OBSTETRICS

Intra-Amniotic Sludge, Short Cervix,and Risk of Preterm DeliveryEmmanuel Bujold, MD, FRCSC, Jean-Charles Pasquier, MD, Jessica Simoneau, Marie-Hélène Arpin,Louise Duperron, MD, FRCSC, Anne-Maude Morency, MD, François Audibert, MDHôpital Sainte-Justine, Université de Montréal, Montréal (Québec)

Abstract

Objective: To evaluate the association between intra-amniotic sludgeseen at cervical ultrasound and preterm delivery.

Method: This retrospective study included women at high risk forpreterm delivery who were referred for second trimester cervicallength measurement by ultrasound between 18 and 32 weeks’gestation. Patients with multiple gestations, cerclage, or pretermlabour were excluded. Ultrasound images were reviewed by twoindependent observers and divided into three groups: (1) noamniotic sludge, (2) light sludge, and (3) dense sludge in theamniotic fluid. The primary outcome measures were delivery within14 days of examination and delivery before 34 weeks’ gestation.Logistic regression analyses were performed to adjust forconfounding factors.

Results: Eighty-nine patients met the inclusion criteria. Meangestational age at presentation was 25.8 ± 4.4 weeks, and meancervical length was 33 ± 12 mm. The prevalence of light and denseamniotic fluid sludge was 10.1% and 5.6%, respectively. Deliverywithin 14 days of examination occurred in four (5.3%) women withno sludge, in two (22.2%) women with light sludge, and in three(60.0%) women with dense sludge (P < 0.01). Delivery before34 weeks occurred in five (6.7%), four (44.4%) and four (80.0%)women, respectively (P < 0.01). Logistic regression analysesdemonstrated that light amniotic fluid sludge, dense sludge, andcervical length of less than 25 mm were all significant andindependent predictors of delivery within 14 days of examinationand delivery prior to 34 weeks.

Conclusion: The presence of amniotic fluid sludge is associated withdelivery within 14 days and delivery before 34 weeks’ gestation.

Résumé

Objectif : Évaluer l’association entre les agrégats intra-amniotiquesconstatés par échographie cervicale et l’accouchement prématuré.

Méthode : La présente étude rétrospective s’est penchée sur desfemmes, courant des risques élevés de connaître unaccouchement prématuré, qui ont été orientées vers des servicesde mesure de la longueur cervicale au deuxième trimestre paréchographie, entre la 18e et la 32e semaine de gestation. Lespatientes présentant des gestations multiples, un cerclage ou untravail préterme ont été exclues. Les images échographiques ontété analysées par deux observateurs indépendants et divisées entrois groupes : (1) aucun agrégat amniotique, (2) agrégats légers et

(3) agrégats denses dans le liquide amniotique. Les critèresd’évaluation primaires étaient l’accouchement dans les 14 jourssuivant l’examen et l’accouchement avant la 34e semaine degestation. Des analyses de régression logistique ont été effectuéespour neutraliser l’effet des facteurs confusionnels.

Résultats : Quatre-vingt-neuf patientes ont satisfait aux critèresd’inclusion. L’âge gestationnel moyen au moment de laprésentation était de 25,8 ± 4,4 semaines et la longueur cervicalemoyenne était de 33 ± 12 mm. La prévalence des agrégats légerset denses dans le liquide amniotique était de 10,1 % et de 5,6 %,respectivement. Un accouchement dans les 14 jours suivantl’examen a été constaté chez quatre (5,3 %) des femmes neprésentant pas d’agrégats, chez deux (22,2 %) des femmesprésentant de légers agrégats et chez trois (60,0 %) des femmesprésentant des agrégats denses (P < 0,01). Un accouchementavant la 34e semaine de gestation a été constaté chez cinq(6,7 %), quatre (44,4 %) et quatre (80,0 %) de ces femmes,respectivement (P < 0,01). Les analyses de régression logistiqueont démontré que la présence d’agrégats légers ou denses dans leliquide amniotique et qu’une longueur cervicale inférieure à 25 mmconstituaient toutes deux des prédicteurs significatifs etindépendants de l’accouchement dans les 14 jours suivantl’examen et de l’accouchement avant la 34e semaine de gestation.

Conclusion : La présence d’agrégats dans le liquide amniotique estassociée à l’accouchement dans les 14 jours suivant l’examen et àl’accouchement avant la 34e semaine de gestation.

J Obstet Gynaecol Can 2006;28(3):198–202

INTRODUCTION

Despite improvements in prenatal care, preterm birth

continues to be a leading cause of perinatal death, and

the rate of spontaneous preterm births has not

decreased in the past 30 years.1,2 Two main factors contrib-

ute to such disappointing figures: the inadequacy of our

therapeutic arsenal and the absence of reliable criteria for

identifying a population at high risk for preterm delivery.

Measurement of cervical length is one of the most recenttools that help in the prediction of preterm birth. Short cer-vical length, usually defined as below the tenth percentile, or< 25 mm measured by transvaginal ultrasound examina-tion, has been strongly related to the risk of preterm birth.3,4

Despite an excellent negative predictive value, the positivepredictive value of this measurement in asymptomaticwomen remains low.4 Moreover, the cause of a short cervix

198 �MARCH JOGC MARS 2006

OBSTETRICS

Key Words: Pregnancy, cervical length, amniotic fluid, ultrasound,preterm birth

Competing Interests: None declared.

Received on December 12, 2005

Accepted on January 10, 2006

has not been elucidated and is probably multifactorial. Pos-sible contributing factors include uterine contractions,inflammation, and other unknown mechanisms, makingeffective therapy even more difficult to develop.5

Recently, Espinoza and colleagues demonstrated that thepresence of intra-amniotic fluid “sludge,” a cluster of par-ticulate matter sitting close to the cervical os, was associatedwith microbial invasion of the amniotic cavity, histologicalchorioamnionitis, and impending preterm delivery inpatients with preterm labour and intact membranes.6

Seventy-one percent of women found to haveintra-amniotic sludge were delivered in the seven days fol-lowing identification, compared with 16% of women with-out intra-amniotic sludge. These very important findingssuggest that this is an index with high positive predictivevalue for preterm birth.

The objectives of this study were (1) to evaluate the pres-ence of intra-amniotic sludge in a subgroup of women athigh risk for preterm delivery but without the diagnosis ofpreterm labour, and (2) to correlate the finding ofintra-amniotic sludge with the risk of subsequent pretermdelivery.

MATERIALS AND METHODS

After approval of our research protocol by the Scientificand Ethics Committees of Hôpital Sainte-Justine, a retro-spective study was conducted by searching our digitallibrary of ultrasound images collected at the AntenatalUltrasound Centre of our institution between January 2004and April 2005. The inclusion criteria were (1) women witha live singleton pregnancy referred for a transvaginalultrasonographic measurement of cervical length and(2) gestational age between 18 and 32 weeks’ gestation. Theexclusion criteria were (1) women with a diagnosis ofpreterm labour or with regular uterine contractions on theday of examination, (2) multiple gestation, (3) fetal anoma-lies or fetal intrauterine growth restriction, (4) cervicalcerclage, and (5) placenta previa.

The ultrasound images were reviewed by two independentsonographers blinded to the obstetrical outcomes to recordthe measured cervical length and the presence or absence ofintra-amniotic fluid sludge. “Amniotic fluid sludge” wasdefined as dense aggregates of particulate matter seen onultrasound in proximity to the internal cervical os, asdescribed previously by Espinoza et al.6 In some cases, theparticulate matter was less dense and clustered, and thesonographers classified these cases as “light sludge.” There-fore, the patients were classified into three groups: (1) noamniotic sludge; (2) light sludge, when the particulate mat-ter was not clustered; and (3) dense sludge, when the partic-ulate matter was clustered in the amniotic fluid (see Figure).

When there was no agreement between the twosonographers, a third independent sonographer blinded tothe obstetrical outcomes was consulted for the appropriateclassification. All transvaginal ultrasounds were performedwith commercially available ultrasound systems (ATLHDI-3500 and ALOKA SSD-5500). Gestational age at thetime of ultrasonographic examination and at delivery wasconfirmed by an ultrasound performed prior to 20 weeks’gestation in all women included in the study.

The medical records of each patient included in the studywere reviewed by an independent observer blinded to theultrasound findings. The following information was col-lected: maternal age, gravidity, parity, previous pretermdelivery, the reason for measuring cervical length, date ofdelivery, gestational age at delivery, birth weight, and pres-ence or absence of histological chorioamnionitis recordedin the pathology report (systematically recorded for alldeliveries in our centre).

Patients with no sludge, light sludge, or dense sludge werecompared. The primary outcomes were delivery within14 days of ultrasound examination and delivery before34 weeks’ gestation. The other adverse outcomes measuredwere the interval between ultrasound examination anddelivery, delivery before 37 weeks’ gestation, andhistological chorioamnionitis. Multiple stepwise logisticregression analyses were performed to adjust for confound-ing factors, including intra-amniotic sludge, short cervicallength (< 25 mm), nulliparity, prior preterm delivery, andadvanced maternal age. The amniotic sludge variable wasmeasured as a categorical variable divided into three groups.These factors were selected prior to study performance.Subsequent regression analyses with 15 mm as a cut-off forcervical length instead of 25 mm were also undertaken. Thechi-square test, Fisher exact test, Student t test, and theKruskal-Wallis test were used when appropriate. SPSS 13.0(SPSS Inc., Chicago, IL) was used for the statistical analyses,and P values < 0.05 were considered to be statisticallysignificant.

RESULTS

During the study period, 89 patients met the inclusion crite-ria. Although the specific reason for the cervical ultrasoundexamination was not available in several cases, we foundthat nine women had a diagnosis of preterm labour,39 women had a diagnosis of threatened preterm labour(with no change in cervical dilatation on examination) thatwas resolved, and 24 (27%) had a history of preterm deliv-ery. Mean gestational age at presentation was 25.8 ± 4.4(mean ± standard deviation) weeks, and mean cervicallength was 33 ± 12 mm. Light sludge was noted in 10.1%(9/89) and dense sludge in 5.6% (5/89) of examinations; 76

Intra-Amniotic Sludge, Short Cervix, and Risk of Preterm Delivery

MARCH JOGC MARS 2006 � 199

(85.4%) women did not show any sign of intra-amnioticsludge on their ultrasound images. The two independentsonographers agreed on the classification of 87 (98%) of89 cases, and they were in complete agreement with theclassification of the five cases with dense sludge. The clini-cal characteristics of the women included in the studyaccording to the presence of amniotic sludge are reported inTable 1. We found that women with intra-amniotic sludgehad shorter cervical length than those without.

Table 2 compares the obstetrical outcomes according to thepresence of amniotic sludge. Women with intra-amnioticsludge, and particularly those with dense intra-amnioticsludge, were more likely to deliver within 14 days of theexamination, before 34 weeks’ gestation, before 37 weeks’gestation, and to have a diagnosis of histologicalchorioamnionitis. Logistic regression analyses demon-strated that the presence of dense intra-amniotic fluidsludge and cervical length < 25 mm were both significantindependent predictors of delivery within 14 days and inde-pendent predictors of delivery < 34 weeks (Table 3). Wesubsequently decided to repeat the multivariable logisticregression analysis with cervical length < 15 mm as theconfounding factor. The presence of dense intra-amnioticsludge remained a significant predictor of delivery< 34 weeks (odds ratio [OR] 23.3; 95% confidence interval[CI] 1.2, 470.1). Finally, Table 4 shows the rate of subse-quent adverse outcomes according to the presence orabsence of both short cervix and intra-amniotic sludge atthe time of transvaginal examination. The combination ofboth factors was associated with a very high rate of deliverybefore 34 weeks’ gestation (75%) and delivery within14 days of examination (50%).

DISCUSSION

Our study reveals that intra-amniotic sludge duringtransvaginal examination is linked with a very high likeli-hood of delivery before 34 weeks’ gestation and also with ahigh risk of delivery within the next 14 days. Moreover,even if the presence of intra-amniotic sludge was associatedwith shorter cervical length, it remained an independent riskfactor related to these adverse outcomes after adjustmentfor short cervical length. Finally, intra-amniotic sludge was arisk factor for delivery before 37 weeks’ gestation andhistological chorioamnionitis. These findings are inagreement with the data of Espinoza and colleagues, whostudied the clinical significance of amniotic fluid sludge in84 patients with preterm labour and intact membranes.6

Intra-amniotic sludge in this subgroup of women was asso-ciated with spontaneous delivery within 48 hours (OR 19.6;95% CI 1.5, 257.4) and within seven days (OR 11.7; 95% CI1.7, 81.6), positive amniotic fluid culture (OR 19.2; 95% CI1.14, 332), and histological chorioamnionitis (OR 8.3; 95%CI 1.3, 50.9). Because intra-amniotic sludge was accompa-nied by microbial invasion of the amniotic cavity, theysuggested that the particulate matter that constitutedintra-amniotic sludge could potentially be bacterialbiofilms, a matrix of polymeric compounds that micro-organisms may produce to protect themselves from hostdefence mechanisms.6,7 Since intra-amniotic microbialinvasion at the time of second trimester amniocentesis hasbeen reported to be more frequent in women who will sub-sequently deliver prematurely,8-11 we believe that thehypothesis of Espinoza and colleagues remains valid in lightof our findings. Amniocenteses, to rule out microbial

OBSTETRICS

200 �MARCH JOGC MARS 2006

Intra-amniotic sludge during second trimester transvaginal ultrasound examination

A) two women with no intra-amniotic sludge, B) two women with light intra-amniotic sludge,

C) two women with dense intra-amniotic sludge (the sludge was delimited on the second

image).

FH: fetal head; LS: light sludge; DS: dense sludge.

invasion of the amniotic cavity, were not performed in ourstudy.

The observed rate of intra-amniotic light or dense sludgebetween 18 and 32 weeks’ gestation in our high-risk popula-tion (15.7%) is also in agreement with the work of Zimmerand Bronshtein, who reported their intra-amniotic findingsfrom 6500 transvaginal ultrasound examinations performedbetween 14 and 16 weeks’ gestation in both low- andhigh-risk women.12 Two hundred six (3.3%) of thesewomen demonstrated solid or semi-solid intra-amnioticmaterial in fetuses without malformation. However, obstet-rical outcomes were not reported in this study population.We did not find any other investigation that examinedobstetrical outcomes related to intra-amniotic sludge.

The limitations of our study include its retrospective nature,the small number of patients with intra-amniotic sludge andthe unavailability of the specific reason for the cervicalultrasound examination. Another potential source of bias isthat the sonographic examinations were obtained for clini-cal use, and managing physicians were not blinded. How-ever, we believe that potential biases were avoided becauseof the following: (1) the sonographers and the research

assistant who performed the chart reviews were all blindedto the findings of the others; (2) definitions ofintra-amniotic sludge, primary outcomes, and potentialconfounders were planned prior to data collection; (3) thetwo independent investigators were in agreement with theclassification on 98% of the cases; and (4) the associationbetween intra-amniotic sludge and preterm birth had notbeen published prior to performing the current study. Thespecific reason for cervical length measurement wasunavailable for approximately one-quarter of the patients,and it is therefore difficult to apply our results directly to aspecific population in clinical practice. However, becausecervical ultrasound examination is not routinely performedin our centre and is usually reserved for patients at higherrisk of preterm delivery, we believe that our study popula-tion represents a reasonably homogeneous subgroup ofwomen at high risk for preterm delivery. Finally, the newdefinition of “light sludge” has both limitations and bene-fits: this intermediate state between the absence of sludgeand dense sludge is not easy to describe, but there wasagreement by both sonographers about its presence, and itwas also associated with an intermediate risk of adverseobstetrical outcomes.

Intra-Amniotic Sludge, Short Cervix, and Risk of Preterm Delivery

MARCH JOGC MARS 2006 � 201

Table 2. Obstetrical outcomes according to the presence of amniotic fluid sludge

No sludgen = 75

Light sludgen = 9

Dense sludgen = 5

P

Gestational age at birth (weeks � SD) 38.2 � 2.8 35.4 � 4.4 27.9 � 7.3 � 0.01

Gestational age at birth

� 34 weeks

� 37 weeks

5 (6.7%)

13 (17.3%)

4 (44.4%)

4 (44.4%)

4 (80.0%)

4 (80.0%)

� 0.01

� 0.01

Interval between examination and birth (days � SD) 87 � 34 76 � 55 36 � 65 0.02

Interval between examination and birth = 14 days 4 (5.3%) 2 (22.2%) 3 (60.0%) � 0.01

Birth weight (grams � SD) 3178 � 694 2,426 � 850 1304 � 1,282 � 0.01

Histological chorioamnionitis 4 (5.3%) 1 (11.1%) 3 (60.0%) � 0.01

SD: standard deviation

Table 1. Clinical characteristics of the study population according to the presence of intra-amniotic sludge

No sludgen = 75

Light sludgen = 9

Dense sludgen = 5

P

Maternal age (years � SD) 32.1 � 4.3 31.9 � 5.3 31.2 � 3.0 0.91

Gravida, median (min, max) 2 (1, 6) 4 (1, 11) 3 (2, 4) 0.12

Parity, median (min, max) 1 (0, 3) 1 (0, 5) 1 (0, 3) 0.34

Previous preterm birth 20 (26.7%) 3 (33.3%) 1 (20%) 0.86

Gestational age at examination (weeks � SD) 25.8 � 3.6 24.2 � 4.5 22.7 � 3.5 0.12

Cervical length at examination (mm � SD) 34 � 10 23 � 11 16 � 14 0.01

SD: standard deviation.

CONCLUSION

Our observations, combined with those of Espinoza andcolleagues,6 indicate that the sonographic finding ofintra-amniotic sludge is an important and independent riskfactor for delivery before 34 weeks’ gestation and forimpending delivery in women with or without the clinicaldiagnosis of preterm labour. The likelihood of microbiolog-ical invasion of the amniotic cavity or subsequenthistological chorioamnionitis warrants further, larger stud-ies to confirm our results and to evaluate the benefits ofamniocentesis and specific therapeutics, such as antibiotics,in the subgroup of patients with intra-amniotic sludge.

REFERENCES

1. Ventura SJ, Martin JA, Curtin SC, Menacker F, Hamilton BE. Births: final data for1999. Natl Vital Stat Rep 2001;49:1–100.

2. McCormack MC. The contribution of low birth weight to infant mortality andchildhood morbidity. N Engl J Med 1985;312:82–90.

3. Rozenberg P, Gillet A, Ville Y. Transvaginal sonographic examination of the cervix inasymptomatic pregnant women: review of the literature. Ultrasound Obstet Gynecol2002;19:302–11.

4. Iams JD, Goldenberg RL, Meis PJ, Mercer BM, Moawad A, Das A, et al. The length ofthe cervix and the risk of spontaneous premature delivery. N Engl J Med1996;334:567–72.

5. Gomez R, Romero R, Nien JK, Chaiworapongsa T, Medina L, Kim YM, et al. A shortcervix in women with preterm labor and intact membranes: a risk factor for microbialinvasion of the amniotic cavity. Am J Obstet Gynecol 2005;192(3):678–89.

6. Espinoza J, Goncalves LF, Romero R, Nien JK, Stites S, Kim YM, et al. Theprevalence and clinical significance of amniotic fluid ‘sludge’ in patients with pretermlabor and intact membranes. Ultrasound Obstet Gynecol 2005;25:346–52.

7. Leid JG, Shirtliff ME, Costerton JW, Stoodley AP. Human leukocytes adhere to,penetrate, and respond to Staphylococcus aureus biofilms. Infect Immun2002;70:6339–45.

8. Gray DJ, Robinson HB, Malone J, Thomson RB Jr. Adverse outcome in pregnancyfollowing amniotic fluid isolation of Ureaplasma urealyticum. Prenat Diagn1992;12:111–7.

9. Cassell GH, Davis RO, Waites KB, Brown MB, Marriott PA, Stagno S, et al. Isolationof Mycoplasma hominis and Ureaplasma urealyticum from amniotic fluid at 16–20weeks of gestation: potential effect on outcome of pregnancy. Sex Transm Dis1983;10:294–302.

10. Horowitz S, Mazor M, Romero R, Horowitz J, Glezerman M. Infection of the amniotic

cavity with Ureaplasma urealyticum in the midtrimester of pregnancy. J Reprod Med

1995;40:375–9.

11. Gerber S, Vial Y, Hohlfeld P, Witkin SS. Detection of Ureaplasma urealyticum in

second-trimester amniotic fluid by polymerase chain reaction correlates with

subsequent preterm labor and delivery. J Infect Dis 2003;187:518–21.

12. Zimmer EZ, Bronshtein M. Ultrasonic features of intra-amniotic “unidentified debris”

at 14–16 weeks’ gestation. Ultrasound Obstet Gynecol 1996;7:178–81.

OBSTETRICS

202 �MARCH JOGC MARS 2006

Table 4. Frequency of delivery before 34 weeks’ gestation and within 14 days of examinationaccording to the presence or absence of short cervix and intra-amniotic sludge

Cervical length � 25 mm Presence of intra-amnioticsludge*

Delivery before 34 weeks’gestation

Delivery within 14 days ofexamination

No No 2/61 (3.3%) 2/61 (3.3%)

Yes No 3/14 (21.4%) 2/14 (14.3%)

No Yes 2/6 (33.3%) 1/6 (16.7%)

Yes Yes 6/8 (75.0%) 4/8 (50.0%)

Prevalence of the outcome 13/89 (14.6%) 9/89 (10.1%)

* Light or dense sludge present at the time of cervical length measurement.

Table 3. Multivariable stepwise logistic regression for predictive factors of delivery before 34

weeks’ gestation and for delivery � 14 days after examination

Factors Odds ratio (95% CI) P

Delivery before 34 weeks’ gestation

Cervical length � 25 mm 7.90 (1.69, 37.05) � 0.01

Amniotic fluid sludge*

Light

Dense

6.82 (1.17, 39.79)

45.99 (3.48, 608.50)

0.03

� 0.01

Delivery � 14 days after examination

Cervical length � 25mm 5.24 (1.02, 27.91) � 0.05

Amniotic fluid sludge*

Light

Dense

2.86 (0.39, 21.10)

17.77 (1.97, 160.26)

0.30

0.01

* Compared with “no sludge”