AMERICAN JOURNAL OF REPRODUCTIVE IMMUNOLOGY 27:117-123 © 1992 MUNKSGAARD
AJ"'Jnanc~ Immunolo2ll
Interleukin-1 ex and Interleukin-1 r3 in Preterm and Term Human Parturition ROBERTO ROMERO, MOSHE MAZOR, FRANCISCO BRANDT, WALDO SEPULVEDA, CECILIA AVILA, DAVIDB. COTTON, AND CHARLES A. DINARELLO Department of Obstetrics and Gynecology, Yale University School of Medicine, New Haven, Connecticut (R.R., M.M., F.B., WS., C.A.J; Department of Obstetrics and Gynecology, Wayne State University School of Medicine, Detroit, Michigan (RR., F.B., WS., D.B.C.); Department ofMedicine, Tufts University School ofMedicine, Boston, Massachusetts (CAD.); and the Intramural Research Branch ofPerinatology, National Institutes ofHealth, Bethesda, Maryland (RR., WS.) ABSTRACT: Interleukin-l (1L-1)has been implicated in the mechanism of human parturition in the setting of infection. The purpose of this study was to determine the effect oflabor (term and preterm) and microbial invasion of the amniotic cavity on amniotic fluid (AF) concentrations IL-la and IL-II3. AF was retrieved by transabdominal amniocentesis from the following groups of women: midtrimester genetic amniocentesis (16 to 18 wk) (N = 15), preterm labor with intact membranes (21 to 36 wk) with or without infection (N = 72), preterm premature rupture of membranes (PROM) (N = 88), and term not in labor or in active labor with or without infection (N = 58). AF was cultured for aerobic and anaerobic bacteria as well as Mycoplasmas. IL-l was measured with a commercially available immunoassay validated for AF (sensitivity: IL-la, 157 pg/ml; 11...-113, 50 pg/ml). All women at midtrimester had undetectable AF IL-la and IL-II3. Among women in preterm labor with positive AF cultures, II...-la and IL-113 were detectable in the AF in 86.6% (13/15) and 100% (15/15), respectively. In contrast, all women with negative AF cultures without labor (N = 36) had undetectable AF IL-la concentrations and 52.7% (19/36) had undetectable AF IL-113 concentrations. Histopathological chorioamnionitis was present in 92.8% (13/14) of patients who had positive AF cultures and detectable IL-l in the AF. IL-l was significantly higher in patients with preterm PROM, labor, and positive AF cultures than in the other subgroups of patients with preterm PROM. Among women at term who were not in labor, 26.3% (5/19) had detectable IL-la and 19.8% (3/19) had detectable IL-113 in the AF. Among those in active labor with positive AF cultures, 44.4% (4/9) had detectable IL-la and 88.8% (8/9) had detectable 11...-113 in the AF. IL-113 was detected more frequently in patients at term with positive AF cultures than in those with negative AF cultures (88.8% [8/9] vs, 38.7% [19/49]; P = 0.0088). We conclude that AF IL-l concentrations are elevated in women with preterm labor an microbial invasion of the amniotic cavity and in women with spontaneous parturition at term. (Am J Reprod Immunol. 1992; 27:117-123.) Key words: Interleukin-l, amniotic fluid, parturition, preterm labor, premature rupture of membranes, prematurity, intra-amniotic infection. INTRODUCTION
Interleukin-I (IL-l) is a term used to describe two polypeptides (lL-la and IL-II3>. which have a wide spectrum Submitted for publication December 24,1991; accepted February 24,1992. Address reprint requests to Roberto Romero, M.D., Wayne State University, School of Medicine, Department of Obstetrics and Gynecology, 4707 St. Antoine, Detroit, MI48201.
of immunological and nonimmunological activities (reviewed in detail in Dinarello'r'), These cytokines have similar molecular weight (approximately 17 kD), but have different isoelectric points (5 for IL-la and 7 for IL-II3). IL-la and IL-113 are the product of two distinct genes, which appear to have evolved from a common ancestor gene. Although these cytokines share less than 30% structural homology, they bind the same cell surface receptors and have similar biological properties. 1,2 Interleukin-l has been implicated as a signal for the onset of parturition in the setting of infection.P:" Indeed, IL-la and IL-113 are produced by human decidual explants in response to bacterial endotoxin," and these cytokines can stimulate prostaglandin production by amnion" and decidua 7 in a dose-dependent manner. Moreover, administration of IL-l to pregnant mice results in preterm parturition." To explore the possible participation oflL-l in the mechanisms of human parturition in the setting of infection, we have previously assayed IL-l in the amniotic fluid (AF) of women with microbial invasion of the amniotic cavity and preterm labor with the use of the DI0.G4.1 bioassay." This bioassay, like others for cytokines, is time-consuming, laborious, semiquantitative, and is affected by the presence of inhibitors. Using this assay method, neutralizing antibodies are required to estimate the relative contributions of IL-la and IL-113 to the total AF IL-l bioactivity. However, interpretation of this type of assay is sometimes difficult because of the complex interaction between the different species of IL-l, other related cytokines active on this bioassay (i.e., IL-6), and inhibitors. Recently, specific immunoassays for IL-la and IL-113 have become available. The purpose of this study was to determine whether changes in the AF concentrations oflL-la and IL-113 occur in the setting oflabor (term and preterm) and microbial invasion of the amniotic cavity. MATERIALS AND METHODS
Study Design A cross-sectional study was conducted according to the results of AF cultures, gestational age, and labor status. Four groups were identified for study purposes. Group 1 consisted of women in the midtrimester of pregnancy (gestational age of16 to 18 wk) who were-undergoing amniocentesis for genetic indications (N = 15). Group 2 consisted of women admitted with preterm labor and intact membranes (N = 72); these patients were subclassified into three subgroups according to the AF culture results and response to tocolysis: women with negative AF cultures who responded to tocolysis and delivered at term (N = 36); women with negative AF cultures who failed tocolysis and delivered prematurely (N = 21), and women with positive AF cultures who delivered a premature neonate (N = 15). Group 3 consisted of women
118
ROMEROET AL.
with premature rupture of membranes (PROM) (N 88); this group was subdivided into four subgroups according to the presence of labor at the time of amniocentesis and the results of the AF culture: women with negative AF cultures and without preterm labor (N = 25); women with negative AF cultures but with preterm labor (N = 21); women with positive AF cultures but without preterm labor (N = 19), and women with positive AF cultures and preterm labor (N = 23). Group 4 consisted of nonlaboring women at term (38 to 39 wk) who had amniocentesis for the assessment offetallung maturity before elective cesarean section (N = 19) and women in spontaneous active labor at term (N = 39). The latter subgroup was subdivided according to the results of AF cultures: women with negative AF cultures (N = 30) and women with positive AF cultures (N = 9). PROM was defined as amniorrhexis before the onset of spontaneous labor. Nitrazine, pooling, and feming tests were performed in all women in labor to exclude leakage of AF into the vagina. Pelvic examinations were not performed at the time of amniocentesis, as this could result in an increased risk of infection. 10 Tocolytic agents were not used in patients with preterm PROM. Preterm labor was defined as the presence of regular uterine contractions with a frequency of at least two every 10 min associated with changes in the cervical Bishop's score. Ritodrine was administered intravenously as the tocolytic agent to all ~atients, according to a protocol described by Caritis.! Amniocentesis was performed before the initiation of therapy. Failure of tocolysis was diagnosed when cervical dilation progressed beyond 5 em or delivery occurred. Retrieval ofAmniotic Fluid AF was retrieved by transabdominal amniocentesis in all women in this study. Amniocentesis is offered in our institution for biological studies to all patients with the diagnosis of preterm labor. Women in active labor at term were ascertained by searching our obstetrical database. They were identified as patients who had undergone amniocentesis because of suspected preterm labor but subsequently delivered (within 12 h of amniocentesis) a newborn who was considered term by pediatric examination. The identification number of patients fulfilling this criterion was cross-referenced with samples stored in our AF bank. A sample of AF was sent for microbiological studies. The remaining fluid was centrifuged at 200g for 10 min at 4°C and stored in polypropylene tubes at - 70°C until assayed. Microbiologic Culture Technique AF was transported to the laboratory in a capped plastic syringe immediately after collection. These conditions precluded air contact with the specimen. Plating occurred within 30 min of collection in all cases. AF was cultured for aerobic and anaerobic bacteria as well as Mycoplasmas using methods previously described. 12 lL-1 a and lL-113 Determinations IL-la was measured with a radioimmunoassay previously described 13 and validated for human AF. The sensitivity of the assay was 157 pg/ml, the intra-assay coefficient of variability was 8.4%, and the interassay coefficient of variation was 12%. IL-l13 was determined with a commercially available immunoradiometric assay
9000
8000 7000
6000 IL-1u (pg/ml)
5000
4000
3000
2000 1000
No Labor No Infection
(n = 36) A
~ Labor No Infection (n = 21)
Labor Infection (n = 15)
B
C
Fig. 1. Amniotic fluid IL-l", concentrations in women with preterm labor and intact membranes. IL-l", was detected more frequently in women with positive amniotic fluid cultures than in women with negative amniotic fluid cultures in the other two groups (group A, median and range = 0; group B, median = 0, range c 157 pg/ml to 0.5 ng/ml; group C, median = 3.5 ng/ml, range '" 157 pg/ml to 8.9 ng/ml; Kruskal-Wallis H = 45.49; P < 0.00001).
validated for AF (Medgenix, Belgium). The sensitivity of the assay was 50 pg/ml, and the intra- and interassay variabilities were 7.1 % and 10.5%, respectively. Statistical Analysis Comparisons between proportions were performed with either a Chi-square or a Fisher's exact test. An unpaired Student's t-test was used to compare continuous variables, which were normally distributed (gestational age, maternal age, etc.) A Wilk-Shapiro test was employed to test the hypothesis that AF IL-l concentrations were normally distributed. A Kruskal-Wallis ANOVA was used to compare AF IL-l concentrations among study groups. Multiple comparisons among groups were conducted with Dunn's test (True Epistat, Epistat Services, Richardson, TX). RESULTS
AF from women in the midtrimester of pregnancy did not contain detectable IL-la or IL-l13. Figures 1 and 2 display AF concentrations ofIL-la and IL-l13 in women
INTERLEUKIN-l IN PRETERM AND TERM PARTURITION
-
1
1000 >
species (N = 4) and Ureaplasma urealyticum (N = 4). Polymicrobial invasion was present in 26.6% (4/15) of patients. None of the patients had clinical chorioamnionitis. One newborn delivered at 26 wk had congenital sepsis. Histopathological chorioamnionitis was detectable in 13 of the 14 (92.8%) patients with positive AF cultures in whom placental examination was performed. All these patients had detectab'.e IL-l[3 in AF. Figures 3 and 4 display AF concentrations oflL-la and IL-l[3 in women with preterm PROM. Women with preterm labor and positive AF cultures had significantly higher AF IL-l concentrations (IL-la, median = 0.62 ng/ml, range ~ 157 pg/ml to 10 ng/ml; IL-l[3, median = 1.085 ng/ml, range = 145 pg/ml to 10.6 ng/ml) than women in the other subgroups (P < 0.00001). Overall, IIAa and IL-1[3 were detected in 82.6% (19/23) and 100% (23/23) of patients with positive AF cultures, respectively, and in only 21% (4/19) and 63.1 % (12/19) ofthose women with positive AF cultures but without labor, respectively. Among women with negative AF cultures, IL-1a and IL-1[3 were detected in 10.8% (5/46) and 30.4% (14/46), respectively.
.
5000 4000 3000 2000
IL-1~
(pg/ml)
119
1000 900 800 700 600 500
20000
400 300 200 100
50 - - - -
0
.t. - -
-~-
- - - - - - - -
No Labor No Infection (n = 36)
A
... 1
10000 - -
----- -- ---- ----
Labor No Infection (n = 21)
Labor Infection
B
C
.
(n = 15)
Fig. 2. Amniotic fluid IL-113 concentrations in women with preterm labor and intact membranes. IL-113 was detected more frequently in women with positive amniotic fluid cultures than in women with negative amniotic fluid cultures in the other two groups (group A, median = 0, range ,,;; 50 mg to 0.206 ng/ml; group B, median = 0, range-s 157 pg/ml to 0.648 ng/ml; group C, median = 10 ng/ml, range = 215 pg/ml to 18 ng/ml; Kruskal-Wallis H = 37.39;P < 0.000001).
in preterm labor with intact membranes. IL-la was detected in 86.6% (13/15) of patients with positive AF cultures for microorganisms (median = 3.5 ng/ml, range ~ 157 pg/ml to 8.9 ng/ml) and IL-l[3 in all cases (15/15) (median = 10 ng/ml, range ~ 50 pg/ml to 18 ng/ml). On the other hand, only 19% (4/21) of women in preterm labor who had negative AF cultures for microorganisms and who delivered prematurely had detectable IL-la and 47.6% (10/21) had detectable IL-l[3 in the AF. Both IL-la and IL-l[3 concentrations were significantly higher in women with microbial invasion of the amniotic cavity than in women who delivered prematurely but had negative AF cultures (IL-la, 86.6% [13/15] vs. 19% [4/21], P = 0.0001; IL-l[3, 100% [15/15] vs. 47.6% [10/21], P = 0.0006; Fisher's exact test). Among women with negative AF cultures who responded to tocolysis, none (0/36) had detectable IL-la and 47.2% (17/36) had detectable IL-l [3 in the AF. The most common microbial isolates in patients with preterm labor and intact membranes were Fusobacterium
3500
IL-1a
3000
(pg/ml)
2500 2000 1500 1000
----..
500
157 -~---~---~------.:.-;,---
o
No Labor Labor No Labor No Infection No Infection Jnfection (n=25) (n=21) (n=19) ABC
La'bor Infection (n=23) 0
Fig. 3. Amniotic fluid IL-la concentrations in women with preterm PROM. IL-la was detected more frequently in women with positive amniotic fluid cultures and preterm labor than in women in the other three groups (group A, median = 0, range s 157 pg/ml to 0.7 ng/ml; group B, median = 0, range « 157 pg/ml to 0.3 ng/ml; group C, median = 0, range ss 157 pg/ml to 2 ng/ml; group D, median = 0.62 ng/ml, range ,,;; 157 pg/ml to 10 ng/ml; Kruskal-Wallis H = 37.18; P < 0.000001).
120
ROMERO ET AL.
"°1
with negative AF cultures (88.8% [8/9] vs. 38.7% [19/49]; P = 0.0088, Fisher's exact test). In contrast, the rate of AF IL-1a detection was not different between patients with positive and negative AF cultures (44.4% [4/9] vs. 32.6% [16/49]; P = 0.7, Fisher's exact test). Among women in labor, those with positive AF cultures had a significantly higher median AF IL-113 concentration than those with negative AF cultures (median = 201 pg/ml, range = 115 pg/ml to 5 ng/ml vs. median = 51.5 pg/ml, range = 51 pg/ml to 2.821 ng/ml, respectively; P = 0.0137, Mann-Whitney U test). Also, IL-113 was detected more frequently in the AF of women in labor with a negative AF than in those not in labor (61.5% [24/39] vs. 15.8% [3119], respectively;P = 0.0017, Fisher's exact test). The most common microorganisms isolated from the nine women in active labor at term were Ureaplasma urealyticum (N = 4; 44.4%) followed by Lactobacillus species, Streptococcus agalactiae, and Mycoplasma hominis (N = 2; 22.2% each). Polymicrobial invasion of the amniotic cavity was present in only two patients. There was not significant difference in cervical dilation between patients with detectable and those with nondetectable AF IL-1a or IL-113.
6000
5000 4000 3000 2000 IL-1~
(pg/ml)
1750
1250 1000 750 500
-.-.
250 - - - - - -
.......
4000 - - - - - - - - - - - -
No Labor Labor No Labor No Infection No Infection Infection (n = 25) (n = 21) (n = 19) ABC
"
3000
Labor Infection (n = 23) 0
Fig. 4. Amniotic fluid IL-113 concentrations in women with preterm PROM. IL-l 13 was detected more frequently in women with positive amniotic fluid cultures and preterm labor than in women in the other three groups (group A, median = 0, range c 50 pg/ml to 0.442 ng/ml; group B, median = 0, range ss 50 pg/ml to 0.947 ng/ml; group C, median = 0.28 ng/ml, range «; 50 pg/ml to 4.7 ng/ml; group D, median = 1.085 ng/ml, range = 146 pg/ml to 10.642 ng/ml; Kruskal-Wallis H = 45.11; P < 0.00001).
2000
900 800 IL-1a (pg/ml)
700 600
The most common microbial isolates in the AF in patients with preterm PROM were Ureaplasma urealyticum (N = 20), Mycoplasma hominis (N = 10), and Bacteroides species (N = 4). Polymicrobial infections were detected in 30.9% (13/42). No significant differences in the proportion of women with microbial invasion due exclusively to Mycoplasmas or polymicrobial infections could be detected between the groups with labor or without labor. Of patients with positive AF cultures and spontaneous onset of labor, placental examination was performed in 16 cases. Histological chorioamnionitis was detected in all placentae. There were no cases of neonatal sepsis in this subgroup of patients. The interval (in hours) between rupture of membranes and amniocentesis was not significantly different among the four subgroups of patients with preterm PROM (medians for intervals in subgroups A, B, C, and D showing in Figures 3 and 4 were 7.5 h, 6.5 h, 11 h, and 11 h, respectively). Figures 5 and 6 display AF concentrations ofIL-1a and IL-113 in women at term. IL-113 was detected more frequently in patients with positive AF cultures than in those
500 400 300 200 157
.
_________________ _l
e
100 ----IIINII--
--liliiii&--
No Labor No Infection
(n =19)
Labor No Infection (n = 30)
•
Labor Infection (n = 9)
A
B
C
Fig. 5. Amniotic fluid IL-la concentrations in women at term (group A, median = 0, range « 157 pg/ml to 0.449 ng/ml; group B, median = 0, range-s 157 pg/ml to 0.627 ng/ml; group C, median = 0, range s 157 pg/ml to 3.592 ng/ml; Kruskal-Wallis H = 1.08;P = 0.58).
.
INTERLEUKIN-l IN PRETERM AND TERM PARTURITION
121
ulating factor 16), IL-1a and IL-ll3 do not seem to be normally present in second trimester AF. Immunoreactive IL-a and IL-ll3 were not detected in most samples of AF obtained from women at term not 3000 in labor. Patients with detectable AF IL-1a (26.3% [5/19]) and IL-ll3 (15.8% [3/19]) had low levels of these cytokines and no clinical or microbiological evidence of 2000 infection (see Figs. 5 and 6). A possible explanation for the detection ofIL-la and IL-ll3 in the AF of these otherwise normal women is the release of these cytokines 1000 by cells normally present in the amniotic cavity during the process of spontaneous cell death/damage. Indeed, we have demonstrated that the cell pellet of term amniIL-1 ~ otic fluid (mainly formed by desquamated keratinocytes) (pg/ml) contains high levels of bioactive IL-1. 3 An alternative 500 explanation to our findings is the existence of an inflammatory process of infectious or noninfectious etiology. A major finding of this study is that microbial inva400 sion of the amniotic cavity is associated with the presence of both IL-1a and IL-ll3 in the AF. Indeed, in each ofthe three populations studied (preterm labor with intact 300 membranes, preterm PROM, and term labor), IL-ll3 was detected more frequently and in higher concentrations in women with a positive AF culture for microorganisms -----r-200 than in those with a negative AF culture. Similar findings were observed for IL-1a in patients with preterm labor (with intact or rupture of membranes). However, 100 we could not document a higher detection rate ofIL-la in the AF of patients with term labor and microbial inva:r.:::: sion of the amniotic cavity. The most likely explanation Labor Labor No Labor for this finding is that microbial invasion of the amniNo Infection Infection No Infection otic cavity occurring during the course of term labor is (n = 9) (n = 30) (n = 19) A B C of short duration and elicits a milder host response than the microbial invasion that occurs in the context of preterm labor with intact membranes. The higher detecFig. 6. Amniotic fluid IL-I(3 concentrations in women at term. IL-l(3 tion rate ofIL-ll3 in comparison with IL-1a in term labor was significantly higher in patients with labor and infection in comprobably reflects the preferential secretion ofIL-ll3 durparison with women in labor but without infection or in nonlaboring ing the course of macrophage activation (lL-ll3:IL-la ratio women (group A, median = 0 pg/ml, range ss 50 pg/ml to 55 pg/ml; group = 9:1). Although the immunoassay for IL-1a used in our B, median = 51.5 pg/ml, range'; 50 pg/ml to 2.821 ng/ml; group C, median = 201 pg/ml, range-s 50 pg/ml to 5.0 ng/ml; Kruskal-Wallis H study was less sensitive than that for IL-ll3 (lL-1a, 157 = 18.79,P = 0.00008). pg/ml; IL-ll3, 50 pg/ml), the data displayed in Figure 5 suggest that this is not a likely explanation for our findings. Indeed, most patients with microbial invasion of Among women with microbial invasion of the amni- the amniotic cavity and detectable IL-ll3 had AF conotic cavity and labor, those in preterm labor had signifi- centrations above 157 pg/ml (66% [6/9]). cantly higher AF concentrations of IL-ll3 than those in It is noteworthy that the detection rate of IL-1a and term labor (median = 10 ng/ml, range = 215 pg/ml to IL-ll3 in the AF of women with microbial invasion of the 18 ng/ml vs. median = 201 pg/ml, range = 50 pg/ml to amniotic cavity and labor was related to gestational age 5 ng/ml; P < 0.0001, Mann-Whitney U test). Similarly, and membrane status. Women with preterm labor and women in preterm labor with microbial invasion of the intact membranes had a higher rate of detectable IL-1a amniotic cavity had a significantly higher median IL-1a and IL-ll3 than women with preterm PROM in labor or concentration than those at term with positive AF cul- women at term in active labor (IL-1a, 86.6% [13/15], 82.6% ture (median = 3.5 ng/ml, range-s 157 pg/ml to 8.9 ng/ml [19/23], and 44.4% [4/9], respectively; P < 0.05); IL-ll3, vs. median = 0, range ss 157 pg/ml to 3.592 ng/ml, re- 100% [15/15], 100% [23/23], and 87.5% [8/9], respectively; spectively; P < 0.001, Mann-Whitney U test). P < 0.05). The explanation for these findings is not readily apparent. We explored the possibility that differences DISCUSSION in qualitative microbiology might be responsible for these Our data indicate that the AF from women in the sec- observations. However, the proportion of patients with ond trimester of pregnancy (16 to 18 wk) does not con- polymicrobial invasion of the amniotic cavity, positive tain detectable immunoreactive IL-1a or IL-ll3. This cultures limited to Mycoplasmas or with other microorfinding is consistent with previous data generated in our ganisms, was not different among these three groups. laboratory with the use ofthe D10.G4.1 bioassay.3,6 Thus, Although it is possible that the inoculum size might be unlike other cytokines that are normal constituents of different among groups, we were unable to test this AF (i.e., interleukin-G'V'" and macrophage-colony stim- hypothesis because the majority of patients did not have > 4000
----
.
-.
122
ROMERO ET AL.
quantitative microbiology. Another possible explanation is that the detection rate ofIL-la and IL-ll3 in the AF is dependent upon the pathway of microbial invasion and the time allowed for microbial-host interaction before amniocentesis. 17 The precise origin of AF IL-la and IL-ll3 is unknown. We have previously demonstrated that the cell pellet of women with microbial invasion of the amniotic cavity contains high levels ofbioactive IL-la and IL-ll3, and this suggests an intra-amniotic source for these cytokines (mainly resident macrophages for IL-ll3 and keratinocytes for IL-l a). 3,6 Another potential source for AF IL-l are cells in the chorio-decidual interface." IL-l produced by this tissue could cross the chorioamniotic membranes and gain access to the amniotic cavity. The virulence of the microorganism and the duration of the microbial invasion of the amniotic cavity may determine the extent of IL-l release by intra-amniotic cells. Patients in whom microbial invasion of the amniotic cavity has been of short duration are likely to have low levels of these cytokines in AF. This is likely to be the case in patients with preterm PROM and term labor. On the other hand, an ascending infection of long duration or with a significant decidual stage may lead to an outpouring ofIL-l into the amniotic cavity. This may be the course of events in the setting of preterm labor with intact membranes and a positive AF culture for microorganisms. Another major finding of this study is that term spontaneous parturition in the absence of demonstrable microbial invasion of the amniotic cavity is associated with detectable Il-Lo and IL-ll3 in the AF (Figs. 5 and 6). These data confirm previous observations made with the use of the DIO.G4.1 bioassay for IL_1 3,6 and is in keeping with similar findings in AF IL_614.15 and tumor necrosis factor (TNF).18Release ofIL-l into theAF during labor may be due to tissue disruption/damage that occurs during parturition. Indeed, labor has been likened to an inflammatory phenomenon.l" and the accumulation of cytokines into the AF may reflect this process. However, we cannot exclude that microbial invasion of the amniotic cavity that escaped detection with microbiological techniques used in our study may account for these findings. The group of patients with PROM provided us an opportunity to dissect the relative contribution of parturition and microbial invasion ofthe amniotic cavity to changes in AF concentrations oflL-la and IL-ll3. In these patients, preterm labor in the absence of microbial invasion of the amniotic cavity was not associated with a significant increase in the detection rate and concentrations oflL-la and IL-ll3 in the AF. On the other hand, microbial invasion of the amniotic cavity in the absence of labor was associated with an increase in the detection rate oflL-ll3 in the AF. The lack of a demonstrable increase in IL-la in this setting may be attributable to the preferential secretion ofIL-ll3 rather than IL-la during the course of macrophage activation. Finally, the combination of preterm labor and microbial invasion was associated with a dramatic elevation in the AF concentrations ofIL-la and IL-ll3. Similar observations have been previously reported in studies ofTNF18 and IL_615 in preterm PROM. Collectively, these data indicate that preterm parturition in the setting of infection is temporally associated with the secretion ofIL-l into the amniotic cavity and support the view that the initiation oflabor may be viewed
as part of the repertoire of host-defense mechanisms elicited in response to intrauterine infection. IL-ll3 was the main form of IL-l detected in the AF in women with intra-amniotic infection. A similar finding has been noted in plasma, cerebrospinal fluid, and synovial fluid of patients with shock, meningitis, or chronic inflammatory joint diseases. 1,2,20-24 However, a novel observation in our study is that IL-la was detected in very high concentrations in the AF of women with intra-amniotic infection. Indeed, the mean IL-ll3:IL-la ratio was 2.3:1 for women with preterm labor and intact membranes and 1.9:1 for women with preterm PROM. These ratios are different from those observed after lipopolysaccharide stimulation of macrophages in vitro and those noted in other biological fluids. The most likely explanation for these findings is that during the course of intrauterine infection, microbial products stimulate not only resident macrophages in the decidua and/or amniotic cavity to produce mainly IL-l~ but also fetal keratinocytes that produce mainly IL-la. 5 Thus, the amniotic fluid IL-ll3: IL-la ratio may reflect the contribution of different maternal and fetal cells to the host-response to infection. Further studies are required to address the specific sources, metabolism, and biological properties of AF IL-l. The detection oflL-l in the AF of women with preterm labor and intra-amniotic infection has considerable biological and clinical implications. First, it supports the notion that cytokines may serve as signals for the initiation or maintenance of parturition in the setting ofinfection. Second, cytokine determinations in AF may have a diagnostic and prognostic value in the identification of patients with intra-amniotic infection or with an intrauterine inflammatory process that will render them refractory to tocolysis. Thirdly, it opens new horizons for novel therapeutic approaches for the treatment of preterm labor. Recently, a naturally occurring receRtor antagonist to IL-l has been purified and cloned. 26, 7 This polypeptide can block IL-l-induced prostaglandin E 2 and collagenase production by synovial cells28 and also can reduce the mortality from endotoxic shock. 29,30 Therefore, the IL-l receptor antagonist or pharmacological agents capable of blocking activation of selective biological activities of the cytokine network (prostaglandin production, myometrial contractility) may play an important role in the treatment of preterm labor associated with inflammation. NOTE ADDED IN PROOFS
Subsequently to the acceptance of this manuscript, we have found that 1) the natural IL-l receptor antagonist is normally present in human amniotic fluid;31 2) human recombinant IL-l receptor antagonist can reduce IL-linduced rrostaglandin production by human amnion and chorion; 1 and 3) pre-treatment with the human natural IL-l receptor antaEmist prevents IL-l-induced-preterm parturition in mice. ACKNOWLEDGMENTS
Supported by a grant from the Walter Scott Foundation for Medical Research. Dr. Romero is the recipient of a Physician Scientist Award from the National Institute of Child Health and Human Development. Dr. Sepulveda is supported by the Fogarty International Center, National
INTERLEUKIN-1 IN PRETERM AND TERM PARTURITION
Institutes of Health. Dr. Dinarello is supported by National Institutes of Health Grant AI 15614. REFERENCES 1. Dinarello CA. Clinical relevance of interleu kin-1 and its multiple biological activities. Bull Inst Pasteur. 1987, 85:267-285. 2. Dinarello CA. Interleukin-1 and its biologically related cytokines. Adv Immunol. 1989; 44:153-205. 3. Romero R, Brody DT, Oyarzun E, Mazor M, Wu YK, Hobbins JC, Durum SK. Infection and labor. III. Interleukin-1: A signal for the onset of parturition. Am J Obstet Gynecol. 1989; 160:1117-1123. 4. Romero R, Durum S, Dinarello CA, Oyarzun E, HobbinsJC, Mitchell MD. Interleukin-1 stimulates prostaglandin biosynthesis by human amnion. Prostaglandins. 1989; 37:13-22. 5. Romero R, Wu YK, Brody DT, Oyarzun E, Duff GW, Durum SK. Human decidua: A source of interleu kin-L Obstet Gynecol. 1989; 73:31-34. 6. Romero R, Parvizi ST, Oyarzun E, Mazor M, Wu YK, Avila C, Athanassiadis AP, Mitchell MD. Amniotic fluid interleukin-1 in spontaneous labor at term. J Reprod Med. 1990; 35:235-238. 7. Mitchell MD, Edwin S, Romero R. Prostaglandin biosynthesis by human decidual cells: Effects of inflammatory mediators. Prostaglandins Leukot Essent Fatty Acids. 1990; 41:35-38. 8. Main EK, Strizki J, Schochet P. Placental production ofimmunoregulatory factors: Trophoblast is a source ofinterleukin-1. In: Miller RK, Thiede HA, eds. Trophoblast Research, Vol 2. New York: Plenum, 1987: 149-160. 9. Romero R, Mazor M, Tartakovsky B. Systemic administration of interleukin-1 induces preterm parturition in mice. Am J Obstet Gynecol. 1991; 165:969-971. 10. Schutte MF, Treffers PE, Kloosterman GJ, Soepatmi S. Management of premature rupture of membranes: The risk of vaginal examination to the infant. Am J Obstet Gynecol. 1983; 146:395-400. 11. Caritis SN. A pharmacologic approach to the infusion ofritodrine. Am J Obstet Gynecol. 1988; 158:380-384. 12. Romero R, ScharfK, Mazor M, Emamian M, Hobbins JC, Ryan JL. The clinical value of gas-liquid chromatography in the detection of intra-amniotic microbial invasion. ObstetGynecol.1988; 72:44-50. 13. Lonnemann G, Endres S, van der Meer JW, Cannon JG, Dinarello CA. A radioimmunoasay for human interleukin-1o.. Measurement ofIL-1o. produced in vitro by human blood mononuclear cells stimulated with endotoxin. Lymphokine Res. 1988; 7:75-84. 14. Romero R, Avila C, Santhanam U, Sehgal PB. Amniotic fluid interleukin-6 in preterm labor. Association with infection. J Clin Invest. 1990; 85:1392-1400. 15. Santhanam U, Avila C, Romero R, Viguet H, Ida N, Sakurai S, Seghal PB. Cytokines in normal and abnormal parturition: Elevated amniotic fluid interleukin-6 levels in women with premature rupture of membranes associated with intrauterine infection. Cytokine. 1991; 3:155-163. 16. Romero R, Oyarzun E, Stanley ER. Macrophage colony-stimulating factor in amniotic fluid. Abstract #439 presented at 36th Annual Meeting of the Society for Gynecologic Investigation, San Diego, 1989. 17. Romero R, Mazor M. Infection and preterm labor. Clin Obstet Gynecol. 1988; 31:553-584.
123
18. Romero R, Manogue KR, Mitchell MD, Wu YK, Oyarzun E, Hobbins JC, Cerami A. Infection and labor. IV. Cachectin-tumor necrosis factor in the amniotic fluid of women with intraamniotic infection and preterm labor. Am J Obstet Gynecol. 1989; 161:336-341. 19. Liggins GC. Cervical ripening as an inflammatory reaction. In: Ellwood DA, Anderson ABM, Embrey Mp, eds. The Cervix in Pregnancy and Labour: Clinical and Biochemical Investigations. Edinburgh: Churchill-Livingston, 1989:1-9. 20. Dinarello CA. The proinflammatory cytokines interleukin-1 and tumor necrosis factor and treatment of the septic shock syndrome. J Infect Dis. 1991; 163:1177-1184. 21. Nouri AME, Panayi GS, Goodman SM. Cytokines and the chronic inflammation of rheumatic disease. I. The presence of interleukin-1 in synovial fluids. Clin Exp Immunol. 1984; 55:295-302. 22. Hopkins SJ, Humphreys M, Jayson MIV.Cytokines in synovial fluid. I. The presence of biologically active and immunoreactive IL-1. Clin Exp Immunol. 1988; 72:422-427. 23. Cannon JG, Tompkins RG, Gelfand JA, Michie HR, Stanford GG, van der Meer JWM, Endress S, Lonnemann G, Corsetti S, Chernow B, Wilmore DW, Wolff SM, Burke JF, Dinarello CA. Circulating interleukin-1 and tumor necrosis factor in septic shock and experimental endotoxin fever. J Infect Dis. 1990; 161:79-84. 24. Ramilo 0, Saez-Llorens X, Mertsola J, Jafari H, Olsen KD, Hansen EJ, Yoshinaga M, Ohkawara S, Nariuchi H, McCracken GH. Tumor necrosis factor-o./cachectin and interleukin 113 initiate meningeal inflammation. J Exp Med. 1990; 172:497-507. 25. Luger TA, Stadler BM, Luger BM, Mathieson BJ, Mage M, Schmidt JA, Oppenheim JJ. Murine epidermal cell-derived thymocyte activating factor resembles murine interleukin 1. J Immunol. 1982; 128: 2147-2152. 26. Eisenberg SP, Evans RJ, Arend WP, Verderber E, Brewer MT, Hannum CH, Thompson RCW. Primary structure and functional expression from complementary DNA of a human interleukin-1 receptor antagonist. Nature. 1990; 343:341-346. 27. Carter DB, Deibel MR, Dunn CJ,'Ibmich CSC, Laborde AL, Slightom JL. Berger AE, Bienkowski MJ, Sun FF, McEwan RN, Harris PKW, Yem AW, Waszak GA, Chosay JG, Sieu LC, Hardee MM, ZurcherNeely HA, Reardon 1M, Heinrikson RL, Truesdell SE, Shelly JA, Eessalu TE, Taylor BM, Tracey DE. Purification, cloning, expression and biological characterization of an interleukin-1 receptor antagonist protein. Nature. 1990; 344:633-638. 28. Arend WP, Welgus HG, Thompson RC, Eisenberg SP' Biological properties of recombinant human monocyte-derived interleukin 1 receptor antagonist. J Clin Invest. 1990; 85:1694-1697. 29. Ohlsson K, Bjork P, Bergenfeldt M, Hageman R, Thompson RC. Interleukin-1 receptor antagonist reduces mortality from endotoxin shock. Nature. 1990; 348:550-552. 30. Eisenberg SP, Brewer MT, Verderber P, Heimdal P, Brandhuber BJ, Thompson RC. Interleukin 1 receptor antagonist is a member of the interleukin 1 gene family: Evolution of a cytokine control mechanism. Proc Natl Acad Sci USA. 1991; 88:5232-5236. 31. Romero R, Sepulveda W, Mazor M, Brandt F, Dinarello CA, Mitchell MD. The natural interleukin-1 receptor antagonist in human amniotic fluid in term and preterm parturition. Am J Obstet Gynecol. 1992; in press. 32. Romero R, Tartakovsky B. The natural interleukin-1 receptor antagonist prevents interleukin-1-induced preterm delivery in mice. Am J Obstet Gynecol. 1992, in press.
AJ"'Jnanc~ Immunolo2ll
Interleukin-1 ex and Interleukin-1 r3 in Preterm and Term Human Parturition ROBERTO ROMERO, MOSHE MAZOR, FRANCISCO BRANDT, WALDO SEPULVEDA, CECILIA AVILA, DAVIDB. COTTON, AND CHARLES A. DINARELLO Department of Obstetrics and Gynecology, Yale University School of Medicine, New Haven, Connecticut (R.R., M.M., F.B., WS., C.A.J; Department of Obstetrics and Gynecology, Wayne State University School of Medicine, Detroit, Michigan (RR., F.B., WS., D.B.C.); Department ofMedicine, Tufts University School ofMedicine, Boston, Massachusetts (CAD.); and the Intramural Research Branch ofPerinatology, National Institutes ofHealth, Bethesda, Maryland (RR., WS.) ABSTRACT: Interleukin-l (1L-1)has been implicated in the mechanism of human parturition in the setting of infection. The purpose of this study was to determine the effect oflabor (term and preterm) and microbial invasion of the amniotic cavity on amniotic fluid (AF) concentrations IL-la and IL-II3. AF was retrieved by transabdominal amniocentesis from the following groups of women: midtrimester genetic amniocentesis (16 to 18 wk) (N = 15), preterm labor with intact membranes (21 to 36 wk) with or without infection (N = 72), preterm premature rupture of membranes (PROM) (N = 88), and term not in labor or in active labor with or without infection (N = 58). AF was cultured for aerobic and anaerobic bacteria as well as Mycoplasmas. IL-l was measured with a commercially available immunoassay validated for AF (sensitivity: IL-la, 157 pg/ml; 11...-113, 50 pg/ml). All women at midtrimester had undetectable AF IL-la and IL-II3. Among women in preterm labor with positive AF cultures, II...-la and IL-113 were detectable in the AF in 86.6% (13/15) and 100% (15/15), respectively. In contrast, all women with negative AF cultures without labor (N = 36) had undetectable AF IL-la concentrations and 52.7% (19/36) had undetectable AF IL-113 concentrations. Histopathological chorioamnionitis was present in 92.8% (13/14) of patients who had positive AF cultures and detectable IL-l in the AF. IL-l was significantly higher in patients with preterm PROM, labor, and positive AF cultures than in the other subgroups of patients with preterm PROM. Among women at term who were not in labor, 26.3% (5/19) had detectable IL-la and 19.8% (3/19) had detectable IL-113 in the AF. Among those in active labor with positive AF cultures, 44.4% (4/9) had detectable IL-la and 88.8% (8/9) had detectable 11...-113 in the AF. IL-113 was detected more frequently in patients at term with positive AF cultures than in those with negative AF cultures (88.8% [8/9] vs, 38.7% [19/49]; P = 0.0088). We conclude that AF IL-l concentrations are elevated in women with preterm labor an microbial invasion of the amniotic cavity and in women with spontaneous parturition at term. (Am J Reprod Immunol. 1992; 27:117-123.) Key words: Interleukin-l, amniotic fluid, parturition, preterm labor, premature rupture of membranes, prematurity, intra-amniotic infection. INTRODUCTION
Interleukin-I (IL-l) is a term used to describe two polypeptides (lL-la and IL-II3>. which have a wide spectrum Submitted for publication December 24,1991; accepted February 24,1992. Address reprint requests to Roberto Romero, M.D., Wayne State University, School of Medicine, Department of Obstetrics and Gynecology, 4707 St. Antoine, Detroit, MI48201.
of immunological and nonimmunological activities (reviewed in detail in Dinarello'r'), These cytokines have similar molecular weight (approximately 17 kD), but have different isoelectric points (5 for IL-la and 7 for IL-II3). IL-la and IL-113 are the product of two distinct genes, which appear to have evolved from a common ancestor gene. Although these cytokines share less than 30% structural homology, they bind the same cell surface receptors and have similar biological properties. 1,2 Interleukin-l has been implicated as a signal for the onset of parturition in the setting of infection.P:" Indeed, IL-la and IL-113 are produced by human decidual explants in response to bacterial endotoxin," and these cytokines can stimulate prostaglandin production by amnion" and decidua 7 in a dose-dependent manner. Moreover, administration of IL-l to pregnant mice results in preterm parturition." To explore the possible participation oflL-l in the mechanisms of human parturition in the setting of infection, we have previously assayed IL-l in the amniotic fluid (AF) of women with microbial invasion of the amniotic cavity and preterm labor with the use of the DI0.G4.1 bioassay." This bioassay, like others for cytokines, is time-consuming, laborious, semiquantitative, and is affected by the presence of inhibitors. Using this assay method, neutralizing antibodies are required to estimate the relative contributions of IL-la and IL-113 to the total AF IL-l bioactivity. However, interpretation of this type of assay is sometimes difficult because of the complex interaction between the different species of IL-l, other related cytokines active on this bioassay (i.e., IL-6), and inhibitors. Recently, specific immunoassays for IL-la and IL-113 have become available. The purpose of this study was to determine whether changes in the AF concentrations oflL-la and IL-113 occur in the setting oflabor (term and preterm) and microbial invasion of the amniotic cavity. MATERIALS AND METHODS
Study Design A cross-sectional study was conducted according to the results of AF cultures, gestational age, and labor status. Four groups were identified for study purposes. Group 1 consisted of women in the midtrimester of pregnancy (gestational age of16 to 18 wk) who were-undergoing amniocentesis for genetic indications (N = 15). Group 2 consisted of women admitted with preterm labor and intact membranes (N = 72); these patients were subclassified into three subgroups according to the AF culture results and response to tocolysis: women with negative AF cultures who responded to tocolysis and delivered at term (N = 36); women with negative AF cultures who failed tocolysis and delivered prematurely (N = 21), and women with positive AF cultures who delivered a premature neonate (N = 15). Group 3 consisted of women
118
ROMEROET AL.
with premature rupture of membranes (PROM) (N 88); this group was subdivided into four subgroups according to the presence of labor at the time of amniocentesis and the results of the AF culture: women with negative AF cultures and without preterm labor (N = 25); women with negative AF cultures but with preterm labor (N = 21); women with positive AF cultures but without preterm labor (N = 19), and women with positive AF cultures and preterm labor (N = 23). Group 4 consisted of nonlaboring women at term (38 to 39 wk) who had amniocentesis for the assessment offetallung maturity before elective cesarean section (N = 19) and women in spontaneous active labor at term (N = 39). The latter subgroup was subdivided according to the results of AF cultures: women with negative AF cultures (N = 30) and women with positive AF cultures (N = 9). PROM was defined as amniorrhexis before the onset of spontaneous labor. Nitrazine, pooling, and feming tests were performed in all women in labor to exclude leakage of AF into the vagina. Pelvic examinations were not performed at the time of amniocentesis, as this could result in an increased risk of infection. 10 Tocolytic agents were not used in patients with preterm PROM. Preterm labor was defined as the presence of regular uterine contractions with a frequency of at least two every 10 min associated with changes in the cervical Bishop's score. Ritodrine was administered intravenously as the tocolytic agent to all ~atients, according to a protocol described by Caritis.! Amniocentesis was performed before the initiation of therapy. Failure of tocolysis was diagnosed when cervical dilation progressed beyond 5 em or delivery occurred. Retrieval ofAmniotic Fluid AF was retrieved by transabdominal amniocentesis in all women in this study. Amniocentesis is offered in our institution for biological studies to all patients with the diagnosis of preterm labor. Women in active labor at term were ascertained by searching our obstetrical database. They were identified as patients who had undergone amniocentesis because of suspected preterm labor but subsequently delivered (within 12 h of amniocentesis) a newborn who was considered term by pediatric examination. The identification number of patients fulfilling this criterion was cross-referenced with samples stored in our AF bank. A sample of AF was sent for microbiological studies. The remaining fluid was centrifuged at 200g for 10 min at 4°C and stored in polypropylene tubes at - 70°C until assayed. Microbiologic Culture Technique AF was transported to the laboratory in a capped plastic syringe immediately after collection. These conditions precluded air contact with the specimen. Plating occurred within 30 min of collection in all cases. AF was cultured for aerobic and anaerobic bacteria as well as Mycoplasmas using methods previously described. 12 lL-1 a and lL-113 Determinations IL-la was measured with a radioimmunoassay previously described 13 and validated for human AF. The sensitivity of the assay was 157 pg/ml, the intra-assay coefficient of variability was 8.4%, and the interassay coefficient of variation was 12%. IL-l13 was determined with a commercially available immunoradiometric assay
9000
8000 7000
6000 IL-1u (pg/ml)
5000
4000
3000
2000 1000
No Labor No Infection
(n = 36) A
~ Labor No Infection (n = 21)
Labor Infection (n = 15)
B
C
Fig. 1. Amniotic fluid IL-l", concentrations in women with preterm labor and intact membranes. IL-l", was detected more frequently in women with positive amniotic fluid cultures than in women with negative amniotic fluid cultures in the other two groups (group A, median and range = 0; group B, median = 0, range c 157 pg/ml to 0.5 ng/ml; group C, median = 3.5 ng/ml, range '" 157 pg/ml to 8.9 ng/ml; Kruskal-Wallis H = 45.49; P < 0.00001).
validated for AF (Medgenix, Belgium). The sensitivity of the assay was 50 pg/ml, and the intra- and interassay variabilities were 7.1 % and 10.5%, respectively. Statistical Analysis Comparisons between proportions were performed with either a Chi-square or a Fisher's exact test. An unpaired Student's t-test was used to compare continuous variables, which were normally distributed (gestational age, maternal age, etc.) A Wilk-Shapiro test was employed to test the hypothesis that AF IL-l concentrations were normally distributed. A Kruskal-Wallis ANOVA was used to compare AF IL-l concentrations among study groups. Multiple comparisons among groups were conducted with Dunn's test (True Epistat, Epistat Services, Richardson, TX). RESULTS
AF from women in the midtrimester of pregnancy did not contain detectable IL-la or IL-l13. Figures 1 and 2 display AF concentrations ofIL-la and IL-l13 in women
INTERLEUKIN-l IN PRETERM AND TERM PARTURITION
-
1
1000 >
species (N = 4) and Ureaplasma urealyticum (N = 4). Polymicrobial invasion was present in 26.6% (4/15) of patients. None of the patients had clinical chorioamnionitis. One newborn delivered at 26 wk had congenital sepsis. Histopathological chorioamnionitis was detectable in 13 of the 14 (92.8%) patients with positive AF cultures in whom placental examination was performed. All these patients had detectab'.e IL-l[3 in AF. Figures 3 and 4 display AF concentrations oflL-la and IL-l[3 in women with preterm PROM. Women with preterm labor and positive AF cultures had significantly higher AF IL-l concentrations (IL-la, median = 0.62 ng/ml, range ~ 157 pg/ml to 10 ng/ml; IL-l[3, median = 1.085 ng/ml, range = 145 pg/ml to 10.6 ng/ml) than women in the other subgroups (P < 0.00001). Overall, IIAa and IL-1[3 were detected in 82.6% (19/23) and 100% (23/23) of patients with positive AF cultures, respectively, and in only 21% (4/19) and 63.1 % (12/19) ofthose women with positive AF cultures but without labor, respectively. Among women with negative AF cultures, IL-1a and IL-1[3 were detected in 10.8% (5/46) and 30.4% (14/46), respectively.
.
5000 4000 3000 2000
IL-1~
(pg/ml)
119
1000 900 800 700 600 500
20000
400 300 200 100
50 - - - -
0
.t. - -
-~-
- - - - - - - -
No Labor No Infection (n = 36)
A
... 1
10000 - -
----- -- ---- ----
Labor No Infection (n = 21)
Labor Infection
B
C
.
(n = 15)
Fig. 2. Amniotic fluid IL-113 concentrations in women with preterm labor and intact membranes. IL-113 was detected more frequently in women with positive amniotic fluid cultures than in women with negative amniotic fluid cultures in the other two groups (group A, median = 0, range ,,;; 50 mg to 0.206 ng/ml; group B, median = 0, range-s 157 pg/ml to 0.648 ng/ml; group C, median = 10 ng/ml, range = 215 pg/ml to 18 ng/ml; Kruskal-Wallis H = 37.39;P < 0.000001).
in preterm labor with intact membranes. IL-la was detected in 86.6% (13/15) of patients with positive AF cultures for microorganisms (median = 3.5 ng/ml, range ~ 157 pg/ml to 8.9 ng/ml) and IL-l[3 in all cases (15/15) (median = 10 ng/ml, range ~ 50 pg/ml to 18 ng/ml). On the other hand, only 19% (4/21) of women in preterm labor who had negative AF cultures for microorganisms and who delivered prematurely had detectable IL-la and 47.6% (10/21) had detectable IL-l[3 in the AF. Both IL-la and IL-l[3 concentrations were significantly higher in women with microbial invasion of the amniotic cavity than in women who delivered prematurely but had negative AF cultures (IL-la, 86.6% [13/15] vs. 19% [4/21], P = 0.0001; IL-l[3, 100% [15/15] vs. 47.6% [10/21], P = 0.0006; Fisher's exact test). Among women with negative AF cultures who responded to tocolysis, none (0/36) had detectable IL-la and 47.2% (17/36) had detectable IL-l [3 in the AF. The most common microbial isolates in patients with preterm labor and intact membranes were Fusobacterium
3500
IL-1a
3000
(pg/ml)
2500 2000 1500 1000
----..
500
157 -~---~---~------.:.-;,---
o
No Labor Labor No Labor No Infection No Infection Jnfection (n=25) (n=21) (n=19) ABC
La'bor Infection (n=23) 0
Fig. 3. Amniotic fluid IL-la concentrations in women with preterm PROM. IL-la was detected more frequently in women with positive amniotic fluid cultures and preterm labor than in women in the other three groups (group A, median = 0, range s 157 pg/ml to 0.7 ng/ml; group B, median = 0, range « 157 pg/ml to 0.3 ng/ml; group C, median = 0, range ss 157 pg/ml to 2 ng/ml; group D, median = 0.62 ng/ml, range ,,;; 157 pg/ml to 10 ng/ml; Kruskal-Wallis H = 37.18; P < 0.000001).
120
ROMERO ET AL.
"°1
with negative AF cultures (88.8% [8/9] vs. 38.7% [19/49]; P = 0.0088, Fisher's exact test). In contrast, the rate of AF IL-1a detection was not different between patients with positive and negative AF cultures (44.4% [4/9] vs. 32.6% [16/49]; P = 0.7, Fisher's exact test). Among women in labor, those with positive AF cultures had a significantly higher median AF IL-113 concentration than those with negative AF cultures (median = 201 pg/ml, range = 115 pg/ml to 5 ng/ml vs. median = 51.5 pg/ml, range = 51 pg/ml to 2.821 ng/ml, respectively; P = 0.0137, Mann-Whitney U test). Also, IL-113 was detected more frequently in the AF of women in labor with a negative AF than in those not in labor (61.5% [24/39] vs. 15.8% [3119], respectively;P = 0.0017, Fisher's exact test). The most common microorganisms isolated from the nine women in active labor at term were Ureaplasma urealyticum (N = 4; 44.4%) followed by Lactobacillus species, Streptococcus agalactiae, and Mycoplasma hominis (N = 2; 22.2% each). Polymicrobial invasion of the amniotic cavity was present in only two patients. There was not significant difference in cervical dilation between patients with detectable and those with nondetectable AF IL-1a or IL-113.
6000
5000 4000 3000 2000 IL-1~
(pg/ml)
1750
1250 1000 750 500
-.-.
250 - - - - - -
.......
4000 - - - - - - - - - - - -
No Labor Labor No Labor No Infection No Infection Infection (n = 25) (n = 21) (n = 19) ABC
"
3000
Labor Infection (n = 23) 0
Fig. 4. Amniotic fluid IL-113 concentrations in women with preterm PROM. IL-l 13 was detected more frequently in women with positive amniotic fluid cultures and preterm labor than in women in the other three groups (group A, median = 0, range c 50 pg/ml to 0.442 ng/ml; group B, median = 0, range ss 50 pg/ml to 0.947 ng/ml; group C, median = 0.28 ng/ml, range «; 50 pg/ml to 4.7 ng/ml; group D, median = 1.085 ng/ml, range = 146 pg/ml to 10.642 ng/ml; Kruskal-Wallis H = 45.11; P < 0.00001).
2000
900 800 IL-1a (pg/ml)
700 600
The most common microbial isolates in the AF in patients with preterm PROM were Ureaplasma urealyticum (N = 20), Mycoplasma hominis (N = 10), and Bacteroides species (N = 4). Polymicrobial infections were detected in 30.9% (13/42). No significant differences in the proportion of women with microbial invasion due exclusively to Mycoplasmas or polymicrobial infections could be detected between the groups with labor or without labor. Of patients with positive AF cultures and spontaneous onset of labor, placental examination was performed in 16 cases. Histological chorioamnionitis was detected in all placentae. There were no cases of neonatal sepsis in this subgroup of patients. The interval (in hours) between rupture of membranes and amniocentesis was not significantly different among the four subgroups of patients with preterm PROM (medians for intervals in subgroups A, B, C, and D showing in Figures 3 and 4 were 7.5 h, 6.5 h, 11 h, and 11 h, respectively). Figures 5 and 6 display AF concentrations ofIL-1a and IL-113 in women at term. IL-113 was detected more frequently in patients with positive AF cultures than in those
500 400 300 200 157
.
_________________ _l
e
100 ----IIINII--
--liliiii&--
No Labor No Infection
(n =19)
Labor No Infection (n = 30)
•
Labor Infection (n = 9)
A
B
C
Fig. 5. Amniotic fluid IL-la concentrations in women at term (group A, median = 0, range « 157 pg/ml to 0.449 ng/ml; group B, median = 0, range-s 157 pg/ml to 0.627 ng/ml; group C, median = 0, range s 157 pg/ml to 3.592 ng/ml; Kruskal-Wallis H = 1.08;P = 0.58).
.
INTERLEUKIN-l IN PRETERM AND TERM PARTURITION
121
ulating factor 16), IL-1a and IL-ll3 do not seem to be normally present in second trimester AF. Immunoreactive IL-a and IL-ll3 were not detected in most samples of AF obtained from women at term not 3000 in labor. Patients with detectable AF IL-1a (26.3% [5/19]) and IL-ll3 (15.8% [3/19]) had low levels of these cytokines and no clinical or microbiological evidence of 2000 infection (see Figs. 5 and 6). A possible explanation for the detection ofIL-la and IL-ll3 in the AF of these otherwise normal women is the release of these cytokines 1000 by cells normally present in the amniotic cavity during the process of spontaneous cell death/damage. Indeed, we have demonstrated that the cell pellet of term amniIL-1 ~ otic fluid (mainly formed by desquamated keratinocytes) (pg/ml) contains high levels of bioactive IL-1. 3 An alternative 500 explanation to our findings is the existence of an inflammatory process of infectious or noninfectious etiology. A major finding of this study is that microbial inva400 sion of the amniotic cavity is associated with the presence of both IL-1a and IL-ll3 in the AF. Indeed, in each ofthe three populations studied (preterm labor with intact 300 membranes, preterm PROM, and term labor), IL-ll3 was detected more frequently and in higher concentrations in women with a positive AF culture for microorganisms -----r-200 than in those with a negative AF culture. Similar findings were observed for IL-1a in patients with preterm labor (with intact or rupture of membranes). However, 100 we could not document a higher detection rate ofIL-la in the AF of patients with term labor and microbial inva:r.:::: sion of the amniotic cavity. The most likely explanation Labor Labor No Labor for this finding is that microbial invasion of the amniNo Infection Infection No Infection otic cavity occurring during the course of term labor is (n = 9) (n = 30) (n = 19) A B C of short duration and elicits a milder host response than the microbial invasion that occurs in the context of preterm labor with intact membranes. The higher detecFig. 6. Amniotic fluid IL-I(3 concentrations in women at term. IL-l(3 tion rate ofIL-ll3 in comparison with IL-1a in term labor was significantly higher in patients with labor and infection in comprobably reflects the preferential secretion ofIL-ll3 durparison with women in labor but without infection or in nonlaboring ing the course of macrophage activation (lL-ll3:IL-la ratio women (group A, median = 0 pg/ml, range ss 50 pg/ml to 55 pg/ml; group = 9:1). Although the immunoassay for IL-1a used in our B, median = 51.5 pg/ml, range'; 50 pg/ml to 2.821 ng/ml; group C, median = 201 pg/ml, range-s 50 pg/ml to 5.0 ng/ml; Kruskal-Wallis H study was less sensitive than that for IL-ll3 (lL-1a, 157 = 18.79,P = 0.00008). pg/ml; IL-ll3, 50 pg/ml), the data displayed in Figure 5 suggest that this is not a likely explanation for our findings. Indeed, most patients with microbial invasion of Among women with microbial invasion of the amni- the amniotic cavity and detectable IL-ll3 had AF conotic cavity and labor, those in preterm labor had signifi- centrations above 157 pg/ml (66% [6/9]). cantly higher AF concentrations of IL-ll3 than those in It is noteworthy that the detection rate of IL-1a and term labor (median = 10 ng/ml, range = 215 pg/ml to IL-ll3 in the AF of women with microbial invasion of the 18 ng/ml vs. median = 201 pg/ml, range = 50 pg/ml to amniotic cavity and labor was related to gestational age 5 ng/ml; P < 0.0001, Mann-Whitney U test). Similarly, and membrane status. Women with preterm labor and women in preterm labor with microbial invasion of the intact membranes had a higher rate of detectable IL-1a amniotic cavity had a significantly higher median IL-1a and IL-ll3 than women with preterm PROM in labor or concentration than those at term with positive AF cul- women at term in active labor (IL-1a, 86.6% [13/15], 82.6% ture (median = 3.5 ng/ml, range-s 157 pg/ml to 8.9 ng/ml [19/23], and 44.4% [4/9], respectively; P < 0.05); IL-ll3, vs. median = 0, range ss 157 pg/ml to 3.592 ng/ml, re- 100% [15/15], 100% [23/23], and 87.5% [8/9], respectively; spectively; P < 0.001, Mann-Whitney U test). P < 0.05). The explanation for these findings is not readily apparent. We explored the possibility that differences DISCUSSION in qualitative microbiology might be responsible for these Our data indicate that the AF from women in the sec- observations. However, the proportion of patients with ond trimester of pregnancy (16 to 18 wk) does not con- polymicrobial invasion of the amniotic cavity, positive tain detectable immunoreactive IL-1a or IL-ll3. This cultures limited to Mycoplasmas or with other microorfinding is consistent with previous data generated in our ganisms, was not different among these three groups. laboratory with the use ofthe D10.G4.1 bioassay.3,6 Thus, Although it is possible that the inoculum size might be unlike other cytokines that are normal constituents of different among groups, we were unable to test this AF (i.e., interleukin-G'V'" and macrophage-colony stim- hypothesis because the majority of patients did not have > 4000
----
.
-.
122
ROMERO ET AL.
quantitative microbiology. Another possible explanation is that the detection rate ofIL-la and IL-ll3 in the AF is dependent upon the pathway of microbial invasion and the time allowed for microbial-host interaction before amniocentesis. 17 The precise origin of AF IL-la and IL-ll3 is unknown. We have previously demonstrated that the cell pellet of women with microbial invasion of the amniotic cavity contains high levels ofbioactive IL-la and IL-ll3, and this suggests an intra-amniotic source for these cytokines (mainly resident macrophages for IL-ll3 and keratinocytes for IL-l a). 3,6 Another potential source for AF IL-l are cells in the chorio-decidual interface." IL-l produced by this tissue could cross the chorioamniotic membranes and gain access to the amniotic cavity. The virulence of the microorganism and the duration of the microbial invasion of the amniotic cavity may determine the extent of IL-l release by intra-amniotic cells. Patients in whom microbial invasion of the amniotic cavity has been of short duration are likely to have low levels of these cytokines in AF. This is likely to be the case in patients with preterm PROM and term labor. On the other hand, an ascending infection of long duration or with a significant decidual stage may lead to an outpouring ofIL-l into the amniotic cavity. This may be the course of events in the setting of preterm labor with intact membranes and a positive AF culture for microorganisms. Another major finding of this study is that term spontaneous parturition in the absence of demonstrable microbial invasion of the amniotic cavity is associated with detectable Il-Lo and IL-ll3 in the AF (Figs. 5 and 6). These data confirm previous observations made with the use of the DIO.G4.1 bioassay for IL_1 3,6 and is in keeping with similar findings in AF IL_614.15 and tumor necrosis factor (TNF).18Release ofIL-l into theAF during labor may be due to tissue disruption/damage that occurs during parturition. Indeed, labor has been likened to an inflammatory phenomenon.l" and the accumulation of cytokines into the AF may reflect this process. However, we cannot exclude that microbial invasion of the amniotic cavity that escaped detection with microbiological techniques used in our study may account for these findings. The group of patients with PROM provided us an opportunity to dissect the relative contribution of parturition and microbial invasion ofthe amniotic cavity to changes in AF concentrations oflL-la and IL-ll3. In these patients, preterm labor in the absence of microbial invasion of the amniotic cavity was not associated with a significant increase in the detection rate and concentrations oflL-la and IL-ll3 in the AF. On the other hand, microbial invasion of the amniotic cavity in the absence of labor was associated with an increase in the detection rate oflL-ll3 in the AF. The lack of a demonstrable increase in IL-la in this setting may be attributable to the preferential secretion ofIL-ll3 rather than IL-la during the course of macrophage activation. Finally, the combination of preterm labor and microbial invasion was associated with a dramatic elevation in the AF concentrations ofIL-la and IL-ll3. Similar observations have been previously reported in studies ofTNF18 and IL_615 in preterm PROM. Collectively, these data indicate that preterm parturition in the setting of infection is temporally associated with the secretion ofIL-l into the amniotic cavity and support the view that the initiation oflabor may be viewed
as part of the repertoire of host-defense mechanisms elicited in response to intrauterine infection. IL-ll3 was the main form of IL-l detected in the AF in women with intra-amniotic infection. A similar finding has been noted in plasma, cerebrospinal fluid, and synovial fluid of patients with shock, meningitis, or chronic inflammatory joint diseases. 1,2,20-24 However, a novel observation in our study is that IL-la was detected in very high concentrations in the AF of women with intra-amniotic infection. Indeed, the mean IL-ll3:IL-la ratio was 2.3:1 for women with preterm labor and intact membranes and 1.9:1 for women with preterm PROM. These ratios are different from those observed after lipopolysaccharide stimulation of macrophages in vitro and those noted in other biological fluids. The most likely explanation for these findings is that during the course of intrauterine infection, microbial products stimulate not only resident macrophages in the decidua and/or amniotic cavity to produce mainly IL-l~ but also fetal keratinocytes that produce mainly IL-la. 5 Thus, the amniotic fluid IL-ll3: IL-la ratio may reflect the contribution of different maternal and fetal cells to the host-response to infection. Further studies are required to address the specific sources, metabolism, and biological properties of AF IL-l. The detection oflL-l in the AF of women with preterm labor and intra-amniotic infection has considerable biological and clinical implications. First, it supports the notion that cytokines may serve as signals for the initiation or maintenance of parturition in the setting ofinfection. Second, cytokine determinations in AF may have a diagnostic and prognostic value in the identification of patients with intra-amniotic infection or with an intrauterine inflammatory process that will render them refractory to tocolysis. Thirdly, it opens new horizons for novel therapeutic approaches for the treatment of preterm labor. Recently, a naturally occurring receRtor antagonist to IL-l has been purified and cloned. 26, 7 This polypeptide can block IL-l-induced prostaglandin E 2 and collagenase production by synovial cells28 and also can reduce the mortality from endotoxic shock. 29,30 Therefore, the IL-l receptor antagonist or pharmacological agents capable of blocking activation of selective biological activities of the cytokine network (prostaglandin production, myometrial contractility) may play an important role in the treatment of preterm labor associated with inflammation. NOTE ADDED IN PROOFS
Subsequently to the acceptance of this manuscript, we have found that 1) the natural IL-l receptor antagonist is normally present in human amniotic fluid;31 2) human recombinant IL-l receptor antagonist can reduce IL-linduced rrostaglandin production by human amnion and chorion; 1 and 3) pre-treatment with the human natural IL-l receptor antaEmist prevents IL-l-induced-preterm parturition in mice. ACKNOWLEDGMENTS
Supported by a grant from the Walter Scott Foundation for Medical Research. Dr. Romero is the recipient of a Physician Scientist Award from the National Institute of Child Health and Human Development. Dr. Sepulveda is supported by the Fogarty International Center, National
INTERLEUKIN-1 IN PRETERM AND TERM PARTURITION
Institutes of Health. Dr. Dinarello is supported by National Institutes of Health Grant AI 15614. REFERENCES 1. Dinarello CA. Clinical relevance of interleu kin-1 and its multiple biological activities. Bull Inst Pasteur. 1987, 85:267-285. 2. Dinarello CA. Interleukin-1 and its biologically related cytokines. Adv Immunol. 1989; 44:153-205. 3. Romero R, Brody DT, Oyarzun E, Mazor M, Wu YK, Hobbins JC, Durum SK. Infection and labor. III. Interleukin-1: A signal for the onset of parturition. Am J Obstet Gynecol. 1989; 160:1117-1123. 4. Romero R, Durum S, Dinarello CA, Oyarzun E, HobbinsJC, Mitchell MD. Interleukin-1 stimulates prostaglandin biosynthesis by human amnion. Prostaglandins. 1989; 37:13-22. 5. Romero R, Wu YK, Brody DT, Oyarzun E, Duff GW, Durum SK. Human decidua: A source of interleu kin-L Obstet Gynecol. 1989; 73:31-34. 6. Romero R, Parvizi ST, Oyarzun E, Mazor M, Wu YK, Avila C, Athanassiadis AP, Mitchell MD. Amniotic fluid interleukin-1 in spontaneous labor at term. J Reprod Med. 1990; 35:235-238. 7. Mitchell MD, Edwin S, Romero R. Prostaglandin biosynthesis by human decidual cells: Effects of inflammatory mediators. Prostaglandins Leukot Essent Fatty Acids. 1990; 41:35-38. 8. Main EK, Strizki J, Schochet P. Placental production ofimmunoregulatory factors: Trophoblast is a source ofinterleukin-1. In: Miller RK, Thiede HA, eds. Trophoblast Research, Vol 2. New York: Plenum, 1987: 149-160. 9. Romero R, Mazor M, Tartakovsky B. Systemic administration of interleukin-1 induces preterm parturition in mice. Am J Obstet Gynecol. 1991; 165:969-971. 10. Schutte MF, Treffers PE, Kloosterman GJ, Soepatmi S. Management of premature rupture of membranes: The risk of vaginal examination to the infant. Am J Obstet Gynecol. 1983; 146:395-400. 11. Caritis SN. A pharmacologic approach to the infusion ofritodrine. Am J Obstet Gynecol. 1988; 158:380-384. 12. Romero R, ScharfK, Mazor M, Emamian M, Hobbins JC, Ryan JL. The clinical value of gas-liquid chromatography in the detection of intra-amniotic microbial invasion. ObstetGynecol.1988; 72:44-50. 13. Lonnemann G, Endres S, van der Meer JW, Cannon JG, Dinarello CA. A radioimmunoasay for human interleukin-1o.. Measurement ofIL-1o. produced in vitro by human blood mononuclear cells stimulated with endotoxin. Lymphokine Res. 1988; 7:75-84. 14. Romero R, Avila C, Santhanam U, Sehgal PB. Amniotic fluid interleukin-6 in preterm labor. Association with infection. J Clin Invest. 1990; 85:1392-1400. 15. Santhanam U, Avila C, Romero R, Viguet H, Ida N, Sakurai S, Seghal PB. Cytokines in normal and abnormal parturition: Elevated amniotic fluid interleukin-6 levels in women with premature rupture of membranes associated with intrauterine infection. Cytokine. 1991; 3:155-163. 16. Romero R, Oyarzun E, Stanley ER. Macrophage colony-stimulating factor in amniotic fluid. Abstract #439 presented at 36th Annual Meeting of the Society for Gynecologic Investigation, San Diego, 1989. 17. Romero R, Mazor M. Infection and preterm labor. Clin Obstet Gynecol. 1988; 31:553-584.
123
18. Romero R, Manogue KR, Mitchell MD, Wu YK, Oyarzun E, Hobbins JC, Cerami A. Infection and labor. IV. Cachectin-tumor necrosis factor in the amniotic fluid of women with intraamniotic infection and preterm labor. Am J Obstet Gynecol. 1989; 161:336-341. 19. Liggins GC. Cervical ripening as an inflammatory reaction. In: Ellwood DA, Anderson ABM, Embrey Mp, eds. The Cervix in Pregnancy and Labour: Clinical and Biochemical Investigations. Edinburgh: Churchill-Livingston, 1989:1-9. 20. Dinarello CA. The proinflammatory cytokines interleukin-1 and tumor necrosis factor and treatment of the septic shock syndrome. J Infect Dis. 1991; 163:1177-1184. 21. Nouri AME, Panayi GS, Goodman SM. Cytokines and the chronic inflammation of rheumatic disease. I. The presence of interleukin-1 in synovial fluids. Clin Exp Immunol. 1984; 55:295-302. 22. Hopkins SJ, Humphreys M, Jayson MIV.Cytokines in synovial fluid. I. The presence of biologically active and immunoreactive IL-1. Clin Exp Immunol. 1988; 72:422-427. 23. Cannon JG, Tompkins RG, Gelfand JA, Michie HR, Stanford GG, van der Meer JWM, Endress S, Lonnemann G, Corsetti S, Chernow B, Wilmore DW, Wolff SM, Burke JF, Dinarello CA. Circulating interleukin-1 and tumor necrosis factor in septic shock and experimental endotoxin fever. J Infect Dis. 1990; 161:79-84. 24. Ramilo 0, Saez-Llorens X, Mertsola J, Jafari H, Olsen KD, Hansen EJ, Yoshinaga M, Ohkawara S, Nariuchi H, McCracken GH. Tumor necrosis factor-o./cachectin and interleukin 113 initiate meningeal inflammation. J Exp Med. 1990; 172:497-507. 25. Luger TA, Stadler BM, Luger BM, Mathieson BJ, Mage M, Schmidt JA, Oppenheim JJ. Murine epidermal cell-derived thymocyte activating factor resembles murine interleukin 1. J Immunol. 1982; 128: 2147-2152. 26. Eisenberg SP, Evans RJ, Arend WP, Verderber E, Brewer MT, Hannum CH, Thompson RCW. Primary structure and functional expression from complementary DNA of a human interleukin-1 receptor antagonist. Nature. 1990; 343:341-346. 27. Carter DB, Deibel MR, Dunn CJ,'Ibmich CSC, Laborde AL, Slightom JL. Berger AE, Bienkowski MJ, Sun FF, McEwan RN, Harris PKW, Yem AW, Waszak GA, Chosay JG, Sieu LC, Hardee MM, ZurcherNeely HA, Reardon 1M, Heinrikson RL, Truesdell SE, Shelly JA, Eessalu TE, Taylor BM, Tracey DE. Purification, cloning, expression and biological characterization of an interleukin-1 receptor antagonist protein. Nature. 1990; 344:633-638. 28. Arend WP, Welgus HG, Thompson RC, Eisenberg SP' Biological properties of recombinant human monocyte-derived interleukin 1 receptor antagonist. J Clin Invest. 1990; 85:1694-1697. 29. Ohlsson K, Bjork P, Bergenfeldt M, Hageman R, Thompson RC. Interleukin-1 receptor antagonist reduces mortality from endotoxin shock. Nature. 1990; 348:550-552. 30. Eisenberg SP, Brewer MT, Verderber P, Heimdal P, Brandhuber BJ, Thompson RC. Interleukin 1 receptor antagonist is a member of the interleukin 1 gene family: Evolution of a cytokine control mechanism. Proc Natl Acad Sci USA. 1991; 88:5232-5236. 31. Romero R, Sepulveda W, Mazor M, Brandt F, Dinarello CA, Mitchell MD. The natural interleukin-1 receptor antagonist in human amniotic fluid in term and preterm parturition. Am J Obstet Gynecol. 1992; in press. 32. Romero R, Tartakovsky B. The natural interleukin-1 receptor antagonist prevents interleukin-1-induced preterm delivery in mice. Am J Obstet Gynecol. 1992, in press.
