Pediatric Nephrology

Pediatr Nephrol (1991) 5:573-577 9 IPNA 1991

Original article Serum oq-microglobulin and [32-microglobulin for the estimation of fetal glomerular renal function S. Nolte*, B. Mueller, and W. Pringsheim Universit~its-Kinderklinik,Mathildenstrasse 1, W-7800 Freiburg, Federal Republic of Germany Received in June 15, 1990; received in revised form in January 30, 1991; accepted February 7, 1991

Abstract. As proteins cannot cross the placenta levels of the microproteins oq-microglobulin (CzlMG) and [32-microglobulin ([~2MG) can be used to assess fetal glomerular renal function, c~IMG, [32MG and creatinine were routinely determined in cord and maternal blood of 133 newborns [gestational age (GA) 2 5 - 4 2 weeks]. Twenty-nine patients with suspected impaired maternal or fetal renal function were studied separately and two fetuses were studied in utero. The mean fetal ~2MG concentration fell from 3.87_+0.56 mg/1 in the 2 5 - 3 1 weeks GA group to 2.60_+ 0.50 mg/1 in the mature newborn group, oqMG concentration fetl from 3.10-+0.51 to 2.25_+0.49 mg/dl. In contrast, the mean maternal ~ I M G concentration rose from 1.73 + 0 . 6 9 mg/1 in the 2 5 - 3 1 weeks G A group to a mean of 1.83 + 0.48 mg/1 in the mature newborn group; o q M G rose from 3.96 + 0 . 5 8 to 4.33_+ 1.6 mg/dl. Maternal and fetal creatinine levels were identical. Fetal microprotein levels fall during intra-uterine development as glomerular filtration rate (GFR) rises. There is no correlation between cord blood and maternal oqMG or [32MG concentrations. In 13 children with urological anomalies only 1 had elevated microprotein levels and he later developed renal insufficiency. Determination of microprotein levels in fetal serum can be used to detect severe renal function disturbances and to estimate GFR independently of maternal renal function.

Key words: c~>Microglobulin -

[32-MicroglobulinCreatinine - Neonate - Fetal kidney function

Introduction Fetal and neonatal renal function are difficult to assess. In most cases normal amounts of amniotic fluid, or sufficient postnatal diuresis and a lack of hyperhydration are indicative of sufficient kidney function. In many situations, such as prenatally diagnosed kidney abnormalities (especially if they are bilateral), or persistent anuria, estimation of glomerular filtration rate (GFR) is of central importance. It may also be important to ascertain if renal damage was prenatal, occurred during or after birth, or was due to any therapeutic measures. Fetal renal function can obviously not be determined by clearance methods. Serum creatinine or urea levels are not valid in utero or in the first hours and days of life because they reflect maternal renal function. As proteins cannot cross the placenta, the levels of microproteins such as cq-microglobulin (~IMG) and [32microglobulin (132MG) can be used to predict fetal glomerular renal function independently of maternal glomerular filtration [ 1]. These microproteins were first detected in the urine of patients with renal tubular damage [2]. They have been measured in fetal urine and amniotic fluid [3, 4] to predict maturity or renal damage, but urine concentrations have not proved to be predictive of overall renal function. In neonates serum ~2MG is a reliable indicator of renal function and allows individualization of digoxin therapy [5]. For evaluation of pre- or postnatal kidney function, serum levels of oqMG and [32MG have not been used, although cord blood samples are easily obtained. The aim of this study was to collect reference values for oqMG and [32MG in the newborn and to compare these with findings in patients with maternal, fetal or neonatal renal disease.

Materials and methods * Present address: Zentmm ftir Kinderheilkunde der Philipps-Universit~it, Deutschhausstrasse 12, W-3550 Marburg, Federal Republic of Germany Offprint requests to:

S. Nolte

Levels of ~2MG, oqMG and creatinine were determined in both cord blood and maternal blood of 133 preterm neonates of 25-42 weeks' gestational age (GA) referred to our hospital. These were divided into four groups: (1) group A, GA 25-31 weeks (mean 29.5 weeks) n = 29; (2) group B, GA 32-34 weeks (mean 33.0 weeks) n = 40; (3) group C,

574 Table 1. Body length (BL), weight (BW), maternal blood (MB) and cord blood (CB) levels of creatinine (Crea), ~2-microglobulin (~2MG) and czl-microglobulin(cqMG) in different gestational age (GA) groups with mean (x-), standard deviation (SD) and standard error of the mean (SEM) A: Prematures, GA 25-31 weeks (mean GA 29.5 weeks) n = 18

SD SEM

BL (cm)

BW (g)

Crea (MB) Crea (CB) (mg/dl)

~2MG (MB) ~2MG(CB) (mg/1)

cqMG (MB) cqMG(CB) (mg/dl)

38.4 3.6 0.8

1271 441 83

0.63 0.22 0.05

1.73 0.69 0.14

3.96 0.58 0.18

0.65 0.24 0.05

3.87 0.56 0.10

3.10 0.51 0.13

B: Prematures, GA 32-34 weeks (mean GA 33.0 weeks) n = 40

SD SEM

BL (cm)

BW (g)

Crea (MB) Crea (CB) (mg/dl)

[32MG(MB) [32MG(CB) (nag/l)

alMG (MB) alMG (CB) (mg/dl)

42.5 2.8 0.4

1723 313 39

0.74 0.42 0.06

2.03 0.39 0.14

4.21 I. 14 0.27

0.75 0.24 0.05

3.40 0.58 0.07

2.57 0.42 0.08

C: Newborns, GA 35-37 weeks (mean GA 36.0 weeks) n = 47

SD SEM

BL (cm)

BW (g)

Crea (MB) Crea (CB) (mg/dl)

[32MG(MB) ~2MG(CB) (rag/l)

oqMG (MB) oqMG(CB) (mg/dl)

45.3 3.0 0.4

2340 520 58

0.73 0.19 0.03

1.95 0.86 0.13

3.40 1.05 0.35

0.76 0.30 0.03

3.26 0.64 0.07

2.52 0.69 0.12

D: Newborns, GA 38-42 weeks (mean GA 38.8 weeks) n = 28

SD SEM

BL (cm)

BW (g)

Crea (MB) Crea (CB) (mg/dl)

~2MG (MB) ~2MG(CB) (rag/l)

alMG (MB) (ZlMG(CB) (mg/dl)

49.7 2.4 0.2

3333 563 41

0.75 0.2 l 0.04

1.83 0.48 0.08

4.33 1.65 0.50

0.72 0.15 0.01

GA 35-37 weeks (mean 36.0 weeks) n = 47; (4) group D, GA 3842 weeks (mean 38.8 weeks) n = 28. Babies with suspected or apparent renal function disturbances were excluded from these groups. Cases of impaired maternal or fetal renal function were studied separately: 4 patients with POTTERs sequence, oligohydramdios and lung hypoplasia, 4 with unilateral multicystic kidneys, 2 with maternal renal transplants, 2 with maternal glomerulonephritis, 2 with maternal impairment of renal function due to pyelonephritis, 2 with a family history of polycystic kidney disease, 11 with urological anomalies of the urinary tract diagnosed antenataily by ultrasound, and 2 with other reasons for neonatal renal insufficiency. Two fetuses were studied in utero, one at a GA of 33 weeks (for hydrops fetalis), the other for multiple malformations and suspected renal agenesia. Creatinine was determined enzymatically (PAP, Boehringer Ingelhelm, FRG) in the clinical laboratory of the University Children's Hospital, ~zMG by radio-immunoassay (Pharmacia, Uppsala, Sweden) and cqMG by a radial immunodiffusion test (Behring, Marburg, FRG). For [~2MG,the inua-assay coefficient of variation was 4.0 at 0.4 rag/1 and 3.1 at 13.4 mg/1(n = 10), the interassay coefficients of variation were 4.5 and 5.2% (n = 12), respectively. For cqMG, the intra-assay coefficient of variation was 7.4 at 0.7 mg/dl and 3.8 at 2.2 mg/dl (n = 15), the interassay coefficients of variation were 4.1 and 3.0 (n = 15). As the data were approximately normally distributed, standard deviation and standard error of the mean were calculated and Student's t-test used for statistical analysis. A P value of less than 0.05 was taken as significant. Estimation of fractiles and their confidence intervals were undertaken to comply with the recommendations for the determination of reference values as published by the International Federation of Clinical Chemistry [6]. Spearman's rank correlation was used to determine the correlation between fetal and maternal values.

2.60 0.50 0.04

2.25 0.49 0.04

Results Table 1 summarizes patient characteristics as well as the results of laboratory determinations. The maternal values correspond to the n o r m a l adult ranges, Creatinine concentrations in cord blood and maternal blood are identical. There is a small but significant rise b e t w e e n groups A and B as pregnancy progresses, but levels are i n d e p e n d e n t of G A (Fig. 1). Fetal [3zMG levels are significantly higher than maternal values (Fig. 2). Fetal concentrations of c q M G are significantly lower than maternal levels and are on the 5th percentile of the normal adult value (Fig. 3). Both ~2MG and c~IMG fall during intra-uterine development. The differences b e t w e e n groups A and B are significant for both c q M G and ~2MG as are differences between groups C and D. Table 2 summarizes the reference ranges obtained for ~ I M G and [3zMG in these age groups; reference limits (fractiles) for a 0.95 reference interval were calculated and confidence intervals for the fractiles obtained. A significant correlation b e t w e e n fetal and maternal values was f o u n d only for creatinine (r = 0.90, P >0.0001, n = 133). In 4 cases with P O T T E R s sequence cord blood values for [32MG were between 14.9 and 16.6 rag/l, and o~IMG 16 and 18 mg/dl, while the serum creatinine levels were normal. Figure 4 shows the results obtained in twins, 1 having n o r m a l kidney function, the other cystic renal dysplasia Potter type II. The newborns with unilateral multicystic

575 Table 2. Reference intervals (0,95 reference interval) defined by fractiles and confidence intervals (0.90 confidence intervals for each of the 0.95 reference limits) of [32MG and ot~MG for the four different GA ~oups A - D

0.9 0.8

0.7

A: GA 25-31 weeks, n = 18

-o 0.6

~2MG: cqMG:

.~0.5

2.76 (2.39-3.13)-4.97 (4.60-5.34) mg/l 2.83 (2.45-3.21)-5.09 (4.71-5.47) mg/dl

B: GA 32-34 weeks, n = 40 '2

0.4

[32MG: (ZlMG:

'~ 0.3 (..)

2.26 (2.00-2.52)-4.54 (4.28-4.80) rag/1 1.75 (1.56-1.94)-3.39 (3.20-3.53) mg/dl

C: GA 35-37 weeks, n = 47 0.2

~2MG: oqMG: A

B

C

D

2.01 (1.75-2.27)-4.51 (4.25-4.77) rag/1 1.16 (0.88-1.44)-3.87 (3.59-4.15) mg/dl

D: GA 38-42 weeks, n = 28

GA group

Fig. 1. Creatinine concentrations ( _+ SEM) in maternal blood (MB, []) and cord blood (CB, [E) in the gestational age (GA)groups A - D . (GA: A, 2 5 - 3 l ; B, 32-34; C, 35-37; D, 38-42 weeks)

-2 2-

' I

o ~E

/ A

I J

2_ B

I C

I 0

GA group

Fig. 2, cq-Microglobulin concentrations (oq,MG) ( 4:_ SEM) in ( [] ) and CB ([]) in the GA groups A - D

E (.9

:t 2-

[32MG: cqMG:

1.62 (1.36- 1.88)-3.58 (3.32-3.84) rag/1 1.29 (1.03 - 1.55)-3.21 (2,95- 3.47) mg/dl

renal dysplasia all had normal microprotein levels. The same is true for the babies born to mothers with impaired renal function. This is illustrated in Fig. 5: a neonate born with a maternal creatinine load has normal levels of (zlMG and [32MG, excess creatinine is excreted in the days after birth. Hence the child can be assumed normal renal function when microprotein levels are normal even though creatinine levels are high. Of the children with antenatally diagnosed anomalies of the urinary tract, 10 had normal microprotein levels and none showed serious impairment of renal function. One term boy with urethral valves had an initial serum 132MG level of 5.5 mg/1 and an OtlMG level of 5 mg/dl; he later developed compensated renal insufficiency. A sibling of 1 of our patients with Potters sequence was born with a serum [32MG level of 7.8 mg/1 and died of recessive polycystic kidney disease at the age of 5 months. One fetus examined in utero at a GA of 32 weeks (for hydrops fetalis) had a ~32MGlevel of 4.2 mg/1 and a (XlMG level of 3.2 mg/dl, which is within the normal range for this GA (Table 1). The second fetus studied for multiple malformations, had a serum ~32MGlevel of 5.9 mg/1 at a GA of 30 weeks; thus renal agenesia was excluded.

,'q

10

,,

3

•

,-q

~2-

~

03

c,l

i m

9 --

\\

n:p

c~

?,1

E o

5

.

A

X'M

B

C

D

GA group

Fig. 3. [32-Microglobulin concentrations (~32MG) (+_ SEM)in MB ( ~ ) and CB ( [] ) in the GA groups A - D

.

.

.

.

MB

I

ca

1I

Fig. 4. Concentrations of ~zMG (N) creatinine (E]) and oqMG (N) in MB and both CB samples of a twin pregnancy. Twin I suffered from multicystic renal dysplasia POTTER IIa. The dotted lines represent the normal reference range in Figs. 4 - 7

576 15-

F o

Z

10-

\\

,N

E 9

c'4

10. \\

5~1

x

"tJ

o 5:

F: 5, (.D MB

CB

2

,,,,

4

Days

1,4B Fig. 5. Concentrations of ~2MG ( ~ ) , creatinine ([]) and oqMG ([]) in MB, CB and on day 2 and 4 after birth. The neonate is born with a maternal creatinine load which is then excreted during the first days of life

20O~

E o ~E

44

15 "O

O4

A

10

i

O3

E (D

MB

CB

1

2

3

4

5

Days

Fig. 6. Concentrations of ~2MG (N), creatinine ([]) and (zIMG ([]) in MB, CB and during the first days of life of a term neonate with renal failure due to bilateral cortical necrosis after a severe placenta praevia haemorrhage In a patient with bilateral cortical necrosis following severe placenta praevia haemorrhage, the normal cord blood microprotein levels exclude a pre-existing renal function disturbance (Fig. 6). All three glomerular parameters rise simultaneously. In the patient illustrated in Fig. 7 microprotein analysis suggested pre-existing renal failure. This patient was the surviving child of a twin pregnancy, in which the other twin died in utero. Thromboplastic material from the dead twin led to bilateral intra-uterine renal vein thrombosis in the surviving child, and thus to renal insufficiency.

Discussion

Many attempts have been made to estimate GFR from plasma concentrations of glomerular markers and body size [7-9]. The rationale is that under steady-state conditions the plasma concentration is inversely related to renal function when a substance is totally excreted in the urine, which is true for inulin, and is almost true for creatinine. The fetal creatinine levels reflect maternal renal function.

CB

Day3

Fig. 7. Concentrations of [3zMG (N), creatinine ([]) and cqMG ([]) in MB, CB and on the 3rd day of life of a neonate presenting with acute oligohydramnios

This is seen in the patients in this study (Figs. 1, 4-7). Thus, after birth, there is no steady state between production and elimination of creatinine, because the newborn has either to excrete a maternal creatinine load (Fig. 5) or to reach his own steady state if his creatinine has been removed in excess before birth (Figs. 4, 7). As microproteins are being reabsorbed and degraded by tubular cell lysosomes [10], it is not possible to calculate clearance by comparing serum levels with urinary excretion. However, it is possible to use the inverse relationship between renal function and serum levels, because it is still true that the excretion rate equals the production rate, even though excretion cannot be measured because of tubular reabsorption and metabolism. High levels of microproteins in the urine indicate a defect of renal tubular reabsorption [1012]. [3zMG and (qMG levels correlate with serum creatinine levels in patients with various degrees of renal insufficiency [8, 13, 14] and during pregnancy [15, 16], but in subjects with normal renal function the correlation does not hold [13, 17], because creatinine production reflects muscle mass [18], ~2MG lymphocyte turnover [10] and CzlMG possibly the hepatic synthetic rate [12, 19-21]; all different individual determinants which can be influenced by many factors. The demonstration of a fall in microglobulin concentration during development does not necessarily reflect an improvement in renal function but could also be modified by the factors mentioned above. In our study, we did not see any influence other than GA on microprotein concentrations, but neonates with severe disease or obvious impairment of renal function were excluded, and numbers were too small to investigate other maternal or fetal factors which could modify lymphocyte activity and [3~MG release, such as prenatal stress, steroid therapy, amniotic infection or gestosis. Taking this into account, our data suggest that 132MG is superior to (zlMG for the evaluation of renal function probably because apart from renal excretion serum 132MG levels predominantly depend on lymphocyte turnover, whereas cqMG reflects the hepatic synthesis rate which seems to have a broader range. This is consistent with the findings of Donaldson et al. [12] who measured oHMG and [3zMG in older children. To our knowledge our study is the first to compare these

577

two parameters in the neonate. The most important finding of our study is that fetal and maternal microprotein levels are independent suggesting that these microproteins do not pass the placental barrier and that fetal serum levels are only influenced by fetal factors. Maternal 132MGlevels rise during pregnancy, and are especially high in women with gestosis [15, 16]; fetal ~2MG levels fall during development. The rise in maternal ~2MG concentration during pregnancy has been shown to relate to glomerular renal function and not to other factors influencing serum ~32MG levels [15]. Tests for I~2MG are now more widely available, partly due to the fact that ~2MG has been isolated as a major component of the amyloid substance in haemodialysed patients [22]. The simple radial immunodiffusion test (Behring, Marburg, FRG) can be performed by inexperienced technicians, and is almost a bedside test. The faster radio-immunoassay or enzyme immunoassay techniques require skilled technical knowledge but are now routinely available in many clinical laboratories. We propose that microprotein determination in fetal blood allows assessment of glomerular renal function to be made independently of maternal kidney function. This enabled us to estimate prenatal kidney function (Figs. 4, 5) and to determine the pre-existing kidney function in a patient with later disturbance of renal function (Fig. 6). This is helpful when renal abnormalities are diagnosed by ultrasound during pregnancy and have to be evaluated at birth [23]. Antenatal diagnosis of renal function could be performed using prenatal cord blood if severe abnormalities of the urinary tract are suspected. In these cases and especially in cases of renal failure during the newborn period, it might be of both medical and legal interest to determine whether renal damage has occurred before, during or after birth. Acknowledgements.Part of this work was presented at the VIII Congress of the International Pediatric Nephrology Association, Toronto 28 August- l September 1989 [24].

References 1. Nolte S, Pringsheim W, Ktinzer W (1986) 132-microglobulin - a parameter of glomerular renal function in the first days of life. Monatsschr Kinderheilkd 134:725 - 728 2. Berggard I, Beam AG (1968) Isolation and properties of a low molecular weight ~32-globulinoccurring in human biological fluids. J Biol Chem 243:4095-4103 3. Burghard R, Pallacks R, Gordjani N, Leititis JU, Hackel/3er B J, Brandis M (1987) Microproteins in amniotic fluid as an index of changes in fetal renal function during development. Pediatr Nephrol I: 574-580

4. Jonasson LE, Evrin PE, Wibell L (1974) Content of ~2-microglobulin and albumin in human amniotic fluid. Acta Obstet GynecoI Scand 53:49-58 5. Ito S, Oka R, Tsuchida A, Yoshika H, Mori Y (1989) Digoxin clearance and serum [32-microglobulin in neonates. J Pediatr 115: 478 -482 6. Solberg HE (for the IFCC and ICSH (1987) Approved recommendation on the theory of reference values. V. Statistical treatment of collected reference values. J Clin Chem Clin Biochem 25:645 - 656 7. Schwartz GJ, Feld LG, Langford DJ (1984) A simple estimate of glomerular filtration rate in full-term infants during the first day of life. J Pediatr 104:849-854 8. Shea BH, Maher JF, Horak E (1981) Prediction of glomerular filtration rate by serum creatinine and 132-microglobulin. Nephron 29: 30-35 9. Zacchello G, Bondio M, Saia OS, Largaiollo G, Vedaldi R, Rubatelli FF (1982) Simple estimate of creatinine clearance from plasma creatinine in neonates. Arch Dis Child 57: 293- 300 10. Revillard JP, Vincent C (1988) Structure and metabolism of beta-2microglobulin. Contrib Nephro162:44-53 11. Engle WD, Arant BS (1983) Renal handling of beta-2-microglobulin in the human neonate. Kidney Int 24: 358- 363 12. Donaldson MDC, Chambers RE, Woolridge MW, Whicher JT (1990) Alphal-microglobulin, ~32-microglobulin and retinol binding protein in childhood febrile illness and renal disease. Pediatr Nephrol 4:314-318 13. Leroy D, Manriat F, Dechaux M, Chopin N, Broyer M, Sachs C (1984) La ~2-microglobuline - index de la filtration glomdrulaire chez l'infant. Arch Fr Pediatr 41: 43 - 47 14. Wibell L, Evrin PE, Berggard I (1973) Serum [32-microglobulin in renal disease. Nephron 10: 320-331 15. Bartl W (1983) Reliability of serum [32-microglobulin as renal parameter in EPH gestosis. Z Geburtshilfe Perinatol 187:178 - 1 8 2 16. Hadagny J, Pacsa AS, Pejtsik B, Csaba IF (1983) Elevated level of beta-2 microglobulin in toxemic pregnancy. IRCS Med Sci 11: 550-551 17. Van Acker KJ, Vlietlinck RF, Steels PM (1984) Estimation of glomemlar filtration rate from ~3z-microglobulin serum levels in children. Int J Pediatr Nephrol 5 : 5 9 - 6 2 18. Sutphen JL (1982) Anthropometric determinants of creatinine excretion in preterm infants. Pediatrics 69:719-723 19. Itoh Y, Enomoto H, Takagi K, Kawai T (1983) Clinical usefulness of serum cq-micro-globulin as a sensitive indicator for renal insufficiency. Nephron 39:69-70 20. Kusano E, Suzuki M, Asano Y, Itoh Y, Takagi K, Kawai T (1985) Human cq-microglobulin and its relationship to renal function. Nephron 41:320-324 21. Vincent C, Marceau M, Blangafin P, Bouic P, Madjar J, Revillard J (1987) Purification of cqmicroglobulin produced by human hepatoma cell lines. Eur J Biochem 165:699-704 22. Gorevic PD, Casey TT, Stone WJ, Di Raimondo CR, Prelli FC, Frangione B (1985) [32-microglobutin is an amyloidogenic protein in man. J Clin Invest 76: 2425- 2429 23. Smith D, Eggington JA, Brookfield DSK (1987) Detection of abnormality of fetal urinary tract as a predictor of renal tract disease. Br Med J 294:27-28 24. Nolte S, Mtiller W, Pringsheim W (1989) Serum cq-microglobulin and [32-microglobulin for estimation of fetal glomerular renal function. Pediatr Nephrol 3: C 126