During the last two decades, it has been widely believed that the primary factor contributing to spontaneous preterm birth is vaginal bacteria, which upon ascending into the uterus cause a decidual, placental, and amniotic infection that prompts a fetal inflammatory response (1, 2). Support for this hypothesis has come from studies documenting an association between preterm labor and preterm prelabor rupture of membranes and the presence of chorioamnionitis at placental examination or by the presence of bacteria in the amniotic fluid or between the membranes (3–5). The presence of bacteria in the amniotic fluid or histologic chorioamnionitis is related to high cytokine levels in the amniotic fluid (1, 5). Some women with signs of intra-amniotic inflammation, such as elevated concentrations of cytokines in their amniotic fluid, do not have documented intra-amniotic infection (6, 7). Women with intra-amniotic inflammation without detectable bacteria in the amniotic fluid are more likely to have bacteria detectable between the membranes than those without intra-amniotic inflammation (3, 4). The ability to delineate between intra-amniotic infection and intra-amniotic inflammation might be due to the lack of a methodology sensitive enough to detect minute amounts of bacteria or the lack of our current understanding as to the possible causes of intra-amniotic inflammation.

Bacterial 16S ribosomal RNA polymerase chain reaction (16S rRNA PCR) techniques, which have been introduced recently, successfully identify bacterial genes even when bacterial cultures are negative. For example, when PCR-based techniques are used Ureaplasma urealyticum is positively identified in the amniotic fluid of women with preterm prelabor of membranes more often than when assayed by conventional bacterial culture (8). PCR-based methods have been used successfully to identify other types of bacteria (9). Still it has been questioned whether every case of intra-amniotic inflammation has bacteria in the amniotic fluid or between the membranes that have not been detected by the bacterial methods used.

One study using conventional bacterial culture techniques found that women in spontaneous labor at term, have amounts of detectable bacteria in amniotic fluid similar to that of women in preterm labor (10). On the other hand, the levels of amniotic fluid cytokines are lower, and the presence of histologic chorioamnionitis is also less common in term spontaneous birth (11). These observations can be interpreted in several ways. One set of inferences suggests that what discriminates between preterm and term labor is not the presence of bacteria in the amniotic cavity, but rather the responses of the mother and fetus.

The authors of a paper in this issue, Miralles et al. have studied intra-amniotic infection and inflammation (12). They analyzed both maternal and fetal tissues/fluids of a limited number of patients from various subgroups of preterm birth. 16S rRNA PCR technique and enzyme linked immuno assay for IL-6 and IL-8 were used to analyze placentas, membranes, amniotic fluid, gastric fluid, bronchoalveolar lavage and umbilical cord blood. Their findings suggest that the intra-amniotic milieu before birth can be studied postnatally in gastric fluid (compartment #1) and bronchoalveolar lavage (compartment #2) of the newborn as well as in the amniotic fluid (compartment #3). In addition, the relationship between PCR positivity in each compartment and histologic chorioamnionitis is evaluated. All cases with histologic chorioamnionitis had a positive 16s rRNA PCR from one of the three compartments. All 8 children from pregnancies with histologic chorioamnionitis whose gastric fluid was available had 16S rRNA identified in the gastric fluid. The authors also found a clear relationship between the inflammatory response, measured as IL-6 and IL-8 levels, in various intrauterine compartments to the presence of a positive test for 16s rRNA bacterial gene.

The findings of Miralles et al. are important from several points of view (12). Gastric fluid appears to be able to serve as a source of information about neonatal exposure to bacteria in utero. When amniocentesis has not been performed, sampling of the gastric fluid might provide information about the intra-amniotic environment. The value of gastric fluid assessment needs to be addressed in a prospective study with a greater number of patients where both compartments are sampled.

The authors report that 70% and 80% of deliveries after preterm prelabor rupture of membranes and preterm labor are associated with microbial rDNA. This is not as high as in one recent study (13), but is higher than the 40–50% cited in a recent review (14). Much of the variability probably reflects the entry criteria for study (preterm birth, preterm labor, preterm labor or prelabor preterm membrane rupture, chorioamnionitis, etc.).

Not all pregnancies with presumed inflammation have identifiable bacteria. This begs the question, if bacteria were never truly present, what is the inflammatory stimulus? New experimental data from mice suggest that augmented production of surfactant protein A by the fetal lung near term causes activation and migration of fetal amniotic fluid macrophages to the maternal uterus, where increased production of interleukine-1 beta activates nuclear factor-kappa B, leading to labor (15). If this observation holds true and is the same in humans, it suggests that delivery can start as an intra-amniotic inflammation without any infectious agent triggering the process. Miralles et al. also found cases (2/10) of spontaneous preterm labor where no bacteria were found in any of the intrauterine compartments which might indicate a nonbacterial etiology for the initiation of labor and delivery in those cases (12). Another possible explanation to cases of spontaneous preterm labor where bacteria are not detected is that viruses could be playing a more important role than previously thought by activating macrophages via the Toll-like receptors (16) and might spread light over the discussion of intra-amniotic infection and inflammation.