|Home | About | Journals | Submit | Contact Us | Français|
The purpose of this work was to examine the relationship between Ureaplasma urealyticum in the placenta and perinatal outcomes in extreme pre-term deliveries and to explore the influence of bacteria co-infection on perinatal outcomes in U urealyticum-positive placentas.
Under sterile conditions, a piece of chorion from 866 singleton deliveries before 28 weeks' gestation was obtained and flash frozen. The tissue was later homogenized and cultured for Ureaplasma and bacteria. Placentas were also examined histologically. Maternal and neonatal data were obtained prospectively. Each infant had three sets of cranial ultrasound scans between days 1 to 4, 5 to 14, and between day 15 and 40th week post conceptional age. CUS findings were by consensus of 2 or 2 of 3 sonologists.
U urealyticum was recovered from 6% and bacteria from 47%; 47% of placentas had no bacteria detectable. Sixty-seven percent of U urealyticum-positive placentas also harbored bacteria. Placentas that harbored U urealyticum only were more likely than sterile placentas to be associated with a higher prevalence of preterm labor and preterm premature rupture of membranes, as well as umbilical cord, fetal vessel, membrane and parenchymal inflammation and to predict intraventricular hemorrhage and echolucent brain lesions. Placentas that harbored U urealyticum only were similar to placentas that harbored bacteria only. The recovery of U urealyticum and bacteria from the same placenta did not enhance the differences between placentas with U urealyticum alone and sterile placentas. Recovery of U urealyticum only was not associated with a higher risk of death before day 7 of life.
The presence of U urealyticum in placental parenchyma before 28 weeks is associated with increased risk of preterm labor and delivery, higher risk of fetal and maternal inflammation, and increased risk of intraventricular hemorrhage and echolucent brain lesions but not of early neonatal death.
Ureaplasma Urealyticum is detected frequently in the uterine cavity following preterm labor and delivery (1-4). Recovery of U urealyticum from the amniotic fluid and placenta is associated with chorioamnionitis (1, 5), elevated amniotic fluid concentrations of proinflammatory cytokines (6, 7), matrix metalloproteinases (8), increased risk of preterm premature rupture of membranes (pPROM) and preterm delivery (2, 9). However, a review of studies of U urealyticum and preterm delivery reveals gaps in 3 important areas. First, none of the studies evaluated the significance of U urealyticum in the placenta in pregnancies that end before 28 weeks; rather, they focused mostly on pregnancies that end before 32 to 36 weeks (2-4, 10). Second, most studies of intrauterine U urealyticum in preterm deliveries cultured amniotic fluid (9), the chorioamnionic surface (3, 11), or combinations of amniotic fluid, amniotic membrane and placenta (2, 12, 13), making comparisons among studies difficult. Third, few studies have evaluated intrauterine U urealyticum in the context of bacteria recovered simultaneously from the uterine cavity (2, 6). We addressed these gaps in the Extremely Low Gestational Age Newborn (ELGAN) Study.
In an effort to test the hypothesis that isolation of U urealyticum from placental parenchyma is associated with increased prevalence of moderate-to-severe placental inflammation, we evaluated placental parenchyma for the presence of U urealyticum and inflammation in a prospective study designed to identify factors associated with neurologic injury in infants delivered before 28 weeks of gestation. We also tested the hypothesis that recovery of U urealyticum from the placenta increases the risk of early perinatal complications including brain lesions on cranial ultrasound (CUS) and early neonatal death in 866 extremely preterm newborn singletons.
The ELGAN study was designed to identify characteristics and exposures that increase the risk of structural and functional neurologic disorders in ELGANs. From March 2002 to August 2004, women delivering between gestational ages 23 to 276/7 weeks at 1 of 14 participating institutions were asked to enroll in the study. The individual institutional review boards approved the enrollment and consent processes. Details of maternal and infant enrollment have been described (14). A total of 1249 mothers of 1506 infants consented to participate in the study; 260 women were either missed or did not consent to participate. We limited this analysis to placentas from singleton pregnancies because placental assignment was not always clear or feasible in multi-fetal gestations, and more than half of the placentas from multi-fetal gestations were fused. Exclusion of multi-fetal placentas and placentas not evaluated histologically and microbiologically resulted in a sample of 866 placentas.
After delivery, a trained research nurse interviewed each mother in her native language, and prospectively reviewed the maternal and neonatal charts to obtain demographic, pregnancy, delivery and neonatal variables as described in detail elsewhere (14). We used structured data collection forms and procedure manuals with a priori defined variables and outcomes. The birth weight z score, defined as the number of SDs above or below the median weight of infants at the same gestational age in a standard data set (15) was noted.
We obtained protocol CUS scans with digitized 7.5- and 10.0-MHz high-frequency transducers, using the anterior fontanel as the sonographic window (16). Details of CUS protocol and interpretation in the ELGAN cohort have been reported elsewhere (17). We performed 3 sets of protocol scans: first scans were obtained between days 1 and 4, second scans between days 5 and 14, and third scans between the day 15 and week 40. In our sample of 866 singletons with both placental histology and microbiology, 506 had all 3 sets of CUS studies.
Two independent readers who were not provided clinical information read all ultrasound scans. Each set was first read by a study sonologist at the institution of the infant's birth and then sent as electronic images to a sonologist at another ELGAN Study institution for a second reading. When the 2 readers differed in their recognition of intraventricular hemorrhage (IVH), moderate-to-severe ventriculomegaly, echodense lesions, and echolucent lesions, the films were sent to a “tie-breaking” reader who was blinded to what the previous readers reported. Interobserver variability in interpretation of the CUS findings was acceptably low, except for echodense lesions (17).
Details of placental handling and processing in the ELGAN cohort have been reported (18, 19). We transported delivered placentas to the sampling room in sterile basins to ensure that contamination did not occur. More importantly, we obtained placenta tissue after separating the amnionic membrane from the chorion with sterile forceps and scissors to expose a sterile field. Thereafter, we used a new set of sterile instruments to obtain a 1-cm3 piece of chorion and underlying tissue from the placenta as described earlier (19). Eighty-two percent of the samples were obtained within 1 hour of delivery. The specimen was flash frozen in a sterile 2 ml cryovial in liquid nitrogen and stored frozen at -80°C until it was shipped on dry ice to a central microbiology laboratory in Boston. Thereafter, pecimens were stored frozen at −80°C until analyzed.
For microbiologic evaluation a portion was obtained from each placenta specimen after thawing and was weighed and homogenized in sterile phosphate buffered saline to achieve a 10-fold weight/volume dilution, as described elsewhere (19). Serial 10-fold dilutions of the homogenates were made in phosphate buffered saline and aliquots of samples and dilutions plated onto selective and nonselective media and incubated in appropriate conditions for aerobic, anaerobic, mycoplasmal and ureaplasmal isolation and identification (19, 20). In an effort to improve the sensitivity of bacteria detection in placenta, we applied polymerase chain reaction (PCR) techniques to homogenized placental specimens using universal bacterial primers and primers specific to U. urealyticum. Although the primer sets and PCR methods were able to detect low copy levels of DNA in vaginal swabs, PCR did not detect any bacterial DNA even in specimens known to contain high numbers of bacteria by culture (19). Onderdonk et al (19) speculated that placental tissues before freezing contain active ingredients that either degrade or block the detection of bacterial DNA in homogenized chorion-decidual specimens.
After sterile removal of the specimen for microbiologic assessment, and collection of specimens for other studies, study site pathologists examined the whole placenta within 24 hrs of delivery, as described elsewhere (18). Representative sections were taken from all abnormal areas, as well as routine sections of umbilical cord, membrane roll, and full thickness sections from the center and a paracentral zone of the placental disk, as per the College of American Pathologists Conference guidelines (21). Study pathologists examined the slides for histologic evidence of abruption, inflammation of the chorionic plate, chorion/deciduas, and fetal vasculitis, as described in detail elsewhere(18).
Information collected on standardized data forms was entered into a central database and analyzed by using Stata Release 10.0 (Stata Corp LP, College Station, TX) statistical software. The generalized form of null hypothesis tested is that recovery of U urealyticum is not associated with any maternal, pregnancy or newborn characteristic. We used Fisher's exact test for pairwise comparisons of discrete variables and present data as column percentages. P values < 0.05 were considered statistically significant.
A total of 866 placentas from 1006 singleton deliveries before 28 weeks of gestation were subjected to histologic and microbiologic evaluations. U. urealyticum was recovered from 52 (6%) and bacteria from 407 (47%; bacteria only), and 407 placentas (47%, no bacteria) had no bacteria detected by culture. Overall, microorganisms were recovered from 459 (53%) of 866 of our cohort.
Thirty-five (67%) of the 52 U urealyticum-positive placentas harbored bacteria (Uu+ bacteria); 27% had 1, whereas 40% had ≥ 2 other organisms isolated along with U urealyticum (Table 1). Aerobic and anaerobic organisms were recovered from similar proportions of U urealyticum-positive placentas (data not shown). Vaginal organisms were recovered from 33%, Mycoplasma sp from 19%, but Chlamydia sp was not isolated from U urealyticum-positive placentas in this sample. In the analysis, we contrast infants with “U urealyticum only” and “Uu+ bacteria” to infants with “bacteria only” and “no bacteria”.
Mothers whose placentas harbored U urealyticum only were more likely to be single compared to mothers with sterile placentas (Tables 2 and and3).3). However, age, years of education, or primigravidity did not vary appreciably among the 4 groups. Compared to women whose placenta harbored no bacteria, women whose placenta harbored U urealyticum only were approximately twice as likely to present with preterm labor or pPROM. They were also more likely to have rupture of membranes > 24 hours but less likely to have been delivered by cesarean section. Isolation of Uu+ bacteria did not enhance the differences between the U urealyticum only and no bacteria groups.
Infants with U urealyticum only did not differ from infants with no bacteria in gestational age, birth weight, or risk of death before 7 days; however, they were less likely to be born at the lower gestational ages and to die before day 7 compared to infants with bacteria only (Table 4). Compared to infants with no bacteria, infants with U urealyticum only had a modestly higher incidence of IVH, an appreciably lower incidence of moderate-to-severe ventriculomegaly and approximately twice the incidence of echolucent white matter brain lesions. The Score for Neonatal Acute Physiology II is an illness severity indicator based on physiologic information gathered within the first postnatal hours. The prevalence of Score for Neonatal Acute Physiology II ≥ 30 was similar in all 4 groups (Table 5).
Placentas with U urealyticum only had increased frequencies of neutrophilic infiltration of fetal vessels in the chorionic plate, umbilical cord vasculitis, and moderate-to-severe inflammation in the chorion and decidua compared with placentas with no bacteria (Table 6). This difference was not enhanced by the simultaneous isolation of bacteria from U urealyticum-positive placentas, and placentas with bacteria only were not appreciably different from placentas with U urealyticum only. Placental vasculopathy, indicated by thrombosis of fetal stem vessels, placental infarction and increased syncytial knots was less likely in placentas with U urealyticum only than in placentas with no bacteria.
Because we cultured placenta parenchyma for U urealyticum as well as bacteria, we were able to show that placentas that harbored U urealyticum and sterile placentas differed in their tendency to be inflamed, to come from pregnancies with different disorders leading to preterm delivery, to predict brain ultrasound lesions in extreme low gestational age newborns. We were also able to show that placentas that harbored U urealyticum alone were similar to placentas that harbored bacteria only and that the isolation of bacteria from U urealyticum-positive placentas did not accentuate the difference between U urealyticum-positive and sterile placentas.
In agreement with reports by others (11, 22), recovery of U urealyticum from the placenta was associated in this sample with an increased risk of fetal vasculitis. Fetal vasculitis has been associated with neonatal sepsis (23), IVH (24), and white matter damage (25). Consequently, our finding that U urealyticum was not associated with a measure of illness severity supports the possibility that relatively benign and not-so-benign forms of umbilical cord vasculitis exist (22). Alternatively, U urealyticum may result in a different pattern of cord inflammation. Umbilical vasculitis that includes arteritis has been associated with a higher rate of neonatal complications than either the absence of cord inflammation or the presence of cord inflammation without arteritis (23). A study of the pattern of umbilical vessel involvement in fetal vasculitis in U urealyticum-positive compared to bacteria-only positive placentas may help explain the observed differences in outcomes.
Intrauterine infection seems to be a major cause of pPROM, spontaneous preterm labor and preterm delivery (4, 26); the lower the gestational age at delivery, the higher the frequency of intrauterine infection. Our finding that pregnancies with placental U urealyticum had an increased risk of pPROM and preterm labor corroborates these previous findings. This might be expected, because U urealyticum in the amniotic fluid and choriodecidual tissue activates the decidua and fetal membranes to produce a number of cytokines (6, 7) and other bacterial products that stimulate prostaglandin synthesis and release (27) and initiate neutrophil chemotaxis and activation, synthesis and release of metalloproteases (8) and other bioactive substances. Prostaglandins stimulate uterine contractions, whereas metalloproteases attack and remodel collagen in the cervix and chorioamnionic membranes, leading to cervical ripening, pPROM and preterm labor (26). However, the precise role of U urealyticum in the pathogenesis of these adverse pregnancy outcomes remains unclear. In one study, isolation of bacteria from the chorioamnionic space, not U urealyticum, was associated with a higher risk of preterm delivery (3) and ≤ 75% of pregnancies with U urealyticum in the amniotic fluid during midtrimester amniocentesis get to term (28, 29). Furthermore, trials of antimicrobial agents active against U urealyticum in colonized women at risk for preterm delivery (30) and in colonized women in preterm labor (31) have shown neither a consistent reduction in rate of preterm labor nor prolongation of pregnancy. In the present report, placentas that harbored bacteria only were associated with earlier gestational age at delivery and higher risk of IVH and death in the first 7 days of life. Previous studies that demonstrated association between intrauterine U urealyticum and adverse pregnancy outcomes did not take into account the possible role of other intrauterine bacteria (2, 4, 9). In our sample, >60% U urealyticum-positive placentas also habored bacteria. This is comparable to studies that reported the recovery of bacterial pathogens from the uterine cavity in 40 – 90% of women with intrauterine U urealyticum infection (2, 32). In a recent study, U urealyticum was recovered from the cord blood in 17% of preterm infants delivered between 23 and 32 weeks of gestation (10). The study also showed that placentas from 90% of infants with positive cord blood culture for U urealyticum and/or Mycoplasma hominis were positive for other organisms. Failure to control for potential confounding by other intrauterine organisms probably contributes to the weak and variable association between U urealyticum and pregnancy outcomes.
Compared to sterile placentas, recovery of U urealyticum only from placenta parenchyma in this report is associated with an increased frequency of fetal vasculitis and CUS evidence of echolucent cerebral lesions. Others have shown that recovery of U urealyticum from cultures of chorionic surface is associated with an increased risk of fetal vasculitis but not of echolucent brain lesions in VLBW infants (11). On the other hand, VLBW infants with fetal vasculitis and intact membranes have been shown to have substantially increased risk of white matter echolucency (25). It has been postulated that the fetal inflammatory response in the later situation is probably less intense and of longer duration, resulting from intrauterine infection with less virulent organisms like U. urealyticum (25, 33). Our finding of an increased frequency of fetal vasculitis and a modestly increased risk of IVH in infants whose placentas harbored U urealyticum lends support to the association between fetal vasculitis and IVH. However, recovery of U urealyticum from the amniotic fluid (9, 34), amnion and placental tissue (35) has not been consistently associated with increased frequency of IVH in VLBW infants. However, none of the studies evaluating the association between intrauterine U. urealyticum and IVH examined placenta or umbilical cord for fetal vasculitis.
The small number of infants with U urealyticum detected in the placental parenchyma is a limitation to the study. The low rate may reflect effectiveness of our efforts to limit sample contamination. A second limitation is our inability to use PCR. We attribute this to interfering substances in the placenta parenchyma (18, 19). However, the culture methods employed appear to be as sensitive as PCR for detecting U urealyticum associated with preterm delivery, based on the results of other investigators (9).
A third limitation is that both our study and comparison groups include cases with different clinical presentation, including spontaneous preterm labor, pPROM and deliveries undertaken for maternal and fetal indications.
The strengths of our study include the large number of infants enrolled and the prospective nature of the study. The gestational-age-defined sample of infants born between 23 and 276/7 weeks of gestation enabled us to evaluate a fairly homogenous high-risk population compared to studies that utilized birth-weight-defined samples, which can have an over-representation of growth restricted newborns (4, 25). Furthermore, the ELGAN study sample afforded us a unique opportunity to compare perinatal outcomes associated with recovery of U urealyticum, alone or with bacteria, with outcomes when the placenta is sterile.
Among placentas delivered before 28 weeks' gestation, the presence of U urealyticum alone was associated with an increased prevalence of pPROM, preterm labor, preterm delivery, and histologic indicators of fetal and maternal inflammation when compared to cases with sterile placentas. The recovery of bacteria from U urealyticum-positive placentas did not enhance these differences with sterile placentas. The precise role U urealyticum in the pathogenesis of intrauterine inflammation and associated adverse pregnancy outcomes remains unclear. Infants whose placentas were positive for U urealyticum had an appreciably increased risk of echolucent white matter lesions and a modest increase in risk of IVH but not increased neonatal illness severity or risk of early neonatal death. Because >60% of U urealyticum-positive placentas also harbored bacteria, future studies of the impact of intrauterine U urealyticum on pregnancy outcomes must include methods to identify and control for the confounding effects of intrauterine bacteria.
This work was funded by a cooperative agreement with the National Institute of Neurological Disorders and Stroke(1 U01 NS 40069-01A2) and a program project grant from the National Institute of Child Health and Human Development (NIH-P30-HD-18655).
Funding Sources: Funded by a cooperative agreement with the National Institute of Neurological Disorders and Stroke (1 U01 NS 40069-01A2) and a program project grant from the National Institute of Child Health and Human Development (NIH-P30-HD-18655).
Financial disclosure – Nothing to disclose
Conflict of interest – None
What's Known on This Subject: Intrauterine infection with Ureaplasma urealyticum is associated with increased risk of adverse perinatal outcomes, including chorioamnionitis, preterm labor and delivery, and neonatal morbidities. Trials of antibiotics against U urealyticum have not consistently reduced the risk of these adverse consequences.
What This Study Adds: U urealyticum-positive placentas frequently harbor bacteria. Placental coinfection with U urealyticum and bacteria did not enhance the adverse effects associated with U urealyticum. Study of U urealyticum and pregnancy outcomes must control for confounding by other intrauterine organisms.