American Journal of Medical Genetics 20: 115-122 (1985)

Fetal Plasma Carbonic Anhydrase 111 in Prenatal Diagnosis of Duchenne Muscular Dystrophy

Robert Heath, Nicholas D. Carter, Stephen Jeffery, Robert J. Edwards, David C. Watts, Rosemary L. Watts, and Charles Rodeck.

Department of Child Health, St. George’s Hospital Medical School, London (R. H., N. D. C., S. J.); Department of Biochemistry, Guy’s Hospital Medical School, London (R. J. E., D. C. W , R. L. W); and Harris-Birthright Research Centre for Fetal Medicine, Department of Obstetrics and Gynaecology, King’s College Hospital, London (C. R.)

Carbonic anhydrase I11 (CAIII), a skeletal-muscle-specific enzyme which is ele- vated in the plasma of Duchenne muscular dystrophy (DMD) patients, was measured by radioimmunoassay in fetal plasma in order to evaluate its application to prenatal diagnosis of DMD. Using fetoscopy, pure fetal blood samples were taken at 17-24 weeks gestation from 25 fetuses at risk for DMD and from 78 control fetuses. Care was taken in the handling and storage of all samples. Normal sons were born in eight cases at risk for DMD. The CAIII levels in the infants were not significantly different from those of the control infants. Pregnancies were terminated in the remaining 17 at-risk cases. The CAIII levels in the fetuses were significantly different (p = 0.0034) from those of the control fetuses, although the distributions overlapped. Based on prior maternal risk, seven affected fetuses were expected in the terminated group; five had CAIII levels at or above the 95th centile of the control range. It is suggested that measurement of CAIU achieves partial discrimination between affected fetuses and their normal at-risk brethren.

Key words: Duchenne muscular dystrophy, carbonic anhydrase 111, prenatal diagnosis.

INTRODUCTION

The introduction of amniocentesis allowed fetal sex determination in pregnan- cies at risk for Duchenne muscular dystrophy (DMD). Subsequently, in attempts to distinguish normal from dystrophic male fetuses, fetal blood was obtained in the second trimester by blind needling [Kan et al, 19771 or at fetoscopy [Hobbins and

Received for publication February 27, 1984; revision received June 23, 1984.

Address reprint requests to N.D. Carter, Department of Child Health, St. George’s Hospital Medical School, London, SW17 ORE, England. Telephone: 01-672-1255.

0 1985 Alan R. Liss, Inc.

116 Heath et a1

Mahoney, 19741. There were two major problems: anterior placentae which made effective sampling difficult, and contamination of fetal blood samples by amniotic fluid and maternal blood. However, Rodeck [ 19801 described an improved technique that overcame these difficulties.

Since the basic biochemical defect in DMD is unknown, indirect tests are used to aid diagnosis, eg serum or plasma creatine kinase (CK) activity, which, even in the preclinical stage of the disease, may be more than 200 times the upper limit of normal [Pennington, 19801. Provided that this effect of the DMD gene is “expressed” early enough in fetal life, an affected fetus might be identifiable by measurement of CK. For such measurements to be authentic and accurate, it is necessary to obtain pure fetal blood uncontaminated by maternal blood or amniotic fluid [Edwards et al, 1984al. Preliminary results of attempts to identify affected fetuses using fetal plasma CK measurement [Edwards et al, 19801 indicated some overlap in the ranges of normal and dystrophic enzyme levels. This has since been confirmed on a larger series of at-risk patients [Edwards et al, 1984bl. We therefore investigated the possibility of using the muscle enzyme carbonic anhydrase 111 (CAIII) as an alterna- tive or corroborative test.

The isoenzyme CAW is specific to skeletal muscle [Jeffery et al, 19801 and to Type I fibres in particular [Shima et al, 19831. It occurs in the cytosol fraction of skeletal muscle, at relatively high concentration [Carter et al, 19791 and increased levels have been found in the plasma of DMD boys [Carter et al, 19831. The enzyme has been shown to occur in muscle early in fetal life [Jeffery et al, 19801. In a preliminary study [Carter et al, 19801, no correlation was found between amniotic fluid CAIII levels and DMD risk.

MATERIALS AND METHODS

Fetal blood samples used in this study were taken by fetoscopy from fetuses between 17-24 weeks [Rodeck, 19801. Blood was collected from each fetus as a series of 200 pl samples.

Purity of Fetal Blood Samples

A sample was deemed to consist of pure fetal blood i f (1) the sampling needle tip was seen intravascularly throughout aspiration, (2) analysis of red blood cell size by the Coulter “Channelyzer” confirmed the absence of maternal red blood cells, and (3) the haematocrit of all samples (except the first of a series which was diluted by fluid in the needle) was constant and about 35 % .

Sample Handling

Each sample of fetal blood was put into a separate plastic 2 ml tube precoated with dried lithium heparin. The tubes were gently shaken to dissolve sufficient heparin to prevent coagulation but without causing significant sample loss through adherence to the tube walls. The tubes were stored on ice until the end of the sampling procedure. 10 pl was removed from each of the samples to determine haematocrit. The blood was then transferred to small plastic micro-centrifuge tubes (350 p1 capacity). The tubes were spun in a micro-centrifuge (Hawksley, England) for 2 minutes. The plasma was removed and stored on ice [Edwards et al, 1984bl.

CAIII assay of the plasma was either started on the same day as sampling or samples were stored in liquid nitrogen for assay at a later date. CAIII is completely stable when stored this way [Heath, 19831.

CAIII and Duchenne Dystrophy 117

CAW Assay

CAIII was measured by radioimmunoassay [Carter et al, 19831; normally this is done in triplicate but the small volumes of fetal plasma available for analysis sometimes restricted the analysis to duplicate tubes at each dilution. Samples were usually assayed at two different dilutions (eg 115 and 1/10). Plasma quality controls were included in each assay.

Sample assay variation was determined in heparinized plasma samples taken from a normal adult male with similar CAIII levels to fetal plasma. The mean intra- assay coefficient of variation was 10%. The mean inter-assay coefficient of variation was 15%.

As a brief check of the stability of CAIII in fetal plasma, two samples were assayed for CAIII before and after a 16-hour incubation at room temperature (20°C). These comprised pooled fetal plasma with overall low and high CAIII concentrations respectively. There was no significant change with time in the CAIII concentrations assayed in either sample.

Control Samples In order to establish normal ranges for CAIII in fetal plasma at 17-21 weeks

gestation, samples were collected from patients undergoing termination of pregnancy (for social reasons or following ultrasound evidence of physical abnormalities) and from those undergoing diagnostic fetoscopy for conditions other than muscle disor- ders. The latter included fetuses with thalassaemia, haemophilia and various other genetic diseases. All blood sampling was performed with the ethical permission of King’s College and Guy’s Hospital Medical Schools and informed patient consent,

Carrier Risk Assessment

For any test applied prior to termination of pregnancy in at-risk mothers there remains the problem of validation, since there is as yet no reliable way of detecting DMD in the aborted fetus [Edwards et al, 19831. In this study, the expected number of dystrophic fetuses was derived from the cumulative carrier risk of the mothers. This was calculated after careful determination of carrier risks for individual patients, following full investigation of families wherever possible. For patients related to an isolated case of DMD where there was inadequate pedigree and CK information, a risk figure based on the residual risk for a series was used for the mother of the isolated case [Edwards et al, 1984bl.

The decision whether or not to terminate a pregnancy was taken throughout on the basis of maternal carrier risk and the wishes of the parents. The experimental nature of the fetal testing was always explained so as to be clearly understood by the family. In cases where at-risk pregnancies proceeded to term, plasma samples ob- tained from the male infant at age 4-5 days were examined for CK activity. Normal plasma CK activity was found in all cases.

RESULTS

The control fetal plasma CAIII data were obtained from fetuses which varied in gestational age, sex, and presence or absence of various disorders. The possible effect of these factors on CAIII levels was evaluated using the Mann-Whitney U-test.

When the data were grouped according to gestational age, there was no signifi- cant difference in plasma CAIII levels between any of the six 1-week age groups,

118 Heath et a1

TABLE I. Comparison of Fetal Plasma CAIII Ranges Between “Normal” Control Fetuses, Various Groups With Non-Muscle Disorders, and Groups At-Risk for DMD*

Statistical Group Median Range fetuses probability

Control fetuses

Plasma CAIII (ng/ml) N ~ . of

“Normal” 46 22-7 1 26

Thalassaernia major 41 20-80 14 p > 0.05 Physical defects 54 50-80 4 p > 0.05

Thalassaemia minor 43 5-69 20 p > 0.05 Sickle cell traitianaernia 38 20-42 4 p > 0.05 Heterogeneous conditions 49 29- 120 10 p > 0.05

Born normal 51 28-84 8 p > 0.05 Pregnancies at risk tor DMD

Terminated 62 48-78 17 p = 0.0034

*Statistical probability was calculated by the Mann-Whitney U-test. “Normal controls” were fetuses from social termination cases or those at risk for non-muscle disorders which were diagnosed as normal.

taken in pairs (p > 0.05). Male and female fetuses also showed no significant difference (p > 0.05). The control fetuses were then subdivided into “normal” fetuses (ie fetuses for social termination and fetuses at-risk for genetic disorders other than DMD, but diagnosed as normal) and fetuses in which other genetic disorders had been found. In Table I, it can be seen that CAIII levels in fetuses with physical defects (thalassaemia major or minor, sickle cell traitlanaemia, plus a heterogeneous group of ten) were not significantly different (p > 0.05) from those in the “normal” control fetuses. All the control results (from 78 fetuses) are combined to give the frequency histogram in Figure l a in which the “normal” control values are indicated by shading. The median value for the control samples is 43 ng/ml and the 95th centile is 75 ng/ml.

Fetuses At Risk for DMD Fetuses at-risk for DMD were grouped according to the outcome of the preg-

nancy. In the first group, normal male infants were born at term and in the second the pregnancies were terminated. Table I1 shows data for individual fetuses in these two groups including carrier at-risk assessment to date, from which the cumulative carrier risks for the groups are calculated. The carrier risks which had to be assumed at the time of referral for fetoscopy were in a number of cases different from those listed, which incorporate information obtained retrospectively [Edwards et al, 1984bl.

Eight fetal plasmas were analysed in the first group, for which the cumulative carrier risk was 1.72, a large proportion of which derives from patient 8, who was ascribed carrier status since she had a high plasma CK activity in pregnancy and also one week after delivery. Figure Ib shows that the values are distributed similarly to those of the control distribution, from which there was no significant difference (p > 0.05); seven of the values fall below the 95th centile of the control distribution.

The second group, where the pregnancies were terminated, comprised 17 fetal plasmas. Only three of the mothers had a carrier risk of less than 0.5 and the

CAUI and Duchenne Dystrophy 119

l 5 1 ~ A I CONTROLS M E D I A N

iN = 781 [ G m I

I n

- - - gSth CENTILE

- Y) w

2 5 - I 4 v1

LL a

i C l PREGNANCY TERMINATED

n n R n 0 I 1 1 . 1 I I I I I I I I I

I 1 0 20 40 60 80 100 120

F E T A L P L A S M A C A - I l l ( n g l m l i

Fig. 1 Distribution of fetal plasma CAIII levels. a. in 78 control fetuses; the 26 “normal” controls are shaded. b. in 8 cases at-risk for DMD where a normal son was born. c. in 17 cases at-risk for DMD where the pregnancy was terminated. The 3 cases where carrier risk < 0.5 are denoted by B .

cumulative figure was 13.42. Hence, the most probable number of affected fetuses in the group was seven. Five of the fetal CAI11 values fell at or above the 95th centile of the control distribution (Fig. lc) and the values for the whole group (Table I) were significantly different from those of the controls (p = 0.0034).

DISCUSSION

Fetal blood sampling by fetoscopy has improved over the years [Rodeck and Nicolaides, 19831 and is associated with minimal maternal trauma and low fetal mortality (2-3 % of all fetoscopies). This has enabled reliable enzyme measurements to be made on pure fetal plasma samples.

The ideal prenatal test would separate affected from unaffected fetuses without any overlap. Figure 1 shows that fetal plasma CAIII levels were increased in more of the at-risk terminated fetuses than would be expected if the distribution were the same as that of the controls. For the group as a whole, the increase, as compared to that of the control fetuses, is highly significant (Table I), even though the expected number of affected fetuses, based on the maternal carrier risk, is only 41 %, so that ten of these 17 would be expected to have values in the normal range. However, the degree of elevation is relatively small, of the same order as that found for fetal plasma CK [Edwards et al, 1984bl and, as with CK, the control and DMD distributions must to some extent overlap. Unfortunately, it is not possible to diagnose DMD in the abortus [Edwards et al, 19831, but from the present data it may be inferred that plasma CAIII

120 Heath et a1

TABLE 11. Fetal Plasma CAIII in Pregnancies At Risk for DMD

Gestational Fetal plasma age CAIII Carrier risk

(weeks) W m l ) to date

Normal son born I . J.A. 18 56 0.05 2. D.B. 21 54 0.08 3. D.D. 22 28 0.3 I 4. E.F. 20 84 0.05 5. R.R. (11) 20 59 0.12 6. D.L. 20 48 0.07 7. G.L. 22 42 0.04 8. M.P. 19 4 1 1 .oo

1.72 __

Pregnancy terminated 9. G.A. 19 61

10. 1 .R . 19 85 11. J.B. 18 78 12. T.C. 24 86 13. A.D. (1) 18 75 14. A.D. (11) 18 68 IS. R.F. (1) 18 52 16. W.H. 20 56 17. D.M. (1) 19 48 18. D.M. (11) 23 74 19. E.P. 22 63 20. D.S. 21 40 21. S.S. 21 44 22. S.U. 20 45 23. J.Wa 23 53 24. J.We 18 92 25. A.W. 20 63

Total

1 .oo 0.53 I .oo 1 .oo I .oo 1 .oo 0.12 0.62 0.20a 0.20a I .oo I .oo 1 .oo 1 .oo 0.75 1 .oo I .oo

13.42

~

level identifies some affected fetuses, although absolute discrimination between af- fected and unaffected fetuses is not possible at this gestational age.

Discrimination might be improved in two ways. First, by simultaneous mea- surement of CAIII and CK, and that of any other muscle proteins found to be useful, the indications might be reinforced. In this respect, myoglobin [Edwards et al, 1984~1, pyruvate kinase [Edwards et al, 1984el and lactate dehydrogenase [Edwards, 19811 have been shown to be of no value. Also in our laboratories, the correlation between fetal plasma CK and CAIII levels is currently being assessed. Second, in view of the low risk of fetal sampling, one might analyse a second sample, say 3-4 weeks after the first, the rationale being that leakage of muscle enzymes might be expected to increase with gestational age in a progressive disease. The leakage phenomenon is probably linked to the process of muscle cell differentiation; the number of differen- tiated fibres increases by up to 3-5-fold over the 4 week period from 16-20 weeks gestation, with the induction of glycolytic and oxidative enzymes [Colling-Saltin, 19781.

CAIII and Duchenne Dystrophy 121

At least in some families, markers or probes for the specific DNA sequence abnormality will be useful to diagnose the disease [Murray et al, 19821. In the meantime, the collection of more data on CAIII and other plasma enzymes that show abnormal levels in DMD during fetal life would seem to offer a basis for the calculation of likelhood ratios (comparable to those used in carrier detection; Sibert et al, 1979) in order to derive more meaningful fetal risk figures than those currently available by the simple process of halving maternal risk.

ACKNOWLEDGMENTS

This work was supported by the Muscular Dystrophy Group of Great Britain.

REFERENCES

Carter, ND, Heath, R, Jeffery, S , Emery, AEH, Burt, D (1980): Amniotic fluid carbonic anhydrase I11 not a predictor of Duchenne dystrophy. Lancet 2: 1381.

Carter, ND, Heath, R, Jeffery, S, Jackson, MJ, Newham, DJ, Edwards, RHT (1983): Carbonic anhydrase 111 in Duchenne muscular dystrophy. Clin Chim Acta 133:201-208.

Carter, ND, Jeffery, S, Shiels, A, Edwards, Y, Tipler, T, Hopkinson, DA (1979): Characterisation of human carbonic anhydrase 111 from skeletal muscle. Biochem Genet 17 (9/10):836-854.

Colling-Saltin, A-S (1978b): Some quantitative biochemical evaluations of developing skeletal muscles in the human fetus. J Neurol Sci 39:187-194.

Edwards, RJ (1981): A biochemical approach towards the prenatal diagnosis of Duchenne muscular dystrophy. PhD thesis. University of London.

Edwards, R , Kouseibati, F, Watts, R, Watts, DC (1980): Prenatal diagnosis of DMD by fetal blood sampling-biochemical results. Abstracts 8th Symposium. Current Research in Muscular Dystro- phy and Allied Neuromuscular Diseases, Newcastle Upon Tyne, 1980. London: Muscular Dys- trophy Group of Great Britain, p. 64.

Edwards, RJ, Rodeck, CH, Watts, DC (1983): Plasma creatine kinase and myoglobin levels, before and after abortion, in human fetuses at risk for Duchenne muscular dystrophy. Am J Med Genet

Edwards, RJ, Watts, DC, Watts, RL, Rodeck, CH (1984b): Creatine kinase estimation in pure fetal blood samples for the prenatal diagnosis of Duchenne muscular dystrophy. Prenatal Diagnosis (in press).

Edwards, RJ, Rodeck, CH, Watts, DC (1984~): The diagnostic value of plasma myoglobin levels in the adult and fetus at-risk for Duchenne muscular dystrophy. J Neurol Sci (in press).

Edwards, RJ, Rodeck, CH, Watts, DC (1984d): Errors in plasma creatine kinase estimates on fetal blood samples resulting from contamination with amniotic fluid and maternal blood: relevance for the prenatal diagnosis of Duchenne muscular dystrophy. Prenatal Diagnosis (in press).

Edwards, RJ, Rodeck, CH, Watts, DC (1984e): Pyruvate kinase fetal plasma and amniotic fluid unsuccessful for the prenatal diagnosis of Duchenne muscular dystrophy. Am J Med Genet (in press).

Emery, AEH, Burt, D, Dubowitz, V, Rocker, I, Donnai, D, Harris, R, Donnai, P (1979): Antenatal diagnosis of Duchenne muscular dystrophy. Lancet 1 :847-849.

Golbus, MS, Stephens, JD, Mahoney, MJ, Hobbins, JC, Haseltine, FP, Caskey, CT, Baker, BQ (1979): Failure of fetal creatine phosphokinase as a diagnostic indicator of Duchenne muscular dystrophy. N Engl J Med 300:860-861.

Hobbins, JC, Mahoney, MJ (1974): In utero diagnosis of haemoglobinopathies-Technic for obtaining fetal blood. N Engl J Med 290: 1065-1066.

Ionasescu, V, Zelleger, H, Cancilla, P (1978): Fetal serum creatine phosphokinase not a valid predictor of Duchenne muscular dystrophy. Lancet 2: 1251.

Jeffery, S , Edwards, Y, Carter, N (1980): Distribution of CAIII in foetal and adult human tissue. Biochem Genet 18 (9/10):843-849.

15 ~475-482.

122 Heath et a1

Kan, YW, Golbus, MS, Trecartin, RF, Filly, RA, Valenti, C, Furbetta, M, Cao, A (1977): Prenatal diagnosis of B-Thalassaemia and sickle cell anaemia. Experiences with 24 cases. Lancet 1 :269- 271.

Mahoney, MJ, Haseltine, FP, Hobbins, JC, Barker, BQ, Caskey, CT, Golbus, MS (1977): Prenatal diagnosis of Duchenne’s muscular dystrophy N Engl J Med 297:968-973.

Murray, JM, Davies, KE, Harper, PS, Meredith, L, Mueller, CR, Williamson, R (1982): Linkage relationship of a cloned DNA sequence on the short arm of the X-chromosome to Duchenne muscular dystrophy. Nature 300:69-71.

Pennington, RJT (1980): Clinical biochemistry of muscular dystrophy. In Walton, JN Br Mastaglia, FL (eds): The Muscular Dystrophies. Br Med Bull 36(2): 123-126.

Rodeck CH (1980): Fetoscopy guided by real-time ultrasound for pure fetal blood samples, fetal skin samples, and examination of the fetus in utero. Br J Obstet Gynaecol 97:449-456.

Rodeck, CH. Nicolaide, KH (1983): Fetoscopy and tissue sampling. Br Med Bull 39:332-337. Shima, K, Tashiro, K. Hibi, N, Tsukada, Y, and Hirai, H (1983): Carbonic anhydrase-111. Immunohis-

tochemical localisation in human skeletal muscle.. Acta Neuropathol (Berlin) 59:237-239. Sibert, JR, Harper, PS, Thompson, RJ, Newcombe, RG (1979): Carrier detection in Duchenne muscular

dystrophy. Evidence from a study of obligate carriers and mothers of isolated cases. Arch Dis Child 54:534-537.

Stengel-Rutowski, L, Scheurbrandt, G , Beckmann, R , Pongratz, D (1977): Prenatal diagnosis of Duchenne’s muscular dystrophy. Lancet 1: 1359-1352.

l?dited by John M. Opitz and James F. Reynolds