a number of rea-
cate any
sons :
(1) Overlap between the normal and abnormal range of maternal A.F.P. levels; such overlap is inevitable with almost any biochemical test, and can only be avoided bv careful attention to assay technique." (2) Assay errors, which can only be avoided by repeat determinations on every sample. (3) Errors in dating the gestation; for example, an A.F.P. which may seem to be normal for the 18th week of pregnancy would be highly abnormal for the 14th week. (4) Closed neural-tube defects, which are often associated with normal levels of A.F.P. in both amniotic fluid and maternal blood. (5) Collection of the sample at an inappropriate stage of gestation; in our study (see fig. 1 and 2) the optimal period seems to be between the 16th and 26th weeks. .
Weeks of pregnancy
The false-negative rate arising from the overlap between normal and abnorThe normal median and 95th centile are shown by solid lines. Levels are probably normal mal populations will be much inup to 14th week of pregnancy, raised between weeks 16 and 24, and normal again in later fluenced by the."action line" which is pregnancy. chosen to divide the two populations: cases described permits, for the first time, some tentative lowering this limit will increase the detection-rate at the observations on the practical use of A.F.P. estimation in expense of an increase in false positives; raising it will Fig. 2-Serial levels of A.F.P. in maternal plasma in 3 women with anen cephalic fetus (
many patients with proven allergy and respiratory asthma as well as patients with allergic gastroenteropathy have serum-IgE levels within normal limits.15 16 The low serum-IgA levels in some of our patients are a little surprising since similar measurements in ulcerative colitis from the same laboratory have shown normal or raised values of serum-IgA." However, there were normal or possibly increased quantities of IgA present in the rectal-biopsy specimens of proctitis patients. Raised serum-IgM levels have also been reported in patients who have had colonic resection because of ulcerative colitis, suggesting that IgM binding occurs in the diseased
P. C. LEIGHTON M. J. KITAU
T. CHARD
Departments of Obstetrics, Gynœcology and Reproductive Physiology, St Bartholomew’s Hospital Medical College and London Hospital Medical College, London EC1A 7BE
Summary
bowel. 18 Our findings of cells with an
increased number of IgE-producof eosinophils in the rectal mucosa, and the clinical improvement in many of the patients treated with D.S.C.G., suggest that an immediate hypersensitivity reaction is important in the pathogenesis of this condition. These features are similar to those which have been described in children with cow’s milk allergy.16 These children had vomiting, diarrhoea, and chronic ill health, and the small-intestinal biopsy specimens contained large quantities of IgE cells.9 10 Allergic reactions involving the respiratory tract in asthma and also rhinitis have similar features. In all of these conditions it is proposed that antigen reacts with a cell-bound antibody, mainly IgE, which causes a breakdown of mast cells and the release of pharmacologically active substances which damage the local tissue. D.S.c.G. has been shown to prevent this immediate-type reaginic reaction as well as Arthus reactions, by inhibiting the breakdown of mast cells which follows antigen/antibody interaction. 19 It is probable that antigens, at present unidentified, stimulate the production of IgE in the rectal mucosa of susceptible individuals and that the subsequent damage to mast cells in the rectal wall is responsible for the inflammatory process. Such antigens may be present in the faeces, since ulcerative procto-colitis often originates in the distal colon or rectum at the site of maximum fsecal stasis. The demonstration of large quantities of IgE in the rectal mucosa, with a clinical response to D.S.C.G., provides strong evidence to support the view that an immediate hypersensitivity reaction is important in the pathogensis of proctitis and possibly other related conditions. an
ing
Y. B. GORDON A. E. LEEK*
excess
A range has been established for
mater-
nal plasma-alpha-fetoprotein (A.F.P.) levels in normal pregnancies (930 women). A.F.P. levels between 10 and 40 weeks gestation were examined in 51 pregnancies associated with fetal neural-tube defect. In 96% of cases (20 anencephalus, 6 spina bifida) examined between 16 and 26 weeks of gestation A.F.P. levels were above the 95th centile of the normal range. It is suggested that measurement of A.F.P. in maternal blood should become a screening test in all pregnancies, and a scheme for the further investigation of patients with abnormal results is described. Introduction
ALPHA-FETOPROTEIN (A.F.P.) is the major circulating protein of the early human fetus. It is also found in amniotic fluid, reaching peak levels at 12-14 weeks,’I and, at lower levels, in maternal blood reaching a peak 32 weeks.23 The finding that A.F.P. levels are often raised in maternal blood in association with a neuraltube defect of the fetus (table i) is an important advance at
TABLE I-LEVELS OF A.F.P. IN MATERNAL BLOOD IN PREGNANCIES ASSOCIATED WITH NEURAL-TUBE DEFECTS: CASES WITH BOTH ANNENCEPHALUS AND SPINA BIFIDA ARE CLASSED AS ANNENCEPHALUS
REFERENCES 1. 2.
Anderson, A. F. R. Am. J. dig. Dis. 1942, 9, 91. Wright, R., Truelove, S. C. Br med. J 1965, ii, 138. 3. Wright, R., Truelove, S. C. ibid. p. 142. 4. Pepys, J. Hypersensitivity Diseases of the Lungs due Dusts.
to
Fungi
and
Organic
Basle, 1969.
5. Altounyan, R. E. C. Acta allerg. 1967, 22, 487. 6. Cox, J. S. G. Nature, 1967, 216, 1328. 7. Stanworth, D. R. Proc. R. Soc. Med. 1969, 62, 971. 8. Heatley, R. V., Calcraft, B. J., Rhodes, J., Owen, E, Evans, B. K Gut, 1975, 16, 559. 9. Shiner, M., Ballard, J., Smith, M. E. Lancet, 1975, i, 136 10. Kilby, A., Walker-Smith, J. A., Wood, C B. S. ibid. p. 531.
*No normal range given; figures based on ranges of Brock et al. tNature of neural-tube defect not specified, classified under anencephalus, Cutoff point for normal range not specified; figures are estimates from data given tRalsed level in anencephalus was associated with spontaneous abortion.
11.
in obstetric
Skinner, J. M., Whitehead, R. J.
12. Lloyd, G., Fox,
H.
clin. Path.
1974, 27, 643.
Pathological Society of Great
Britain and
Ireland, 1974,
abstr. 19. 13.
Gelzayd,
E.
A., Sumner, C. K., Kirsner, J. P. Gastroenterology, 1968, 54,
334. 14. Gelzayd, E. A., Sumner, C. K., Fitch, F. W., Kirsner, J. P. ibid. p. 341. 15. Johansson, S. G. O. Lancet, 1967, ii, 951. 16. Freier, S., Berger, H. ibid. 1973, i, 913. 17. Bolton, P. M., James, S. L., Newcombe, R G, Whitehead, R H., Hughes,
since it presents the possibility of a screening programme leading to the early diagnosis and termination of these abnormal pregnancies. We have established a normal range for A.F.P. concentrations in maternal blood and present the results in 51 cases of neural-tube defect.
practice
L. E. Gut, 1974, 15, 213. 18. Gelernt, I. M., Present, D. H., Janowitz, H. D. ibid. 1972, 13, 21. 19. Pepys, J., Hargreave, F. E., Chan, M., McCarthy, D S. Lancet, 1968, 134.
ii,
*Present address: Institute of Cellular vain,
Brussels, Belgium.
Pathology, Catholic
Universe of Ion-
1013 Patients and Methods
TAR! F 11-MATERNAI PLASMA-A.F.P. I FVPI S IN NORMAl I’RH;NANC)
blood-samples were collected from 1322 pregnant part of a prospective study in the City of London and Hackney, Winchester, Southampton, and Wycombe districts. In the first district samples were collected at every visit throughout pregnancy; in the other three districts one sample was collected at the booking clinic. 930 pregnancies (3143 samples) could be defined as normal-i.e., the period of gestation was certain, there was no evidence of a threatened abortion, and the patient delivered a single live-born fetus without congenital abnormality. In the last trimester of pregnancy only3 patients without obstetric complications as previously defined 4400
women as
included. women in whom the fetus had a neural-tube defect were examined, and 96 blood-samples were assayed for A.F.P. The group included 8 patients who had had a previous infant with a neural-tube defect. In 18 cases (13 anencephalus, 5 spina bifida) termination was carried out. All cases were first identified by a raised maternal-blood A.F.P. A.F.P. was estimated by radioimmunoassay.3 The procedure is simple, rapid, and has a between-assay variation of 6%. The were
51
targeted to yield optimal precision in 40-640g/l plasma; the detection threshold is 10
dose-response curve the range
was
:. Results The levels of A.F.P. in maternal blood in normal pregnancy increase to a peak at 32 weeks’ gestation and subsequently fall to term (table u, fig. 1). 45 normal women had levels above the 95th centile between weeks 10 and 26 and provided a second blood-sample within 4 weeks; only 17 (38%) had a raised A.F.P. the second time. Of the 34 cases of anencephalus in which maternal blood was examined, 26 had A.F.P. levels greater than the 95th centile at some time during the pregnancy (fig. 1). In all 8 with normal levels the blood-samples were collected before the 16th week or after the 26th week. In the 3 pa-
I
I
I
I
*704 of these observations at 20 weeks of pregnancy .3 ject of a previous publication
or more
-
have been the sub-
tients in whom the pregnancy continued into the third trimester the levels were transiently raised between the 16th and 28th weeks of pregnancy and subsequently returned to the normal range (fig. 2); in 1 case A.p.p. levels before the 16th week were normal and subsequently rose. There were 12 cases of open spina bifida, 6 associated with maternal A.F.P. levels greater than the 95th centile at some time during the pregnancy (fig. 1). In 3 of the remaining cases samples were collected once at 10 weeks’ gestation, and in 2 more cases blood was examined only in late pregnancy; in the 1 case with a normal level of A.F.P. at 16 weeks the exact nature of the defect was not clearly recorded at the time of spontaneous abortion at 22 weeks, and it could have been a closed lesion. None of the 5 cases with a closed neural-tube defect had maternal blood levels of A.F.P. above the 95th centile. Thus, the detection-rate for neural-tube defects on the basis of maternal blood A.F.P. between weeks 16 and 26 was 97% for anencephalus and 92% for open spina bifida (table in). Discussion
Fig. 1-Median lines).
and 95th centile of maternal
plasma-A.F.P. throughout
normal pregnancy
(solid
Single or serial levels of maternal A.F.P. are shown for patients with a fetus having either anencephalus (8) or open spina bifida (o). Serial levels in individual cases are joined by interrupted lines.
This study confirms and extends previous findings on the use of maternal-blood
A.F.P.
determinations in the
early diagnosis of neural-tube defects of the fetus (table i). The number of
’
1014
further line of potential hazardous investigation. False-negative results may occur for a number of rea-
cate any
sons :
(1) Overlap between the normal and abnormal range of maternal A.F.P. levels; such overlap is inevitable with almost any biochemical test, and can only be avoided bv careful attention to assay technique." (2) Assay errors, which can only be avoided by repeat determinations on every sample. (3) Errors in dating the gestation; for example, an A.F.P. which may seem to be normal for the 18th week of pregnancy would be highly abnormal for the 14th week. (4) Closed neural-tube defects, which are often associated with normal levels of A.F.P. in both amniotic fluid and maternal blood. (5) Collection of the sample at an inappropriate stage of gestation; in our study (see fig. 1 and 2) the optimal period seems to be between the 16th and 26th weeks. .
Weeks of pregnancy
The false-negative rate arising from the overlap between normal and abnorThe normal median and 95th centile are shown by solid lines. Levels are probably normal mal populations will be much inup to 14th week of pregnancy, raised between weeks 16 and 24, and normal again in later fluenced by the."action line" which is pregnancy. chosen to divide the two populations: cases described permits, for the first time, some tentative lowering this limit will increase the detection-rate at the observations on the practical use of A.F.P. estimation in expense of an increase in false positives; raising it will Fig. 2-Serial levels of A.F.P. in maternal plasma in 3 women with anen cephalic fetus (
