Chorioamnionitis
What is chorioamnionitis?
Literally, the term means inflammation in the chorion and the amnion, but it is almost never used in that restricted sense. There are really three definitions of chorioamnionitis. The first is a conceptual one. Chorioamnionitis is an ascending infection of the intrauterine cavity during pregnancy. The concept is that microorganisms, primarily bacteria, ascend from the vagina through the cervix into the intrauterine cavity where they initiate an inflammatory response in the mother and fetus. The second definition is clinical and is usually some description of the clinical consequences of the intrauterine inflammation such as maternal fever, leukocytosis and uterine tenderness. Technology has advanced the ability to detect intrauterine inflammation by using amniotic fluid samples to culture for microorganisms or to directly assay inflammatory mediators and white blood cells. When I use the term clinical chorioamnionitis, I will be referring to women with the conventional findings not to cases defined by amniotic fluid criteria. The third definition can be called histologic chorioamnionitis which is defined as acute inflammatory cells, i.e. neutrophils, in the fetal surface of the placenta, or the amnion and/or chorion of the membranes. The location of the neutrophils can be defined as amniocentric, meaning that the neutrophils have left the maternal blood in a direction toward the amnion cavity, but there is no requirement that they actually reach the amnion. There may be evidence of fetal inflammation in the umbilical cord or in the fetal surface vessels of the placenta. This definition requires looking at microscope slides of samples of the placenta, membranes and umbilical cord. The surprising discovery is that histologic chorioamnionitis is present not just in women with clinical chorioamnionitis but in a large number of women without clinical symptoms of chorioamnionitis[1]. More importantly, women with asymptomatic histologic chorioamnionitis have a similar risk for complications including preterm labor, fetal and maternal infection and inflammatory injury of the infant, as women with clinical chorioamnionitis.
Evaluation of amniotic fluid can prenatally detect more cases of intrauterine inflammation than can the criteria of conventional clinical chorioamnionitis. The goal would be concordance of the three definitions so that all patients with intrauterine inflammation from ascending intrauterine infection could be diagnosed prior to microscopic examination of the placenta. Even if this goal is achieved, histologic chorioamnionitis still has its own insights to reveal. The purpose of this chapter is to explore them. We will start with a little history.
The state of understanding of chorioamnionitis in 1980
In 1980 a CIBA symposium was published on chorioamnionitis. Back then I was just starting my training in pathology. Now sorting through my papers I found my Xeroxed copies with numerous underlines and comments. This was my introduction to the topic, and I think it is also a useful starting point for this chapter because the symposium was the start of a turning point in the understanding of the significance of chorioamnionitis.
In the symposium, Dr. Blanc gives a summary of the microscopy of chorioamnionitis as understood at that time. He is thoughtful and thorough, and covers many of the observations that this chapter will discuss in more detail. My one paragraph summary follows:
Dr. Blanc notes that pathologists see inflammation but not infection, and that many cultures of amniotic fluid do not identify microorganisms even though the placenta shows histologic chorioamnionitis[2]. Histologic chorioamnionitis was common but infant sepsis was not, and he wondered if the emphasis of study should be the efficacy of the infant’s defense. Chorioamnionitis was associated with the duration of rupture of membranes (based on clinical chorioamnionitis), but notes that it can be present with intact membranes. He speculates that if infection occurred before rupture, bacteria and inflammation in the membranes should be diffuse, but after rupture the inflammation should spread in a centrifugal manner up from the cervix . He notes that leukocyte migration through the chorionic surface of the placenta is inhibited by thick barriers compared to thin areas consistent with a chemotactic substance in the amniotic fluid. Similarly the fetal migration of neutrophils in the umbilical cord starts with the vein and is directed toward the amnion surface of the cord. He notes the fundamental uniqueness of chorioamnionitis in that there is both a fetal and maternal response to the same infection, and that the fetal response occurs from large vessels due to the lack of a capillary bed in the cord or placental surface. He comments that the cells in the “exudates” are maternal based on sexing the cells from a personal communication. He wonders about the effect of the inhaled or swallowed inflammatory cells. With a few exceptions the pathology gives no clues as to the microorganism species. The exceptions are Candida plaques on the cord, the villous necrosis and acute inflammation in Listeria, and possible brown spots in Corynebacterial infection. He reports his experience with fetal Group B streptococcal infection that shows a modest or absent inflammation in the placenta. Finally, he speculates that the amniotic fluid including neutrophil response must be important in surveillance against infection, and that poor nutrition might compromise this defense.
In rereading Dr. Blanc I realized that many things I thought I had discovered for myself he had already presented in this symposium, and that I must have read them first. I will expand on many of his observations in the question format of the chapter. There was another pathologist, Dr. Naeye, who presented at the same symposium whose perspective was based on having been the placental pathologist for the Perinatal Collaborative Study of the National Institute of Neurological and Communicative Disorders and Stroke which followed more than 56,000 children from gestation to 7 years of age and covered 12 university affiliated hospitals in different geographic regions. He published the placental data in numerous journal articles and in two monographs. Below is my two paragraph summary of his Ciba symposium presentations.
Dr Naeye argued that risk factors for histological chorioamnionitis included undernutrition , and coitus during pregnancy[3]. However, the presentation fails to prove his thesis because of the numerous confounders. For example, the data on coitus did not lookat the incidence of chorioamnionitis but the incidence of deaths due to amniotic fluid infection (defined by histological chorioamnionitis) stratified by maternal age, not gestational age. When questioned he admitted that the neonatal deaths were due to prematurity, but opined that the fetal deaths were due to sepsis. However, the criteria for fetal death by sepsis were not presented.
A third paper at the symposium presented by Dr. Ross looked at the etiology of chorioamnionitis[4]. At that time a major concern was the relationship between neonatal pneumonia and chorioamnionitis. He reported that the bacterial correlation with chorioamnionitis and pneumonitis was high, but surprisingly the microorganisms were often not very pathogenetic, with ureaplasma being the most common either alone or in culture with other organisms. The evidence that chorioamnionitis was an ascending infection from the vagina was that the organisms cultured from the fetal lung or placenta were those found in the vagina, that the first twin was more likely to have chorioamnionitis than the second, and the membranes over the os often showed the most advanced inflammation. In considering why women might be more susceptible to amniotic infection they demonstrated a significant association in 60 women with higher maternal weight gain and protein intake with stronger in vitro antibacterial properties of the amniotic fluid. They do not describe any test for chorioamnionitis in the tested fluid. They also found a significant correlation with higher cord blood cholesterol and globulin levels and higher antibacterial amniotic fluid. Zinc levels were examined and did not prove to be a deciding factor in anti-bacterial strength of amniotic fluid. He reviews Dr. Naeye’s evidence on the association of coitus within the month before delivery and the incidence of chorioamnionitis. He found the same association in South Africa. Also in South Africa there was a high incidence of vaginal trichomonas infection (approximately 50%), but he did not specifically correlate this with chorioamnionitis.
These authors were asking important new questions about the cause and consequences of chorioamnionitis. During this period the study of chorioamnionitis was further directed onto a new course by two factors. The first was the routine use of antibiotics in preterm infants which made the microscopic diagnosis of chorioamnionitis less important as a predictor of infection in the infant. As a friend who was a pediatric resident from 1974-6 put it: When he started, the attending physician questioned why a preterm infant was given antibiotics since they clearly had hyaline membrane disease to why did you not give this preterm infant antibiotics since hyaline membrane disease can not be distinguished from congenital pneumonitis. The second factor was the discovery that histologic chorioamnionitis was associated with preterm labor. In his 1983 monograph Naeye states boldly “Amniotic fluid infections appear to be the most frequent cause of preterm delivery in all of the studies that we have conducted[5].” I visited him somewhere within 1981 to 1982 to consider working there, and he was emphatic that chorioamnionitis was the cause of preterm labor, and the task ahead was to successfully eliminate that infection.
Somewhat later, 1988, a paper by Dr. Bejar and colleagues in San Diego demonstrated a correlation of ultrasound detected periventricular leukomalacia in preterm infants with purulent amniotic infection and with inflammation of the umbilical cord[6]. This paper redirected thinking about chorioamnionitis. The current focus of most investigation of chorioamnionitis is on its role in preterm labor and the effect of inflammation on the infant.
Discussion and Literature Review
Where are the microorganisms in chorioamnionitis?
If we imagine that a bacterium has finally made the long journey to the interior cervical os, it will suddenly be faced with new choices. It could enter the oxygen rich decidua but that has maternal capillaries with defenses and a dense extracellular matrix, or it could travel in the less oxygenated but unobstructed potential space between the amnion and chorion connective tissue, or it could crawl along the amnion epithelium or plunge into the anaerobic amniotic fluid. No doubt its normal chemotaxis for pH, oxygen or nutrients, its physical motility, and its receptors for binding the surface of human cells will all factor into the pathway the bacterium pursues. Individual species may respond differently. In order to properly sample the intrauterine space to detect microorganisms and to treat infection, we need to answer the question, where are the microorganisms.
As pathologists we are accustomed to not always being able to find the microorganisms on a microscopic slide despite evidence of infection, for example in endometritis. The majority of microscope slides showing chorioamnionitis do not demonstrate microorganisms. In the cases that do, they may have grown after delivery, so called overgrowth, as the placenta was being held prior to fixation in formalin which would kill the bacteria. There are nonetheless reasons to believe that these microscopically visible microorganisms provide information about where they were surviving in the living tissue. The fact that the bacteria have a single morphology, for example all the same shaped rods, and are not a mixed morphology, is evidence that they are not simply a sample of different flora from vaginal contamination during delivery. As further evidence that the overgrowth is not contamination, I examined a placenta that I knew was fixed in formalin immediately after delivery, and that placenta demonstrated numerous filamentous bacterial rods in the chorion connective tissue. Thin bacterial rods typically penetrate into the connective tissue of the amnion and chorion (Fig 1). Cocci are found in a different location, characteristically on the surface of the amnion (Fig 2)[7]. In some cases of Group B Streptococcal infection, cocci may be widely distributed, and in second trimester fetal loss can even be seen in the fetal chorionic vessels with an acute villitis (Fig 3). Pseudohyphae of the yeast Candida are characteristically found in plaques of neutrophils on the umbilical cord surface (Fig 4)[8]. In the special case of subacute necrotizing funisits (umbilical cord inflammation) microorganisms can be found in Wharton’s jelly.
If a bacterial species prefers the amnion surface, then amniotic fluid may be the most sensitive way to sample it, but if the bacterial species is one that prefers to burrow deeper into the membranes, then culture beneath the amnion may be the most sensitive method to detect that species. Culture swabs are often inserted beneath the amnion in vaginally delivered placentas in an attempt to avoid surface contamination from vaginal passage. This technique may preferentially sample microorganisms that reside in this location, and the manipulation required might in itself sometimes result in surface contamination. The most valid way to compare surface to subamnion culture would be to evaluate placentas delivered by Cesarean delivery, handled aseptically, and cultured immediately. A more feasible but less strict approach is to compare cultures of amniotic fluid taken soon before delivery with subamnion placental cultures from vaginal delivery. One such study demonstrated 28 cases in which the membrane culture was positive and the fluid culture negative, and 7 in which the fluid culture was positive and the membrane culture was negative[9]. Sixteen were positive in both fluid and membrane. An even more extreme technique obtained cultures through the surface of the placenta and swabbed the subchorionic fibrin facing the intervillous space. This deep technique had positive cultures in 17 of 33 cases, compared to 23 positive cultures from the placental surface. The deep cultures did not grow organisms in cases without clinical chorioamnionitis, but surface swabs did[10]. I find the logic of the study weak as the swab is actually sampling the maternal blood. Another study used the swab between amnion and chorion to correlate microbiologic culture with histologic chorioamnionitis and found that bacterial cultures were often positive without histological chorioamnionitis[11]. This surprising result could have been due to the increased manipulation of the membranes producing contamination in the cultures, but also could have been because of too strict a definition of histologic chorioamnionitis. A study from the same first author of the previous study found that cultures from amniocentesis correlated with elevated cytokines in the amniotic fluid culture, but swab cultures taken at delivery beneath an amnion flap did not[12].
What can we conclude from studies of placental cultures? I think that vaginal contamination remains a possible confounder in these cultures even if every attempt is made to aseptically lift the amnion flap. As a pathologist I commonly see vernix and other material that has found its way most likely during labor between the amnion and chorion (Fig 5)[13]. Bacteria might also contaminate this space in a similar manner during delivery. Contaminating bacteria may also be able to penetrate layers with different success depending on species in the interval between delivery and sampling. Finally, I think from the examples of direct staining of organisms that culturing beneath the amnion may selectively culture certain species that reside in that space rather than on the amnion surface. The bottom line is that any placenta microbiological culture study has a potential for bias or error, and the results must be interpreted cautiously.
Only infectious causes of chorioamnionitis have been plausibly proposed, but it remains a frustrating fact that even very sensitive molecular techniques have failed to find bacteria in the amniotic fluid in all cases of chorioamnionitis or of elevated amniotic fluid Il-6 cytokine as a marker of inflammation[14, 15].There are possible explanations, such as species not having the commonly probed gene sequence, or that the cytokine elevations were not from chorioamnionitis. It also follows from the discussion in the previous two paragraphs that the microorganism could be in the membranes and not in the fluid. Conversely, even if we accept that all chorioamnionitis has an infectious etiology, it does not necessarily follow that the organism detected in the amniotic fluid is actually the organism that initiated the inflammatory response, as the population of organisms might undergo a change over time. The concepts of chorioamnionitis without infection or of a progression of microorganisms perhaps from mycoplasma to less motile bacteria remain possibilities that we are not currently proven or disproven.
Why do we seldom see microorganisms in the tissue sections of chorioamnionitis? One explanation is that the density of the organisms is very low and a single microorganism is difficult to distinguish from other material in the histologic background. Some investigators have had greater sensitivity using a tissue gram stain, but I at least have not found this helpful to find organisms[16]. The Gram stain can characterize the cell wall of bacteria as either gram positive or negative, but is not very specific or sensitive stain for bacteria. Investigators have localized bacteria that were applied to membranes in vitro by using transmission or scanning electron microscopy, but this technique in clinical specimens would be less sensitive than optical microscopy [17]. Other possible reasons for the rarity of microorganisms in sections are that a) they do not stain on routine stains and can therefore not be seen, b) they have been internalized by inflammatory cells and are partially digested, and c) they are not present in the inflamed tissue.
If the problem is a paucity of microorganisms, a better detection technique might identify them in tissue. I was consulted on a legal case in which in situ PCR (used to amplify and detect specific DNA or RNA on a tissue section) was used to amplify a genomic sequence common to many bacteria. On one slide standing out brightly was a morphologically perfect single bacterium. However there was no inflammation and the bacterium was located in the intervillous space which in life would have had flowing maternal blood. This sole microorganism was likely a contaminant. Yet it vividly exemplified the potential of the technique. To optimize the technique the placenta and membranes should be fixed immediately after delivery in 4% paraformaldehyde to stop bacterial movement and reproduction, and to optimize the preservation for DNA amplification. The amplified DNA is detected on the slide by binding a chemical reaction system that produces a marker pigment, and since bacteria have cytoplasmic rather than nuclear DNA, the morphology of the bacteria should become evident. More specific probes could identify individual species including those such as the Clostridialis species in bacterial vaginosis that are not usually cultured[18].
If the microorganisms have been ingested by inflammatory cells the immunochemical identification of specific bacterial antigens within inflammatory cells could potentially detect this footprint of infection. This technique would need to be tested in an experimental system, but could potentially answer the question of where the organisms are (or were)even if they are not intact.
One way in which microorganisms could not be present in the membranes but still be driving an inflammatory response is if they were present in the endocervix and the mediators of inflammation were being distributed to the amnion. The same detection techniques presented above would then need to be applied to endocervical biopsies.
How do organisms get to the intrauterine cavity?
The conceptual answer is that the organisms ascend from the vagina where they are normal or at least transient flora. Culture studies support this concept as the organisms cultured are the expected vaginal organisms or gastrointestinal flora known to colonize the vagina in some individuals. A long tradition in obstetrics recognizes that prolonged labor and prolonged ruptured membranes increases the risk of clinical chorioamnionitis. As a new obstetrics and gynecology resident in 1975, I bought a book “Controversies in Obstetrics and Gynecology” which featured a chapter demonstrating that prolonged labor led to an increase in clinical chorioamnionitis [19]. Since then a few studies have also confirmed that the incidence of histological chorioamnionitis also increases with increasing duration of labor and of membrane rupture [20-22]. The most dramatic evidence for the ascending route of chorioamnionitis was the historical discovery by Semmelweiss that hand washing prior to examining the pregnant patient reduced the incidence of often lethal childbed fever.
A supporting piece of evidence for the ascending route of infection is that the twin situated over the cervical os often shows grossly or microscopically an increased intensity of inflammation compared to the other twin (Fig 6). This phenomenon has also been demonstrated clinically by amniocentesis [23].
This evidence for ascending infection suggests that the cervix provides a highway for vaginal flora to enter the intrauterine cavity impeded only by the time of transit. However, from a microorganism’s standpoint, the ascent may not be so simple. The cervix is filled with mountainous ridges raining down viscous mucus, drone like antibodies, chemical warfare, and white cell snipers. The cervical mucus plug evokes the image from the past of cotton plugs in test tubes and Pasteur flasks that kept airborne bacteria out of the cultures (fig 7). The mucus plug has been studied including by proteomics, and there is likely more to learn [24]. Cervical cytokines have been measured suggesting a low response is predictive of preterm labor and presumably associated chorioamnionitis[25]. Neutrophils in the cervix have been suggested as playing a role in collagenolysis with cervical dilatation[26]. The cervix is often examined by pathologists following Cesarean hysterectomy, but there are no biopsy or specimen studies specifically of the interaction of cervical histology and chorioamnionitis (fig 8).
If the cervix were an open highway, eventually any motile microorganism should find its way to the uterus. The evidence is against this made rush of microorganisms reaching the uterus. Studies have generally found a single species or a limited number of species in intrauterine cultures [27, 28]. When microorganisms are seen by direct histology, they are usually of one morphologic type. The species cultured vary from study to study and patient to patient, although some, such as the highly motile ureaplasma, predominate at least in those studies that provided the special cultural needs or DNA primers for PCR detection for this microorganism [29, 30]. Differences in the microorganisms cultured from study to study likely reflect regional and temporal variation in vaginal flora. This variation does not explain why multiple species are not the rule in an ascending infection especially since cultures show many different single species from the vaginal flora are capable of causing chorioamnionitis. One explanation for finding only a single species in a given patient is that only a rare microorganism actually succeeds in reaching and colonizing the intrauterine cavity. This culling of organisms may be due to host factors and barriers, specific signaling differences between host and microorganism or competition between microorganisms. This single microorganism hypothesis if true is more evidence that the ascent is more like a trek in hostile territory than an open highway.
What initiates preterm chorioamnionitis?
Often repeated in the medical literature is the assertion that chorioamnionitis causes preterm labor. There is little doubt clinically or experimentally that intrauterine inflammation leads to labor, the question is what causes the initial infection. Teleologically and evolutionarily, once chorioamnionitis is present, expulsion of the pregnancy may provide the best hope for survival of the mother and her ability to reproduce successfully in the future. It is an accepted clinical maxim that tocolytic therapy is ineffective if the patient has chorioamnionitis. Histologic chorioamnionitis, as already discussed, is more prevalent by far in preterm versus term labor. Yet the effect of antibiotic therapy is minimal in preventing preterm labor and delivery even with preterm premature rupture of membranes. If chorioamnionitis is simply the ascent of vaginal flora through the cervix, and if all women have a vaginal flora and cervix, why do they not all have premature labor. Clearly something must make a subgroup of patients more susceptible to preterm chorioamnionitis. Microorganisms must be surmounting the normal barriers blocking their ascent if chorioamnionitis is initiating preterm labor.
There is ample evidence that microorganisms can reach the intrauterine cavity despite intact membranes and a closed cervix [31, 32]. It is possible that a clinically closed cervix may not always be a sufficient barrier to microorganisms. They may also be able to enter through the membranes at the internal os even if intact. Yet it is easy to imagine chorionamnionitis as a simple ascending infection in the woman with prolonged labor or prolonged rupture of membranes. In a woman presenting with preterm labor and histological chorioamnionitis more of an explanation for her particular susceptibility is needed. Perhaps she is just unlucky and statistically some microorganisms ascend through the imperfect cervical barrier with a given probability of reaching the uterus over time. The problem with this idea is that there are epidemiologic risk factors for preterm labor and preterm chorioamnionitis, making a random process unlikely. A previous preterm pregnancy is a strong risk factor for a subsequent preterm delivery, and allotypical differences in inflammatory response may account for some increased susceptibility to chorioamnionitis. The allotypical differences quantitatively can not account for the many random women with preterm labor.
Another way to try to understand the relationship between chorioamnionitis and preterm labor is to ask what initiates preterm chorioamnionitis? One potential answer is a process that alters the cervix such as cervical labor (effacement and dilatation) that could hasten or abet ascending infection. In the sheep, hormonal manipulation can produce cervical labor without uterine contractions[33]. Evidence of a similar process in humans is that a short cervical canal on vaginal ultrasound has an increased risk of preterm delivery [34-36] . Along the same line of reasoning, most placentas that I examine from patients with the clinical diagnosis of incompetent cervix demonstrate chorioamnionitis and this also correlates with clinical studies [37].
If preterm chorioamnionitis is a consequence of preterm cervical changes that break down the cervical defenses, then patients with preterm labor may be no more susceptible intrinsically to chorioamnionitis than patients at term with an open cervix. It also follows that the best treatment of preterm chorioamnionitis would be to stop the cervical changes. Perhaps this is the basis of successful treatment of prevention of preterm labor with vaginal progesterone which would quiet labor and stop cervical changes [38]. In opposition to this hypothesis, the question must arise why patients whose cervix is dilated for weeks in late pregnancy do not have clinical infection or rapid onset of labor. The answer may be that the timing of cervical changes prior to chorioamnionitis and preterm labor is equally prolonged. This would be a valuable explanation since the more time between the onset of cervical changes and chorioamnionitis the longer the potential treatment interval. Of course, even if progesterone is effective it may not be the initiating factor in cervical changes. The common clinical use of prostaglandin applied vaginally to induce labor suggests that cervical dilation can occur without changes in maternal progesterone. Still unclear to me is why Cytotec is given vaginally compared to orally. How does the effect concentrate locally rather than being absorbed into the blood stream? There appears to be a significant gap in our understanding of cervical labor and its relationship to vaginal flora.
One way that might explore changes in the cervix is to implant biosensor in the cervix for compliance, inflammatory mediators, and bacterial products. This would require some experimental work in animals, but microscopic sized biosensors could be applied to cerclage stitches or put in the cervix with induction of labor. As discussed, endocervical biopsies if processed for maximum information could also add information about the processes that allow bacterial ascent in the cervix.
There are at least three other theories to explain the initiation of chorioamnionitis in preterm labor beside cervical changes. One is that the composition of vaginal flora is critical. Another postulates that endometrial inoculation with organisms prior to pregnancy may be the origin of chorioamnionitis. The third envisions bacteria hitching a ride through the cervix on a more aggressive” species” including trichomonas and human sperm.
Do variants in vaginal flora initiate chorioamnionitis?
Different microbes in the vaginal flora have increased either the incidence of preterm labor or the risk of neonatal infection with chorioamnionitis. Both bacterial vaginosis and ureaplasma in the vaginal flora have been found to favor preterm labor although the effect is not large [39]. Perturbations of vaginal flora based on vaginal cytology that included categories of atypical gram positive rods, and of unexplained leukorhea (neutrophils) also had some predictive value for preterm labor[40]. However, the vaginal flora is proving to more complex than the diagnosis of bacterial vaginosis suggests. The colonization of vaginal flora with colonic bacteria in women with recurrent urinary tract infection, and the association of urinary tract infection with preterm labor suggest that host factors as well as the species of microorganism might influence the incidence of chorioamnionitis. Treatment of asymptomatic bacteriuria in pregnancy reduces the incidence of preterm labor [41, 42]. If bacterial flora variation can predispose to chorioamnionitis, the mechanism has not been elucidated. Certain flora might duplicate the inflammatory action of vaginal misoprostol (Cytotec), a prostaglandin analog in inducing labor or provide more microorganisms to actively assault the cervix. The role of vaginal flora and its potential interaction with the cervix remain incompletely understood aspects of chorioamnionitis.
.
Is chorioamnionitis initiated by microorganisms chronically in the endometrium?
Not surprisingly, pathogenic microorganisms, E. coli and Group B streptococci, were not found in genetic amniocenteses by the PCR technique [43]. Some investigators however have made a case that vaginal flora microorganisms particularly mycoplasmas can be chronically present in the endometrium based on culturing them from amniotic fluid from second trimester genetic amniocentesis [44-46]. These positive cultures are predictive of preterm premature rupture of membranes or of preterm labor and delivery. These studies contradict the usual tenet that the healthy endometrium is sterile. Unlike most body surfaces no endogenous flora or biofilm was expected in the uterus, in part because of menstrual shedding and the cervical barrier, but also because microscopically plasma cells are not present. The gut and other mucosal surfaces in contact with the external environment do have plasma cells. Plasma cells which produce antibody localize where antigen is presented to B cells as typically occurs at the site of infection and the lymph nodes draining it, or in mucosal surfaces exposed to microorganisms. The endometrium is usually devoid of plasma cells, and finding any in the endometrium has been considered diagnostic of chronic infection in the endometrium. This chronic endometritis is usually in response to microorganisms that aggressively enter the uterus in normal women such as chlamydia or gonorrhea. The theory that some women have an endometrial flora, and therefore are at risk of preterm labor and chorioamnionitis is interesting, but I think less than proven. Interconceptional antibiotics have not prevented preterm labor even in patients with positive endometrial culture or plasma cells on endometrial biopsy at the start of the study and may have increased the incidence of adverse outcome (including preterm labor) [47]. A study of women with positive fetal fibronectin done as a routine screening test and amniocentesis fluid culture did find any positive cultures, not even in the patients with elevated Il-6 cytokine who went on to have preterm labor[48].
I infrequently find plasma cells in the decidua of the fetal membranes, but the patients generally do not have chorioamnionitis, or pelvic inflammatory disease (Fig 9). (The decidua is the endometrium which undergoes dramatic histologic change with pregnancy and which like deciduous leaves sheds with the placenta and membranes.) These plasma cells could be in response to fetal chorionic antigens rather than infection. There is a commonly recognized plasma cell infiltration that occurs in approximately 25% of placentas in the decidua beneath the placenta, called chronic deciduitis [49, 50](fig 10). Like plasma cells in the membranes, these decidual plasma cells do not correlate with chorioamnionitis or other evidence of maternal infection. Immunoglobulin staining of these cells was positive for IgG [50] (fig 10). I wish I knew to what antigen they are responding, but I suspect it is more likely a fetal antigen related to implantation rather than infection. A not so simple study would be to recover the cells and create a hybridoma and in turn use the hybridoma antibodies to localize the antigens in the tissue or test them against bacterial antigens. This experiment would require banking a lot of tissue at delivery for potential plasma cell recovery, and would not be an easy undertaking.
There is another interesting form of acute inflammatory lesion in the decidua that is not associated with chorioamnionitis. This lesion manifests as discrete foci of decidual necrosis with much basophilic debris and neutrophils not associated with a generalized membrane inflammation (fig 11). The lesion seems to be more common in preterm labor and spontaneous fetal loss, but this may reflect curettage resulting in a more extensive decidual sample. The lesions can also be seen in normal term pregnancy. Speculatively, they may be due to decreasing progesterone levels as decidual necrosis and neutrophils have been described in pregnant sheep following oopherectomy[51]. The sheep pregnancy is dependent on the progesterone from the corpus luteum in the ovary, unlike human pregnancy in which the placenta takes over the progesterone production. The sharp focality of the lesions has never been well explained, and a local response to infection can not be totally excluded.
More sensitive techniques for localizing microorganisms in tissue could help in proving a role for them in early pregnancy, both in looking at decidual lesions, and perhaps in looking at induced and spontaneous abortion specimens. The amniotic fliuid culture studies have some risk of contamination, but it is possible that highly motile mycoplasma such as ureaplasma commonly enter the uterine cavity perhaps even transiently. Their role in preterm labor could be due to chronic endometrial persistence or hark back to an abnormally weakened cervical barrier to ascending infection.
Are there vectors for chorioamnionitis?
Another method for vaginal flora to ascend the cervix would be to attach to a vector that has the ability to transverse the cervix as part of its pathogenicity or function. The usual suspects are those that cause sexually transmitted diseases such as chlamydia or gonorrhea but also motile microorganisms like trichomonads or human sperm (a microorganism by analogy)[52]. The evidence presented by Dr. Naeye and colleagues for sperm has already been mentioned in the historical introduction. Dr. Naeye and Dr. Ross provided clinic patients in South Africa, primarily Zulu tribe members, with information on the benefit of sexual abstinence during late pregnancy and also offered them condoms if they did have coitus[53]. They then had a two trained personnel obtain sexual histories from the patients as to the timing of any coitus, female orgasm, and the use of condoms. They studied all placentas, but separated inflammation in the membranes from inflammation in the chorionic plate of the placenta which they called amniotic infection. There is no basis for this distinction. The clearest result was that having coitus without a condom within 2 days of delivery significantly increased the incidence of inflammation in the membranes (8 of 23, P<0.001) compared to all other groups including no coitus (17 of 188) and coitus with a condom with 2 days of delivery (2 of 19). As the authors admit this was not a randomized study and therefore factors besides not using the condom may have biased their result. However, the role of spermatozoa in chorioamnionitis has been less than exhaustively studied. Epidemiologic studies have difficulty in obtaining accurate sexual information[54]. An experimental model would be useful to see if in fact sperm can be a carrier of vaginal flora into the uterus of the pregnant animal.
As mentioned by Dr. Ross in the Ciba symposium, the high incidence of trichmonas infection was correlated with a high incidence of perinatal infection. Tritrichomonas foetus is an important cause of infectious fetal loss in cows and is characterized by a giant cell pneumonitis [55, 56]. I have seen an unexplained giant cell pneumonitis in a 21 week gestation human fetus from a mother with severe trichomonas vaginitis (fig 12). Others have also noted unexplained giant cell pneumonitis in human fetuses [57]. Trichomonas vaginalis, a common cause of human vaginitis has not been directly implicated in chorioamnionitis, but giant cell chorioamnionitis is an occasional finding which I initially attributed to fetal squames with premature rupture of membranes [58] (fig 13). Immunostaining of both giant cells in the fetal lung and chorionic membrane might be able to detect trichmonads if they are present in the tissue [59]. I have also seen a placenta from a 37 week gestation, not autopsied , stillborn infant whose mother according to the midwife had a treated severe trichomonas vaginitis (fig 13). The placenta showed chorionic calcifications that could have represented foci of necrosis from inflammation.
The role of chlamydia and gonorrhea is seldom mentioned in the context of chorioamnionitis, despite the ability of these microorganisms to enter the pelvic organs, although at least one study of chorioamnionitis included a case with gonorrhea [14] and studies of gonorrhea in pregnancy have generally noted the increased incidence of chorioamnionitis[60, 61]. Chlamydia in vaginal culture in one study was correlated with preterm labor[62]. Early in my career I cultured all fetal lungs at autopsy on “chocolate agar”, a media that allowed the culture of gonorrhea. From more than 100 cultures, I had only one positive result, but that third trimester fetus had chorioamnionitis, pneumonitis, and died in utero with evidence of septic shock including adrenal hemorrhages. The public health measure that requires the application of antibiotic to infant eyes to prevent chlamydia conjunctivitis does suggest that sexually transmitted disease might have a contributory role in chorioamnionitis, but there is little direct evidence for this hypothesis. These two microorganisms may directly cause chorioamnionitis but the fact that they are capable of penetrating the nongravid cervix also makes them candidates to carry or pave the way for other microorganisms.
Can chorioamnionitis be caused by blood borne microorganisms?
Not all chorioamnionitis is necessarily from ascending infection. Maternal bacteremia does appear to reach the fetus and perhaps secondarily cause chorioamnionitis. The most dramatic example is Listeria monocytogenes which ingested by the mother leads to bacteremia, and can cause micro and macro abscesses and chorioamnionitis in the placenta and infection of the fetus [63, 64] (fig14). This and other blood borne infections of the fetus such as Haemophilus influenza, Actinomyces meyeri and Pseudomonas aeruginosa may be associated with chorioamnionitis. It is difficult to prove that the chorioamnionitis is due to the blood borne infection, since chorioamnionitis could still be from an unrelated ascending infection or sometimes even that the infection was blood borne[65-67]. Another piece of evidence for blood borne transmission of bacteria to the fetus has been provided by Dr. Kim Boggess and colleagues who used an array of bacterial antigens from dental organisms and found that some fetuses had antibodies to those organisms[68]. These dental bacteria must have found their way to the fetus from asymptomatic bacteremia from the oral cavity. There was no evidence that they had produced chorioamnionitis, but they did show that non-pathogenic blood borne bacteria can reach the fetus and their presence was associated with first and second trimester bleeding and or a small for gestational age infant[69, 70]. Experimentally in pregnant sheep, the injection of lipopolysaccharides from dental pathogens directly into the amniotic fluid does produce chorioamnionitis. Oral microbial flora are occasionally reported to cause chorioamnionitis [71-74]. The treatment of periodontal disease at least in some studies was not associated with preterm labor and treatment had no effect on preterm labor although chorioamnionitis was not studied [75, 76]. Perhaps another line of evidence for blood borne chorioamnionitis is the rare chorioamnionitis with a non-vaginal organism. I examined the placenta in a case of pneumococcal pneumonia in a newborn, and although days later, a culture of the placental surface grew pneumococcus.
One benefit of living in Kentucky is that I was able to see firsthand the pathology of Mare Reproductive Loss Syndrome, a well-studied example of blood borne chorioamnionitis (obviously a disease of horses not people)[77]. The anatomy of the horse is such that the allantoic cavity is against the cervix and then the amniotic cavity. In ordinary vaginal flora type chorioamnionitis in horses both membranes are inflamed. In MRLS, only the amniotic cavity is inflamed, consistent with a blood borne rather than ascending chorioamnionitis. MRLS was first associated with the epidemic rise of Eastern Tent caterpillars. While horses do not eat them, when they are swarming they can get swallowed, and there setae, sharp hair-like bristles, produce gastrointestinal ulcerations. Field tests demonstrated that horses kept in caterpillar free fields did not get the syndrome. The microorganisms that were cultured from the placentas were typical of horse cecal flora, unlike the usual equine chorioamnionitis which like human disease is due to vaginal flora. The mares often developed a pericarditis or uniocular uveitis as well as chorioamnionitis. The simplest explanation for these findings is that the caterpillar induced ulcers led to a cecal bacteremia and that the microorganisms were eliminated except in privileged sites in which the maternal defenses could not clear them such as the anterior eye chamber, pericardial space and amnion fetal space. The concept that microorganisms can survive in certain privileged sites is not new, and the amniotic cavity may be another such potential site[78].
The above evidence for blood borne chorioamnionitis shows that this route of infection is at least conceivable. I examined a stillborn infant with findings that could be interpreted as a blood borne chorioamnionitis with some analogies to MRLS [79])(fig15). The placenta demonstrated extensive necrosis and dystrophic calcification of the fetal chorionic membrane and of the umbilical vein which could be the result of an old necrotizing chorioamnionitis. The infant died shortly after birth due to fetal akinesia sequence with rigid joints, pulmonary hypoplasia and a short umbilical cord. There was extensive periventricular white matter and subependymal gliosis of the lateral ventricles which was a clear cause of the fetal paralysis in utero (fetal akinesia). The periventricular gliosis likely was associated with ventriculitis because of the loss of ependymal cells and marked subependymal gliosis, so direct infection of the brain can not be excluded. No viral inclusions were found, and stains of the umbilical cord for spirochetes were negative. The infant had pericarditis and adhesions of one eye to the lid , both privileged sites of infection in the mares in MRLS.
I speculated that this infant had a resolved chorioamnionitis much earlier in pregnancy. As will be discussed later there is an association of severe chorioamnionitis and periventricular leukomalacia in premature infants, thus an episode of chorioamnionitis might be the cause of the brain lesion. The pattern of calcifications in the fetal membranes and umbilical cord are in the same locations in which neutrophils accumulate with chorioamnionitis. Healing or resolving chorioamnionitis is a nonexistent lesion presumably because the intrauterine inflammation promotes labor and delivery of the placenta before such resolution could occur. The analogy of my case to MRLS in no way suggests the intervention of caterpillars, but only of a fetal blood borne infection causing inflammation in the fetal membranes, much as Listeria is associated with membrane inflammation in almost all cases. The infant also had a cleft lip and palate and a very small but normally formed foot on one side. A karyotype was normal. These two lesions are more difficult to explain. I can not prove that the membrane lesion is a resolved chorioamnionitis, but no one has proposed a better explanation. If this is a case of resolved chorioamnionitis, the implication is that chorioamnionitis, likely blood borne, does not necessarily lead to labor. If this hypothesis were true, then the relationship between preterm labor and chorioamnionitis may be more complex than chorioamnionitis causing preterm labor.
Could vaginal flora reach the amniotic cavity via the blood due to some change in vaginal ecology or vaginal/cervical trauma? Could lymphatics ever transmit vaginal flora to the amniotic cavity. Except for the onset of chorioamnionitis with a closed cervix and intact membranes, there is no evidence to suggest or refute such mechanisms.
Can chorioamnionitis be caused by direct inoculation of microorganisms into the uterus?
There are other uncommon mechanisms of infection of the intrauterine cavity in pregnancy. There is a case report of a fistula between the ileum and the uterus which resulted in recurrent fetal loss as well as maternal pneumonia, bacteremia and chorioamnionitis[80]. I was involved in reporting a case in the newsletter for the Study Group for Complications of Perinatal Care in which chorioamnionitis followed a genetic amniocentesis. The mother became septic and died of adult respiratory distress syndrome. The frozen amniotic fluid grew bowel flora but did not have neutrophils. We inferred that the amniocentesis needle had penetrated the colon prior to entering the uterus. I examined another case of fulminant chorioamnionitis with Staphlococcus aureus following genetic amniocentesis which from the timing was reasonably attributed to inoculation of skin flora during the procedure. I have also examined one case of septic coat hanger induced abortion with fulminant sepsis. There were clostridia visible in the curettage specimen. The commonality of these cases was the fulminant maternal infection compared to the usual chorioamnionitis, despite arguably small innoculations. This may have been due to the pathogenicity of the microorganisms or perhaps the directness of the route. There is also a bias since insignificant post amniocentesis infections would not have come to my attention in a pathology laboratory. Still, when induced abortion was illegal in the United States, there was a high infectious mortality with illegal abortions.
Where is the inflammation?
As a pathologist, I identify chorioamnionitis by the presence of neutrophils, a white cell that is relatively easy to identify with routine H&E sections, and is a recognized marker of acute inflammation. Neutrophils are normally circulating in the blood, but enter the tissue in response to acute inflammatory signals (chemotaxsis) secreted by lymphocytes and macrophages or from tissue injury. There are three sources of blood neutrophils that contribute to placenta and membrane inflammation: the maternal intervillous circulation, the maternal capillary bed in the decidua of the membranes, and the fetal circulation in the superficial chorionic and umbilical cord vessels. The microscopic evidence for chorioamnionitis is that neutrophils have migrated into the tissue in an amnio-centric direction. Most body tissues have a capillary bed from which neutrophils can easily migrate outward, and this is true of the decidualized endometrium. However, there are no capillary beds in the chorionic and amniotic membranes, on the surface of the placenta or in the umbilical cord. As a result, the migration of neutrophils often appears as a linear wave starting from the blood source.
Neutrophils attracted toward the amniotic cavity from the intervillous circulation must travel through the fetal surface of the placenta en route. The first barrier they encounter is the subchorionic fibrinoid, a layer of fibrin and any residual matrix from the subchorionic chorion epithelium that is on the underside of the connective tissue of the placental surface. Neutrophils often appear to marginate against the blood fibrinoid interface suggesting that as in capillaries the signal causes them to adhere and marginate, otherwise the bulk blood flow beneath the surface would dilute and carry away the chemotactic communication before the neutrophil could respond. The neutrophils accumulate in this fibrinoid and show a relatively even disbursement along the surface. The reasonable presumption is that they are responding to a chemotactic stimulus in the amniotic fluid. Neutrophils in the fibrinoid may be absent beneath overlying large vessels, any layer of persistent chorionic epithelial cells or squamous metaplasia of the amnion surface. This appears to be a kind of shadowing from the chemotactic signal. The neutrophils can continue to penetrate the fibrinoid and enter the chorionic connective tissue and progress on to the amnion (fig 16).
In the fetal membranes, the neutrophil migration begins in the capillary bed of the decidua. Neutrophils are often seen to accumulate at the junction of the chorionic epithelium and connective tissue, but in other cases they have passed into the chorion connective tissue or completed the trip to reach the amnion (fig). In male infants, these neutrophils have been shown to be of maternal origin by direct in situ hybridization for double X chromosomes [81].
The fetal blood in the chorionic surface vessels of the placenta and in the umbilical cord has no capillaries, just arteries and veins with thick muscular media. Typically, neutrophils tend to migrate from the subendothelium to the smooth muscle of the media in the most direct path to the surface facing the amnion. In the umbilical cord they will have to traverse some distance of Wharton’s jelly usually in an arc projecting from the vessel (fig).
From autopsies of infants, often those dying because of delivery before viability, the lung airspaces and the stomach and even distal gastrointestinal tract may be filled with neutrophils (Fig). From the histology alone, it may not be possible to know if the neutrophils traveled inward from the amniotic fluid or were a local fetal response to chemotactic stimuli or microorganisms. I usually hedge the diagnosis as “aspirated and ingested infected amniotic fluid”.
Is there variation in intensity of inflammation among patients?
The simple answer is yes. By intensity, the pathologist is usually judging the density of neutrophils and or the degree of necrosis of surrounding tissue and of neutrophils. Some schemes had tried to semi-quantify the density of neutrophils, but fine gradations are difficult to do under ordinary reading of the slide ([82]). Descriptive qualifiers can also be helpful. In some cases neutrophils pile up forming a mass of pus that can be seen grossly and microscopically, usually with evidence of neutrophil necrosis, and this inflammation can be called purulent (fig). In other cases, there is disproportionate necrosis of the chorionic epithelium, a necrotizing inflammation (fig). Karyorhectic cells are often present and may be either neutrophils or chorionic epithelium (Fig). Karyorhexsis is evidence of apoptosis which has been demonstrated in chorioamnionitis by the TUNEL technique[83]. In fetal vessels, there may be evidence of mural injury and laminar thrombus (Fig). In the umbilical cord there may be a concentric ring of necrosis in Wharton’s jelly called subacute necrotizing funisitis (which will be discussed later in more detail). One study claimed to show the PCR determined mycoplasma load correlated with the intensity of histologic inflammation in preterm premature rupture of membranes, but my reading is only that on average the load of mycoplasma is higher in cases with than without histological chorioamnionitis, but there was no direct one to one correlation with the actual intensity of the neutrophil infiltration[84]. Interestingly, many cases were positive without histological chorioamnionitis, which suggests that either due to timing or load, there was no inflammatory response to the microorgansims. Funisitis had the highest correlation with mycoplasmal load.
Not only can variation occur between the intensity of different cases, but it can vary within a single placenta. The intensity of the fetal response may be less or even more intense than the maternal response. An area of the membrane may be markedly more intense than another. This can be particularly marked over the cervical os. The inflammation may be confined to this area, suggesting that the chemotactic stimulus was confined to this area, perhaps by ascending micro-organisms, or the area over the os may be the only area without inflammation as if suspended over an open internal os with no source of neutrophils from an underlying decidua(Fig). Inflammation of the umbilical cord can vary from the fetal to the placental end[85].
The clinical significance of inflammatory intensity is weak. It does not show a correlation with maternal symptoms such as fever, at least in ordinary practice it is not uncommon to have a fever with minimal inflammation in the subchorionic fibrin, but on the other hand to have no fever in the mother with purulent chorioamnionitis. Inflammation may be more intense with infection with pyogenic organisms, but Group B streptococcus may show a muted maternal response (ref)(fig). Evidence of high grade fetal inflammation, usually with evidence of injury to the vessel wall, may be associated with a higher risk of white matter injury in very premature infants (ref) (fig).
What does staging of chorioamnionitis mean?
The evaluation of intensity has been called grading analogous to the pathologist’s approach to malignancies. Staging is also analogous but this is not the progressive spread of a malignancy but the progression of neutrophils from the blood source to the amniotic cavity. By describing how far the neutrophils are from their blood source on a projected trip to the amnion, the pathologist can identify a stage of progression. The problem becomes tricky in correlating the progression from three different sources of neutrophils. The maternal migration is usually considered separately from the fetal. Since a microscope slide shows a single window of time at delivery, a progression can not be seen in a single case but is inferred from multiple cases. Consider, if some cases have neutrophils only at point A , and all cases that have neutrophils at point B have them at A, and all cases that have neutrophils at point C have them at points A and B, then A B and C appear to be points on a line of progression. In practice point A is maternal inflammation is the subchorionic intervillous fibrinoid and point B is the chorionic epithelium , and point C is the amnion (Fig). The decidua is not included, but may show isolated diffuse neutrophils in some cases with neutrophils otherwise limited to the subchorionic intervillous fibrinoid (fig). In the fetus, the usual progression is the umbilical vein and then the umbilical arteries with each progressing into the Wharton’s jelly (fig). The involvement of the surface chorionic vessels is usually an independent component of staging. Staging is easier to see than to explain, at least for a pathologist.
Does stage relate to the timing of infection ?
In legal testimony, I have heard experts give estimates as to the duration of intrauterine infection based on the staging of chorioamnionitis. Such an inference must assume that any infection produces a similar level of chemotactic stimulus starting at a time zero and progressing in a uniform manner in all patients. This is a lot to believe without experimental evidence, and on the face of it, it seems very improbable. There are many parameters, such as the size of inoculum of microorganisms, the species of the microorganism and the host inflammatory response, that will vary and could have profound effects on the timing of neutrophil migration.
Another problem with interferences about timing is that neutrophils can travel fairly quickly on a microscopic level[86]. This can be demonstrated by applying the neutrophils to one side of fetal membrane in vitro (ref). In vivo experiments demonstrate a relatively rapid course when organisms or lipopolysaccharide is injected directly into the amniotic cavity or “subchorionic space” (ref). (Subchorionic injections are likely in the decidua.) The natural time course of infection based on clinical chorioamnionitis has been measured as a time of onset from the rupture of membranes or the beginning of a stage of labor. When I was a resident, the general conclusion was that labors longer than 24 hours needed to be delivered because of the risk of infection. The clinical observations are consistent with a relatively rapid development of chorioamnionitis over 24 to 48 hours.
If neutrophils can respond rapidly to infection, then it seems improbable that they would so often be caught by histology at early stages of their progression in the membranes. Since neutrophils appear to most often accumulate at certain anatomic “barriers” namely the junction of the chorion connective tissue with the chorion epithelium and just below the amnion epithelial basement membrane, these barriers might be slowing the migration sufficiently that they are captured en route on the histologic section. This also seems improbable. Unless the barrier were like a dam and suddenly all the neutrophils came through at its rupture, then there should still be progression of neutrophils seen visibly. There is seldom histological evidence of a “dam break”. Possibly, it is not the neutrophils but the chemotactic signal that accumulates at different depths of the membrane.
A simple explanation for the appearance of staging is that the neutrophils stop when they encounter the microorganisms. The evidence for this battlefront scenario is indirect. The occasional visible localization of cocci on the amnion surface and of filiform bacterial rods in the connective tissue is certainly consistent with this latter explanation. These visible organisms could reflect some postpartum overgrowth during storage of the placenta, but in one case that was fixed immediately after delivery, there were plentiful bacteria visible with gram stain. If this were true, it would have a puzzling implication that there was a relatively even distribution of bacteria entering through the surface of the membranes in many cases even when they can not be visualized by on microscopic sections.
The dynamic features of cytokine production and distribution, of microorganism movement and distribution and of neutrophils chemotaxsis are still incompletely understood. Therefore, one can not assume that neutrophils at different stages were on their way to the amnion surface, but were interrupted by delivery. Any inference about timing from histological staging is premature.
Whose neutrophils are in the amniotic fluid?
Elevated counts of white blood cells in the amniotic fluid have been used as a criterion for chorioamnionitis [87].. At least going in the reverse direction from fetal to maternal side, neutrophils in vitro can cross the amnion[88]. However, this does not insure that they do. Neutrophils are only rarely seen passing through the amnion basement membrane, although rarely ulcerations are present (Fig). They usually appear to line up behind the amnion epithelial basement membrane. Three is some evidence that fetal neutrophils may be the main source of amniotic fluid cells. In a study of amniocentesis in mothers in preterm labor at least three cases had elevated amniotic neutrophils without histologic chorioamnionitis, so the neutrophils logically had to be fetal [87]. Assays for sex chromosomes have demonstrated that the majority of fetal pulmonary and amniotic fluid neutrophils are fetal in origin [89, 90]. Neutrophils are rare at the umbilical cord surface, so it is unlikely that they are entering the amniotic fluid directly from the umbilical or chorionic vessels. Chorioamnionitis may be associated with a decreased neutrophil count in the infant, suggesting depletion of neutrophils(ref). In cases with a heavy neutrophil infiltration in the length of the cord, the volume of neutrophils may be sufficient to account for the low infant count. However, in most infants that also seems unlikely. It is possible that in some cases a fetal pulmonary contribution to the amniotic fluid neutrophils could be quantitatively significant. Quantitation of total fetal neutrophil volumes in different placental compartments compared to the ability of the fetus to produce sufficient neutrophils and the effects on infant white cell counts has not be done.
Is there chronic chorioamnionitis?
There is a pathological entity designated as chronic chorionitis but it is unlikely to be related to the usual acute chorioamnionitis (refs) . Histologically, it consists of an accumulation of lymphocytes in the chorion, usually where there is deficient chorionic epithelium ([91]). From the histologic picture it is not possible to distinguish between lymphocyte destruction of the chorion versus lymphocyte attraction due for example to exposed HLA antigens in areas that were denuded of epithelium for another reason. There is usually no direct evidence of necrosis or apoptosis associated with the lymphocytes, and there is no strong clinical correlation. The lesion is usually only present in a portion of the membranes varying from minimal to a majority.
The lesion that is not seen is an organizing or healing stage of chorioamnionitis. The reasonable presumption is that the acute inflammation leads to labor, and that even if effective antibiotic therapy or effective natural defenses were initiated they would not be successful in stopping preterm labor and delivery of the membrane before healing could start. This was the case in a study of preterm premature rupture of membranes in which antibiotics successfully delayed delivery a mean of a few days, but showed no difference in the histology of the placenta between treated and untreated mothers (ref). As presented under the discussion of blood borne chorioamnionitis, I have seen one possible exception that did not demonstrate actual healing but did demonstrate an extensive dystrophic calcification of the chorion. Dystrophic calcification is one response to necrosis of a lipid rich tissue, and conceivably the membrane had suffered a necrotizing chorioamnionitis earlier in the pregnancy.
Where does the chemotactic stimulus come from?
A common histologic observation is that chorioamnionitis is fairly evenly distributed over the surface of the placenta and membranes. It may be less uniform over the umbilical cord, but the effect is small and may have causes other than an evenly distributed chemotactic stimulus. Neutrophils will migrate toward numerous chemotactic stimuli including methylated bacterial proteins, cytokines from mononuclear inflammatory cells, and arachidonic acid metabolites such a leukotrienes. Elevations of the latter two chemicals have been measured in the amniotic fluid with preterm labor and with chorioamnionitis (ref)s. In rabbits and monkeys that have had bacteria injected into the intrauterine cavity, proteomics demonstrated peaks for S100 protein family proteins as a potentially important component of the inflammation[92]. Human infected amniotic fluid, as well as the monkey, demonstrated the S100 protein family (Calgranulin B) and a unique fragment of insulin growth factor binding protein[93]. Whatever the origin or composition, the disbursement of the chemotactic stimulus within the amniotic fluid is the most direct way to obtain an even distribution of neutrophils over the surface of the placenta and membranes.
The question remains where does the chemotactic signal come from in chorioamnionitis? The stimulus could come from a focal area of infection or a generalized one. It could from inflammatory cells patrolling the connective tissue of the amnion and chorion. These monocytic cells show accumulation of meconium pigment just a few hours following exposure to meconium in the amniotic fluid. While bacteria on the amnion surface may be handled differently from meconium pigments, these macrophages are in a position to receive signals of infection, but it is unknown whether they could reverse transport a signal across the amnion into the fluid. Another possibility is that the fetal surfaces such as lung and possibly other fetal surfaces could secret cytokines into the fluid.
Knowledge of the details of the control of the chemical mediators of inflammation could be important to understanding the potential harmful effects of chorioamnionitis. How can a signal in the amniotic fluid or membranes have an effect on the relatively distant myometrium? How are inflammatory mediators transported into the fetal circulation? How can mediators of inflammation be suppressed to allow time for treatment of infection or replacement of hormones to reverse the pathway to preterm labor?
Is all chorioamnionitis the result of infection?
An early study in rabbits failed to produce chorioamnionitis with the intraamniotic injection of human meconium, rabbit amniotic cellular debris, gastric juice or acidified amniotic fluid[94]. Only various bacteria produced inflammation. The body’s innate immune system responds to bacteria via Toll like receptors in innate immunity and through specific immune responses to specific antigens from bacteria. In experimental models pure bacterial endotoxin can produce chorioamnionitis, but this unlikely to occur in nature. The failure to find a non-infectious cause does not prove there are no non-infectious conditions causing chorioamnionitis. If all chorioamnionitis could be shown to be associated with microorganisms, that would be very persuasive evidence that all chorioamnionitis was infectious. This has not be easy demonstrate. There is an imperfect correlation of positive cultures and histologic chorioamnionitis with both positive cultures not showing inflammation, and histologic chorioamnionitis present without a positive culture. The techniques of obtaining the culture have varied. Some have used amniotic fluid from amniocentesis or drawn from intrauterine pressure monitoring catheters[95], some have cultured the surface of the chorion after lifting the amnion[96-98], and others have cultured the placenta in various ways. The laboratory identification techniques have varied with only some for example capable of cultivating ureaplasma and mycoplasma (ref). A few have used nucleic acid amplification using primers to bacterial genes that were common to most bacteria(ref). The failure to prove an infection is not evidence that it is not present. The technique or sample could be faulty or the means of identifying an unknown microorganism inadequate. The possibility that a subset of chorioamnionitis is not infectious can not be excluded, but no plausible alternative mechanism has been proposed.
Bottom line on chorioamnionitis pathology
The routine histological diagnosis of maternal or fetal acute inflammation has good reproducibility between pathologists, and with the same pathologist on reevaluating the same case [64-66]. The finer details of neutrophil location and intensity there is greater variability between observers [82, 99]. Based on evaluation of reproducibility, the Society for Pediatric Pathology has produced a reasonable list of histologic features for clinical research[82]. In my own evaluation of cases, I use an even more simplified system. If there is inflammation only in the placental plate, I diagnose early chorioamnionitis. Finding this degree of inflammation is very common. Any other stage or grade of maternal inflammation is termed simply chorioamnionitis. Fetal inflammation is indicated as funisitis and then in parenthesis umbilical phlebitis or arteritis. I do note as a separate modifier if any inflammation shows the microscopic equivalent of pus, or extensive tissue necrosis. Necrosis or fibrin in fetal chorionic vascular inflammation is also diagnosed. Any remaining description is recorded in the microscopic description included in the pathological report.
I sample the membranes as a strip stuffed into a cassette based on a quick roll of the tissue. This appears adequate. For a study of preterm premature rupture of membranes, I and colleagues made careful membrane rolls orthogonally to each other starting from the rupture site. Except for changes confined to the likely rupture site based on marking the whole margin with black ink, there was no information gained from the careful 4 orthogonal rolls[100]. I was however very surprised by a paper using this same 4 directional sampling that found a linear increase in the number of cases diagnosed as chorioamnionitis based on how many of the samples were examined for a given case[101]. Not only does this run counter to my experience, but it implies a non-uniformity of chorioamnionitis that is certainly not evident on the routine microscope slide. If this observation is true, it would have profound implications for the spread of the inflammatory signal in the amniotic fluid, suggesting that the inflammation was based on a local area of infection or inflammation. The slides from my study were returned to NIH, and can not easily be accessed to test the same hypothesis by blindly diagnosing each slide, but someone should do the study. I am not persuaded by this small study (maximum 19 cases of chorioamnionitis) to change my sampling to include more membrane area.
Ascending infection occurs in a complex microscopic environment with real physical barriers, flows and distances. There is likely a constant parrying of offense and defense. Ecological factors are complex such as oxygen and nutrient levels, coexisting microorganisms, maternal history with the microorganisms etc. The pathologist sees the end result as a topology of neutrophil chemotaxsis. This identification of neutrophils in the membranes or umbilical cord is one of the more reliable placental observations made by pathologists in terms of consistency between observers[99, 102, 103]. Inferences about timing, etiology or mechanisms are limited. However, a standardized approach is useful for further study of clinical correlations with the inflammation. The Society for Pediatric Pathology has tested criteria for concordance among pathologists[82]. These are a reasonable set of histologic criteria and nosology for clinical studies. There is also a need for continued studies and refinement of identifying microorganisms in tissue, as well as cells and compartments critical in producing the inflammatory response and perhaps communicating that inflammation to the mother and her uterus.
What diseases are associated with chorioamnionitis?
From the historical section it is clear that congenital infant infection including pneumonitis, sepsis and meningitis are important and logical consequences of untreated infants. Maternal endometritis is less clearly associated with histological chorioamnionitis, since it is the contamination with pathogenic organisms that may be more important than evidence of inflammation. The association of preterm labor even with clinically silent histologic chorioamnionitis is well documented. Finally, the harmful effects, not of pathogenic infection, but of the inflammatory response itself on the fetus needs to be considered. These three correlations of infection, preterm labor and fetal inflammatory response are interrelated.
Does histologic chorioamnionitis predict fetal infection with a pathogen?
In the 1980 Ciba symposium was a presentation showing infection was an important cause of infant mortality at least in one area of South Africa. Drs. Tafari and Liungh-Wadstrom presented their experience in an area of South Africa with very high perinatal mortality 22 such deaths per thousand deliveries[104]. These infections greatly increased mortality after 34 weeks of gestation but the mortality below that gestation was high from non-infectious causes. Nearly 80% of the infections occurred without premature rupture of membranes. The cultures of fetal blood often included gastrointestinal pathogens including E. Coli (20%), Pseudomonas aeruginosa (11%), Enterobacter (11%), as well as Acinetobacter calcoaceticus (33%) and Staphylococcus epidermidis (15%). Group B Streptococcus was not present at the 1% level. In the early 1980’s the recovery of ureaplasma from the chorionic surface beneath the amnion was associated with perinatal mortality [105, 106]. Other studies could not confirm a specific correlation of Ureaplasma with mortality or preterm labor[107, 108]. In the later in the decade the concern was for an association of colonization with ureaplasma and chronic lung disease in the newborn .The problem with these observations was that the small motile ureaplasma are the most common microorganisms recovered in chorioamnionitis. The effects of fetal inflammation or of preterm labor associated with the ureaplasma infections were not always adequately separated from the infection itself.
Dr. Ronald Gibbs was lead author on several important papers studying the microbiology of amniotic fluid in patients with clinical chorioamnionitis which he and Dr. Duff summarized in 1991 article [95]. An early study of microbial culture of amniotic fluid obtained via a transvaginal pressure monitor demonstrated “virulent “ organisms at 100 or more colony units per ml in 70% of women with clinical chorioamnionitis compared to 8% in controls. A follow up study analyzed 408 cases and found that in fluid with multiple organisms some organisms were associated with each other, namely Gardnerella vaginosis, M. hominis, and anaerobes, suggestive of flora of bacterial vaginosis. A second cluster were enterococci, E. coli and gram negative rods, suggesting bowel flora. Low birth weight infants were more likely to have positive cultures for anaerobes such as Bacteroides bivus and Fusobacterium species. Group B Streptococcus and E. coli accounted for 20% of cultures, but 67% of maternal and neonatal bacteremia. No neonatal and only 2 cases of maternal bacteremia cultured anaerobes. Group B streptococcus accounted for 60 positive cultures of the 408, with 4 cases of neonatal sepsis and 2 septic deaths. E. coli had 33 positive cultures, 6 neonatal sepsis and one septic death. The other 315 cultures had 9 cases of sepsis and one septic death. Using the 70% to extrapolate to 580 patients with clinical chorioamnionitis, the risk of neonatal sepsis was 3%. If intrapartum antibiotic treatment eliminates the Group B streptococcal cases, then the risk is 2.5%. If the incidence of histologic chorioamnionitis alone were double the rate of clinical chorioamnionitis, then the risk of sepsis with histologic chorioamnionitis would be around 1.25%.
There is no question that chorioamnionitis caused by a pathogenic bacterial species can cause severe or lethal perinatal infection. Logically, congenital infection starts in utero and should demonstrate chorioamnionitis unless the organism is acquired by vaginal inoculation during delivery. The more difficult question to answer is: what is the positive predictor value of histologic chorioamnionitis for neonatal infection? Does the diagnosis offer any additional clinical value beyond the appropriate clinical measures and bacterial cultures? The negative predictive value for the mother can be diagnostically important, at least anecdotally, as I have had one case of seeming clinical chorioamnionitis with no histologic chorioamnionitis that proved to be maternal appendicitis. Even this scenario has practical problems as I have also often seen maternal fever associated with only very mild inflammation of the subchorionic intervillous fibrinoid. We can ask more specific questions about histologic chorioamnionitis and infection.
A question I have been asked by neonatologists is: should antibiotics be stopped in a preterm infant after three days if there is no histologic chorioamnionitis? I do not know the answer. There have been attempts in the past to incorporate the placental examination into the protocol for neonatal antibiotic treatment including rapid gram stain and even immune-fluorescent anti-bacterial stains and frozen sections of fetal placental surface for histologic chorioamnionitis[16, 109, 110]. These studies do not have application in current practice in which antibiotics are administered to complicated deliveries.
Do any features of the histologic inflammation in chorioamnionitis predict neonatal infection?
For the pathologist to see severe inflammation measured by grade and stage in infants without clinical infection is common. There may be some correlation between pathogenic organisms and the intensity of chorioamnionitis, but there is no evidence that the pathologic features of chorioamnionitis have sufficient positive or negative predicative value to be useful. In one study there was more fetal morbidity/mortality with chorioamnionitis if there was a positive culture for a pathogen than if there was only histologic chorioamnionitis and a negative culture[97].
Daily histologic examination of placentas confirms that there is a wide range of difference between maternal and fetal response in individual placentas. It is unusual for the fetal response to be more intense than the maternal. In general the more intense the maternal response, the more intense will be the fetal response [1, 111]. The one exception may occur with Group B streptococcal infection, but in this case cocci will usually be present. A masked antigenicity for Group B streptococcus for the mother has been proposed as a mechanism.(ref)
A study comparing amniocentesis results for culture and interleukin 6 levels with interleukin 6 levels in cordocentesis found elevated IL-6 in 10 infants with negative amniotic cultures, of which 4 of 9 had histological chorioamnionitis[112]. Ten fetuses with positive cultures had normal IL-6 levels. These results demonstrate that a fetus may have an elevated inflammatory cytokine without detectable infection or maternal inflammation. It also demonstrates that even with intrauterine infection there may be no fetal response. A lack of fetal inflammation does not mean that the fetus is not infected. In the case of a stillborn infant the presence of fetal inflammation is helpful to prove that the infection occurred prior to the death of the infant. Chorioamnionitis as a maternal response can occur with infection after the death of the infant.
Are there specific features of chorioamnionitis related to specific microorganisms?
The most microorganism specific lesion of the placenta is candida infection which forms small yellow plaques on the surface of the cord. Microscopically, the plaques are composed of neutrophils and Candida hyphae often visible on H&E stained slides, but more easily found with fungal stains such as methenamine silver(fig). The gross lesion may be subtle and overlooked, but if the microscopic surface shows neutrophil plaques then fungal stains are indicated to identify the yeast. If surface neutrophils are only directly adjacent to an intensely inflamed umbilical vessel, I may skip the special stain. The lesion is neither highly associated with a history of candida vaginitis or with candida septicemia in the infant, although the latter may occur with such infection (ref). Most infants appear to be unscathed including many premature infants.
Group B beta hemolytic streptococcus may be seen as an overgrowth of cocci frequently on the amnion surface. Paradoxically, severe neonatal infection may occur with minimal inflammation in the placenta, even when there is prominent overgrowth. This may reflect the rapid invasiveness of the infection in the infant or perhaps a muted response to the pathogen by the mother because of specific features of this microorganism. On the other hand particularly in previable infants the fetal response to Group B streptococcus may be intense (fig). The pattern of fetal inflammation more intense than maternal may suggest streptococcal infection, but is not useful diagnostically.
Fortunately with the legalization of elective abortion, septic abortion from clostridia is rare. I have only seen on case which was from a coat hanger induced abortion. The gram positive rods are easily identified by the pathologist, but the progress of the disease is so rapid that the tissue diagnosis is usually too late to be of any clinical value.
Obviously, the most specific placental finding is to identify the microorganism on the slide. This is not possible with routine staining in most cases. As a fellow I tried making a direct gram stain of the amnion and attaching it to a microscope slide. In most cases I could not identify any bacteria, let alone even predict gram positive or negative organisms. In the preterm premature rupture of membrane study that I participated in, the statistician asked if I had any secondary analysis, and I asked if there were any correlations with the histologic identification of bacteria. The numbers were too small but for that sample unpublished data found no correlation.
The one situation in which the identity of specific microorganism species is clinically helpful occurs when blood borne infection secondarily infect the membranes or umbilical cord. Herpes simplex virus and toxoplasmosis may directly infect the amnion and be identifiable. More frequently organisms are found in the umbilical cord in a lesion named subacute necrotizing funisitis.
What is subacute necrotizing funisitis?
This lesion is a response to microorganisms in Wharton’s jelly of the umbilical cord. The gross appearance is distinctive creating a “barber pole” line of white opacity that follows the vascular helix of the vessels at a distance (fig). The exterior will appear at least focally transparent and yellow. On cross section, or on a microscope slide, there is an arc of necrosis, usually calcified, between the cord surface and the vessel. [113-116] (fig). This appearance of subacute nercrotizing funisitis has been likened to an Ouchterlony plate. This plate consists of a gel with circular wells in which antigen is placed in one well and antibody to the antigen in an adjacent well. As antigen and antibody diffuse toward each other there is point at which the relative concentrations cause aggregates of antigen-antibody complex to precipitate. The precipitation forms a white arc between the wells. This analogy is not only visually but also dynamically apt. Wharton’s jelly has no capillaries or lymphatics and is the equivalent of the Ouchterlany gel. The vessels act like circular wells in the plane of cross section, and the outside of the cord acts as the edge of the well against the amniotic fluid. In the usual funisitis, chemotactic stimuli entering beneath the amniotic skin of the cord diffuse inward, neutrophils respond with an outward migration from the vessels. In necrotizing funisitis, microorganisms are traveling in Wharton’s jelly inward from the surface circumference. When they encounter the outward migrating neutrophils, they will create a battle line with degranulation of neutrophil enzymes that produces the chalky necrosis equivalent of Ouchterlony precipitation. This encounter can sometimes be seen directly on the microscope slide (fig).
The reported series of this lesion found an association with stillbirth, prolonged premature rupture of the membranes and an increased risk of bacterial infection in the newborn. The incidence was on the order of 1 per 1000 deliveries. Our incidence of the lesion in unselected pregnancies is only 1 in 3000. It seems likely that earlier delivery in preterm premature rupture of membranes may be reducing the incidence and mortality of the lesion, although we do not have enough data to prove it. A prospective study of preterm neonates compared 5 with necrotizing funisitis to 13 without and found a significant increase in measures of chronic lung disease at the p0.03 level[117]. Cervical cultures did identify pathogenic bacteria in 4 of the mothers (Bacteroides fragilis, Bacteroides bivius, E coli, Klebsiela pneumoniae, Enterococcus sp. and coagulase negative enterococci). Perhaps counter-intuitively, the cord blood neutrophil count was significantly elevated in the necrotizing funisitis group. Not only bacteria but candida and other organisms can produce the lesion. In one report culture of the Wharton’s jelly grew a dental organism Actinomyces meyeri, a dental organism[67]. I have sent Wharton’s jelly samples for microbiologic culture from umbilical cords with subacute necrotizing funisitis and never had a positive culture, but this technique could be useful in proving infection with unusual species. The reports suggest that subacute necrotizing inflammation may be the result of a longer duration of infection, but the ability of the microorganism to penetrate into Wharton’s jelly could also be a factor.
There is another lesion with necrosis in the umbilical cord that was carefully described by Navarro and Blanc who had first named subacute necrotizing funisitis, and they called the lesion chronic funisitis[118]. The lesion is characterized by necrosis of the vessel media often with calcification, a lymphocyte and plasma cell infiltration, and villitis. (fig) Cases of funisitis with neovascularization (granulation tissue) likely also fall into this category since medial necrosis is likely to result in the outgrowth of endothelium (fig). The dynamic of this lesion is different than subacute necrotizing funisits. The organisms are escaping from the umbilical blood vessels into the relative safety of the capillary free Wharton’s jelly only to be pursued by already primed effector lymphocytes which cause tissue destruction starting with encounters with microorganisms in the vessel wall. As a result the bands of necrosis are more likely to be circumferential rather than arcs between the vessels and the surface, and there is less likely to be a viable area between the vessel and the band of necrosis. Neutrophils may be present, but unlike subacute necrotizing funisitis, they are not the majority inflammatory cell. This chronic lesion was associated with viral inclusions in the initial report.
The published literature however does not always clearly separate subacute necrotizing funisitis from chronic funisitis. Controversy about the significance of subacute necrotizing funisitis occurred with the publication of a paper suggesting that all cases were due to syphilis. Fojaco et. al. found 16 cases with necrotizing funisitis, all with congenital syphilis, from a database of 465 perinatal autopsies[119]. The database had 35 cases of congenital syphilis (32 stillbirths, 3 neonatal deaths) which results in 46% with necrotizing funisitis, but no cases of necrotizing funisitis without syphilis. Their autopsy database is weighted with a high incidence of syphilis. Of the 16 cases, all had some serological evidence of syphilis (although in 4 this was only a reactive RPR). Ten cases were stained with the Steiner stain, and only 4 showed spirochetes. However, the illustrations of the lesion are those of chronic funisitis which is more common with syphilis [120, 121].
Chronic funisitis is usually the result of blood borne infection, not ascending infection. Often, especially with syphilis, Wharton’s jelly is the practical place to find large numbers of blood borne microorganisms compared to other areas of the placenta. However, the distinction between chronic and subacute necrotizing funisitis is not always clear. Schwartz in reviewing the pathology of congenital syphilis and subacute necrotizing funisitis comments that in three cases spirochetes were found beneath the surface epithelium of the cord suggesting a “transamnionic epithelial portal of entry of spirochetes into the umbilical cord[120]. A reported case of Herpes simplex virus and subacute necrotizing funisitis demonstrated virus beneath the cord epithelial surface[122]. Often it is not possible to know whether intrauterine HSV infection is an ascending infection or a blood borne infection that spreads to amnion and umbilical cord from the fetal skin. However, these cases suggest that a blood borne infection reaching the amniotic fluid could then produce subacute necrotizing funisitis.
.
Bottom Line on subacute necrotizing funisitis
With typical subacute necrotizing funisitis and acute chorioamnionitis without lesions of candida funisitis, the infection is likely bacterial. This might warrant more careful observation of the infant for infection. With chronic funisitis the pathologist needs to use special staining to identify blood borne infection, typically HSV, syphilis or toxoplasmosis and the clinician should search for such infection in the infant. (fig) With subacute necrotizing funisitis without visible bacteria, finding atypical histologic features, villitis, or clinical suspicion, would also justify looking further for specific blood borne infection. If in doubt about candida funisitis, a fungal stain can usually resolve the diagnosis.
Does chorioamnionitis cause preterm labor?
The correlation of preterm labor with an increased incidence of histologic chorioamnionitis is well established (refs and ranges). Discovering this association was a major advance in the understanding of preterm labor that had required the input of tissue pathology. As I finished my fellowship and was looking for a position I visited Dr. Naeye a pathologist involved in this research at Hershey in 1981. He was adamant that preterm labor was no longer an area of research. It was due to chorioamnionitis, it was simply an infection. Unfortunately it did not to prove to be so simple. Antibiotic treatment of chorioamnionitis did not prevent preterm labor(ref). Even prenatal antibiotics given to women with HIV did not decrease the incidence of preterm labor [123]. Antibiotic treatment did not prevent preterm labor when used between pregnancies [44, 124]. Chorioamnionitis was certainly important in driving preterm labor. Terbutaline treatment was not effective if chorioamnionitis was present. A meta-analysis of published studies found that early clindamycin treatment in women with bacterial vaginosis can reduce preterm delivery prior to 37, but not 33 weeks of delivery[125]. However, the clindamycin treatment had no effect on other outcome measures related to prematurity. The meta-analysis does not prove that bacterial vaginosis causes preterm labor.
The role of inflammation in labor is often complex. For example, a study of women with short cervix using amniotic fluid MMP-8, neutrophil collagenase, as a marker of inflammation, found that 4 of 10 women with elevation of MMP-8 versus 3 of 35 without this marker delivered within 1 week, a significant difference[126]. However women with and without inflammation both had an increased risk of preterm delivery. There was no placental examination reported. Did all the patients delivering in the 1 week interval show chorioamnionitis, or did some deliver despite not having infection? The use of MMP-8 as a marker is ambiguous. In some Cesarean hysterectomy specimens the cervix shows acute inflammation, and the placental membranes may show inflammation just over the cervical os. Could this cervical inflammation elevate amniotic fluid markers of neutrophil inflammation? ( A subsequent similar study used an array of inflammatory markers and has similar results, but the same questions remain[127].) The study could be taken as evidence that factors causing short cervix perhaps with cervical inflammation cause preterm labor, or it could be interpreted that inflammation increases the risk of preterm labor in this setting but does not cause it. On the other hand if only those patients who developed chorioamnionitis in the context of short cervix delivered early, then chorioamnionitis could be part of the chain of causation, and preventing or effectively treating the chorioamnionitis, might prevent the preterm labor.
Even if chorioamnionitis can produce preterm labor it begs the question what causes chorioamnionitis. This problem has be discussed under the question “How do microorganisms get to the the intrauterine cavity?” One answer was that the microorganisms were already in the endometrium of some women. This leads to the next question of why are bacteria present in only some women. Other studies have tried to show that having elevated ureaplasma concentration[128, 129] or bacterial vaginosis in the vaginal flora is the predisposing factor[130, 131]. The problem with all such studies is that the vaginal flora is complex and changing and they beg the question why a certain bacterial flora is present. The underlying first cause remains elusive.
Since preterm labor is a risk factor for preterm labor, it would be interesting to know if recurrent spontaneous preterm labor in subsequent pregnancies had the same histology, particularly to know if chorioamnionitis was recurrent. I had collected a set of slides to test this hypothesis many years ago as I was transitioning positions, and never completed the study. However, I was hopeful that someone else had done the study when I saw a paper subtitled Placental histology in recurrent spontaneous and indicated preterm birth”[132]. The paper was content to demonstrate that spontaneous preterm labor had very significantly increased rate of chorioamnionitis compared to indicated preterm labor, presumably mostly due to toxemia or intrauterine growth retardation. Of the subjects, 83 did have a prior spontaneous preterm labor, and 71 had spontaneous preterm labor on the subsequent pregnancy which was the placenta studies histologically, but there was no histological one to one comparison with the earlier pregnancy. The histology for the earlier pregnancies did not appear to have been reviewed. The degree of recurrence of histological chorioamnionitis in recurrent spontaneous labor seems to be still unknown.
Direct evidence that chorioamnionitis causes preterm labor has been obtained by measuring inflammatory mediators or culturing microorgansims from amniotic fluid to predict preterm delivery. One problem common to some studies is that the amniocentesis was obtained at presentation for preterm labor, and the results used to predict the interval until preterm delivery (PROM)[12, 133-135]. This only proves that elevated cytokines and positive cultures favor delivery not that they initiated the process. They may also be taken as evidence that current tocolyitic therapy was not successful in the presence of chorioamnionitis. The results still elucidate some aspects of the relationship between preterm delivery and chorioamnionitis. In one study only the fetal level of Il-6 not the amnion level of IL-6 was correlated with preterm delivery within 7 days in mothers with preterm PROM suggesting that it may not be chorioamnionitis per se, but the fetal inflammatory response that produces preterm labor[136]. The serum also demonstrated a rise in plasma cortisol, but not of dehydroxyepiandrosterone sulfate[137]. The implication is that inflammation produces fetal stress response without reducing maternal estriol levels, or at least the substrate for those levels. In the sheep, a corticosteroid induced decrease in fetal estrogen production leads to labor, but this mechanism does not occur in primates, at least in the serum level changes seen in sheep,
While not provable, the survival of the mother by expelling an infected pregnancy should have a selective advantage for propagation of the species. Given a preterm infection, what mechanism developed to expel the infant, i.e. produce preterm labor? Labor is usually defined as effective uterine contractions. On a cellular level this occurs because of a profound phenotypical change in the smooth muscle cell of the uterus. Gap junctions develop between cells, the channels generating electric change, and oxytocin receptors become prominent. The easiest way to send a signal to the uterus would be via maternal blood. Another way might be a propagation of paracrine signals perhaps starting in the cervix, or in the endometrium. The maternal blood is accessible to the fetus presumably thorough the villous circulation. A signal in the amniotic fluid would either be via the fetus or directly across the membranes. The latter implies a primary role for the myocytes just beneath the uterus. One study suggested that bacterial phospholipase A2 could produce prostacyclins from amniotic phospholipids as a direct paracrine mechanism of producing labor[138]. Another study of experimental chorioamnionitis in the monkey demonstrated loss of catabolic Prostaglandin Dehydrogenase in the chorion membrane epithelium, which could increase the amount of prostaglandin reaching the decidua[139]. However the severity of the chorioamnionitis caused necrosis of the chorion which is more extreme than most chorioamnionitis, and may not a mechanism in less intense infection. Cyclo-oxygenase inhibitors which block prostaglandin production have been used to inhibit preterm labor (but have a potentially harmful effect in that decreasing fetal prostaglandins may result in contraction of the ductus arteriosus producing elevated right sided heart pressures). Unlike terbutaline, cyclo-oxygenase inhbitors such as indomethacin, may inhibit preterm labor even with infection. In experimental bacterial endotoxin (lipopolysaccharide) produced preterm birth in mice, cyclo-oxygenase inhibitors significantly decreased preterm birth[140]. Since prostaglandins even inserted in the vagina can induce labor, this is a plausible signal for initiating labor with infection, but the complexity of the inflammatory system may have other interactions with myometrium.
Have animal models clarify the role of chorioamnionitis and preterm labor?
A study injecting 106 colony forming units of group B streptococcus intra-amniotically into monkeys demonstrated a rise of inflammatory mediators in 12 to 15 hours, and preterm labor after just 36 hours[141]. The placentas had severe chorioamnionitis and funisitis. There was extensive fetal infection. Of course this is data for one particularly pathogenic species and there is no way to know how the dose compares to natural inoculation. However in these circumstances the process is relatively rapid. This research extended to include chorio-decidual injection of streptococci and the measurement of fetal and maternal steroid production. The route of injection was not important. The important result was that infection elevated fetal adrenal steroid output, but not maternal estrogens. The latter were elevated in spontaneous parturition. To test the effects of different cytokines, they were intra-amniotically injected individually into the monkey and interleukin 6 and 8 elicited membrane neutrophils, but did not initiate labor. Interleukin 1 beta and to a lesser extent tumor necrosis factor did elicit labor and inflammation. While this is a non-physiologic dosing, the results suggest that fetal membrane inflammation alone does not produce labor.
Other animal models of infection have not been particularly enlightening. A rabbit model injection of Gardnerella vaginalis into the amniotic fluid of rabbits did not produce preterm labor, did not show histological inflammation in the placenta or fetal tissue, and did not produce a change in “brain score” different from the sham injected controls[142] . A model of endotoxin injection in pregnant sheep did produce inflammation, after 6 days with intra-amniotic infusion, and after 2 days with infusion beneath the membranes although the effect was predominantly near the pump, but there was no preterm labor[143].
The monkey experimental streptococcal model also demonstrated a different fetal endocrine profile than spontaneous labor. Liggen’s well known model of the labor in the lamb initiated by fetal cortisol elevation with fetal maturation which in turn suppresses progesterone production and increases the estrogen to progesterone level in maternal blood which signals myometrial cells to transform into labor producing cells. In the primate, the fetal adrenal with its predominance of dehydroepiandrosterone (DHEAS) production, the precursor to estriol, made the signaling of fetal maturity to the mother’s uterus less direct. In the monkey model spontaneous labor is preceded by a marked elevation of DHEAS in fetal blood, but not maternal[144]. In the streptococcal model of preterm labor, there is no elevation of DHEAS in the fetal blood, although there is some increase in fetal cortisol. This implies that infection uses a different pathway in initiating maternal labor than does fetal maturation.
One problem with this research is that it is not aimed at the question of what produces chorioamnionitis in the natural state. Research could test theories of cervical manipulation or of other stresses to the mother or fetus that have been associated with preterm labor to understand better the mechanism. Just the fact that manipulation of the primate uterus for fetal surgery induces labor provides a way to look at the onset of cervical changes and bacterial movement without injecting microorganisms. There are many basics questions that need answering,
Using monkeys is necessary because only primates have the same fetal adrenal axis and the same type of hemochorial placenta as human beings. Most of the work that I have cited is from the Oregon Primate Research Center which certainly has been creative and important. I think such studies needed to be expanded, more open to outside collaboration, and well-funded. Such studies are expensive and are likely to be opposed by animal rights activists, but some this must be balanced against the massive human suffering that our failure to prevent preterm labor causes to both the children and the parents. Even with the incredible advances in the care of the preterm newborn that I have seen in my career, I still have performed over 1000 autopsies whose underlying cause of death is prematurity. Many of these died because they were born before whatever the current neonatal treatments considered viable. Certainly some carefully planned research would pay back if it advances our understanding of the underlying causes of prematurity.
The most evident conclusion of this animal research is thatt he direct injection of microorganisms into pregnant uteri in animals induces labor and chorioamnionitis. This is seldom the route of infection in humans, so the conclusion is useful, but does not answer the question how do woman acquire chorioamnionitis in the usual clinical setting of preterm labor.
Why do some women get chorioamnionitis?
Chorioamnionitis is an inflammation in an anatomic area, like pneumonia. You get pneumonia because you have been exposed to powerful pathogen in the air like Legionella, or you have a condition that defeats your immune defenses like cigarette smoking, hypoventilation or influenza or perhaps you have a very specific physical accident such as aspiration of irritating or infected material. The risk factors for chorioamnionitis as discussed often center around the loss of the cervical barrier as in prolonged labor and perhaps even with preterm labor, but these are not the only risk factors Even if loss of the cervical barrier is a key initiating event, not everyone with prolonged labor or ruptured membranes suffers chorioamnionitis.
There are many studies of risk factors for preterm labor, and since preterm labor associates with an increased incidence of chorioamnionitis, it is tempting to use preterm labor as a surrogate marker for chorioamniontis. In doing so, there is the problem that a substantial number of women with preterm labor do not have chorioamnionitis, and that studies of preterm labor are not usually stratified for histologic chorioamnionitis. I will however occasionally use this shortcut, see below.
Epidemiologic studies have found many risk factors for chorioamnionitis which need confirmation, for example alcohol consumption, anti-B antibodies and anemia in one study of clinical chorioamnionitis[145]. A study of preterm infants found two interleukin 6 genotypes were more prevalent with histologic chorioamnionitis, but another with fetal septicemia[146]. This study reflects the idea that Dr. Blanc had in the 1980 Ciba symposium, that a strong inflammatory response may be protective of neonatal infection in chorioamnionitis. Epidemiologic studies of preterm labor must cope with complex confounders that can profoundly alter the conclusion, as in a study of histological chorioamnionitis and race in which the raw data found an significant excess in African-Americans that was substantially reduced when various cofactors were considered[147]. Getting to causation from epidemiologic associations is often a tangled path.
Can pathological examination of the placenta contribute to the understanding of the relationship between chorioamnionitis and preterm labor?
Pathologists in general approach use a different inferential process to look at causation. The specimen of a single patient is examined in detail to try to understand the process that caused the disease. Pathologists have compared the placentas of women with preterm labor with and without chorioamnionitis, and these two groups appear to have a fundamental difference. Those without chorioamnionitis often have advanced villous maturation consistent with utero-placental ischemia while those with chorioamnionitis appear less mature [148, 149]. My own experience is concordant with this observation even in patients without any clinical evidence of preeclampsia. This finding suggests that utero-placental ischemia can initiate labor.. It is possible that the ischemic group would have other differences such as serum Flts-1 level or smaller abdominal girth for gestation.
This observation does not explain why those without utero-placental ischemia get chorioamnionitis. Perhaps the most salient pathological observation as discussed in the historical section, but confirmed by others, it that chorioamnionitis is associated with preterm gestation, and from later studies peaks at 20 to 24 weeks as a percentage of placentas showing histological chorioamnionitis[150]. This gestational age peak could reflect a counter process of more early deliveries due to ischemia as gestational age progresses.
The pathology can be used to monitor the degree of inflammation in relation to new treatments for preterm labor. In the preterm premature rupture of membrane study in which I examined the membranes there was no difference in patients with delay of an average of three days to delivery with antibiotic treatment and those without. This implies that if antibiotics did anything to the inflammatory response they delayed it, but did not eliminate it or show evidence of healing. The study also demonstrated patients with or without antibiotics could have preterm premature rupture of membranes and preterm delivery without chorioamnionitis.
The pathology of the membranes can also be useful in stratifying other risk factors. I had started but didn’t finish looking at the placentas of woman with recurrent preterm labor. Since preterm labor is a recognized risk factor for preterm labor, I wanted to know if there was a pattern of recurrence of preterm labor with or without chorioamnionitis. Fortunately other investigators did complete the study (ref).
At one point when I was at the University of Cincinnati I participated in a brain storming session for a program project grant for preterm labor. An engineer presented biosensors as perhaps being of use to us. Following up on that a group of us proposed using such sensors to record cervical changes such as pH and compliance first in sheep with hormonally induced cervical labor and then in patients undergoing cerclage. The sensor would be included in the cerclage suturing, and possibly could even radio out a signal. For various reasons the project did not occur, but the idea of being able to look at cervical changes and in some way measuring intrauterine inflammation or time to delivery to establish the relationship of cervical change to chorioamnionitis seemed promising in understanding preterm labor. Sensors now could look identify the presence of specific biomarkers in the cervix.
A major problem with membrane pathology is that the results are not available until after delivery. This is akin to the observation that the autopsy provides the most information, only too late. What is needed is a marked of chorioamnionitis in the blood that can specifically identify intrauterine inflammation. Perhaps the S100 protein family that responded to streptococcal injections in monkeys can be measured in maternal serum. This segues into the next question.
What is the fetal inflammatory response?
The answer is both obvious and subtle. The obvious is that the fetus responds to the infection in the amniotic cavity, so that is a fetal inflammatory response. The subtle is that the fetal response has its own peculiarities that in some circumstances could cause fetal injury. The fetus not only has incompletely developed organs, but many functions are qualitatively different from the older infant. While the inflammatory response in the fetus in utero is similar to the neonate the consequences may be different because of the intrauterine physiology. The pathologist may be able to contribute to the understanding of fetal inflammation by the autopsy examination of previable infants who are normal except for preterm delivery. Not usually considered in the fetal inflammatory response is that some bacteria can produce septic shock with ischemic organ injury or death in utero. These effects will be considered in a separate discussion of fetal shock.
The migration of neutrophils to the infection may stress the marrow and liver capacity to produce mature neutrophils resulting in blood neutropenia. The up-regulation of corticosteroid production causes accelerated involution of immature fetal thymocytes in the thymus, and stimulation of the permanent cortex of the adrenal and perhaps decreased fetal cortical production of DHEA, the precursor to placental estriol.
Is infection the cause of the fetal inflammatory response?
Clearly the localization of fetal neutrophils in the umbilical cord and surface vessels of the placenta is a direct marker of a fetal inflammatory response in chorioamnionitis. Many fetuses showing a maternal histologic chorioamnionitis do not show a fetal neutrophil response. This does not mean that fetal inflammation is not occurring especially as microorganisms can enter the fetal lung and gut. One study did show some looseness in the relationship between fetal inflammation and amniotic infection. This study used umbilical cord blood levels of the cytokine interleukin 6 (Il-6) as a marker of fetal inflammation obtained by cordocentesis of patients with preterm labor or preterm premature rupture of membranes and found that some fetuses had an elevated Il-6 despite negative amniocentesis cultures and even absent histological chorioamnionitis. Other patients with positive bacterial cultures of the amniotic fluid did not have elevated IL-6. Interestingly some fetuses with positive Il-6 and positive cultures on admission did not deliver until 7 days later, the outer limit of the study. In experimental inoculation of intraamniotic Group B Streptococci, Il-6 was not as sensitive measure of the fetal response as interleukin 1B, interleukin-1 receptor antagonist or tumor necrosis factor[141, 151]. A study of amniotic fluid from women in very early preterm labor found a correlation of tumor necrosis factor elevation with elevated heat shock protein and with mycoplasma detected by gene amplification[152] A study injecting lipopolysaccharide directly into fetal lung or gastrointestinal tract demonstrated different effects on inflammatory mediators[153]. As in the adult the fetal inflammatory response may depend on many factors, and generalizations may not be as important as the specifics of the response in an individual patient or disease.
The current interest in the fetal inflammatory response however usually is often focused on the potential for fetal brain injury. This susceptibility is due not only to the immaturity of the brain but also to the unique fetal circulation. Unoxygenated fetal blood enters the two umbilical arteries from the two iliac arteries. These arteries merge at the placental surface (anatsomosis of Hyrtl) and then branch becoming end vessels in stem villi of the placenta. Within the capillaries of the villi, blood is oxygenated and transported upward ending in a single umbilical vein which enters the ductus venosus and inferior vena cava. The anatomy of the fetal heart is such that the foramen ovale is almost continuous with the inferior vena cava which is covered by a membrane septum but flaps freely allowing blood to flow more or less directly from the umbilical vein into the left ventricle. The result is that the oxygenated blood from the umbilical vein is pumped from the left ventricle to the aorta and goes preferentially to the brain and upper extremities. The blood from the superior vena cava is ejected from the right ventricle into the pulmonary artery but the high resistance in the unventilated lung drives the blood into the ductus arteriosis a shunt vessel that joins the aorta just below the left subclavian artery. As a result the deoxygenated blood from the head and arms is distributed to the viscera, lower extremities and back to the placenta. During normal development this benefits the brain with the most oxygenated blood. As is discussed with chorionic vessel thrombi, this arrangement does not protect the brain from substances or emboli coming in from the placenta. With chorioamnionitis this would include inflammatory cytokines coming from inflamed vessels in the placenta and umbilical cord. In the non-fetal circulation, cytokines and other inflammatory mediators would pass to the right side of the heart and flow to the lung capillaries where they would be deactivated. The foramen ovale would be closed because left atrial pressures would now be higher than right atrial pressures, and the ductus arterosis would have actively constricted closed.
One of the serious complications of preterm birth is periventricular leukomalacia or leukoencephalopathy, basically injury to the white matter of the cerebral cortex. This injury can be macroscopic necrosis of regions of the whiter matter that in the process of healing become fluid filled cysts or areas of calcification in the brain. This gross lesion can be seen directly by the pathologist at autopsy and can be seen my cranial ultrasonography of the infant. The seminal study relating chorioamnionitis and perventricular leukomalacia was done by Dr. Bejar and colleagues who found a significant association of periventricular leukomalacia found by ultrasound examination on all premature infants with clinically purulent amniotic fluid and with inflammation of the umbilical cord (funisitis). The funisitis did not have a high predictive value for leukomalacia but the association proved the concept. These studies were refined by others who found an association of white matter injury in preterm infants with higher grades and stages of chorioamnionitis and with markers of fluid or cord blood cytokines. These associations were not perfect with both brain injury without inflammation and with many cases of inflammation without brain injury.
Prior to seminal studies descrbied above, pathologists had attributed neonatal white matter necrosis to ischemia based on the white matter being the end circulation of the branches of arteries dispersed over the cortical surface. If the fetus became hypotensive, this would be the area of the brain most vulnerable to hypoxia.In my fellowsip I was introduced to an influential paper by the pediatric neuropathologist Dr. Banker correlating neonatal white matter injury with asphyxia. The neuropathologists from the perinatal collaborative study found that the white matter lesions were coagulative necrosis, that is consistent with infarctions. In primates under special conditions of chronic asphyxia white matter lesions could be experimentally produced in fetal monkeys. Since chronic asphyxia can produce severe white and gray matter injury any experimental study must avoid brain ischemia or fetal shock. At least one model of lipopolysaccharide induced chorioamnionitis in the sheep did not measure pressures and described brain lesions akin to those of asphyxia[154].
The change in the accepted cause of the lesion in the recent decades probably has two sources. First , there is the improved survival of many younger gestation infants but with a persistent percentage with some neurologic impairment. The second source is the advances in brain imaging that allow physicians to see more subtle and more diffuse white matter loss than was possible with autopsy studies. These changes focused attention on the specific risk of young premature infants for white matter injury. The novel theory that inflammatory mediators could cause the injury perhaps by a direct effect on immature oligodendroglia, the precursors of mature white matter, led to new research and insight. So far this research has not led to the prevention of white matter injury in premature infants, and as with all research raises new questions.
When does white matter injury become irreversible?
This question could be asked in other way such as what is the progression of white matter injury? I tried to compare white matter in previable infants dying in labor or with a few hours of delivery between those with and without chorioamnionitis. I used our normal H&E stain, a PAS stain which can show injury, and a glial fibrillary protein immunostain to look for early evidence of scarring. While these stains were easily available, they are not particularly specific or sensitive to any early changes in the cells, so I was not totally surprised to not find any lesions that I could recognize except for an occasional hemorrhage in the germinal matrix in any of these infant brains. I did not publish such totally negative results because they were expected. In hundreds of autopsies of otherwise normal infants delivered because of preterm labor or rupture of membranes before viability, brain lesions are very rare. The one case that I did find of brain injury in a previable infant had evidence of prolonged infection with follicle development in the spleen and involution of the thymus. (fig). I tried a second tactic and collected a small aliquot of cord blood within 30 minutes of delivery on infants with birth weight less than 1500 grams. This was considered a waste specimen. The institutional review board also allowed that we could we could wait for the first ultrasound results in 24 hours and then label the tube either positive or negative based on those results and as long as all identifying information was discarded at that point, we did not have to ask for research consent. This seemed like a fair trade off, since we had been given the opportunity for one proteomics comparison between a positive and a negative case to see if there were differences that would justify applying for funding for a full study. Thirty specimens later all the blood was negative (no white matter lesion) and we cancelled the study. The moral is that there early lesions are not easily detected. It does not mean they are not there and irreversible, but the morphologic intactness of most brains gives hope that any effects of chorioamnionitis might be reversible.
Does chorioamnionitis cause other injury to the infant?
The lung is an obvious target for injury from chorioamnionitis. In previable infants the lungs often have neutrophils. If bacteria are also present, some inflammatory injury could be occurring in the lung even without an overt neonatal pneumonia. In an autopsy study, infantsless than 24 weeks of gestation with chorioamnionitis were more likely to have bronchus associated lymphoid aggregates[155]. However, there are other studies suggesting that inflammation, probably via the adrenal cortical stimulation, advances lung maturation and that chorioamnionitis is associated with less severe neonatal lung disease.
In addition, chorioamnionitis is usually associated with acute involution of the thymus which is reasonably interpreted as evidence of an increase in circulating corticosteroid. The acute involution of the thymus is recognized by the starry sky appearance of the thymus which refers to the phagocytosis of lymphocytes within the cortex leaving clear areas in the otherwise deep blue lymphocyte background. This lesion can be produced experimentally in neonatal rabbits by injection of steroid. The lymphocytes being phagocytized are immature thymocytes CD4 and positive. The elimination of these cells occurs normally over a longer period. The accelerated acute involution with stress has no known immune effect, and cardiac surgeons often discard the thymus during neonatal surgery without any apparent consequences. Chronic stress such as with rh alloimmunization and fetal hydrops the thymus may be completely involuted, that is devoid of cortical lymphocytes, at the time of birth. Other effects of this endogenous corticosteroid elevation are unknown, but corticosteroids are often given as a bolus to mothers in preterm labor to accelerate lung maturation.
Neutrophils are found in the stomach in most previable infants with chorioamnionitis, and they may also be found much further even into the colon. If bacteria and inflammatory cells are present could they interfere with normal development of gut flora and increase the risk of necrotizing enterocolitis. The evidence is against this hypothesis.
Chorioamnionitis appears to increase hematopoiesis in the fetus. Particularly in the preterm infant but even at term hematopoiesis is a function of the liver. Myelopoiesis occurs in the triads, and at midgestation is very prominent, but at least one study has found an increase with chorioamnionitis. Erythro and megakaryo poiesis occurs in the liver sinusoids. An elevation of nucleated red cells is with fetal inflammation is usually attributed to up-regulation of erythropoietin. However in a mouse model of intraamniotic injection of lipopolysaccharide there was an increase in nucleated red cells six hours later without an increase in erythropoietin[156]. As with asphyxia, injury to the endothelium of the sinusoids can not be discounted as a mechanism for premature release of immature erythroblastic cells. Chorioamnionitis in at least one study was associated with an increase in subcapsular hemorrhage of the liver which could also be explained as endothelial injury. The significance of subcapsular hemorrhage must be approached with caution as they can be inadvertently produce at autopsy my pressure on the fetal liver. In some cases these is evidence of surrounding tissue injury to support an in vivo origin. Severe toxemia, a disease with endothelial injury may also have large liver hemorrhages.
Other organ damage:
Some studies have found a correlation of fetal histological inflammation and NEC[157]
1. van Hoeven, K.H., et al., Clinical significance of increasing histologic severity of acute inflammation in the fetal membranes and umbilical cord. Pediatr Pathol Lab Med, 1996. 16(5): p. 731-44.
2. Blanc, W.A., Pathology of the placenta and cord in ascending and in haematogenous infection. Ciba Found Symp, 1980. 77: p. 17-38.
3. Naeye, R.L., Disorders of the Placenta, Fetus and Neonate, Diagnosis and Clinical Significance. 1992, St. Louis: Mosby Year Book.
4. Ross, S.M., et al. The genesis of amniotic fluid infection. in Ciba Foundation Symposium 77. 1980.
5. Naeye, R.L. and N. Tafari, Risk Factors in Pregancy of the Fetus and Newborn. 1983, Baltimore: Williams & Wilkins. 361.
6. Bejar, R., et al., Antenatal origin of neurologic damage in newborn infants. I. Preterm infants. Am J Obstet Gynecol, 1988. 159(2): p. 357-63.
7. Vigorita, V.J. and T.H. Parmley, Intramembranous localization of bacteria in beta-hemolytic group B streptococcal chorioamnionitis. Obstet Gynecol, 1979. 53(3 Suppl): p. 13S-15S.
8. Qureshi, F., et al., Candida Funisitis: A Clinicopathologic Study of 32 Cases. Pediatric and Developmental Pathology, 1998. 1(2): p. 118-24.
9. Andrews, W.W., et al., Amniotic fluid interleukin-6: correlation with upper genital tract microbial colonization and gestational age in women delivered after spontaneous labor versus indicated delivery. Am J Obstet Gynecol, 1995. 173(2): p. 606-12.
10. Aquino, T.I., et al., Subchorionic fibrin cultures for bacteriologic study of the placenta. Am J Clin Pathol, 1984. 81(4): p. 482-6.
11. Hillier, S., et al., Microbiologic Causes and Neonatal Outcomes Associated with Chorioamnion Infection. American Journal of Obstetrics and Gynecology 1991. 165(4): p. 955-961.
12. Hillier, S.L., et al., The relationship of amnionic fluid cytokines and preterm delivery, amniotic fluid infection, histologic chorioamnionitis, and chorioamnion infection. Obstet Gynecol, 1993. 81: p. 941-8.
13. Jacques, S. and F. Qureshi, Subamnionic vernix caseosa. Pediatr Pathol, 1994. 14: p. 585-93.
14. Markenson, G.R., et al., The use of the polymerase chain reaction to detect bacteria in amniotic fluid in pregnancies complicated by preterm labor. Am J Obstet Gynecol, 1997. 177(6): p. 1471-7.
15. Jacobsson, B., et al., Microbial invasion and cytokine response in amniotic fluid in a Swedish population of women with preterm prelabor rupture of membranes. Acta Obstet Gynecol Scand, 2003. 82(5): p. 423-31.
16. Evaldson, G.E., A.S. Malmborg, and C.E. Nord, Premature rupture of the membranes and ascending infection. Br J Obstet Gynecol, 1982. 89: p. 793-801.
17. Galask, R.P., et al., Bacterial attachment to the chorioamniotic membranes. Am J Obstet Gynecol, 1984. 148(7): p. 915-28.
18. Fredricks, D.N., T.L. Fiedler, and J.M. Marrazzo, Molecular identification of bacteria associated with bacterial vaginosis. N Engl J Med, 2005. 353(18): p. 1899-911.
19. Reid, D.E. and C.D. Christian, eds. Controvery in Obstetrics and Gynaecology. second ed. Vol. 2. 1974, W.B. Saunders Company: Philadelphia.
20. Fox, H. and F.A. Langley, Leukocytic infiltration of the placenta and umbilical cord. A clinico-pathologic study. Obstet Gynecol, 1971. 37(3): p. 451-8.
21. Becroft, D.M., J.M. Thompson, and E.A. Mitchell, Placental chorioamnionitis at term: epidemiology and follow-up in childhood. Pediatr Dev Pathol, 2010. 13(4): p. 282-90.
22. Maudsley, R.F., et al., Placental inflammation and infection: a prospective bacteriologic and histologic study. Am J Obstet Gynecol, 1966. 95: p. 648-659.
23. Romero, R., et al., Infection and labor. VI. Prevalence, microbiology, and clinical significance of intraamniotic infection in twin gestations with preterm labor. Am J Obstet Gynecol, 1990. 163(3): p. 757-61.
24. Lee, D.C., et al., Protein profiling underscores immunological functions of uterine cervical mucus plug in human pregnancy. J Proteomics, 2011. 74(6): p. 817-28.
25. Simhan, H.N. and M.A. Krohn, First-trimester cervical inflammatory milieu and subsequent early preterm birth. Am J Obstet Gynecol, 2009. 200(4): p. 377 e1-4.
26. Junqueira, L., et al., Morphologic and histochemical evidence for the occurrence of collagenolysis and for the role of neutrophilic polymorphonuclear leukocytes during cervical dilatation. Am J Obstet Gynecol, 1980. 138: p. 273-281.
27. Garite, T.J. and R.K. Freeman, Chorioamnionitis in the preterm gestation. Obstet Gynecol, 1982. 59(5): p. 539-45.
28. Romero, R., et al., Labor and infection. II. Bacterial endotoxin in amniotic fluid and its relationship to the onset of preterm labor. Am J Obstet Gynecol, 1988. 158(5): p. 1044-9.
29. Martinez, M.A., et al., Occurrence and antimicrobial susceptibility of Ureaplasma parvum (Ureaplasma urealyticum biovar 1) and Ureaplasma urealyticum (Ureaplasma urealyticum biovar 2) from patients with adverse pregnancy outcomes and normal pregnant women. Scand J Infect Dis, 2001. 33(8): p. 604-10.
30. Witt, A., et al., Increased intrauterine frequency of Ureaplasma urealyticum in women with preterm labor and preterm premature rupture of the membranes and subsequent cesarean delivery. Am J Obstet Gynecol, 2005. 193(5): p. 1663-9.
31. Miller, J.M., M.J. Pupkin, and G.B. Hill, Bacterial colonization of amniotic fluid from intact fetal membranes. Am J Obstet Gynecol, 1980. 136: p. 796-804.
32. Wahbeh, C.J., et al., Intra-amniotic bacterial colonization in premature labor. Am J Obstet Gynecol, 1984. 148(6): p. 739-43.
33. Stys, S., et al., Hormonal Effects on Cervical Compliance in Sheep, in Dilatation of the Uterine Cervix, F. Naftolin and P. Stubblefield, Editors. 1980, Raven Press: New York. p. 147-156.
34. Holst, R.M., et al., Cervical length in women in preterm labor with intact membranes: relationship to intra-amniotic inflammation/microbial invasion, cervical inflammation and preterm delivery. Ultrasound Obstet Gynecol, 2006. 28(6): p. 768-74.
35. Park, K.H., et al., Transvaginal ultrasonographic measurement of cervical length in predicting intra-amniotic infection and impending preterm delivery in preterm labor: a comparison with amniotic fluid white blood cell count. J Perinat Med, 2008. 36(6): p. 479-84.
36. Naim, A., et al., Changes in cervical length and the risk of preterm labor. Am J Obstet Gynecol, 2002. 186(5): p. 887-9.
37. Romero, R., et al., Infection and labor. VIII. Microbial invasion of the amniotic cavity in patients with suspected cervical incompetence: prevalence and clinical significance. Am J Obstet Gynecol, 1992. 167(4 Pt 1): p. 1086-91.
38. Progesterone and preterm birth prevention: translating clinical trials data into clinical practice. Am J Obstet Gynecol, 2012. 206(5): p. 376-86.
39. Klebanoff, M.A., et al., Is bacterial vaginosis a stronger risk factor for preterm birth when it is diagnosed earlier in gestation? Am J Obstet Gynecol, 2005. 192(2): p. 470-7.
40. Verstraelen, H., et al., Modified classification of Gram-stained vaginal smears to predict spontaneous preterm birth: a prospective cohort study. Am J Obstet Gynecol, 2007. 196(6): p. 528 e1-6.
41. Romero, R., et al., Meta-analysis of the relationship between asymptomatic bacteriuria and preterm delivery/low birth weight. Obstet Gynecol, 1989. 73(4): p. 576-82.
42. Elder, H.A., et al., The natural history of asymptomatic bacteriuria during pregnancy: The effect of tetracycline on the clinical course and the outcome of pregnancy. Am J Obstet Gynecol, 1971. 111: p. 441-462.
43. Straka, M., et al., Rapid detection of group B streptococcus and Escherichia coli in amniotic fluid using real-time fluorescent PCR. Infect Dis Obstet Gynecol, 2004. 12(3-4): p. 109-14.
44. Espinoza, J., O. Erez, and R. Romero, Preconceptional antibiotic treatment to prevent preterm birth in women with a previous preterm delivery. Am J Obstet Gynecol, 2006. 194(3): p. 630-7.
45. Perni, S.C., et al., Mycoplasma hominis and Ureaplasma urealyticum in midtrimester amniotic fluid: association with amniotic fluid cytokine levels and pregnancy outcome. Am J Obstet Gynecol, 2004. 191(4): p. 1382-6.
46. Gerber, S., et al., Detection of Ureaplasma urealyticum in second-trimester amniotic fluid by polymerase chain reaction correlates with subsequent preterm labor and delivery. J Infect Dis, 2003. 187(3): p. 518-21.
47. Tita, A.T., et al., Clinical trial of interconceptional antibiotics to prevent preterm birth: subgroup analyses and possible adverse antibiotic-microbial interaction. Am J Obstet Gynecol, 2007. 197(4): p. 367 e1-6.
48. Yoon, B.H., et al., The frequency and clinical significance of intra-amniotic inflammation in patients with a positive cervical fetal fibronectin. Am J Obstet Gynecol, 2001. 185(5): p. 1137-42.
49. Khong, T.Y., et al., Chronic deciduitis in the placental basal plate: definition and interobserver reliability. Hum Pathol, 2000. 31(3): p. 292-5.
50. Bendon, R. and M. Miller, Plasma cells in the basal decidua adherent to the placenta. Placenta, 1990. 11: p. 369-370.
51. Staples, L., et al., Migration of leukocytes into the uterus after acute removal of ovarian progesterone during early pregnancy in the sheep. Placenta, 1983. 4: p. 339-350.
52. Keith, L.G., et al., On the causation of pelvic inflammatory disease. Am J Obstet Gynecol, 1984. 149(2): p. 215-24.
53. Naeye, R.L. and S. Ross, Coitus and chorioamnionitis: a prospective study. Early Human Develop, 1982. 6: p. 91-97.
54. Naeye, R.L., Coitus and associated amniotic-fluid infections. N Engl J Med, 1979. 301(22): p. 1198-200.
55. Rhyan, J.C., L.L. Stackhouse, and W.J. Quinn, Fetal and placental lesions in bovine abortion due to Tritrichomonas foetus. Vet Pathol, 1988. 25(5): p. 350-5.
56. Rhyan, J.C., et al., Immunohistochemical detection of Tritrichomonas foetus in formalin-fixed, paraffin-embedded sections of bovine placenta and fetal lung. J Vet Diagn Invest, 1995. 7(1): p. 98-101.
57. Sgrignoli, A.R., D.R. Yen, and G.M. Hutchins, Giant cell and lymphocytic interstitial pneumonia associated with fetal pneumonia. Pediatr Pathol, 1994. 14(6): p. 955-65.
58. Bendon, R.W. and M.B. Ray, The pathologic findings of the fetal membranes in very prolonged amniotic fluid leakage. Arch Pathol Lab Med, 1986. 110(1): p. 47-50.
59. Bennett, B.D., J. Bailey, and W.A. Gardner, Jr., Immunocytochemical identification of trichomonads. Arch Pathol Lab Med, 1980. 104(5): p. 247-9.
60. Edwards, L.E., et al., Gonorrhea in pregnancy. Am J Obstet Gynecol, 1978. 132(6): p. 637-41.
61. Sarrel, P.M. and K.A. Pruett, Symptomatic gonorrhea during pregnancy. Obstet Gynecol, 1968. 32(5): p. 670-3.
62. Gravett, M.G., et al., Independent associations of bacterial vaginosis and chlamydia trachomatis infection with adverse pregnancy outcome. JAMA, 1986. 256: p. 1899-1903.
63. Klatt, E.C., et al., Epidemic perinatal listeriosis at autopsy. Hum Pathol, 1986. 17(12): p. 1278-81.
64. Topalovski, M., S.S. Yang, and Y. Boonpasat, Listeriosis of the placenta: clinicopathologic study of seven cases. Am J Obstet Gynecol, 1993. 169(3): p. 616-20.
65. Campognone, P. and D.B. Singer, Neonatal sepsis due to nontypable Haemophilus influenzae. Am J Dis Child, 1986. 140(2): p. 117-21.
66. Insoft, R.M. and A. Sola, Perinatally aquired pseudomonas infection: a newly recognized maternal risk factor. Am J Perinatol, 1995. 12: p. 25-26.
67. Wright, J.R., Jr., et al., Necrotizing funisitis associated with Actinomyces meyeri infection: a case report. Pediatr Pathol, 1994. 14(6): p. 927-34.
68. Boggess, K.A., et al., Fetal immune response to oral pathogens and risk of preterm birth. Am J Obstet Gynecol, 2005. 193(3 Pt 2): p. 1121-6.
69. Boggess, K.A., et al., Antepartum vaginal bleeding, fetal exposure to oral pathogens, and risk for preterm birth at <35 weeks of gestation. Am J Obstet Gynecol, 2006. 194(4): p. 954-60.
70. Boggess, K.A., et al., Maternal periodontal disease in early pregnancy and risk for a small-for-gestational-age infant. Am J Obstet Gynecol, 2006. 194(5): p. 1316-22.
71. Kostadinov, S. and H. Pinar, Amniotic fluid infection syndrome and neonatal mortality caused by Eikenella corrodens. Pediatr Dev Pathol, 2005. 8(4): p. 489-92.
72. Edwards, C., C.H. Yi, and J.L. Currie, Chorioamnionitis caused by Capnocytophaga: case report. Am J Obstet Gynecol, 1995. 173(1): p. 244-5.
73. Abadi, M. and J. Abadi, Actinomyces chorioamnionitis and preterm labor in a twin pregnancy: A case report. Am J Obstet Gynecol, 1996. 175: p. 1391-2.
74. Reed, C.S., R.G. Feldman, and R.M. Cochrane, Amnionitis caused by Propionibacterium acnes. Case report. Br J Obstet Gynaecol, 1988. 95(6): p. 626-7.
75. Michalowicz, B.S., et al., Treatment of periodontal disease and the risk of preterm birth. N Engl J Med, 2006. 355(18): p. 1885-94.
76. Srinivas, S.K., et al., Periodontal disease and adverse pregnancy outcomes: is there an association? Am J Obstet Gynecol, 2009. 200(5): p. 497 e1-8.
77. Powell, D.G., A. Troppman, and T. Tobin, eds. Proceedings of the First Workshop on Mare Reproductive Loss Syndrome. Vol. 1. 2002, Maxwell H. Gluck Equine Research Center, University of Kentucky: Lexington, KY. 134.
78. Bendon, R. and M. Sebastian. Chorioamnionitis in the human with some comparison to the horse in mare reproductive loss syndrome. in Proceedings of a workshop on the equine placenta. 2003. Maxwell H. Gluck Equine Research Center: University of Kentucky.
79. Bendon, R.W. and M.M. Sebastian, Stillborn infant with calcified chorionic epithelium, corneal scarring, and pericarditis. Fetal Pediatr Pathol, 2009. 28(6): p. 274-8.
80. Shaw, E., et al., Recurrent chorioamnionitis and second trimester abortion because of an enterouterine fistula. Obstet Gynecol, 1995. 86: p. 639-41.
81. Steel, J.H., et al., Maternal origin of inflammatory leukocytes in preterm fetal membranes, shown by fluorescence in situ hybridisation. Placenta, 2005. 26(8-9): p. 672-7.
82. Redline, R.W., et al., Amniotic Infection Syndrome: Nosology and Reproducibility of Placental Reaction Patterns. Pediatr Dev Pathol, 2003.
83. Murtha, A.P., R. Auten, and W.N. Herbert, Apoptosis in the chorion laeve of term patients with histologic chorioamnionitis. Infect Dis Obstet Gynecol, 2002. 10(2): p. 93-6.
84. Kacerovsky, M., et al., The microbial load with genital mycoplasmas correlates with the degree of histologic chorioamnionitis in preterm PROM. Am J Obstet Gynecol, 2011. 205(3): p. 213 e1-7.
85. Katzman, P.J. and L.A. Metlay, Fetal inflammatory response is often present at early stages of intra-amniotic infection, and its distribution along cord is variable. Pediatr Dev Pathol, 2010. 13(4): p. 265-72.
86. Tan, J. and W.M. Saltzman, Topographical control of human neutrophil motility on micropatterned materials with various surface chemistry. Biomaterials, 2002. 23(15): p. 3215-25.
87. Romero, R., et al., Amniotic Fluid White Blood Cell Count – A Rapid and Simple Test to Diagnose Microbial Invasion of the Amniotic Cavity and Predict Preterm Delivery. American Journal of Obstetrics and Gynecology, 1991. 165: p. 821-830.
88. Russo, R.G., et al., Polymorphonuclear leukocyte migration through human amnion membrane. J Cell Biol, 1981. 91(2 Pt 1): p. 459-67.
89. Sampson, J.E., et al., Fetal origin of amniotic fluid polymorphonuclear leukocytes. Am J Obstet Gynecol, 1997. 176(1 Pt 1): p. 77-81.
90. Scott, R.J., D. Peat, and C. Rhodes, Investigation of the fetal pulmonary inflammatory reaction in chorioamnionitis; using an in situ Y chromosome marker. Pediatr Pathol, 1994. 14: p. 997-1003.
91. Gersell, D., N. Phillips, and K. Beckerman, Chronic chorioamnionitis: A clinicopathologic study of 17 cases. Int J Gynecol Pathol, 1991. 10: p. 217-229.
92. Klein, L.L., et al., Detection of intra-amniotic infection in a rabbit model by proteomics-based amniotic fluid analysis. Am J Obstet Gynecol, 2005. 193(4): p. 1302-6.
93. Gravett, M.G., et al., Diagnosis of intra-amniotic infection by proteomic profiling and identification of novel biomarkers. Jama, 2004. 292(4): p. 462-9.
94. Lauweryns, J., et al., Intrauterine pneumonia. An experimental study. Biol Neonate, 1973. 22(3): p. 301-18.
95. Gibbs, R. and P. Duff, Progress in pathogenesis and management of clinical intraamniotic infection. American Journal of Obstetrics and Gynecology 1991. 164(5): p. 1317-1326.
96. Krohn, M.A., et al., The severity of fetal membrane infection and pregnancy complications. Ann Epidemiol, 1993. 3: p. 78-85.
97. Zhang, J.M., F.T. Kraus, and T.I. Aquino, Chorioamnionitis: a comparative histologic, bacteriologic, and clinical study. Int J Gynecol Pathol, 1985. 4(1): p. 1-10.
98. Onderdonk, A.B., et al., Colonization of second-trimester placenta parenchyma. Am J Obstet Gynecol, 2008. 199(1): p. 52 e1-52 e10.
99. Bendon, R.W., et al., Histologic features of chorioamnion membrane rupture: development of methodology. Pediatr Pathol Lab Med, 1997. 17(1): p. 27-42.
100. Bendon, R.W., et al., Fetal membrane histology in preterm premature rupture of membranes: comparison to controls, and between antibiotic and placebo treatment. Pediatr Dev Pathol, 1999. 2(6): p. 552-0558.
101. Winters, R. and B.L. Waters, What is adequate sampling of extraplacental membranes?: a randomized, prospective analysis. Arch Pathol Lab Med, 2008. 132(12): p. 1920-3.
102. Simmonds, M., et al., Intraobserver and interobserver variability for the histologic diagnosis of chorioamnionitis. Am J Obstet Gynecol, 2004. 190(1): p. 152-5.
103. Grether, J.K., et al., Reliability of placental histology using archived specimens. Paediatr Perinat Epidemiol, 1999. 13(4): p. 489-95.
104. Tafari, N. and A. Ljungh-Wadstrom. Consequences of amniotic fluid infection: early neonatal septicemia. in Ciba Foundation symposium 77. 1980.
105. Kundsin, R.B., S.G. Driscoll, and P.A. Pelletier, Ureaplasma urealyticum incriminated in perinatal mortality and morbidity. Science, 1981. 213: p. 474-476.
106. Kundsin, R.B., et al., Association of ureaplasma urealyticum in the placenta with perinatal morbidity and mortality. N Eng J Med, 1984. 310: p. 941-5.
107. Naessens, A., et al., Postpartum bacteremia and placental colonization with genital mycoplasmas and pregnancy outcome. Am J Obstet Gynecol, 1989. 160(3): p. 647-50.
108. Madan, E., M.P. Meyer, and A.J. Amortegui, Isolation of genital mycoplasmas and Chlamydia trachomatis in stillborn and neonatal autopsy material. Arch Pathol Lab Med, 1988. 112(7): p. 749-51.
109. Relier, J.P., L. Helffer, and J.C. Larroche, Approach to Materno-Fetal contamination in a neonatal intensive care unit An argument for the non-prophylactic use of antibiotics in premature infants. Paediatrician, 1976. 5: p. 278-291.
110. St Geme, J.W., Jr., et al., Perinatal bacterial infection after prolonged rupture of amniotic membranes: an analysis of risk and management. J Pediatr, 1984. 104(4): p. 608-13.
111. Fuksman, R.B. and N.G. Mazzitelli, Second-trimester histopathological placental findings in maternal-fetal inflammatory response syndrome. Pediatr Dev Pathol, 2009. 12(1): p. 42-6.
112. Gomez, R., et al., The fetal inflammatory response syndrome. Am J Obstet Gynecol, 1998. 179(1): p. 194-202.
113. Perrin, E.V.D. and J. Kahn-VanderBel, Degeneration and calcification of the umbilical cord. Obstet Gynecol, 1965. 26: p. 371-376.
114. Navarro, C. and W. Blanc, Subacute necrotizing funisitis. J Pediatr, 1974. 85: p. 689-697.
115. Craver, R.D. and V.J. Baldwin, Necrotizing funisitis. Obstet Gynecol, 1992. 79(1): p. 64-70.
116. Jacques, S.M. and F. Qureshi, Necrotizing funisitis a study of 45 cases. Hum Pathol, 1992. 23: p. 1278-1283.
117. Matsuda, T. and et.al, Necrotizing funisitis: Clinical significance and association with chronic lung disease in premature infants. Am J Obstet Gynecol, 1997. 177: p. 1402-7.
118. Navarro, C. and W.A. Blanc, Chronic viral funisitis. J Pediatr, 1977. 91(6): p. 967-73.
119. Fojaco, R.M., G.T. Hensley, and L. Moskowitz, Congenital syphilis and necrotizing funisitis. JAMA, 1989. 261: p. 1788-1790.
120. Schwartz, D.A., et al., Pathology of the umbilical cord in congenital syphilis: analysis of 25 specimens using histochemistry and immunofluorescent antibody to Treponema pallidum. Hum Pathol, 1995. 26: p. 784-791.
121. Knowles, S. and T. Frost, Umbilical cord sclerosis as an indicator of congenital syphilis. J Clin Pathol, 1989. 42(11): p. 1157-9.
122. Heifetz, S.A. and M. Bauman, Necrotizing funisitis and herpes simplex infection of placental and decidual tissues: study of four cases. Hum Pathol, 1994. 25(7): p. 715-22.
123. Goldenberg, R.L., et al., The HPTN 024 Study: the efficacy of antibiotics to prevent chorioamnionitis and preterm birth. Am J Obstet Gynecol, 2006. 194(3): p. 650-61.
124. Andrews, W.W., et al., Interconceptional antibiotics to prevent spontaneous preterm birth: a randomized clinical trial. Am J Obstet Gynecol, 2006. 194(3): p. 617-23.
125. Lamont, R.F., et al., Treatment of abnormal vaginal flora in early pregnancy with clindamycin for the prevention of spontaneous preterm birth: a systematic review and metaanalysis. Am J Obstet Gynecol, 2011. 205(3): p. 177-90.
126. Vaisbuch, E., et al., Patients with an asymptomatic short cervix (<or=15 mm) have a high rate of subclinical intraamniotic inflammation: implications for patient counseling. Am J Obstet Gynecol, 2010. 202(5): p. 433 e1-8.
127. Kiefer, D.G., et al., Amniotic fluid inflammatory score is associated with pregnancy outcome in patients with mid trimester short cervix. Am J Obstet Gynecol, 2012. 206(1): p. 68 e1-6.
128. Abele-Horn, M., et al., High-density vaginal Ureaplasma urealyticum colonization as a risk factor for chorioamnionitis and preterm delivery. Acta Obstet Gynecol Scand, 2000. 79(11): p. 973-8.
129. Minkoff, H., et al., Risk factors for prematurity and premature rupture of membranes: A prospective study of the vaginal flora of pregnancy. Am J Obstet Gynecol, 1984. 150: p. 965-72.
130. Leitich, H., et al., Bacterial vaginosis as a risk factor for preterm delivery: a meta-analysis. Am J Obstet Gynecol, 2003. 189(1): p. 139-47.
131. Hauth, J.C., et al., Early pregnancy threshold vaginal pH and Gram stain scores predictive of subsequent preterm birth in asymptomatic women. Am J Obstet Gynecol, 2003. 188(3): p. 831-5.
132. Goldenberg, R.L., et al., The Alabama Preterm Birth Project: placental histology in recurrent spontaneous and indicated preterm birth. Am J Obstet Gynecol, 2006. 195(3): p. 792-6.
133. Yoon, B.H., et al., The clinical significance of detecting Ureaplasma urealyticum by the polymerase chain reaction in the amniotic fluid of patients with preterm labor. Am J Obstet Gynecol, 2003. 189(4): p. 919-24.
134. Yoon, B.H., et al., Clinical implications of detection of Ureaplasma urealyticum in the amniotic cavity with the polymerase chain reaction. Am J Obstet Gynecol, 2000. 183(5): p. 1130-7.
135. Cherouny, P.H., et al., The presence of amniotic fluid leukoattractants accurately identifies histologic chorioamnionitis and predicts tocolytic efficacy in patients with idiopathic preterm labor. Am J Obstet Gynecol, 1992. 167: p. 683-8.
136. Romero, R., et al., A fetal systemic inflammatory response is followed by the spontaneous onset of preterm parturition. Am J Obstet Gynecol, 1998. 179(1): p. 186-93.
137. Yoon, B.H., et al., An increase in fetal plasma cortisol but not dehydroepiandrosterone sulfate is followed by the onset of preterm labor in patients with preterm premature rupture of the membranes. Am J Obstet Gynecol, 1998. 179(5): p. 1107-14.
138. Bejar, R., et al., Premature labor. II Bacterial sources of phospholipase. Obstet Gynecol, 1981. 57: p. 479-482.
139. Giannoulias, D., et al., Localization of prostaglandin H synthase, prostaglandin dehydrogenase, corticotropin releasing hormone and glucocorticoid receptor in rhesus monkey fetal membranes with labor and in the presence of infection. Placenta, 2005. 26(4): p. 289-97.
140. Lee, P.R., et al., Therapeutic effect of cyclo-oxygenase inhibitors with different isoform selectivity in lipopolysaccharide-induced preterm birth in mice. Am J Obstet Gynecol, 2003. 189(1): p. 261-6.
141. Gravett, M.G., et al., An experimental model for intraamniotic infection and preterm labor in rhesus monkeys. Am J Obstet Gynecol, 1994. 171(6): p. 1660-7.
142. McDuffie, R.S., Jr., et al., Chronic intrauterine and fetal infection with Gardnerella vaginalis. Am J Obstet Gynecol, 2002. 187(5): p. 1263-6.
143. Moss, T.J., et al., Chorioamnionitis induced by subchorionic endotoxin infusion in sheep. Am J Obstet Gynecol, 2003. 189(6): p. 1771-6.
144. Gravett, M.G., et al., Fetal and maternal endocrine responses to experimental intrauterine infection in rhesus monkeys. Am J Obstet Gynecol, 1996. 174(6): p. 1725-31; discussion 1731-3.
145. Aly, H., et al., ABO phenotype and other risk factors associated with chorioamnionitis. J Pediatr, 2008. 153(1): p. 16-8.
146. Reiman, M., et al., Interleukin-6 polymorphism is associated with chorioamnionitis and neonatal infections in preterm infants. J Pediatr, 2008. 153(1): p. 19-24.
147. Dammann, O., A. Leviton, and E.N. Allred, What explains away the increased risk of histological chorioamnionitis in African-American mothers of very-low-birthweight infants? Developmental Epidemiology Network Investigators. Paediatr Perinat Epidemiol, 2000. 14(1): p. 20-9.
148. Arias, F., et al., Placental histology and clinical characteristics of patients with preterm premature rupture of membranes. Obstet Gynecol, 1997. 89: p. 265-71.
149. Arias, F., et al., Maternal placental vasculopathy and infection: two distinct subgroups among patients with preterm labor and preterm ruptured membranes. Am J Obstet Gynecol, 1993. 168(2): p. 585-91.
150. Sebire, N.J., R.D. Goldin, and L. Regan, Histological chorioamnionitis in relation to clinical presentation at 14-40 weeks of gestation. J Obstet Gynaecol, 2001. 21(3): p. 242-5.
151. Witkin, S.S., et al., Induction of interleukin-1 receptor antagonist in rhesus monkeys after intraamniotic infection with group B streptococci or interleukin-1 infusion. Am J Obstet Gynecol, 1994. 171(6): p. 1668-72.
152. Jean-Pierre, C., et al., Extracellular 70-kd heat shock protein in mid-trimester amniotic fluid and its effect on cytokine production by ex vivo-cultured amniotic fluid cells. Am J Obstet Gynecol, 2006. 194(3): p. 694-8.
153. Kramer, M.S., et al., Etiologic determinants of abruptio placentae. Obstet Gynecol, 1997. 89(2): p. 221-6.
154. Gavilanes, A.W., et al., Chorioamnionitis induced by intraamniotic lipopolysaccharide resulted in an interval-dependent increase in central nervous system injury in the fetal sheep. Am J Obstet Gynecol, 2009. 200(4): p. 437 e1-8.
155. Gould, S.J. and P.G. Isaacson, Bronchus-associated lymphoid tissue (BALT) in human fetal and infant lung. J Pathol, 1993. 169(2): p. 229-34.
156. Ernst, L.M., et al., Inflammation-induced preterm birth in a murine model is associated with increases in fetal macrophages and circulating erythroid precursors. Pediatr Dev Pathol, 2010. 13(4): p. 273-81.
157. Andrews, W.W., et al., The Alabama Preterm Birth study: polymorphonuclear and mononuclear cell placental infiltrations, other markers of inflammation, and outcomes in 23- to 32-week preterm newborn infants. Am J Obstet Gynecol, 2006. 195(3): p. 803-8.
obstetrical pathology 
Leave a Reply