Effect of amnionitis on the complement system of preterm infants

Ear& Human Development, 21(1990) 59-69 Elsevier Scientific Publishers Ireland Ltd.

59

EHD 001016

Effect of amnionitis on the complement system of preterm infants Hiroyuki Kitajima”, Masanori Fujimuraa, Toru Takeuchi”, Akira Miyanob, Masahiro Nakayama’, Tomio Fujitad, Shirou Imai” and Akira Shimizub Departments of ‘Neonatology, bCIinical Laboratory, ‘Pathology, Waternal Internal Medicine and ‘Obstetrics, Osaka Medical Center and Research Institute for Maternal and Child Health, Izumi, Osaka, 590-02 (Japan) Accepted for publication 17th July 1989

Summary The development of the complement system was studied by quantitation of total hemolytic complement activity (CHJ, Clq, C3, C4, and C3 split product (C3d) in cord plasma of nine human fetuses (17-22 weeks of gestation), 110 preterm (24-36 weeks of gestation) and 30 term neonates. The complement levels were analyzed in relation to various illnesses of preterm infants. Histological examination of the placenta revealed a higher incidence of amnionitis in the placenta of less than 34 gestational weeks. In cases without amnionitis, there were significant correlations between complement levels and gestational age. In cases with amnionitis, the complement system was activated even in infants of less than 28 weeks gestation. The complement levels correlated with the extent of the inflammation in the placentas and umbilical cords except for Clq. In infants with Wilson-Mikity syndrome, complement levels other than Clq were significantly elevated in comparison with those of infants with respiratory distress syndrome. In the group of preterm infants without amnionitis, no differences were found between infants with intrauterine growth retardation and those with growth appropriate for gestational age. complement system; preterm infants; amnionitis; respiratory Wilson-Mikity syndrome; intrauterine growth retardation.

distress syndrome;

Introduction It is well known that the complement

system is important

in the host defense,

Correspondence to: Dr. H. Kitajima. 0047-6374/90/$03.50 Printed and Published in Ireland

0 1990 Elsevier Scientific Publishers Ireland Ltd.

60

especially in extremely premature infants of less than 30 weeks gestation when serum immunoglobulin levels are still very low. Adinolfi examined the development of each component in fetuses of early gestation (less than 24 weeks) 111. All complement components appear in fetal serum by 18 weeks gestation and increase thereafter. The level of each component in cord sera of term infants is near or greater than half the level in maternal sera, except for C9 [2]. A few studies have reported complement levels of preterm neonates. However, the subjects were ah more than 30 weeks gestation and the number of samples was small [8,10]. The levels of complement components in extremely premature neonates still remain to be elucidated. Preterm delivery of less than 35 weeks gestation has been significantly associated with amnionitis [6]. No studies were available except for ours regarding the effect of amnionitis on the fetal complement system [14]. There is no direct method using neonatal blood samples to define intrauterine infection which affects fetuses and neonates; various indirect methods are inadequate, such as complete blood counts, differentials of the white blood cells and measurements of acute-phase reactants (Creactive protein, haptoglobin. We chose to examine each placenta histopathologically, which could confirm ascending or descending infection to the fetus. Wilson-Mikity syndrome (WMS) is a chronic lung disease in premature infants [21], but its etiology is unknown. We found that the majority of such cases had high IgM levels in the cord blood, suggesting the presence of intrauterine infection [l 11. Their placentas showed subacute or chronic inflammation of placental membranes and umbilical cords [ 121. In the present study, we investigated the development of the complement system of fetuses in early gestation and the effects of amnionitis on complement production and/or consumption. In addition, the complement levels in the cord plasma were analyzed in relation to various illnesses such as respiratory distress syndrome (RDS), WMS and intrauterine growth retardation (IUGR) of preterm infants. Materials and methods The subjects of the study were 9 fetuses and 140 neonates, including 110 preterm infants born at less than 37 weeks of gestation, and 30 normal term neonates. There were 24 infants with RDS, 12 with WMS and 11 with IUGR in the preterm infant group. Excluded were those infants with congenital anomalies and with sepsis and/ or meningitis. Gestational age (GA) was counted from the first day of the last menstural period, or assessed by fetal crown-rump length using ultrasonography. Newborn infants were assessed by the method of Dubowitz [9], and neonatal assessment was assigned as GA if the neonatal and obstetrical GA differed by more than 2 weeks. The diagnosis of RDS was made on the clinical signs of grunting, cyanosis, retraction breathing and by the characteristic pattern of chest X-ray. The assessment of pulmonary surfactant was made using the lecithin/ sphyngomyelin ratio [19], shake test (71 and microbubble test [18] of amniotic fluid and gastric aspirates. Wilson-Mikity syndrome was diagnosed by the clinical course and the findings of

61

chest X-ray according to Wilson and Mikity [21]. On the basis of applicable growth standards for Japanese [16], infants were classified as IUGR if the birth weight was below - 1.5 SD.‘The number of infants is shown according to GA, sex and neonatal illness in Table I. Infants of less than 24 weeks gestation included one 17.9, one 20.0, one 21.4, two 22.0, two 22.9 and two at 23.0 weeks of gestation, and were all born alive. Cord blood was collected into tubes containing EDTA sodium salt (final concentration about 5 mM) according to Brandslund et al. [5], and plasma was separated within 1 h of sampling and stored at - gO°C until analysis. CH, was determined by a modified method of Mayer et al. adapted to one-point measurement [13], using a commercial assay kit (Japan Lyophilization Laboratory, JLL, Tokyo, Japan). Clq was determined by single radial immunodiffusion [22], using antiClq antiserum (Behring). C3 and C4 were measured using Behring Nor Partigen plates (Hoechst Japan, Tokyo, Japan). Reference sera (Behring) were used for each component. C3d was determined by crossed immunoelectrophoresis with a slight modification of the method of Brandslund et al. [4]. The details of the method were reported by Miyano et al.[14]. At first, C3d was expressed by arbitrary units using normal adult serum after in vitro activation at 37OC for 4 days as a reference. The C3d level of the standard serum was defined as 500 munits/ (one munit/l corresponded to 0.024 mg/dl of the Behring reference sera). The unit of C3d is expressed as mg/dl in this text. Plasma samples from 12 healthy young adults were examined in the same manner. All samples were assayed in duplicate. In addition, we also determined the levels of C-reactive protein, haptoglobin, IgG, IgA and IgM in the cord serum. The placenta was fiied in 20% buffered formalin and cut into slices. Standard blocks of amnion, villi and cord were taken and embedded. Histological sections were stained with hematoxylin and eosin, and if necessary, PAS or Gram stains were TABLE I Distribution of patients. Gestation

Diseases

Sex

(Weeks) Male

Female

Total

<24 24-26 27-28 29-30 31-32 33-34 35-36 37-38 >38

6 8 11 9 a 13 7 10 10

3 9 16 5 11 11 2 2 a

9 17 27 14 19 24 9 12 18

Total

82

67

149

RDS

WMS

IUGR

24

12

11

62

added. Inflammation was estimated from the extent of leukocyte infiltration in the placenta by the method of Blanc [3] and in the umbilical cord by the method of Nakayama [IS]. In the placenta, the maternal polymorphs first marginate in the subchorionic intervillous space (intervillositis: grade I) and migrate towards the amniotic cavity through the chorion (chorionitis: grade II) and reach the amnion (chorioamnionitis: grade III) in severe and prolonged infections. In the umbilical cord, the fetal leukocyte infiltration is confined to endothelium (grade I) and to vascular wall (grade II) and advanced to Wharton’s jelly (grade III). Statistical analysis was performed by Student’s t-test and x2 test.

Table II shows the incidence rate of amnionitis according to GA in our subjects. A high incidence of amnionitis (over 60%) was seen at less than 28 weeks of gestation. Figure 1 shows each complement level in various GA. Number of samples in each GA is shown in Table II. Each complement level in infants with amnionitis was higher than that in infants without amnionitis, except for those at less than 24 weeks of gestation (Fig. 1). Infants with amnionitis at 28 weeks of gestation showed elevated levels of CH,,, C3 and C4, and the values were comparable with those of fullterm neonates without amnionitis. C3d levels were higher in the same group than those of full-term neonates and adults. No difference of Clq level was found in cases with and without amnionitis. In cases without amnionitis, significant correlations were noted between complement levels (CH,,,C4,C3,Clq and C3d) and gestational age. Clq was the most stable component among them. We analyzed relationships between the complement levels and the extent of inflammation in placentas and umbilical cords of the infants at less than 30 weeks of gestation (Table III and Fig. 2). Except for Clq, each complement level and C3d level increased with the extent of inflammation. The data of grade I in the umbilical artery are not shown, because of the small number of samples (only 2 cases). TABLE II Incidence of amnionitis. Gestation (weeks)

Incidence of amnionitis (@IO)

Amnionitis (-)

(+)

Total

<24 24-26 27-28 29-30 31-32 33-34 35-36 37-38 >38

3 5 10 9 14 20 9 12 18

6 12 17 5 5 4 -

9 17 27 14 19 24 9 12 18

66.7 70.5 63.0 35.7 26.3 16.7 -

Total

100

49

149

32.9

Clq

4

9 O24262830323438384oA GW

c3

C3d

T

.:.II P 11 -%, 30323436 GW

4

I 30323436384Oc;IA

384oA

GW

Fig. 1. Total hemolytic activity (CH,J and levels of each component (Clq, C4 and C3), and C3 split product (C3d) in each gestation period. 0, Infants with amnionitis; 0, infants without amnionitis; Cl, normal adults. Values represent mean f SD. Statistical differences are expressed as *P < 0.05, l*P < 0.02, or l**P< 0.01 by Student’s t-test.

TABLE III Extent of inflammation in the placentas and cords of infants of less than 30 weeks gestation. Total

Grade (extent of inflammation)

PlacentaI membranes Umbilical veins Umbilical artery

0

I

II

III

21 27 37

8 4 2

9 9 5

18 16 10

56 56 56

‘The extent of inflammation was represented as grade number by the method of Blanc (in the placenta) or Nakayama (in the umbilical vessels).

64

-

c1q

8 Wdl

I

Placental membranes(M) Umbilical veins(V) Umbilical arteries(A)

GRODE

GRRDE

36

wdl

-

-

c3

c4

I

l-*co ys 1II:M.U) -* (0 vs 111:&x ** (I vs 1II:ll)

80

l l

a3

-c3d

2

ma/d

****(O vs III:R.U.Al *= (I v3 1II:Ml ** (0 vs 1I:M)

l

*-co M 1II:rl.U.A) * (I vs 11I:f9

l

60

40 10 28

0

0

0

I GRADE

II

III

0

I GRADE

II

III

0

I

II

III

GRADE

Fig. 2. Relationships between the complement levels and the extent of inflammation in the placentas and cords of the infants with less than 30 weeks of gestation. Each bar represents the mean level of total hemolytic activity (CHJ and of each component (Clq, C4 and C3) and C3 split product (C3d). Grade gives the extent of inflammation by the method of Blanc [3] or Nakayama [IS]. Statistical differences are expressed as lP < 0.05. **P < 0.02, ***P < 0.01, or ****P < 0.001 by Student’s t-test. In the parentheses, each M. V or A means the statistical differences between the values of each grade in the placental menbranes (M). umbilical cord veins (V) or umbilical cord artery (A).

Figure 3 shows complement levels of patients with RDS, WMS and IUGR. In infants with RDS, complement levels of CH,,, C4 and C3 were lower than other groups and the levels of Clq and C3d were in the normal range. Only three cases with RDS had higher complement levels. In contrast, the complement levels of most of the infants with WMS were much higher than the normal levels, except for Clq. Complement levels were also studied in relation to the neonatal respiratory illnesses of RDS and WMS. Table IV shows the clinical data of infants with RDS and WMS. The mean birth weight of infants with RDS was greater than that of infants with WMS. There was no difference in the incidence of prolonged rupture of the membrane (PROM), but significant differences were seen in the extent of inflammation of the placenta and umbilical cord vessels in the two groups. CRP increased in

65

GW

’

242628302234383840

GW

GW

GW

. GW

Fig. 3. The levels of total hemolytic activity (CHJ and of each component (Clq, C4 and C3), and C3 split product (C3d) in each patient. 0, RDS; l , WMS; A. IUGR. Solid lines represent f S.D. from mean for each component of infants without amnionitis for each gestation period.

the cord serum of only one patient with WMS. Haptoglobin increased to above 1.3 mg/dl in 8 patients with RDS and 4 patients with WMS. IgA increased to above 2.2 mg/dl in 3 patients with RDS and 9 patients with WMS. There were no significtit differences of the levels of CRP, haptoglobin (RDS, 4.9 + 11.3 mg/dl; WMS, 16.5 f 28.5 mg/dl), IgA (RDS, 4.7 f 8.4 mg/dl; WMS, 14.1 & 24.0 mg/dl) and IgG (RDS, 357 + 246 mg/dl; WMS, 532 * 199 mg/dl) between infants with RDS and WMS. The levels of haptoglobin and IgA of infants with WMS tended to be higher than those of infants with RDS. In infants with WMS, IgM was markedly elevated (142 f 106 mg/dl), whereas in infants with RDS, IgM was 12.9 +: 12.5 mg/dl (P < 0.01). The complement levels and IgM are shown in Fig. 4. Infants with WMS had significantly higher complement levels and IgM than those with RDS, except for Clq. Complement levels of preterm IUGR infants were compared with 30 AGA infants without amnionitis. No difference was found (Table V).

66

TABLE IV Patient profiles of the neonates with RDS or WMS.

No. of patients Sex (male/female) Gestation (week.+ Birth weight (gy PROMb <24h >24h

RDS

WMS

P

24 9115 29.0 f 3.3 1195 f 482

12 l/5 21.1 f 1.4 947 i 192

< 0.05

17 4 3

5 3 4

Iflammation of placenta and cord Grade [3] or [15] 0 I

II

III

0

I

II

III

Placental membranes Umbilical veins Umbilical artery

1 2 1

0 0 0

0 2 6

0 1 0

5 2 0

7 I 6

20 21 23

2 1 0

< 0.001 < 0.001 < 0.001

Gestation and birth weight are expressed as mean f S.D. bathe cases of PROM were classified by the duration of rupture of membrane until delivery. Statistical analyses were determined by Student’s t-test and 2 test.

. .

C3d

c3 m

CH50

l(

. .

l

II

2 o-

pra ..

klM mKH

20 O-

"T

T

1.o-

i RW

O-

1C )(1-

RW

Ob RW

Fig. 4. Total hemolytic activity (CHJ and levels of each component (Clq, C4 and C3). C3 split product (C3d) and IgM in infants with RDS (R, open bars) and WMS (W. closed bars). Values represent mean rt S.D. Statistical differences are expressed as *P< 0.05, l*P< 0.01, or l**P< 0.001 by Student’s t-test.

TABLE V Comparison of the complement levels in IUGR infants and AGA infants of the same gestation without amnionitis. IUGR

No.of patients Gesation (weeks) Birth weight (g) CH, (units/ml) Clq (mg/dl) C4 (mg/dl) C3 (mg/dl) C3d (mg/dl)

AGA

11 33.0 + 2.1 1131 zt 363 14.0 5.67 12.0 45.2 0.583

f -c f f f

4.4 1.93 7.1 17.0 0.353

P

30 32.6 f 1.3 1631 -c 386 12.9 5.47 13.0 41.8 0.593

f + f f f

< 0.05

3.5 1.58 8.2 15.8 0.324

Results are expressed as mean f S.D. Statistical analyses were determined by Student’s t-test.

Discussion

The complement levels were studied of infants mainly from 24 to 34 weeks gestation. Pathology of the placentas revealed a high incidence of amnionitis in the preterm births. Significant differences of complement levels were noted in relation to the presence of amnionitis. In spite of inflammation of the placenta and cord, most neonates with amnionitis had a normal course in the neonatal period. No serious infection was seen in these neonates, but the complement levels were higher than those of infants without amnionitis. Therefore we divided all neonates into two groups; those with and without amnionitis. It was found that amnionitis stimulated production and activation of the complements, because the levels of CH,,, Clq, C4 and C3 in infants with amnionitis were higher than those of infants without amnionitis. Furthermore, we found that the complement levels correlated with the extent of the inflammation in the placentas and umbilical cords, except for Clq. High C3d levels indicate the activation of the complement system. Our results show that even premature infants of less than 28 weeks produced a moderate amount of complements, which were activated in vivo with the presence of amnionitis. Recently Stabile et al. reported the complement levels of the fetus during the second trimester of normal pregnancy [20]. They examined 55 matched fetal and maternal plasma and expressed the complement levels in arbitrary units. The range of C3 at 15 to 28 weeks gestation was 6 to 21 units in the fetus. If we could assume that the values of adult controls are the same in our and their experiments, the range of C3 became 5.7 to 20.0 mg/dl. The range of C4 at the same gestation was 2 to 17 units in the fetus and became 0.6 to 5.1 mg/dl. the relations of complement levels to GA were as follows; correlation coefficient (r) of C3 was 0.57 (P < 0.0001) and of C4 was 0.52 (P < 0.0001). In our data, mean levels of C3 and C4 of infants without amnionitis at 27 to 28 weeks gestation was 28.6 and 5.6 mg/dl, respectively. Correlation coefficients of C3 and C4 of 105 infants without amnionitis were 0.63 (P <

68

0.001) and 0.51 (P < 0.001). Both data could be comparable. According to their data, however, the complement levels of early gestation (less than 26 weeks) were higher than those of later gestation (Fig. 1) and their values. Neonates with RDS might have been physiologically immature, or they were actually “non-stressed” babies for their gestational age, because their complement vaIues seemed to be lower than those of infants without amnionitis, especially the levels of C4 and C3. In contrast, the complement levels of WMS cases were significantly higher than those of infants without amnionitis, except for Clq. In all cases of WMS, chronic or subacute amnionitis in the placenta was found. High IgM level as well as high levels of complement components in the cord blood might be related to the etiology of WMS. Only three cases of RDS had higher complement levels than those of infants without amnionitis. In two cases, acute inflammation was noted in their placentas. The third case was a live-born twin with antepartum death of one twin. This neonate died on the fourth day of life. Autopsy revealed bilateral cortical necrosis of the kidneys, massive infarcts of spleen and cerebral necrosis with middle cerebral arterial occulusion. Venous anastomoses were detected in the monochorionic-diamniotic placenta. However, no inflammatory changes were found in the placenta. Therefore, not only acute amnionitis but also fetal coagulopathy may activate the complement system. Some reports [8,17] have revealed that the complement levels of term infants with IUGR were the same as those of AGA infants of the same GA. Our results are consistent with this conclusion. In the infants of term IUGR and also preterm IUGR, the complement levels were not affected by intrauterine malnutrition. Acknowledgments The authors thank Miss U. Higashi and Miss K. Umebayashi for excellent technical assistance. References Adinolfi, M. (1977): Human complement. Onset and site of synthesis during fetal life. Am. J. Dis. Child., 131, 1015-1023. Adinolfi, M. and Beck, SE. (1975): Human complement C7 and C9 in fetal and newborn sera. Arch. Dis. Child., SO.562-564. Blanc, W.A. (1981): Pathology of the placenta, membranes and umbilical cord in bacterial, fungal and viral infections in man. In: Perinatal Diseases, pp.67-1323. Editor: N. Kaufmann. Williams and Wilkins, Baltimore. Brandslund, I., S&ted, H.C., Svehag, SE. and Tiesner, B. (1981): Double-decker rocket immunoelectrophoresis for direct quantitation of complement C3 split products with C3d specificities in plasma. J. Immunol. Methods, 44,63-71. Brandslund, I., Teisner, B., Petersen. P.H. and Svehag, S.E. (1984): Development and clinical application of electroimmunoassays for the direct quantitaion of complement C3 split products C3c and C3d. Stand. J. Clin. Lab. Invest. (Suppl. 168). 44,57-73. Chellam, V.G. and Rushton, D.I. (1985): Chorioamnionitis and funiculitis in the placentas of 200 births weighing less than 2.5 kg. Br. J. Obstet. Gynaecol.. 92,808-814. Clements, J., Platzker, A., Tierney, D. et al. (1972): Assessment of respiratory-distress syndrome by a rapid test for surfactant in amniotic fluid. N.Engl. J. Med., 286.1077-1081.

69

8 9 10 11 12

13 14

15 16 17 18 19

20

21 22

Drew, J.H. and Arroyave, C.M. (1980): The complement system of the newborn infant. Biol. Neonate, 37,209-217. Dubowitz, L., Dubowitz, V. and Goldberg, C.(1970): Clinical assessment of gestational age in the newborn infant. J. Pediatr., 77, l-10. Fireman, P., Zuchowiski, D.A. and Taylor, P.M. (1%9): Development of human complement system. J. Immunol., 103,25-31. Fujimura, M., Takeuchi, T., Ando, M. et al(l983): Elevated immunoglobulin M levels in low birthweight neonates with chronic respiratory insufficiency. Early Hum. Dev., 9,27-32. Fujimura, M., Takeuchi, T., Kitajima, H. and Nakayama, M. (1985): Elevated serum IgM of the neonate and chronic inflammation of the placenta with subseqent development of Wilson-Mikity syndrome. Biol. Neonate, 47,251. Kitamura, H., Inai, S. and Nagaki, R. (1983): A simple procedure for the titration of total hemolytic complement activity. One point method. Japan J. Clin. Chem., 12, 143-147. Miyano, A., Nakayama, M., Fujita, T., Kitajima, H., Imai, S. and Shim& A. (1987): Complement activation in fetuses: Assessment by the levels of complement components and split products in cord blood. Diagn. Immunol., 5,86-90. Nakayama, M. and Shim& I. (1988): diagnosis of chorioamnionitis. Sanfujinkachiryo, 56,410413 (in Japanese). Nishida, H., Sakanoue. M., Kurachi, K., Asada, A. and Funakawa, H. (1984): Fetal growth curve of Japanese. Acta Neonat. Jap., 20,90-97. Notarangelo, L.D., Chirico, G., Chiara, A. et al. (1984): Activity of classical and alternative pathways of complement in preterm and small for gestational age infants. Pediatr. Res., 18,281-285. Pattle, R.E., Krataing. C.C., Parkinson, C.E. et al. (1979): Maturity of fetal lungs tested by production of stable microbubbles in amniotic fluid. Brit. J. Obstet. Gynaecol., 86,615-622. Quinlivan, W.L.G., Reynolds, W.F., Mar&e, T. et al. (1973): An evaluation of multiple tests and the lecithin/sphingomyelin ratio for determining gestational age. Am. J. Obstet. Gynecol. 116: 1147 -1151. Stabile, I., Nicolaides, K.H., Bach, A. et al. (1988): Complement factors in fetal and maternal blood and amniotic fluid during the second trimester of normal pregnancy. Br. J. Obstet. Gynaecol. 95,281-285. Wilson, M.G. and Mikity, V.G. (1960): A new form of respiratory disease in premature infants. Am. J. Dis. Child., 99,489-499. Yonemasu, K., Kitajima, H., Tanabe, S., Ochi, T. and Shinkai, H. (1978): Effect of age on Clq and C3 levels in human serum and their presence in colostrum. Immunology, 35,523-529.