Benirschke and Kaufmann feel that chorioamnionitis (which they use as synonymous with acute chorioamnionitis) is always (their bold) caused by infection (3). Other forms of injury to the placenta, such as infarction, will also illicit an acute inflammatory reaction but will be localized to the area of injury. Acute chorioamnionitis is the most common diagnosis rendered on the placenta. Despite this, no universally accepted definition, grading, or staging system has been developed.
Acute Chorioamnionitis: Gross and Microscopic
The gross appearance of the membranes is a very poor indicator of inflammation (Fig. 1). The membranes are usually macroscopically normal. Infection may result in slight clouding (loss of translucency) of the membranes, granularity or a dull appearance, which may be obscured by formalin fixation. Only severe inflammation will result in grossly thickened, cloudy membranes that may be yellow or green because of the pyocyanin within the white blood cells obscuring the fetal vessels on the chorionic plate. Infection may result in membrane edema.
The amount of inflammation used to define chorioamnionitis differs from publication to publication. It has been described generally as "locally dense or had necrosis" (14) or specifically as one focus of at least 5 neutrophils (4) or 10 or greater neutrophils in 10 nonadjacent fields at 400x magnification within free or attached membranes (15,16). Defined criteria for examination of membranes in preterm premature rupture of membranes (PPROM) have been reported (17). An attempt has been made to come to a consensus for the definition for both grade and stage of acute inflammation (18).
Naeye used three histological stages of chorioamnionitis (I, II, III) to indicate sub-chorionic inflammation, chorionic inflammation, and full-thickness inflammation of both chorion and amnion, respectively (19). Early infection may be confined to the maternal decidua near the cervical os, exemplified by diffuse sprinkling of maternal neutrophils that is often associated with some decidual necrosis. Membrane rupture occurs overlying the cervical os, probably because of unequal stretch or unequal wall
tension or possibly because of a loss of decidual blood supply in that area (20). Some acute decidual inflammation is seen in 85% of term deliveries, especially near the zone of membrane rupture (2). Inflammation marginated at the junction of the cellular and fibroblastic chorion indicates migration toward the amniotic cavity and is a feature of intra-amniotic fluid infection (4) (Fig. 2). Maternal neutrophils marginate within the subchorionic fibrinoid prior to migrating in the chorion and amnion on their way toward the amniotic fluid cavity (Fig. 3A). Chellam and Rushton suggested that this is equivalent to margination of inflammatory cells within a vascular space and therefore
may represent the earliest maternal response to infection within the amniotic cavity (21). The inflammatory infiltrate progresses through the chorion (chorionitis) and amnion (chorioamnionitis) (Fig. 3B) and eventually into the amniotic fluid. Because inflammation is chemotactically attracted to stimulus within the amniotic fluid, thickening of the membranes either by squamous metaplasia or placement of the chorionic plate vessels will focally interrupt this process.
Romero and Mazur proposed a four-stage clinical progression of ascending infection (22). Stage I is an overgrowth of organisms in the vagina or cervix. Stage II is localized inflammation of the intrauterine cavity localized to the decidua. Stage III is intra-amniotic infection. Stage IV is infection of the fetus through breathing or swallowing the contaminated amniotic fluid or through cutaneous infection, including conjunctiva. It would be appropriate to add a stage V to this scheme to indicate systemic fetal infection (sepsis or meningitis).
Duration of Infection
Location and appearance of the inflammation has led to assumptions about duration of infection. In general, the inflammatory response is neutrophilic regardless of the duration of infection. Intact neutrophils have led some to estimate infection of less than 24 hours because the average life span of neutrophils outside the vascular space is 24 hours (23,24). The problem with this correlation is that the neutrophils that migrate into the avascular tissues of the amnion may not disintegrate as quickly as in ischemic tissues as studied in acute myocardial infarction. Naeye reported that maternal inflammation is subchorial intervillositis in the first 48 hours of an infection, and that only in the next few subsequent days do the neutrophils migrate into the chorionic plate (19). The maternal inflammatory cells continue their progress toward the amniotic fluid cavity. The time necessary for complete progress from the maternal space to the intra-amniotic fluid space has been estimated at 5-7 days but may occur more quickly with highly virulent organisms (19). Dating chorioamnionitis with any degree of precision is felt to be imprudent (1).
Cultures of the placenta to assess intrauterine infection are of limited value, in part because of antibiotic use during labor and especially because of contamination with a vaginal delivery. There is no correlation between organisms isolated from the cervical canal or the placenta and fetal infection (25). Micro-organisms were not isolated significantly more often from placentas in cases with chorioamnionitis than those without (26). There was a correlation between increased severity of membrane inflammation and fetal vascular inflammation with increasing incidence of positive cultures (26). In preterm labor, bacteria can be isolated from amniotic fluid with intact membranes or from placental cultures with variable success (27). Approximately 70% of placental cultures are positive when there is histological chorioamnionitis (27). Any cultures should be sent directly from labor and delivery. There is a decreased recovery of organisms with refrigeration and a prolonged interval between delivery and cultures. It has been proposed that cultures are most reliable when obtained from the space between the amnion and chorion because surface contamination is eliminated. Cultures for aerobes, anaerobes, Mycoplasma, and Ureaplasma will increase the yield.
Microscopic Identification of Bacteria Within the Placental Membranes
It is actually the exception rather than the rule to find the causative organism within the placental membranes. Group B streptococcus may show significant colonization of the umbilical cord and membranes with little or no maternal inflammation (28). Fetal vasculitis is present in 0-58% (29,30), may be more prolific than the maternal inflammation, and is frequently eosinophilic (2). Fusobacterium sp. is a long, thin, filamentous, Gram-negative bacterium. The organisms have a "hair-on-end" appearance and characteristically cause necrotizing amnionitis. Listeria monocytogenes is a small, Gram-positive coccobacillus that frequently colonizes the subamniotic tissues and
Fig. 3. (opposite page) (A) Subchorionic margination of maternal neutrophils may be the earliest indication of actual infection within the amniotic fluid. (B) Acute chorioamnionitis, full-thickness inflammation, is caused by infection and not other forms of amnion stimulus.
amnion epithelium in large numbers. Special stains for organisms rarely are positive if they are not already apparent on the routine hematoxylin and eosin (H&E) stain.
PROM, PPROM, and Preterm Labor and Premature Delivery
Prematurity is strongly associated with chorioamnionitis, accounting for at least 50% of preterm deliveries. The incidence of clinical chorioamnionitis decreases with gestational age, is nearly 100% at 24 weeks, 73% prior to 30 weeks, and 3-4% of term deliveries (2,31-33). The actual incidence of histological chorioamnionitis is unknown. Preterm birth occurs in 11% of all pregnancies in the United States, with 50% caused by premature onset of labor, 33% caused by PROM, and the remaining 20% caused by specific maternal or fetal considerations (34). Prematurity is increasing, mostly because of the increased number of multiple fetal pregnancies from assisted reproductive technology. Prematurity is responsible for the majority of neonatal morbidity, mortality, and long-term neurological sequelae (34). Histological chorioamnionitis is the most important predictor of infection in premature newborns (35). More women with PROM have histological chorioamnionitis than those without PROM (36), but positive amni-otic fluid cultures occur in a significant number of women with premature labor and intact membranes (37), indicating that rupture is not a prerequisite for intra-amniotic infection.
The histology of PPROM was not different between those who received antibiotics and those who did not, suggesting that the membrane changes preceded the antibiotics
(38). In preterm delivery, there is no increase in maternal inflammation from rupture of membranes to delivery, but fetal inflammation increases with a greater latency period
(39). PROM because of bacteria may be secondary to production of bacterial proteases and can degrade collagen and weaken membranes (40), or the inflammatory response may produce cytokines or prostaglandins, which stimulate labor (41). The release of steroids in response to inflammation paradoxically results in the amnion stimulating prostaglandin production (11).
Bacterial vaginosis (BV) is characterized by changes in the microbiological flora of the vagina, where there is a transition from Lactobacillus sp to Gram-negative and anaerobic bacteria. The most characteristic organisms are Gardnerella vaginalis, Chlamydia, Myocoplasma, and anaerobes such as Bacteroides andMobiluncus (42,43). There are no studies correlating the clinical diagnosis of BV and placental pathology. BV is associated with second trimester fetal loss and premature delivery (44,45). Pla-cental changes often include severe chorioamnionitis, which is frequently necrotizing and polymicrobial (43).
Of meconium-stained placentas, 64% have significant inflammation of the fetal membranes, thought to be the etiology of the meconium spill (19). Although several investigators have implied that the presence of meconium is a risk factor for infection, it is mostly likely the other way around (46,47), possibly because of an alteration in the antimicrobial properties of the amniotic fluid (48). Meconium may be passed because of postmaturity (6%); decreased uteroplacental blood flow; or decreased fetal-placen-tal blood flow, usually because of cord compression (6%), birth asphyxia (1%), and unknown causes (25%) (19). Kaspar et al. reported that 20% of their cases of meco-
nium had associated chorioamnionitis, 25% had fetal vascular thrombosis, 38% had infarcts, and 14.5% had villitis (48), suggesting a less-important role of ascending infection.
Most investigators agree that meconium itself does not cause inflammation of the membranes. It has been suggested that interleukin (IL)-8 within the meconium may be the etiology for neutrophilic inflammation in the lungs of babies with meconium aspiration syndrome (33), as it does in in vitro studies (49). No similar study has been associated with placental inflammation.
Villous edema is commonly seen in conjunction with chorioamnionitis, especially with preterm delivery; 25% occurs in preterm placentas vs 11% at term (50). Edema may be focal or diffuse, mild or severe. It occurs as a fetal response rather rapidly after the "stress" and then disappears over the next 24 hours or so. Postedematous placentas may have residual Hofbauer cell hyperplasia. The associated increased morbidity with edematous villi may result from an increased diffusion distance between the maternal intervillous space and fetal capillaries. This must be considered in the context of the initial stressor, which is usually chorioamnionitis, and prematurity.
The second most common etiology of placental abruption is chorioamnionitis; the first is maternal preeclampsia. Chorioamnionitis with abruption is especially well documented in preterm gestations (51). This form of abruption is usually marginal, beginning at the placental edge nearest the cervix. Large areas of decidual necrosis and acute inflammation may be identified in this area and may or may not have features of abruption.
Ohyama et al. described a specific form of membrane inflammation that was associated with prematurity and development of chronic lung disease and accounted for 54% of placentas with chorioamnionitis (52). Subacute chorioamnionitis was defined as a mixed degenerative neutrophil and mononuclear cell infiltrate. Amnion necrosis was present in 40%. Subacute chorioamnionitis was felt to be an indicator of persistent intrauterine inflammation, possibly caused by organisms with low virulence. Rarely, an antibiotic-treated acute chorioamnionitis may have mixed acute and chronic inflammation.
Chronic inflammation of the fetal membranes has rarely been described in conjunction with viral infections including herpes simplex virus (HSV) (53), rubella (54), and toxoplasmosis (55). Chronic chorioamnionitis tends to be most often associated with nonspecific chronic inflammation elsewhere within the placenta, such as villitis of undetermined etiology (VUE) (56). It is usually focal, rarely involves the amnion connective tissue, and does not result in necrosis of the amnion epithelium (57). Chronic chorioamnionitis is often associated with chronic inflammatory lesions elsewhere within the placental or decidual tissues but can occur as an isolated phenomenon.
Was this article helpful?