Infections as a Cause of PTB –the Amniotic Fluid Sludge – an Ultrasound Marker for Intra-Amniotic Infections and a Risk Factor for Preterm Birth

1. Introduction

Preterm birth (PTB), defined as delivery before 37 completed weeks of gestation is the leading cause of severe neonatal morbidity and mortality world-wide and accounts for about 50% of neonatal deaths [1,2]. Of the estimated 15 million babies born preterm globally, 5% are before 28 weeks, 10% between 28 and 32 weeks and 84% are between 32 and 36 weeks [3]. It is estimated that of all the deaths in the first 5 years of life, about a million (18%) are in children born preterm [4] The overall PTB rate is about 11% but this varies from 4-20% depending on the WHO Region/Country - the highest being in low-income countries. [3,5,6]

Although the exact cause of spontaneous preterm birth (SPTB) is unknown in as many as 50% of cases, it is generally accepted that it is multifactorial in a large number [3,7]. Infections appear to be a common pathway in approximately 25% of cases (and probably much higher in deliveries before 23 weeks at 79% and then declining to about 11% for deliveries between 21 and 34 weeks) [8,9]. The presentation of these infections varies from obvious features of choroamnionitis (such as fever, uterine tenderness, maternal leucocytosis and tachycardia, fetal tachycardia and foul smelling/malodourous vagina discharge) to no obvious clinical features. Confirmation of microbial invasion of the amniotic cavity (MIAC) has traditionally been either from cultures of amniotic fluid and tissues and/or histologically on examination of fetal membranes, umbilical cord (funisitis) and the decidua [10,11,12]. Several studies have demonstrated the presence of infections in the amniotic fluid, fetal membranes and the cervico-vaginal secretions in women presenting in spontaneous preterm labour (SPTL) [11,12,13]. In those with a short cervix, the evidence is overwhelming [14,15] with typical organisms identified including Ureaplasma urealyticum, Gardnerella vaginalis, Candida albicans and Fusibacterium spp [16,17]. For diagnosing intra-amniotic infections, opinions vary on whether amniocentesis should be routinely performed on all those presenting in SPTL for this [13] or to only screening the vagina/cervix for infections. Tests on the amniotic fluid, when obtained, include those able to generate rapid results (such as the quantification of glucose, interleukin 6 and MMP-8 levels, white blood cells and microscopy) that could guide management and cultures that may require time to generate results. The increasing use of molecular biology diagnostic approaches (such as multiplex PCR test) has improved the rapidity of diagnosis and increased the isolation of micro-organisms that would otherwise not have been identified from routine cultures [17,18].

While the evidence is robust for the role of infections in those in SPTL with or without a short cervix [17] this is less so in those who are asymptomatic and with a normal cervix who may eventually progress to deliver preterm. Some of these women may have sub-clinical chorioamnionitis. Early identification of these women and timely institution of interventions such as (antibiotics) may reduce the risk of progression to preterm labour, not only in the symptomatic but also in the asymptomatic women. A key challenge for clinicians is whether there are features that may be indicative of a possible infection prior to shortening and dilatation of the cervix or contractions in this population. An ideal setup will be one where these women at risk are identified and timely interventions to interrupt the process of preterm labour instituted. This can only be achieved with reliable and specific markers of the causes of SPTB. Ultrasound scan can identify a population with cervical changes and allow/enable interruptions (e.g. cervical cerclage) that have indeed been shown to reduce/delay SPTB. Another possible ultrasound marker, that for MIAC- a process that has been shown to predispose to SPTL, is the amniotic sludge [19]. This review brings together the evidence for considering the AF sludge as a marker for MIAC and the need to treat this with antibiotics.

2. The Amniotic Fluid (AF) Sludge and Its Constitution

Particulate materials in the amniotic cavity are a common ultrasound scan finding. They are in general evenly distributed in the amniotic cavity and are thought to represent desquamated fetal cells, vernix caseosa, meconium and in some cases blood [20,21,22,23]. This can also be pathological, where there is excessive desquamation as in the case of congenital ichthyosis [24]. These particulate materials have been reported in about 4% of scans performed between the first and second trimesters [25,26] and rising to about 88% by 35 weeks [27]. The AF sludge on the other hand is a dense aggregate of highly echogenic material accumulating above the cervix, which is present in about 1% of uncomplicated term pregnancies [19] and rising to about 23.5% in high risk-populations (spontaneous preterm labouring women with intact membranes) [28]. Espinoza and colleagues proposed this term to describe this free-floating hyperechogenic material and associated it with an increased risk of SPTB [19].

The location of the AF sludge and its association with microbial invasion of the amniotic cavity is highly suggestive of an infective/inflammatory process (involved in its formation) [29]. Microorganisms may reach the amniotic cavity by breaching the membranes (when they ascend from the lower genital tract) or transplacentally reaching through a haematogenous spread [30]. On reaching the amniotic cavity, the micro-organisms provoke an inflammatory response in the epithelial cells of the fetal skin, the membranes and the umbilical cord [31]. As a result, there is a significant increase in pro-inflammatory cytokines (typically IL-1, IL-6 and IL-8 and TNF-a), prostaglandins, chemokines, and an increased expression of matrix metalloproteinases [29,31]. The increase expression of cytokines/chemokines in the amniotic cavity stimulates the migration of neutrophils across the decidua and the chorioamniotic membrane into the amniotic cavity resulting in an increased white cell population and enhanced antimicrobial activity [29,31,32]. The fact that some studies have failed to isolate micro-organisms and that the AF sludge has also been found in pregnancies that progress to term, suggest that the infective pathogenesis may not be applicable in all cases. Some have suggested that intra-amniotic inflammation can be as a result of exposure to "danger signals" produced by cells undergoing stress/damage or death [33,34,35,36]. This is supported by the fact that sterile inflammation is more common in women presenting in SPTL and pre-term premature rupture of fetal membranes (PPROM) compared to the classical microbiological inflammation [33,34,35,36]. Of note is also the fact that conventional cultures are only able to accurately identify infection in a proportion of cases. The increasing use of modern approaches to identify infective organisms (PCR techniques) has increased isolation rates [18,19].

In an interesting report, Romero et al. not only performed an amniocentesis from the sludge but were able to physically observe it in-vitro. The sludge was described as resembling pus on naked -eye examination. Gram stain from this and other studies has shown the presence of a variety of organisms (Mycoplasma hominis, Streptococcus mutans and Aspergillus flavus). The key question therefore is what is the precise composition of this AF sludge? It has been postulated that progressive infection induces an intense inflammatory response and that the inflammatory cells from this response (neutrophils) combine with the micro-organisms to form the particulate material that is visible as the AF sludge [36].

When micro-organisms breach the feto-maternal barrier, the consequence depends on how large the dose of micro-organisms reaching the fetal membranes is - with the risk of infection/inflammation greatest with the highest micro-organism load [36]. These micro-organisms (bacteria) can exist in one of two forms-singly (in the planktonic form) or organised in biofilms or both [36]. Following invasion of the amniotic cavity, bacteria change their phenotype to protect themselves from the host response (which includes the generation of inflammatory cells (neutrophils and monocytes) and the production of anti-microbial peptides and other mediators which can kill or injure the bacteria [36]. The bacteria achieve this by aggregating themselves in building-like structures known as biofilms. These biofilms make the micro-organisms resistant to attack by the macrophages, natural or synthetic antibiotics and anti-inflammatory mediators [37,38,39,40]. The bacteria in these biofilms are less likely to elicit an inflammatory response [28]. The formation of these intra-amniotic biofilms would partly explain why the intra-amniotic infection tends to be chronic and perhaps why they are difficult to treat as these biofilms are relatively resistant to antibiotics treatment [37,38,39,40,41]. The balance between the proportion of bacteria that are in the planktonic form and those in the biofilms likely determines the course of the infection and the likelihood of positive cultures from amniotic fluid sampling. Planktonic bacteriuria are more likely to be culture positive as opposed to those in the biofilm.

3. Imaging for the AF Sludge

The best approach to identifying the AF sludge is with a transvaginal ultrasound scan with the woman lying in the lithotomy position after emptying her bladder. The probe should be placed in the anterior fornix and adjusted to obtain a sagittal view of the entire cervical length. The length of the cervix should be measured preferably thrice and the average obtained. AF sludge is diagnosed from the presence of dense aggregate of highly echogenic particulate matter in close proximity to the internal cervical os [19,42]. This material may scatter with fetal movements or abdominal pressure but should re-accumulate within a few seconds. Where facilities are available the VOCAL software can be used to measure the volume of the AF sludge. We recommend that at least 3 minutes is spent assessing the sludge because of the dynamic nature of this marker. Figure 1 shows the AF sludge in four of our patients.

4. The AF Sludge and Intra-Amniotic Infections

An investigation of the AF sludge aspirated by amniocentesis showed this to be positive for the micro-organisms Streptococcus mutans, Mycoplasma hominis, Ureplasma urealyticum and Aspergillus flavus [43]. Interestingly in a study by Yoneda and colleagues [44] using polymerase chain reaction (PCR), AF "sludge" was present in 18.1% (19/105) of patients. However, there was a similar positive microorganisms rate in the women with a sludge and those without (31.6% versus 38.4%), but a significantly higher level of amniotic fluid interleukin-8 (15.2ng/ml vs 5.8ng/ml; p=0.005) and a higher frequency of histological chorioamnionitis in those with a sludge (52.6% vs 23.3%; p=0.01). In another study of 25 women with an AF sludge, examination of amniotic fluid collected after amniotomy showed that it was frequently associated with an inflammatory process [45]. Gill et al. in a cohort of 62 women with a short cervix and an AF sludge (AFS) who had undergone amniocentesis showed the rate of intramniotic inflammation to be 31.4% vs 3.7% in those who delivered <32 weeks compared to those who delivered after and furthermore, histological chorioamnionitis was significantly more common in the former group (75% vs 32%) [46]. Interleukin-8 was shown to have the highest sensitivity and specificity for intraamniotic inflammation and histological chorioamnionitis in this cohort. In an earlier study, Kusanovic and colleagues [28] found a higher positive culture rate in those with a sludge (33.3% vs 2.5%; P=0.003) and a higher frequence of histological chorioamnionitis (77.8% vs 19%; P<0.001). The findings of histological chorioamnionitis and funisitis were in 4 cases presenting with cervical insufficiency and an AF sludge [47]. Despite all these studies showing a high frequency of infection/inflammation in the presence of AF sludge, Ventura et al in a study of 16 cases failed to show any difference in these parameters although the women with an AF sludge delivered earlier [48]. Taken together, these case series appear to suggest that the AF sludge is a proxy or indeed a marker for microbial invasion of the amniotic cavity which in some cases may manifest as intra-amniotic infection and or inflammation and therefore a risk factor for PTB. There is need for more studies to investigate just how good a marker of preterm birth AF sludge is. Table 1 summarises the studies that have investigated the association between the AF sludge and intra-amniotic infection/inflammation (19,28,44,47,48,49,50.

5. The AF Sludge and an Ultrasound Marker for Spontaneous Preterm Labour?

Since intra-amniotic infections are generally associated with SPTB, and the AF sludge is thought to represent MIAC, the key question is whether the AF sludge could be an ultrasound marker of PTB? Identification of the AF sludge in the first half of pregnancy has been shown to be associated with inflammation/infection while in late pregnancy, it is thought to reflect maturation of the fetus (i.e. dominated by vernix caseosa, fetal squames and meconium) [20,21,51,52,53]. Espinoza et al [19] in a retrospective study of 84 women (19 or 22.4%) with intact membranes and spontaneous preterm labour, and 298 (1% with AF sludge) at term with intact membranes, concluded that the presence of AF ‘sludge’ in the spontaneous preterm labour and intact membranes group is a risk factor for MIAC, histological chorioamnionitis and impending spontaneous preterm delivery. Several other case series (mostly retrospective) have not only shown that those with the AF sludge are at greater risk of SPTL but that this risk is much greater in those with a short cervix. In a large retrospective study of 281 asymptomatic women who underwent cervical length measurement and screening for the AF sludge, Kusanovic et al [28] showed the AF sludge to be present in 66 cases. The shorter the cervical length, the higher the presence of AF sludge. The SPTB rate was higher in those with an AF sludge and this significant difference was maintained for PTB <28, <32 and <35 weeks. The combination of a cervical length of <25mm and the presence of an AF sludge conferred an odds ratio of 14.8 and 9.9 for spontaneous delivery at <28 and <32 weeks respectively [28]. Although some of studies have concluded that the presence of an AF sludge is an independent risk factor for the occurrence of SPTB, these observations have not been universal. There has to the best of our knowledge been only one prospective study [49]. Table 2 is a summary of the reports on the association between amniotic fluid sludge and SPTB [19,25,26,28,42,48,55,56,57].

6. Will Treatment Improve Outcome?

Since there is strong evidence associating the AF sludge with infection/inflammation, it follows that treatment with antibiotics may reduce the risk of PTB. This was the basis of several studies that have investigated the effect of antibiotics in women at risk of SPTB based on the ultrasound finding of an AF sludge. The results from various studies on the efficacy of antibiotics in this regard have been inconclusive. These studies (reviewed below) include case reports and retrospective and prospective studies. The first report of the use of antibiotics in women with AF sludge was reported by Himaya et al [49]. They describe a woman diagnosed with a sludge at 15⁺⁶ weeks with a cervical length of 33mm treated with intravenous ampicillin-gentamicin and oral azithromycin combination following an amniocentesis and culture of Staphylococcus warneri at 22 weeks. Following treatment, a second amniocentesis was performed, and culture was negative. She progressed to deliver at term. A historical controlled observational study by Hatanaka et al. [58] was reported in which women with an AF sludge diagnosed before 2012 and who were not treated were compared with those diagnosed after 2012 -2015 and treated with antibiotics. The women were divided into 2 groups - those at low risk were given oral clindamycin (300mg every 6 hours) and cephalexin (500 mg every 6 hours) for 7 days while the high-risk group was given intravenous clindamycin 600 mg every 8 hours and cefazolin (1g every 8 hours) for 5 days followed by oral treatment for 5 days. They showed that there was a reduction in the PTB rate in the treated group compared to the untreated group (13.2% vs 38.5%). Cuff et al. [57] in a retrospective cohort study of 97 women with the sludge compared outcome between those who had either been given oral azithromycin 500 mg on day 1 followed by 250 mg orally for 4 days or moxifloxacin 400 mg orally taken daily for 5 days and those who had not been treated. They showed that sonographic resolution of the AF sludge occurred in 34% of those who had been treated and 43% of those who had not - a difference that was not statistically significant. Furthermore, there were no differences in the rates of SPTB in both groups. Pustotina [59] undertook a prospective study of 29 women with an AF sludge (some with a short cervix and symptomatic) in which they offered vaginal clindamycin and other combinations (cefoperazone, clavulonate+amoxicillin and in some cases combined with indomethacin, progesterone and IV sulbactam) and showed that antibiotic treatment eliminated the AF sludge and prevented SPTB in all the cases. Jin et al [60] in a retrospective study of 58 women with a sludge diagnosed at 15-23 weeks and treated with a combination of IV ceftriaxone, clarithromycin and metronidazole showed a lower level of preterm birth following treatment with disappearance of the sludge in 51.7% of cases. More recently Giles et al. [61] reported on a retrospective cohort of women with the AF sludge treated and compared outcomes to those not treated. The antibiotic used was azithromycin. Interestingly the overall spontaneous preterm birth rate was higher in the treatment group but there were no differences in neonatal morbidity. Table 3 summarises all these studies [57,58,59,60,61,62,63]. Two meta-analyses [64,65] and a review of the literature by Luca et al. [66] concluded that while the AF sludge is a marker of preterm birth, there are no robust data on the benefit of antibiotics in this group. More recently there have been reports on cases where the AF sludge was identified and antibiotics administered and followed until it disappeared, and the pregnancy progressed to term [63].

A major factor that has been suggested as a potential confounder responsible for the variable results is the different antibiotic regimens used and routes of their administration. Additionally, the studies have been very heterogenous varying from those on asymptomatic low-risk/high-risk women to those on women with symptoms (i.e. presenting with uterine activity/contractions) or a combination of both. The two studies that showed no difference in treatment used azithromycin as the main antibiotic compared to the others that showed a difference (all of which used a variety of antibiotics including intravenous clindamycin and a cephalosporin). From these data it would seem that the biofilm is less likely to be penetrated by antibiotics administered orally as these may not achieve high levels in the blood. It would therefore be reasonable to recommend that these women are offered intravenous antibiotics. We have treated a number of cases in our unit (N=21) over the last 18 months with the AF sludge with a combination of intravenous clindamycin, metronidazole and ceftriaxone (given for one weeks) and in 14 out of the 21 (67%) cases, the AF sludge resolved, and pregnancies progressed to term or delivery was delayed by an average of 2 weeks in those who delivered preterm. We believe that considering how common this finding is in women at risk of SPTL, there should be randomised controlled trials on the efficacy of intravenous clindamycin combined with a cephalosporin and metronidazole to determine if such a regimen will reduce the risk of SPTB and therefore neonatal morbidity and mortality. This is the combination that appears to be most effective from the case series and covers most of the spectrum of organisms that have been isolated from the various studies.

Finally, it could be argued that leaving the fetus in-utero with MIAC would increase morbidity. The fact that the antibiotics led to resolution of the AF sludge in many cases and that the case series reviewed have not reported increased morbidity in the neonates (if anything delaying birth was associated with better outcomes) is reassuring in this context. Prior to commencing women on broad spectrum antibiotics, the potential of bacteria resistance must be discussed.

7. Conclusion

There is no doubt that intra-amniotic infections are central to a significant proportion of preterm births. Some of these manifests as obvious clinical infections and in most cases with uterine contractions and a short/dilated cervix. In some cases, however, the infections may be subclinical/asymptomatic. Identifying such cases is challenging but the presence of an AF sludge has been shown in case series to be a marker of such infections. The evidence linking the amniotic sludge with SPTB although predominantly from case series is moderately robust although more data are required. What is uncertain is whether treatment with antibiotics does indeed lead to the resolution of the AF sludge and prevention /delay of PTB. The potency of the antibiotics depends on the type of antibiotics and how they are administered. The data reviewed here suggests that oral azithromycin is not effective while a combination of parenteral ceftriaxone, metronidazole and clindamycin appear to be effective. While we feel there is enough to suggest treating these women with this combination, there is a need for randomized controlled trials on the efficacy of this regimen to generate robust evidence. Until such studies are undertaken, clinicians must continue to counsel women diagnosed with an AF sludge on the pros and cons of antibiotics and the limited data on efficacy.