SHORT REPORT

Resolution of acute cervical insufficiency after antibiotics in a case withamniotic fluid sludge

Lami Yeoa,b,c, Roberto Romeroa,b,d,e,f,g, Tinnakorn Chaiworapongsab,c, Robert Paraa,b,c, Jeffrey Johnsona,b,c,David Kmakb, Eunjung Junga,c, Bo Hyun Yoona,h and Chaur-Dong Hsua,c,i

aPerinatology Research Branch, Division of Obstetrics and Maternal-Fetal Medicine, Division of Intramural Research, Eunice KennedyShriver National Institute of Child Health and Human Development, National Institutes of Health, U.S. Department of Health andHuman Services, Bethesda, MD, Detroit, MI, USA; bDetroit Medical Center, Detroit, MI, USA; cDepartment of Obstetrics andGynecology, Wayne State University School of Medicine, Detroit, MI, USA; dDepartment of Obstetrics and Gynecology, University ofMichigan, Ann Arbor, MI, USA; eDepartment of Epidemiology and Biostatistics, Michigan State University, East Lansing, MI, USA;fCenter for Molecular Medicine and Genetics, Wayne State University, Detroit, MI, USA; gDepartment of Obstetrics and Gynecology,Florida International University, Miami, FL, USA; hDepartment of Obstetrics and Gynecology, Seoul National University College ofMedicine, Seoul, Republic of Korea; iDepartment of Physiology, Wayne State University School of Medicine, Detroit, MI, USA

ABSTRACTCervical insufficiency generally refers to a condition in which there is mid-trimester cervical dila-tation or protruding chorioamniotic membranes in the absence of uterine contractions. Suchcondition is a risk factor for spontaneous mid-trimester abortion or early preterm birth, and isassociated with adverse neonatal outcomes. Both intra-amniotic infection and inflammationascertained by amniocentesis have been identified in patients with cervical insufficiency, and arepoor prognostic factors. A subset of patients with intra-amniotic inflammation will have no dem-onstrable microorganisms detected via cultivation or molecular methods, and therefore repre-sent cases of sterile intra-amniotic inflammation. Amniotic fluid sludge (free-floatinghyperechogenic material within the amniotic fluid in close proximity to the uterine cervix) iden-tified on sonography is a biomarker for intra-amniotic infection and inflammation. Recent evi-dence suggests that intra-amniotic infection, as well as sterile intra-amniotic inflammation canbe treated successfully using antimicrobial agents. We report a unique case in which administra-tion of antibiotics in the presence of mid-trimester cervical insufficiency, sterile intra-amnioticinflammation, and amniotic fluid sludge was associated with resolution of the cervical findings,as demonstrated on both sonographic and speculum examination. The patient successfullyunderwent elective cesarean delivery at 36-2/7weeks of gestation. This case illustrates that anti-biotic therapy may be effective despite the presence of several high-risk pregnancy conditions,and that successful outcome is possible.

ARTICLE HISTORYReceived 14 December 2020Accepted 22 January 2021

KEYWORDSAmniocentesis; CRP;interleukin-6; short cervix;sterile intra-amnioticinflammation

Introduction

Cervical insufficiency generally refers to a condition in

which there is mid-trimester cervical dilatation or pro-

truding chorioamniotic membranes in the absence of

uterine contractions. It is thought that cervical suffi-

ciency/insufficiency is a continuum [1], and is syn-

dromic in nature caused by various factors [2].

Unfortunately, cervical insufficiency is a risk factor for

spontaneous mid-trimester abortion or early preterm

birth [3]. Intra-amniotic infection has been reported in

8–52% [4–11], and intra-amniotic inflammation in 81%

[7] of such patients when ascertained by

amniocentesis. Moreover, intra-amniotic inflammation,regardless of amniotic fluid culture results, is a risk fac-tor for impending preterm delivery and adverse out-comes [7]. Indeed, Lee et al. reported that whencervical insufficiency and intra-amniotic inflammationare present, 50% will deliver preterm within 7 days [7].A subset of patients with intra-amniotic inflammationwill have no demonstrable microorganisms detectedvia cultivation or molecular methods, and thereforerepresent cases of sterile intra-amniotic inflamma-tion [12–17].

Amniotic fluid sludge describes free-floating hypere-chogenic or particulate material within the amniotic

CONTACT Lami Yeo lyeo@med.wayne.edu; Roberto Romero prbchiefstaff@med.wayne.edu Perinatology Research Branch, NICHD, NIH, DHHS,Hutzel Women’s Hospital, 3990 John R, 4 Brush, Detroit, MI 48201, USA

Supplemental data for this article is available online at https://doi.org/10.1080/14767058.2021.1881477.

This work was authored as part of the Contributor’s official duties as an Employee of the United States Government and is therefore a work of the United States Government.In accordance with 17 U.S.C. 105, no copyright protection is available for such works under U.S. Law.

THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINEhttps://doi.org/10.1080/14767058.2021.1881477

fluid in close proximity to the uterine cervix, which isdemonstrable on transvaginal sonography. Sludge is abiomarker for intra-amniotic infection and inflamma-tion [18], and is an independent risk factor for pretermprelabor rupture of membranes (PROM), as well asspontaneous preterm delivery [19–21].

The conventional view has been that in the settingof cervical insufficiency, intra-amniotic infection/inflam-mation cannot be successfully treated [10]. Yet, recentevidence suggests that intra-amniotic infection, as wellas sterile intra-amniotic inflammation can be treatedsuccessfully with antimicrobial agents [10,13,22,23].We report herein a case in which administration ofantibiotics in the presence of mid-trimester cervicalinsufficiency, sterile intra-amniotic inflammation, andamniotic fluid sludge was associated with resolutionof cervical findings as demonstrated on both sono-graphic and speculum examination. When the anti-biotic regimen was initiated, the maternal C-reactiveprotein concentration was elevated, but normalizedsix days later during treatment. The patient success-fully underwent elective cesarean delivery at 36 2/7weeks of gestation due to a history of prior myo-mectomy. Histologic examination of the placenta dem-onstrated no acute inflammatory lesions or funisitis.This unique case illustrates that antibiotic therapy maybe effective despite the presence of several high-riskpregnancy conditions, and that successful outcomeis possible.

Case description

A 39-year-old woman, gravida 2, para 0010 wasreferred to our institution for a fetal anatomy surveyat 20 3/7weeks of gestation due to advanced maternalage, multiple fibroids, and an echogenic intracardiacfocus noted on the office ultrasound exam. The gesta-tion was dated by a first trimester ultrasound examin-ation at 5weeks. The patient’s past obstetrical andmedical history was significant for a prior ectopicpregnancy, anemia, mild asthma, and partial thyroi-dectomy, as well as a history of myomectomy withentry into the endometrial cavity and a vertical uterineincision. At 7 6/7weeks of gestation, she receivedvaginal metronidazole for bacterial vaginosis.

For the Case presented herein, she was examinedat Detroit Medical Center/Wayne State University, andthe Perinatology Research Branch of NICHD, NIH,DHHS. The patient was enrolled in a research protocolapproved by the Institutional Review Board of NICHD,NIH, and by the Human Investigation Committee ofWayne State University. She provided written informed

consent for the use of sonographic images forresearch purposes.

On transabdominal sonographic examination, theuterine cervix appeared short; therefore, a transvaginalexamination was performed, confirming a short cer-vical length (7.3mm) with funneling and amniotic fluidsludge present (Figure 1). The funnel width and lengthwere 13.8 and 29.2mm, respectively. In addition, morethan 15 uterine leiomyomas were identified (largestmeasured 4.4� 4.9� 5.5 cm).

Due to the sonographic findings, the patient wassent to the obstetrical triage unit. She gave a historyof occasional cramping, but denied bleeding or leak-ing of fluid. Digital exam revealed the uterine cervixto be closed and 70% effaced. Speculum examshowed a small amount of white discharge in thevaginal vault, but there was absent bleeding or pool-ing of fluid. No contractions were evident on the toco-dynamometer. Urine culture obtained that dayeventually returned as positive for group B streptococ-cus (10–100,000 CFU/mL). The patient was counseledon the availability of amniocentesis testing to deter-mine the microbial status of the amniotic cavity. Shewas discharged home with instructions to take vaginalprogesterone nightly, and returned four days later asan outpatient at 21weeks of gestation for transabdomi-nal amniocentesis. Fetal karyotype and microarray ana-lysis were normal. Amniotic fluid analysis showed anegative Gram stain, white blood cell count of 2 cells/mm3, and glucose of 24mg/dl. Amniotic fluid culturesfor aerobic and anaerobic bacteria, as well asUreaplasma urealyticum and Mycoplasma hominis werenegative. However, the amniotic fluid interleukin-6 (IL-6), a pro-inflammatory cytokine, was elevated at3.055 ng/ml (normal <2.6 ng/ml), consistent with

Figure 1. Sonographic cervical evaluation at 20 3/7 gesta-tional weeks. On transabdominal ultrasound, the cervixappeared short; therefore, a transvaginal sonographic examin-ation was performed. The cervical length is 7.3mm with fun-neling and amniotic fluid sludge noted (white arrow).

2 L. YEO ET AL.

sterile intra-amniotic inflammation [15]. It is note-worthy that at a later time, the amniotic fluid super-natant was also analyzed by PCR for Candida spp., andreturned negative.

Hospital course

Due to the amniocentesis results, the patient wasadmitted to the hospital at 22weeks of gestation. Shecomplained of vaginal discharge, but denied leakingof fluid, bleeding, or contractions (also confirmed ontocodynamometer). However, speculum exam showedthe cervix to be 3 cm dilated with bulging membranes,and amniotic fluid sludge was visibly “passing by themembranes,” consistent with cervical insufficiency. Thepatient was counseled that in the presence of cervicalinsufficiency with intra-amniotic inflammation, theprognosis is poor, with 50% of women experiencing apreterm delivery within 7 days [7]. She was also coun-seled about the study by Oh et al, in which theauthors reported that in patients with cervical insuffi-ciency and intra-amniotic infection/inflammation,administration of antibiotics was followed by reso-lution of the intra-amniotic inflammatory process orintra-amniotic infection in 75% of patients, and wasassociated with treatment success in 59% of cases[10]. After discussion, the patient agreed to receive tri-ple antibiotic therapy: (1) ceftriaxone 1 gram IV every24 h; (2) clarithromycin 500mg oral every 12 h; and (3)metronidazole 500mg IV every 8 h [10]. She had beentaking vaginal progesterone since 20 3/7weeks of ges-tation, and this medication was continued in-house.

The patient was placed in Trendelenburg position,although this was not followed consistently. To evalu-ate for the presence of maternal systemic inflamma-tion or infection, the following labs were ordered atadmission (before antibiotic administration) and seri-ally during the hospital stay: (1) C-reactive protein(Figure 2); (2) procalcitonin; (3) complete blood countwith differential; (4) ferritin; and (5) coagulation panel(fibrinogen, D-dimer, prothrombin time, and partialthromboplastin time). Genital cultures on admissionwere negative for yeast, but later yielded numerousgroup B streptococcus. In addition, the urine cultureon admission showed group B streptococcus(>100,000 CFU/ml).

Based on recommendations by neonatology, ante-natal corticosteroids were administered at 22 2/7 to22 3/7weeks. At 22 3/7weeks of gestation (two weeksafter the initial cervical sonographic examination), thepatient underwent a repeat transvaginal ultrasoundexamination, and this now demonstrated no measur-able cervical length. The chorioamniotic membraneswere bulging past the external cervical os into thevagina (Figure 3, Supplementary Movie 1 andSupplementary Movie 2). Moreover, a large quantity ofamniotic fluid sludge was visualized within the bulg-ing membranes (Figure 3, Supplementary Movie 1 andSupplementary Movie 2). Due to such findings, vaginalprogesterone was discontinued; altogether, the patientreceived this only between 20 3/7 and 22 5/7weeksof gestation.

During the hospital course, the patient received tri-ple antibiotics for 11 days total (i.e. 22 to 23 3/7weeksof gestation). As seen in Figure 2, maternal C-reactiveprotein concentrations were elevated on the same dayantibiotics were initiated, but became normal six dayslater during treatment, and remained as such duringthe hospital stay, even after antibiotics were discontin-ued. The other laboratory results obtained seriallywere not consistent with maternal systemic inflamma-tion or infection. The patient was counseled aboutrepeat amniocentesis testing to monitor the microbio-logic and inflammatory status of the amniotic cavity.The plan was to use this information (as well as sono-graphic examination results) to determine the durationof antibiotic therapy. However, the patient complainedof several side effects (i.e. nausea, dyspepsia, nightsweats) and requested discontinuation of all antibiot-ics. Repeat amniocentesis was not performed at thediscretion of her providers. Three days after antibioticswere discontinued, transvaginal sonogram at 23 6/7weeks of gestation still demonstrated sludge; how-ever, the cervical length was surprisingly longer in

Figure 2. Maternal C-reactive protein concentrations duringhospital stay (22 to 24 1/7 gestational weeks). Just before anti-biotics were started, the initial C-reactive protein concentrationwas elevated (16.48mg/l) at 22weeks, and increased further(33.24mg/l) at 22 3/7weeks. However, subsequent C-reactiveprotein concentrations became normal six days later duringantibiotic treatment, and remained as such even after antibiot-ics were discontinued (normal <10mg/l; dotted line).

THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINE 3

length (21.5mm) with mild funneling present (Figure4, Supplementary Movie 3). Moreover, on ultrasound,the membranes were no longer visualized bulgingthrough the external cervical os into the vagina.Continued in-patient management was recommendeduntil 32weeks; however, the patient decided to leavethe hospital against medical advice at 24 1/7weeksof gestation.

Outpatient course

Approximately one week later, the patient presentedto the hospital at 25weeks of gestation complaining ofpelvic pressure. Speculum exam revealed white dis-charge in the vaginal vault and the cervix appeared1–2 cm dilated, but no bulging membranes werevisualized. Light microscopy showed clue cells andthere was a positive whiff test, consistent with

bacterial vaginosis. The patient was discharged homewith oral metronidazole for 7 days.

At 25 3/7weeks of gestation, the estimated fetalweight was noted to be at the 11th percentile, andamniotic fluid volume was normal. Transvaginal cervicallength was 12.7mm with funneling and sludge againseen; however, the quantity of sludge appeared muchless than previously (Figure 5, Supplementary Movie 4).

The patient continued obstetrical care with her pro-vider, and she underwent an ultrasound examinationat 33 2/7weeks of gestation. The estimated fetal weightwas noted to be at the 5th percentile, and theabdominal circumference measured 3weeks behinddates. However, the amniotic fluid volume and fetalDoppler velocimetry (umbilical artery, umbilical vein,middle cerebral artery, ductus venosus) were withinnormal limits. The transvaginal cervical length was20.9mm, and there was no evidence of funneling or

Figure 3. Sonographic cervical evaluation at 22 3/7 gestational weeks. (A) Transvaginal sonographic examination revealed nomeasurable cervical length with chorioamniotic membranes bulging past the external cervical os into the vagina. A large quantityof amniotic fluid sludge is present within the bulging membranes (also see Supplementary Movie 1 and Supplementary Movie 2).(B) Same findings as A; in addition, the anterior-posterior diameter of the bulging membranes measured 24.9mm.

Figure 5. Sonographic cervical evaluation at 25 3/7 gesta-tional weeks. Transvaginal cervical length was 12.7mm withfunneling and the presence of amniotic fluid sludge (collectionjust proximal to the internal cervical os); however, the quantityof sludge appeared much less than seen previously (also seeSupplementary Movie 4).

Figure 4. Sonographic cervical evaluation at 23 6/7 gesta-tional weeks. Transvaginal cervical length was 21.5mm withmild funneling present. Amniotic fluid sludge was still identi-fied (also see Supplementary Movie 3). It is noteworthy thatbulging membranes past the external cervical os into thevagina are no longer present.

4 L. YEO ET AL.

dynamic changes (Figure 6, Supplementary Movie 5).In addition, no amniotic fluid sludge could be identi-fied below the presenting vertex. The patient wascompletely asymptomatic and was doing well.

Delivery

Due to the history of prior myomectomy, the patientunderwent an elective cesarean delivery at 36 2/7weeks of gestation. Just prior to the delivery, a trans-vaginal ultrasound was performed, and the cervicallength was 19.7mm with mild funneling distal to thechorioamniotic membranes (Figure 7, SupplementaryMovie 6). No amniotic fluid sludge was identifiedbelow the presenting vertex. The patient had receivedone prophylactic IV dose of azithromycin and cefazolinprior to delivery. A viable female neonate weighing

2560 grams was born, with Apgar scores of 8 and 9 (1and 5min, respectively). Umbilical cord arterial andvenous blood gases were within normal limits. Duringthe cesarean section, an attempt was made to collectamniotic fluid from the exposed amniotic sac to evalu-ate the microbial status of the amniotic cavity.However, due to decreased amniotic fluid, this wasnot successful. The amniotic fluid was noted to beclear. The placenta was manually extracted and notedto be adherent to the right posterior uterine wall;therefore, curettage was performed.

Histologic examination of the placenta showed noacute inflammatory lesions or funisitis, no lesions asso-ciated with maternal vascular or fetal vascular malper-fusion, and marginal intervillous fibrin deposition waspresent. However, chronic lymphoplasmacytic deciduitiswas identified. Gross examination revealed the fetalsurface to be blue-gray in color, and extraplacentalmembranes were translucent (Figure 8). The maternalsurface was unremarkable. Serial sections of the pla-centa revealed diffuse fibrosis around the periphery,collectively occupying less than 5% of the totalcut surfaces.

Neonatal course

At delivery, the neonate had a temperature of 36.5 �C,and in the nursery, the temperature was unstable,ranging from 36 to 36.3 �C. Due to the history of posi-tive maternal urine and genital cultures for group Bstreptococcus, neonatal temperature instability, andconcern for possible neonatal infection, ampicillin andgentamicin were initiated. However, labs (i.e. completeblood count, blood cultures, and C-reactive protein)obtained from the neonate on the day of delivery(before the first dose of antibiotics) were within nor-mal limits. Repeat complete blood count and C-react-ive protein labs were also normal. The neonataltemperature stabilized, antibiotics were discontinuedafter a 3-day course, and the neonate was dischargedhome on day of life four. As of this writing (twomonths after delivery), the patient and neonate aredoing well.

Discussion

The main message of this report is that administrationof antibiotics in the presence of mid-trimester cervicalinsufficiency, sterile intra-amniotic inflammation, andamniotic fluid sludge was associated with resolutionof the cervical findings, as demonstrated on bothsonographic and speculum examination. Although the

Figure 6. Sonographic cervical evaluation at 33 2/7 gesta-tional weeks. Transvaginal cervical length was 20.9mm andthere was no evidence of funneling or dynamic changes. Noamniotic fluid sludge could be identified below the presentingvertex (also see Supplementary Movie 5).

Figure 7. Sonographic cervical evaluation at 36 2/7 gesta-tional weeks on the day of delivery. Transvaginal cervicallength was 19.7mm with mild funneling distal to the cho-rioamniotic membranes. No amniotic fluid sludge could beidentified below the presenting vertex (also seeSupplementary Movie 6).

THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINE 5

patient was asymptomatic, we offered an amniocen-tesis when a sonographic short cervix and sludgewere identified at 21weeks of gestation, in order todetermine the microbial status of the amniotic cavity.Prior observations have suggested an associationbetween short cervix and intra-amniotic infection[24,25]. Moreover, investigators have also reportedthat a mid-trimester sonographic short cervix(�25mm) is associated with intra-amniotic inflamma-tion, and such patients are at risk for adverse preg-nancy outcomes [26–29]. For example, inasymptomatic women with a mid-trimester sono-graphic short cervix (�15mm), the prevalence ofintra-amniotic inflammation and negative amnioticfluid culture was 22.2% [27]. In such patients, therewas a higher rate of preterm delivery within 7 dayscompared to those without intra-amniotic inflamma-tion (40 vs. 5.7%; p¼ 0.016) [27].

Cervical insufficiency and intra-amniotic infection/inflammation

Cervical insufficiency generally refers to a condition inwhich there is mid-trimester cervical dilatation or pro-truding chorioamniotic membranes in the absence ofuterine contractions. Unfortunately, such condition is arisk factor for spontaneous mid-trimester abortion orearly preterm birth [10]. Studies have reported an8–52% prevalence of intra-amniotic infection (ascer-tained by amniocentesis) in patients with cervicalinsufficiency [4–11]. Moreover, intra-amniotic inflam-mation has been identified in patients with cervicalinsufficiency [7,30,31], and has been associated withpreterm delivery, as well as adverse pregnancy andneonatal outcomes [7,30].

A subset of patients with intra-amniotic inflamma-tion will have no demonstrable microorganismsdetected via cultivation or molecular methods, andtherefore represent cases of sterile intra-amnioticinflammation, which applies to the case herein[12–17]. The inflammatory response detected in amni-otic fluid may be due to a sterile inflammatoryresponse activated by non-microbial endogenous dan-ger signals, such as cellular alarm signals from dis-tressed and injured cells or tissues [15,16,32–35].Indeed, sterile intra-amniotic inflammation is morecommon than intra-amniotic infection in asymptom-atic patients with a sonographic short cervix [15], andthe same has been reported for patients with cervicalinsufficiency [7]. Lee et al. reported that a very highpercentage (81%) of patients with acute cervical insuf-ficiency have intra-amniotic inflammation demon-strated by amniocentesis (regardless of amniotic fluidculture results), and that the prognosis is poor, with50% delivering preterm within 7 days [7]. The sameauthors also reported that in patients with intra-amni-otic inflammation but with negative amniotic fluid cul-tures, 95% of patients had a preterm birth, and theperinatal mortality rate was 63% [7]. Therefore, basedon such data, our patient underwent amniocentesisand the results were consistent with sterile intra-amni-otic inflammation.

Amniotic fluid sludge is an indicator of microbialinvasion of the amniotic cavity and intra-amnioticinflammation

Amniotic fluid sludge (free-floating hyperechogenic orparticulate material within the amniotic fluid in closeproximity to the uterine cervix) identified on

Figure 8. Gross examination of the placenta. The fetal surface is shown in the left image. Extraplacental membranes were translu-cent (right image).

6 L. YEO ET AL.

transvaginal sonography represents a sign that micro-bial invasion of the amniotic cavity and an inflamma-tory process are occurring [18]. It is feasible thatprogressive infection induces an intense inflammatoryresponse, and that the combination of inflammatorycells with microorganisms leads to sludge formation.The observation that the shorter a cervical length is,the greater the likelihood of sludge has been inter-preted to indicate that intra-amniotic infection/inflam-mation will eventually lead to a short cervix [19].Moreover, sludge is an independent risk factor for pre-term PROM, as well as spontaneous preterm deliv-ery [19–21].

For our case, antibiotic therapy was initiated at22weeks of gestation and sonographic examination3 days later showed a large quantity of sludge.Antibiotics were administered for a duration of 11 daysuntil 23 3/7weeks of gestation. Yet, it is noteworthythat at 25 3/7weeks of gestation, the quantity ofsludge had diminished greatly. Moreover, by 33 2/7and 33 6/7weeks of gestation, no sludge could bevisualized below the presenting vertex. It is unclear ifsuch finding was a false-negative (see below) or ifthere was true resolution of sludge by the third tri-mester. Although the diminished quantity and thendisappearance of sludge cannot be attributed directlyto antibiotic therapy (i.e. causation), it is possible,since no other treatment was administered except forvaginal progesterone. The patient received such medi-cation when the cervical length was 7.3mm; however,during the course of vaginal progesterone therapy,the cervix progressed to 3 cm dilatation on speculumexam, and transvaginal sonogram demonstrated nomeasurable cervical length with membranes bulgingpast the external cervical os into the vagina.Therefore, vaginal progesterone was not effective inthis case. Such observations are not surprising.Fonseca et al. noted that vaginal progesteronereduced the rate of spontaneous preterm delivery<34weeks of gestation by only 25% in patients with acervical length of 6–10mm, but the effect size was75% in those with a cervical length between 11 and15mm [36]. A possible explanation for these results isthat women with a very short cervix are more likely tohave intra-amniotic inflammation and may be lessresponsive to progesterone, as was the caseherein [26,27,37].

Several points regarding amniotic fluid sludge arenoteworthy. First, the “resolution” of sludge observedin the current case may not occur all the time and inall patients, such as those who have intra-amnioticinfection. Microbial biofilms have been identified in

the presence of intra-amniotic infection with sludge[38]. Such biofilms are communities of sessile microor-ganisms formed by cells that are attached irreversiblyto a substratum or interface, or to each other andembedded in a hydrated matrix of extracellular poly-meric substances [38]. This is relevant, because bac-teria in these biofilms are more resistant toantimicrobial agents and to the host response toinfection [39–41].

Second, we theorize that sometimes the“disappearance” of sludge is a false-negative finding.Based upon a breadth of experience in our unit evalu-ating sludge via transvaginal ultrasound throughoutgestation, there may be several reasons for a false-negative result: (1) the fetal vertex, particularly withadvancing gestational age in the third trimester, mayprohibit the collection of sludge in close proximity tothe internal os (i.e. the vertex acts as a “plug”); (2) it ispossible that maternal positional changes may influ-ence the likelihood that sludge settles into the loweruterine segment. For example, patients who havebeen lying supine (e.g. in the hospital bed) may notdemonstrate sludge in the lower uterine segment ontransvaginal sonography, because they simply havenot been upright; (3) an active fetus (and/or perhapsan active mother) can disperse the free-floating hyper-echogenic material throughout the amniotic cavity,such that it does not collect in the lower uterine seg-ment; and (4) sonographic examinations capture onlya moment in time; sludge may still be present, but forvarious reasons just described is not evident at thetime of exam. To confirm the presence or absence ofsludge, we always tap gently on the maternal abdo-men in the area of the lower uterine segment toobserve dispersion of any hyperechogenic or particu-late material (see Supplementary Movie 2,Supplementary Movie 5, Supplementary Movie 6).

Antibiotics are effective in treating patients withintra-amniotic infection/inflammation and cervicalinsufficiency

The conventional view is that intra-amniotic infectionand/or intra-amniotic inflammation in patients withacute cervical insufficiency cannot be treated success-fully, and that delivery is inevitable. For our patient,the sonographic findings (i.e. membranes bulging pastthe external cervical os into the vagina) at 22 3/7weeks of gestation were very concerning, especiallyin the presence of sludge. Yet, Lee et al. observedunexpectedly that some patients with bulging mem-branes and intra-amniotic inflammation [defined as an

THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINE 7

elevated amniotic fluid concentration of matrix metal-loproteinase-8 (MMP-8) >23 ng/ml] who were treatedwith parenteral antibiotics for more than 7 days ultim-ately delivered after 34weeks of gestation [7]. Of thewomen who underwent repeat amniocentesis, in allcases the MMP-8 concentration was lower in thesecond amniotic fluid sample, suggesting that eitherantibiotic therapy eradicated a microbial stimulus forinflammation, or that the natural history of this condi-tion includes a reversible phase [7].

Subsequently, Oh et al. described a series ofwomen with cervical insufficiency (16–27.9weeks ofgestation) and intra-amniotic infection/inflammationwho were treated with three antimicrobial agents (cef-triaxone, clarithromycin, and metronidazole) [10].Cervical dilatation was determined by visual examin-ation during sterile speculum examination (vs. transva-ginal sonography), and intra-amniotic inflammationwas defined as IL-6> 2.6 ng/ml. After antibiotic ther-apy, follow-up amniocentesis was routinely offered tomonitor the microbiologic and inflammatory status ofthe amniotic cavity. The authors reported that: (1) 75%of women had objective evidence of resolution of theintra-amniotic infectious/inflammatory process aftertreatment; and (2) there was overall treatment success(i.e. resolution of intra-amniotic infection/inflammationor delivery �34weeks) in 59% of cases [10]. Therefore,this report provided strong evidence that antibioticscan be effective in treating intra-amniotic infection/inflammation in patients with cervical insufficiency. Itis especially remarkable that for three patients withbulging membranes who received antibiotics (but notcervical cerclage placement), there was biochemicalevidence of successful treatment of intra-amnioticinflammation, as well as spontaneous reduction ofmembranes back into the uterine cavity while receiv-ing antibiotic therapy. Indeed, the authors reportedthat this was an unexpected observation made duringthe course of their study. We also surprisinglyobserved resolution of cervical insufficiency in ourpatient, as demonstrated on sonographic examinationat 23 6/7weeks of gestation (Figure 4, SupplementaryMovie 3), as well as speculum examination at 25weeksof gestation.

Our patient received triple antibiotic therapy for11 days only, because she experienced symptomaticside effects. Yet, she also received oral metronidazoleto treat bacterial vaginosis for an additional 7 days(25� 26weeks of gestation) when the cervix appeared1� 2 cm dilated on speculum exam. In Oh et al.’sstudy, the use and discontinuation of antibiotics wereleft to the discretion of the treating clinician. For the

patients who had confirmed resolution of intra-amni-otic inflammation and delivered �34weeks, themedian (range) duration of the antibiotic regimen was27 (7–66) days [10]. Metronidazole was administeredfor a maximum of 4weeks. Therefore, in contrast toOh’s study, our patient received a much shorter courseof antibiotic treatment. In addition, the amniotic fluidIL-6 concentration was 3.055 ng/ml (normal <2.6 ng/ml). Oh et al. reported that patients with cervical insuf-ficiency who received antimicrobial agents and whodelivered within 1week of amniocentesis had a signifi-cantly higher median (range) amniotic fluid IL-6 con-centration than those who remained undelivered forat least 1week after amniocentesis [47.7 (5.7� 49.3)vs. 9.4 (2.8� 35.4) ng/mL; p¼ 0.005] [10]. Therefore,one can speculate that despite the shorter duration ofantibiotics that our patient received, perhaps therewas a successful outcome because the amniotic fluidIL-6 concentration was only slightly above the 2.6 ng/ml cutoff. Ideally, the duration of antibiotic treatmentcan be directed clinically via a follow-up amniocentesisto determine the microbial and inflammatory status ofthe amniotic cavity.

Why did we choose the specific antibiotic regimenfor our case? Erythromycin or azithromycin has beenshown to be frequently inadequate in eradicatingintra-amniotic infection with Ureaplasma spp., perhapsdue to limited transplacental passage leading to inad-equate antimicrobial activity in the amniotic fluid [42].In contrast, clarithromycin is not only effective fortreating intra-amniotic infection by Ureasplasma spp.[13], but this agent also has a much higher rate oftransplacental passage [42]. Moreover, in patients withpreterm PROM and either intra-amniotic infection orsterile intra-amniotic inflammation, IV clarithromycin isassociated with attenuation in the intensity of theintra-amniotic inflammatory response [23]. Ceftriaxonewas administered due to its enhanced coverage ofaerobic bacteria and high rate of transplacental pas-sage [43,44], while metronidazole was utilized becauseit has a powerful effect against anaerobic bacteria.Finally, our group has previously described the suc-cessful eradication of intra-amniotic inflammation and/or infection in at least 33% of patients with pretermPROM after administration of this specific antibioticregimen [22].

Another interesting question is the mechanisms bywhich antibiotics can treat intra-amniotic inflammation[10]. This is not entirely clear. One possibility is thatclarithromycin suppresses intra-amniotic inflammationcaused by danger signals by down-regulating theexpression of pro-inflammatory transcription factors

8 L. YEO ET AL.

(e.g. activator protein-1 and nuclear factor-ŒB), whichcan induce the production of pro-inflammatory cyto-kines [10]. Another possibility is that antibiotics areeffective against microorganisms that were notdetected via cultivation techniques [15,17,45–47].

For the case herein, we do not know if intra-amni-otic inflammation was truly eradicated, because arepeat analysis of amniotic fluid was not performedafter antibiotic treatment. However, several points arenotable which may provide indirect support that thiswas the case. First, on the day antibiotic therapy wasinitiated, the maternal C-reactive protein concentrationwas elevated, but became normal six days later duringtreatment and remained as such, even after antibioticswere discontinued. Second, neonatal labs (i.e. com-plete blood count, blood cultures and C-reactive pro-tein) obtained on the day of delivery were withinnormal limits. Finally, placental histologic examinationdid not show evidence of acute inflammatory lesionssuch as acute chorioamnionitis, funisitis, or chorionicvasculitis. Indeed, sterile intra-amniotic inflammation isknown to be associated with acute histologic cho-rioamnionitis in 40–60% of cases [15–17,48,49].Instead, chronic lymphoplasmacytic deciduitis was iden-tified, and this may not be surprising. Chronic decidui-tis is characterized by the presence of lymphocytesand plasma cells in the basal plate of the placenta,and chronic microbial infection or immune mecha-nisms have been implicated in its etiology [48,50]. Yet,chronic lymphoplasmacytic deciduitis has also beenreported in 9% of placentas delivered from womenwith term pregnancies and a normal outcome [51].

Conclusions

In conclusion, intra-amniotic inflammation is a poorprognostic factor in the presence of cervical insuffi-ciency. Therefore, we propose that for patients pre-senting with acute cervical insufficiency, the presenceof intra-amniotic inflammation (and/or infection)should be considered, and amniocentesis should beoffered to determine the microbial and inflammatorystatus of the amniotic cavity, particularly when amni-otic fluid sludge is present. Recent evidence suggeststhat antibiotics can be effective in treating intra-amni-otic infection/inflammation in the setting of cervicalinsufficiency. We report herein a case in which admin-istration of antibiotic therapy in the presence of mid-trimester cervical insufficiency, sterile intra-amnioticinflammation, and amniotic fluid sludge was associ-ated with resolution of the cervical findings. Therefore,based on scientific evidence and the outcome of our

case, a viable option is to offer antibiotic therapy suchas the regimen described herein to such patients.

Acknowledgments

Dr. Romero has contributed to this work as part of his offi-cial duties as an employee of the United StatesFederal Government.

Disclosure statement

No potential conflict of interest was reported bythe author(s).

Funding

This report was supported, in part, by the PerinatologyResearch Branch, Division of Obstetrics and Maternal-FetalMedicine, Division of Intramural Research, Eunice KennedyShriver National Institute of Child Health and HumanDevelopment, National Institutes of Health, U.S. Departmentof Health and Human Services (NICHD/NIH/DHHS); and, inpart, with Federal funds from NICHD/NIH/DHHS underContract No. HHSN275201300006C.

References

[1] Iams JD, Johnson FF, Sonck J, et al. Cervical compe-tence as a continuum: a study of ultrasonographiccervical length and obstetric performance. Am JObstet Gynecol. 1995;172(4):1097–1103.

[2] Romero R, Espinoza J, Erez O, et al. The role of cer-vical cerclage in obstetric practice: can the patientwho could benefit from this procedure be identified?Am J Obstet Gynecol. 2006;194(1):1–9.

[3] Oh KJ, Romero R, Park JY, et al. Evidence that anti-biotic administration is effective in the treatment of asubset of patients with intra-amniotic infection/inflammation presenting with cervical insufficiency.Am J Obstet Gynecol. 2019;221(2):140.e1–140.e18.

[4] Romero R, Gonzalez R, Sepulveda W, et al. Infectionand labor. VIII. Microbial invasion of the amniotic cav-ity in patients with suspected cervical incompetence:prevalence and clinical significance. Am J ObstetGynecol. 1992;167(4 Pt 1):1086–1091.

[5] Goodlin RC. Cervical incompetence, hour-glass mem-branes, and amniocentesis. Obstet Gynecol. 1979;54:748–750.

[6] Mays JK, Figueroa R, Shah J, et al. Amniocentesis forselection before rescue cerclage. Obstet Gynecol.2000;95:652–655.

[7] Lee SE, Romero R, Park CW, et al. The frequency andsignificance of intraamniotic inflammation in patientswith cervical insufficiency. Am J Obstet Gynecol.2008;198(6):633.e1-8–633.e8.

[8] Bujold E, Morency AM, Rallu F, et al. Bacteriology ofamniotic fluid in women with suspected cervicalinsufficiency. J Obstet Gynaecol Can. 2008;30(10):882–887.

THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINE 9

[9] Airoldi J, Pereira L, Cotter A, et al. Amniocentesis priorto physical exam-indicated cerclage in women withmidtrimester cervical dilation: results from theexpectant management compared to physicalexam-indicated cerclage international cohort study.Am J Perinatol. 2009;26:63–68.

[10] Oh KJ, Lee SE, Jung H, et al. Detection of ureaplasmasby the polymerase chain reaction in the amnioticfluid of patients with cervical insufficiency. J PerinatMed. 2010;38:261–268.

[11] Lisonkova S, Sabr Y, Joseph KS. Diagnosis of subclin-ical amniotic fluid infection prior to rescue cerclageusing gram stain and glucose tests: an individualpatient meta-analysis. J Obstet Gynaecol Can. 2014;36(2):116–122.

[12] Shim SS, Romero R, Hong JS, et al. Clinical signifi-cance of intra-amniotic inflammation in patients withpreterm premature rupture of membranes. Am JObstet Gynecol. 2004;191(4):1339–1345.

[13] Yoon BH, Romero R, Park JY, et al. Antibiotic adminis-tration can eradicate intra-amniotic infection or intra-amniotic inflammation in a subset of patients withpreterm labor and intact membranes. Am J ObstetGynecol. 2019;221(2):142.e1–e22.

[14] Yoon BH, Romero R, Moon JB, et al. Clinical signifi-cance of intra-amniotic inflammation in patients withpreterm labor and intact membranes. Am J ObstetGynecol. 2001;185(5):1130–1136.

[15] Romero R, Miranda J, Chaiworapongsa T, et al. Sterileintra-amniotic inflammation in asymptomatic patientswith a sonographic short cervix: prevalence and clin-ical significance. J Matern Fetal Neonatal Med. 2015;28(11):1343–1359.

[16] Romero R, Miranda J, Chaiworapongsa T, et al.Prevalence and clinical significance of sterile intra-amniotic inflammation in patients with preterm laborand intact membranes. Am J Reprod Immunol. 2014;72(5):458–474.

[17] Romero R, Miranda J, Chaemsaithong P, et al. Sterileand microbial-associated intraamniotic inflammationin preterm prelabor rupture of membranes. J MaternFetal Neonatal Med. 2015;28(12):1394–1409.

[18] Romero R, Kusanovic JP, Espinoza J, et al. What isamniotic fluid ’sludge’? Ultrasound Obstet Gynecol.2007;30(5):793–798.

[19] Kusanovic JP, Espinoza J, Romero R, et al. Clinical sig-nificance of the presence of amniotic fluid ‘sludge’ inasymptomatic patients at high risk for spontaneouspreterm delivery. Ultrasound Obstet Gynecol. 2007;30(5):706–714.

[20] Adanir I, Ozyuncu O, Gokmen Karasu AF, et al.Amniotic fluid “sludge”; prevalence and clinical signifi-cance of it in asymptomatic patients at high risk forspontaneous preterm delivery. J Matern FetalNeonatal Med. 2018;31(2):135–140.

[21] Tsunoda Y, Fukami T, Yoneyama K, et al. The pres-ence of amniotic fluid sludge in pregnant womenwith a short cervix: an independent risk of pretermdelivery. J Matern Fetal Neonatal Med. 2020;33(6):920–923.

[22] Lee J, Romero R, Kim SM, et al. A new antibiotic regi-men treats and prevents intra-amniotic inflammation/

infection in patients with preterm PROM. J MaternFetal Neonatal Med. 2016;29(17):2727–2737.

[23] Kacerovsky M, Romero R, Stepan M, et al. Antibioticadministration reduces the rate of intraamnioticinflammation in preterm prelabor rupture of themembranes. Am J Obstet Gynecol. 2020;223(1):114.e1–e20.

[24] Hassan S, Romero R, Hendler I, et al. A sonographicshort cervix as the only clinical manifestation of intra-amniotic infection. J Perinat Med. 2006;34:13–19.

[25] Rust OA, Atlas RO, Reed J, et al. Revisiting the shortcervix detected by transvaginal ultrasound in thesecond trimester: why cerclage therapy may not help.Am J Obstet Gynecol. 2001;185(5):1098–1105.

[26] Kiefer DG, Keeler SM, Rust OA, et al. Is midtrimestershort cervix a sign of intraamniotic inflammation? AmJ Obstet Gynecol. 2009;200(4):374.e1-5–374.e5.

[27] Vaisbuch E, Hassan SS, Mazaki-Tovi S, et al. Patientswith an asymptomatic short cervix (<or ¼ 15mm)have a high rate of subclinical intraamniotic inflam-mation: implications for patient counselling. Am JObstet Gynecol. 2010;202:433.e1–8.

[28] Keeler SM, Kiefer DG, Rust OA, et al. Comprehensiveamniotic fluid cytokine profile evaluation in womenwith a short cervix: which cytokine(s) correlates bestwith outcome? Am J Obstet Gynecol. 2009;201:276.e.1–e6.

[29] Kiefer DG, Keeler SM, Rust O, et al. Amniotic fluidinflammatory score is associated with pregnancy out-come in patients with mid trimester short cervix. AmJ Obstet Gynecol. 2012;206(1):68.e1–e6.

[30] Jung EY, Park KH, Lee SY, et al. Non-invasive predic-tion of intra-amniotic infection and/or inflammationin patients with cervical insufficiency or an asymp-tomatic short cervix (�15mm)). Arch Gynecol Obstet.2015;292(3):579–587.

[31] Diago Almela VJ, Martinez-Varea A, Perales-Puchalt A,et al. Good prognosis of cerclage in cases of cervicalinsufficiency when intra-amniotic inflammation/infec-tion is ruled out. J Matern Fetal Neonatal Med. 2015;28(13):1563–1568.

[32] Matzinger P. The danger model: a renewed sense ofself. Science. 2002;296(5566):301–305.

[33] Ogura Y, Sutterwala FS, Flavell RA. The inflamma-some: first line of the immune response to cell stress.Cell. 2006;126(4):659–662.

[34] Pedra JH, Cassel SL, Sutterwala FS. Sensing pathogensand danger signals by the inflammasome. Curr OpinImmunol. 2009;21(1):10–16.

[35] Gotsch F, Romero R, Chaiworapongsa T, et al.Evidence of the involvement of caspase-1 underphysiologic and pathologic cellular stress duringhuman pregnancy: a link between the inflammasomeand parturition. J Matern Fetal Neonatal Med. 2008;21(9):605–616.

[36] Fonseca EB, Celik E, Parra M, Fetal MedicineFoundation Second Trimester Screening Group, et al.Progesterone and the risk of preterm birth amongwomen with a short cervix. N Engl J Med. 2007;357(5):462–469.

[37] Romero R, Nicolaides K, Conde-Agudelo A, et al.Vaginal progesterone in women with an

10 L. YEO ET AL.

asymptomatic sonographic short cervix in the mid-trimester decreases preterm delivery and neonatalmorbidity: a systematic review and metaanalysis ofindividual patient data. Am J Obstet Gynecol. 2012;206(2):124.e1-19–124.e19.

[38] Romero R, Schaudinn C, Kusanovic JP, et al. Detectionof a microbial biofilm in intraamniotic infection. Am JObstet Gynecol. 2008;198(1):135.e1–135.e5.

[39] Donlan RM. Role of biofilms in antimicrobial resist-ance. Asaio J. 2000;46(6):S47–S52.

[40] Stewart PS, Costerton JW. Antibiotic resistance of bac-teria in biofilms. Lancet. 2001;358(9276):135–138.

[41] Davies D. Understanding biofilm resistance to antibac-terial agents. Nat Rev Drug Discov. 2003;2(2):114–122.

[42] Witt A, Sommer EM, Cichna M, et al. Placental pas-sage of clarithromycin surpasses other macrolide anti-biotics. Am J Obstet Gynecol. 2003;188(3):816–819.

[43] Kafetzis DA, Brater DC, Fanourgakis JE, et al.Ceftriaxone distribution between maternal blood andfetal blood and tissues at parturition and betweenblood and milk postpartum. Antimicrob AgentsChemother. 1983;23(6):870–873.

[44] Lamb HM, Ormrod D, Scott LJ, at al. Ceftriaxone: anupdate of its use in the management of community-acquired and nosocomial infections. Drugs. 2002;62(7):1041–1089.

[45] DiGiulio DB, Romero R, Kusanovic JP, et al. Prevalenceand diversity of microbes in the amniotic fluid, the

fetal inflammatory response, and pregnancy outcomein women with preterm pre-labor rupture of mem-branes. Am J Reprod Immunol. 2010;64:38–57.

[46] DiGiulio DB, Romero R, Amogan HP, et al. Microbialprevalence, diversity and abundance in amniotic fluidduring preterm labor: a molecular and culture-basedinvestigation. PLoS One. 2008;3(8):e3056.

[47] Rowlands S, Danielewski JA, Tabrizi SN, et al.Microbial invasion of the amniotic cavity in mid-trimester pregnancies using molecular microbiology.Am J Obstet Gynecol. 2017;217(1):71.e1–e5.

[48] Kim CJ, Romero R, Chaemsaithong P, et al. Acute cho-rioamnionitis and funisitis: definition, pathologic fea-tures, and clinical significance. Am J Obstet Gynecol.2015;213(4 Suppl):S29–S52.

[49] Romero R, Miranda J, Chaiworapongsa T, et al. A novelmolecular microbiologic technique for the rapid diagno-sis of microbial invasion of the amniotic cavity andintra-amniotic infection in preterm labor with intactmembranes. Am J Reprod Immunol. 2014;71(4):330–358.

[50] Khong TY, Bendon RW, Qureshi F, et al. Chronic decid-uitis in the placental basal plate: definition and interob-server reliability. Hum Pathol. 2000;31(3):292–295.

[51] Romero R, Kim YM, Pacora P, et al. The frequency andtype of placental histologic lesions in term pregnan-cies with normal outcome. J Perinat Med. 2018;46(6):613–630.

THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINE 11