doi:10.1111/jog.12666

J. Obstet. Gynaecol. Res. Vol. 41, No. 7: 1009–1017, July 2015

Blind vaginal fetal fibronectin swab for prediction of preterm birth Amr H. Farag1, Magdeldin M. Mohammed1, Mohamed I. Ellaithy1 and Hossameldin A. Salama2 1 Obstetrics and Gynecology Department, Faculty of Medicine, Ain Shams University, Cairo and 2Obstetrics and Gynecology Department, Elmahalla General Hospital, Mahalla, Egypt

Abstract Aim: To assess the accuracy of blind vaginal swab in diagnosis of preterm labor (PTL) and prediction of subsequent occurrence of preterm birth (PTB). Methods: Eligible women who presented at 24–34 weeks of gestation with threatened PTL had their cervicovaginal secretions sequentially sampled for fetal fibronectin (fFN) using two types of swabbing techniques. The first swab was a blind vaginal swab collected without the aid of speculum, while the second one was a routine cervicovaginal swab. All participants were followed up until delivery. Results: fFN in both swabs was significantly higher in women who delivered before term compared with women who delivered at term. On regression analysis, r2 = 0.735 and 0.785 for blind vaginal and cervicovaginal swabs respectively, while on receiver operating characteristic analysis the area under curve was 0.965 and 0.977, respectively, without a statistically significant difference. Using an fFN cut-off of ≥0.05 μg/dL, the sensitivity, specificity, positive predictive value and negative predictive value of blind vaginal swab to predict PTB were 53.9%, 91.9%, 70.0% and 85.0%, respectively; while those for cervicovaginal swab were 58.3%, 94.7%, 77.8% and 87.8%, respectively. Conclusions: In women presenting at 24–34 weeks of gestation with threatened PTL, blind vaginal fFN swab is as effective as cervicovaginal swab to predict PTB, but it is easier to perform and does not require doctor supervision. Key words: blind vaginal swab, cervicovaginal swab, fetal fibronectin, preterm birth, preterm labor.

Introduction Spontaneous preterm birth (PTB), before 37 weeks gestation, occurs in 11.1% of all live births worldwide.1 It represents the second most common direct etiological factor of child death before the age of 5 years,1 and many of the surviving preterm infants suffer from serious morbidities such as bronchopulmonary dysplasia, intraventricular hemorrhage, retrolental fibroplasia, neurodevelopmental problems and cognitive difficulties.2

Fetal fibronectin (fFN), a complex adhesive glycoprotein, is a member of the extracellular matrix family at the maternal–fetal interface. During pregnancy, it is expressed in the extracellular matrix located in the chorio-decidual junction between the maternal decidua and fetal membranes, as well as in the uterus and placenta.3–5 It is detectable during the first 22 weeks of pregnancy, where it may reflect normal growth of trophoblasts and placenta. It is typically absent at 22–34 weeks gestation, and then normally detectable after 34 weeks. It is thought to be trophoblast glue that

Received: June 10 2014. Accepted: November 16 2014. Reprint request to: Dr Mohamed Ibrahem Ellaithy, Building 14, Block 14, Alwaha District, Nasr City, Cairo 11528, Egypt. Email: [email protected]

© 2015 The Authors Journal of Obstetrics and Gynaecology Research © 2015 Japan Society of Obstetrics and Gynecology

1009

A. H. Farag et al.

promotes cellular adhesion at the uterine–placental and decidual fetal membrane interface. It is released into cervicovaginal secretions when the extracellular matrix of the chorionic/decidual interface is disrupted.3–6 Fetal fibronectin has been well validated as a valuable biomarker that is used widely in the risk stratification of symptomatic women presenting with threatened preterm labor (PTL) with intact membranes.7 It has a strong negative predictive value (NPV),8 so that women at low risk for imminent PTB can be identified and discharged in order to minimize unnecessary inpatient observation and resource use,7 and to preserve antenatal steroids and tocolytics for use in the high-risk women to optimize neonatal outcome.9,10 One of the drawbacks of this test is that it requires, as originally described, a speculum examination and sampling of cervicovaginal secretions under direct visualization; this makes the test difficult for some women and necessitates the presence of trained professional medical staff.11 The blind vaginal sampling technique has been described previously, and some considered it as efficient as the standard speculum-directed technique while others did not.11–14 The main objective of the current study was to compare the accuracy of blind vaginal and cervicovaginal fetal fFN swabs to predict PTB among symptomatic women presenting with threatened PTL.

Methods This prospective study was carried out at Ain Shams University Maternity Hospital, in Cairo, Egypt, after being approved by the local institutional ethics and research committee. This is one of the largest maternity hospitals in Egypt and has almost 18 000 deliveries yearly. Young women presenting with threatened PTL at 24–34 weeks of gestation were consecutively recruited for comparison of the accuracy of blind vaginal and cervicovaginal fFN swabs to predict PTB. Written informed consent was obtained from each participant after full counseling. Gestational age was calculated based on the date of first day of the last menstrual period (LMP) and confirmed on early midtrimester ultrasound. The diagnosis of threatened PTL was based on the presence of regular uterine contractions at ≤8 min apart as confirmed on abdominal examination and cardiotocography, whether or not associated with cervical changes (dilatation < 4 cm and/or effacement ≤ 50%).

1010

Women with gestational age > 34 weeks, multi-fetal pregnancy, established PTL, preterm pre-labor rupture of fetal membranes (PPROM), antepartum hemorrhage (placental abruption, placenta previa or vasa previa), non-reassuring fetal status, medical and obstetric comorbidities that can lead to PTB and/or vaginal douching or sexual intercourse in the preceding 24 h were excluded from the study. Women included in this study underwent history taking, standard examination, and trans-abdominal ultrasound to detect number, viability, presentation and gestational age of the fetus and location of the placenta; and also laboratory investigations (complete blood count, renal function testing, random blood glucose and urine analysis). Women were examined in the dorsal lithotomy position with good illumination. Two sequential swabs were used to collect the samples. The first Dacron swab (blind vaginal swab) was inserted into the vagina and directed toward the posterior fornix where it was left in place for at least 30 s after digital separation of labia without the aid of a speculum, followed by insertion of sterile non-lubricated speculum to detect the cervical changes and for collection of the second swab. The second Dacron swab (cervicovaginal swab) was placed in the posterior vaginal fornix and slightly rotated for 10 s to absorb cervicovaginal secretions. The specimens were put in sterile plastic containers (Eppendorfs), which were stored at −20°C until testing. The swabs were collected before any digital examination and without the use of lubricants. Qualitative as well as quantitative assessment of fFN was done using enzyme-linked immunosorbent assay (ELISA) containing the Adeza fFN Kit FDC-6 monoclonal antibody used to detect fFN. The samples were incubated in micro-titer wells coated with FDC-6 monoclonal antibody specific for fFN. fFN concentration ≥ 0.05 μg/mL was considered positive.11 Digital examination was done after collection of the swabs to assess the cervical changes and fetal membranes, to exclude women with established PTL and/or PPROM. In order to validate the laboratory results and readings, the clinicians were trained to perform the fFN testing, and the laboratory personnel performing the assay were blinded to patient identity, clinical history, and the final pregnancy outcome. Reproducibility within the assay was evaluated and the overall intraassay coefficient of variation was 5.7%. Assay-to-assay reproducibility within one laboratory was evaluated and the overall inter-assay coefficient of variation was 6.9%. The limit of detection of fFN was 0.025 ng/mL.

© 2015 The Authors Journal of Obstetrics and Gynaecology Research © 2015 Japan Society of Obstetrics and Gynecology

Blind vaginal fFN swab

Figure 1 Participant flowchart. fFN, fetal fibronectin; PTL, preterm labor.

The participants as well as the staff dealing with them were kept double blinded to the blind vaginal and cervicovaginal swabs results until after delivery. Dexamethasone (12 mg, IMI, two consecutive doses, 24 h apart) and tocolysis (if indicated) were given to all participants according to the Ain Shams University Maternity Hospital threatened PTL management protocol. All participants were then followed up until delivery and only women with spontaneous PTB were included in the final statistical analysis. The primary outcome measure was the accuracy of blind vaginal fFN swab to predict PTB. After delivery, the collected

data on admission were confirmed, reviewed and statistically analyzed to assess the accuracy for each test to predict PTB. The required sample size was calculated using G Power version 3.17 for sample size calculation (Heinrich Heine Universität; Dusseldorf; Germany), setting the α error probability at 0.05, power (1–β error probability) at 0.99%, and effective sample size (w) at 0.50. The effective size (w) was calculated as follows: 2 w = χ 2 N , where χ is the chi-squared test and N is the total sample size. The number of participants needed to produce a statistically acceptable figure was 49.

© 2015 The Authors Journal of Obstetrics and Gynaecology Research © 2015 Japan Society of Obstetrics and Gynecology

1011

A. H. Farag et al.

Table 1 Subject characteristics vs presence of PTB Variables

Preterm birth group (n = 7)

P-value

25.0 ± 4.4 27.8 ± 3.4 2 (0–3) 7 (16.3) 30 (69.8) 13 (30.2)

23.7 ± 3.4 29.3 ± 3.1 1 (0–3) 5 (71.4) 5 (71.4) 2 (28.6)

0.80 0.66 0.60 0.006* 1.00

2 (4.7) 0.043 ± 0.007

7 (100) 0.077 ± 0.006

Blind vaginal fetal fibronectin swab for prediction of preterm birth.

To assess the accuracy of blind vaginal swab in diagnosis of preterm labor (PTL) and prediction of subsequent occurrence of preterm birth (PTB)...
474KB Sizes 4 Downloads 11 Views