Prediction of Spontaneous Preterm Birth

Abstract

Preterm birth is a leading cause of neonatal morbidity and mortality. It is a major goal in obstetrics to lower the incidence of spontaneous preterm birth (SPB) and related neonatal morbidity and mortality. One of the principal objectives is to discover early markers that would allow us to identify subgroups of patients at high risk of SPB and, secondly, to manage those patients with an appropriate and effective strategy.

The purpose of this thesis was to provide data on prediction of spontaneous preterm birth. The studies described in this thesis aim to evaluate (1) which predictive tests are appropriate to use to identify women at risk for SPB; (2) what factors influence the predictive value of a diagnostic test in general and of cervical length assessment in particular; and (3) what strategies are used regarding threatened SPB in the specific Dutch obstetric care.

In chapter 2 we described the mechanisms contributing to SPB, which - mainly - have been identified during the past decade. SPB may result from one of the four primary pathogenic mechanisms, or from a combination: (1) activation of the maternal or fetal HPA axis; (2) inflammation (of ascending genital tract, systemic or chorion/decidual); (3) decidual hemorrhage and (4) pathologic distention of the uterus. The review in the second part of chapter 2 describes the usefulness of new biochemical fluid markers to identify women at risk for SPB. There are no current data to support the use of salivary estriol, home uterine monitoring and for other biochemical markers (e.g., corticotropin-releasing hormone (CRH), maternal serum ?-fetoprotein (MSAFP), interleukine-6 (Il-6) and granulocyte colony-stimulating factor (G-CSF)) larger studies are needed. In the general population no other risk factors than the obstetric history has proven to be effective. In the population at risk transvaginal ultrasonography to determine cervical length and fetal fibronectin testing, or a combination of both is likely to be useful in determining women at high risk for preterm labor. However, their clinical usefulness may rest primarily with their high negative predictive value, thereby avoiding unnecessary intervention. Only treatment with oral antibiotics of women with BV who are at risk for SPB has resulted in a reduction of SPB. Prophylactic treatment in women at risk for SPB, but without BV is not indicated.

In chapter 3 we described a study regarding the usefulness of possible novel markers matrix metalloproteinase (MMP)-1 and MMP-9 in the prediction of SPB. MMP-1 levels in cervicovaginal secretions were low and did not change during either preterm or term labor. MMP-9 levels increased during term and preterm labor, with highest values in laboring patients with ruptured membranes, MMP-9 is not a useful predictor of preterm birth, but does play an important role in membrane rupture.

In chapter 4 and 5 we described the association of socio-demographic factors on the change of cervical length All women had a slight decrease of cervical length across gestation. Women who were African-American, under stress, or working as skilled manual laborers demonstrated significant shortening of the cervix during gestation. African-Americans delivered on average earlier (38.4 weeks) than the other race groups. All pregnancies displayed progressive cervical shortening except Asians. African-American women had shorter cervices across all 3 gestational age intervals even after controlling for preterm delivery. It suggests that shortening of the cervix is particularly a risk indicator for preterm birth in African-American women, since they already have shorter cervices to begin with. Measuring cervical length before 25 weeks gestation might not be of substantial contribution to the prediction of preterm birth in other ethnical groups.

In chapter 6 we evaluated the role of transvaginal ultrasonographic cervical length measurements for women with cervical incompetence. We determined that there was an increase in cervical length after cerclage. However, the degree of lengthening after cerclage was not predictive of term delivery.We alsoshowed that serial cervical length measurement in the late second or early third trimester predicted preterm birth.

Since we were interested how predictors of SPB are implemented in the Dutch obstetric care. We developed a questionnaire to study opinions and management regarding threatened preterm birth, mutual co-operation and the ‘VIL’ between first line obstetrical practitioners (midwives and general practitioners practicing obstetrics) and gynecologists (Chapter 7). There appeared to be consensus among the three professional groups about the way that risk selection takes place. In chapter 8 we assessed the use of ‘modern’ diagnostic methods such as transvaginal sonographic measurements of cervical length, examination of bacterial vaginosis (BV) and fetal fibronectin (FFN) by Dutch gynecologists. Cervical length measurements are used by most of the Dutch gynecologists, even as testing and treating for BV. Fetal fibronectin testing - according to literature the best marker available now - is seldomly been used. Gynecologists in general hospitals without residency programs prescribe tocolytics and repeat corticosteroid treatment more often compared to the gynecologists in hospitals with residency programs. They prescribe less antibiotics for women with threatened preterm birth. Apparently 40% of Dutch gynecologists ‘treat’ preterm premature rupture of membranes (PPROM) with tocolytics while there are no contractions.On some points the disciplines involved in obstetric care should adopt their guidelines to use the evidence from the literature. Use of FFN in the Dutch setting should be evaluated in the future.