- Email: [email protected]
Management of a woman with a previous spontaneous preterm birth
CASE-BASED LEARNING
Management of a woman with a previous spontaneous preterm birth
Preterm birth may be considered as indicated (iatrogenic) or spontaneous. Indicated birth accounts for at least one-third of preterm deliveries and occurs when complications arise that put mother or fetus at risk unless delivery is expedited, for example severe fetal growth restriction, massive antepartum haemorrhage, or severe pre-eclampsia. The remainder of preterm births are considered spontaneous (including those preceded by preterm pre-labour rupture of membranes). Spontaneous preterm labour can be initiated by a range of pathology, including infection, uterine stretch, placental ischaemia, haemorrhage, and other inflammatory causes. This is likely to explain why, despite considerable healthcare advances, successful prediction and prevention of spontaneous preterm birth has remained so elusive: as preterm birth represents a final common pathway of a heterogeneous group of conditions, each predictor or intervention is likely to only identify or improve outcomes for a subset of cases. There are many risk factors for preterm birth; however a prior history of spontaneous preterm birth is the most significant and consistently identified clinical risk factor, and the identification of women with a history of preterm birth gives an opportunity to optimise care for subsequent pregnancies. This article will review and discuss risk assessment and management of singleton pregnancies where there is a history of a prior preterm birth.
Charlotte Oyston Katie Groom
Abstract Preterm birth is an important cause of neonatal morbidity and mortality and has long term adverse health consequences. Worldwide, close to 15 million babies are born preterm each year, and there is no sign that the rate of preterm birth is slowing. A history of a previous spontaneous preterm birth is a significant risk factor for a subsequent spontaneous preterm birth; identifying these women provides an opportunity to optimise care in future pregnancies. Antenatal identification of women at the highest risk of preterm birth is challenging, as tests that accurately identify asymptomatic women who go on to deliver preterm are lacking. Furthermore, the short- and long-term benefits of interventions such as cerclage and progesterone remain unclear. Research is underway to develop biomarkers that can accurately predict women who will deliver preterm. However, without effective strategies that diminish rates of preterm birth and improve perinatal outcomes, the clinical role of these tests is less well defined.
Case 1 Pre-conception risk prediction for women with a previous preterm birth Mrs PT is a 32-year-old G3P1 who attends your clinic prepregnancy. Her notes document an elective early surgical termination in her first pregnancy, and a previous delivery at 27 weeks, with her baby spending 9 weeks in the neonatal unit. She is considering becoming pregnant again and is extremely anxious. She wants to know the likelihood of this baby being born preterm. Risk scoring tools based on clinical history alone perform poorly. However, details of the circumstances leading to the previous preterm birth may aid prediction and so a full history and case note review of the previous births is important to assess future risk. There may be areas where there is opportunity to reduce or eliminate risk factors for recurrent preterm birth as well as other pregnancy complications.
Keywords cervical cerclage; cervical length; fetal fibronectin; preterm birth; prevention; progesterone
Introduction Preterm birth is defined as birth occurring prior to 37 completed weeks of gestation. Preterm birth is a leading cause of perinatal morbidity and mortality: worldwide almost 15 million babies are born preterm, and over one million die from complications of prematurity each year. Those that survive the neonatal period have greater rates of respiratory disease, visual impairment, deafness, and neurodevelopmental disability compared to their term peers. Those born preterm are also more likely to develop non-communicable diseases such as obesity, type II diabetes, hypertension and cardiovascular disease in adulthood. Despite extensive ongoing research into causes, predictors and treatments for preterm birth, the rate of preterm birth has not fallen, even increasing in some countries.
Number and timing of previous deliveries The risk of recurrent spontaneous preterm birth is dependent on the aetiology of the previous preterm birth, the number of previous preterm births, and the severity of prematurity. The likelihood of having a preterm birth in a given pregnancy increases with the number of previous preterm births, with a reducing chance of delivery after 33 weeks of 83, 86 and 68% after 1, 2 or 3 prior preterm births. Risk of preterm birth also increases as the gestational age of previous preterm births declines, with the greatest risk for those with a previous delivery at less than 28 weeks’ gestation. If the previous preterm delivery was an indicated delivery, the risk of recurrence will be related to the recurrence risk of the condition leading to the indicated delivery. A history of previous indicated preterm birth is also associated with an increased risk of spontaneous preterm birth in subsequent pregnancy. Presumably this is due to a shared aetiology
Charlotte Oyston MB ChB BMedSci (Hons) Dip OMG PhD Advanced Trainee in Obstetrics and Gynaecology, Middlemore Hospital, Auckland, New Zealand. Conflicts of interest: none declared. Katie Groom MB BS BSc FRANZCOG PhD CMFM Associate Professor of Maternal and Perinatal Health, The Liggins Institute, University of Auckland; Maternal Fetal Medicine Subspecialist, Department of Obstetrics and Gynaecology, National Women’s Health, Auckland City Hospital, Auckland, New Zealand. Conflicts of interest: none declared.
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
353
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
women with uterine anomalies may also have a cervical anomaly which contributes to cervical insufficiency in pregnancy.
(such as placental ischaemia) between pregnancy complications such as pre-eclampsia and preterm birth. The timing of the last pregnancy should also be reviewed, as risk of early delivery increases with reducing interpregnancy interval. An analysis of 6181 women with a previous preterm birth demonstrated that after adjusting for confounding factors, those with an interpregnancy interval of less than 6 months had the highest increase in risk of preterm delivery (44% increase in risk of recurrent preterm birth), those with an interval of 6e12 months had a modest increase, and there was no increase in risk of recurrent preterm birth with an interpregnancy interval of 12e18 months, compared to the risk of those with an interval of 18 months or more.
Use of assisted reproductive technologies Several systematic reviews have shown an increase in risk of preterm delivery in women undergoing IVF treatment compared to those who conceive naturally, with the risk of preterm delivery doubling for delivery prior to 37 weeks and tripling for delivery prior to 33 weeks for singleton pregnancies. It is unclear whether this increase in risk results from the fertility treatment itself, underlying maternal factors associated with infertility, or iatrogenic bias in the care of these pregnancies. Demographic, socioeconomic, psychosocial and lifestyle factors Women who are obese or significantly under weight prepregnancy are at increased risk of preterm delivery, as well as other pregnancy complications. A large meta-analysis of data from over one million individuals found that women from developed countries with a pre-pregnancy body mass index (BMI) of less than 18.5 kg m2 are more likely to have a preterm birth e either spontaneous or indicated d compared to women with a BMI in the normal range (RR 1.22). Women with a BMI of 35 kgm2 also appear to be at increased risk of indicated preterm birth (OR 1.5e1.8), which may be partially explained by an increased prevalence of chronic disease (such as hypertension or diabetes) in this population. Race and ethnicity are important risk factors for preterm birth. After adjusting for risk factors such as age, education and parity, black women have a two- to three-fold higher risk of preterm birth compared to non-black women. The aetiology of this association is unclear but is possibly interlinked with other lifestyle factors and/or social disparities; rates of preterm birth are increased amongst socioeconomically deprived women, women who have high levels of anxiety or perceived stress or low levels of support, are at the extremes of maternal age, and those who smoke, consume alcohol, and use illicit drugs during pregnancy.
History of cervical surgery or trauma/uterine instrumentation Women undergoing excisional treatment for cervical dysplasia have an increased risk of preterm birth, although the mechanism(s) mediating this increase in risk are unclear. One hypothesis is that these procedures reduce the mechanical support of the cervix which may then increase susceptibility to cervical insufficiency, or spontaneous loss of the cervical mucous plug leading to ascending infection. Cervical conisation has the most consistent and strongest association with preterm birth. Women with a history of conisation are 2e3 times more likely to have a preterm birth than healthy controls, and those conceiving within 2e3 months of conisation or large conisation (>1 cm depth) may have the greatest risk. Meta-analyses also suggest an increased risk of preterm birth with other excisional therapies e such as large loop excision of the transformation zone (LLETZ) (RR 1.70 compared to healthy controls). Although many studies within the meta-analyses use healthy controls as the comparison group, an increased risk of preterm birth with all excisional therapies remains d albeit attenuated d when women undergoing cervical treatment are compared to those who had been diagnosed with precancerous lesions, but not treated. This suggests that the increase in risk may be in part due to the underlying pathology of the precancerous lesion, as well as the cervical procedure itself. Cervical dilation and curettage for miscarriage or termination of pregnancy also increase the risk of spontaneous preterm birth in subsequent pregnancies. While some large cohort studies have found no significant difference in rates, more recent good quality studies have demonstrated a small but significant increase in risk of spontaneous preterm birth after these procedures, with odds ratios of up to 1.8 and evidence that the risk of preterm birth in subsequent pregnancies rises with the number of procedures performed.
Pre-conception strategies to reduce risk of preterm birth What advice should be given and what medical or surgical interventions are available pre-conception that could reduce the risk of preterm delivery? There are few pre-conception interventions that improve outcomes in preterm birth. Many risk factors for preterm birth are non-modifiable, and those that can be modified only minimally increase absolute risk and are poorly predictive of subsequent preterm birth. When assessing the impact of strategies to prevent preterm birth, it is important to assess impact on perinatal morbidity and mortality, as in some instances, preterm birth may represent a mechanism through which the fetus escapes an adverse intrauterine environment; prolonging these pregnancies may be associated with neonatal harm rather than benefit.
Presence of congenital uterine anomalies Congenital uterine anomalies are found more commonly in women with a history of second trimester loss or preterm birth, with a prevalence of up to 25% (compared to 6% in women from an unselected population). The risk of preterm birth varies with the type of uterine anomaly; women with a uterine didelphys or a septate uterus have the highest risk with up to 33% of pregnancies ending in preterm delivery. It has long been thought that the increased rate of preterm birth results from a reduced volume and distensibility of the uterine cavity; however it is most likely that
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
Pre-conception advice and medical interventions Ideally, counselling and assistance regarding optimal pregnancy spacing, the benefits of maintaining a healthy weight, and abstinence from smoking, alcohol and drugs should be provided
354
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
in the preconception period. Smoking cessation (and limiting passive exposure) is of particular importance, as this is the largest modifiable risk factor for preterm birth. At an individual level, providing behavioural counselling with nicotine replacement therapy (safer than continuing to smoke, and associated with increased rates of cessation) is the most effective strategy for smoking cessation. Incentive schemes (where incentives such as cash or vouchers are used to encourage cessation), also improve cessation rates. At a population level, clinicians should support and promote smoke-free environments: rates of preterm birth reduce by an estimated 10% after introduction of smokefree legislation. Medical management of chronic medical conditions such as diabetes and hypertension should be optimised; although this may not be associated with reductions in rate of spontaneous preterm birth, optimal control of chronic conditions may lower the risk of additional pregnancy complications, and therefore the rates of indicated preterm birth, as well as potentially leading to better long-term outcomes for mother and baby.
transabdominal cerclage, compared to those who had a repeat vaginally-placed cerclage (6 vs 12.5%). The MAVRIC trial is the first randomised trial comparing abdominal with vaginal cerclage. Preliminary data suggests that for women with a previously unsuccessful vaginally-placed cervical cerclage, a transabdominally-placed cerclage is superior to repeat vaginallyplaced cerclage in reducing risk of delivery prior to 32 weeks and fetal death.
Case 2 Antenatal tests for prediction of preterm birth Ms S is a 24-year-old G3P2 with a history of two previous preterm deliveries at 32 weeks. You see her in clinic at 12 weeks and she asks if there are any tests she can have that will detect whether she will go into labour early. What tests can be used in pregnancy to help stratify the risk of preterm delivery in asymptomatic women? Accurately identifying women who are at high risk of delivering preterm provides an opportunity to optimise the timing and administration of interventions such as antenatal corticosteroids, MgSO4 and arrange for transfer to a centre with tertiary neonatal facilities. A negative test also provides reassurance for those who are unlikely to deliver preterm. Tests that are currently available in clinical practice for the prediction of spontaneous preterm birth include sonographic cervical length measurement, and the vaginal biomarkers, fetal fibronectin (fFN), placental alpha macroglobulin-1 (PAMG-1) and cervical phosphorylated insulinlike growth factor binding protein-1 (phIGFBP-1). Of the biomarkers tests, only fetal fibronectin has published data to support use in asymptomatic women.
Pre-conception surgical interventions As uterine anomalies are associated with increased risk of preterm birth, it seems logical that normalising uterine anatomy would be associated with a lowering of preterm delivery. For women with a septate uterus, there is evidence from observational studies that pregnancy outcomes (including preterm birth rates) may be improved with hysteroscopic resection of the septum. However, given the paucity of adequately powered randomised studies comparing outcomes of hysteroscopic resection with no treatment, clear recommendations as to which women are appropriate candidates for this procedure cannot be made. Similarly, other surgeries aimed at normalising anatomy in women with different congenital uterine malformations have been described, but evidence is limited to case reports and small observational studies. Where surgery is considered for women with a uterine anomaly, the benefits and risks should be carefully considered on a case by case basis. Cervical cerclage is a procedure where a non-absorbable suture is placed around the cervix to reduce the risk or preterm delivery. The mechanism of action is thought to be through increasing mechanical support, or through reducing the risk of ascending infection through cervical lengthening or retention of the mucous plug. Cerclage is typically performed vaginally in the second trimester for at-risk pregnancies (the indications for cerclage placement during pregnancy are discussed in more detail in the section ‘Antenatal strategies to reduce the risk of preterm birth’ below). Abdominal placement of a cervical suture via laparotomy or laparoscopy is sometimes carried out preconception in women with a very short or scarred cervix (where transvaginal placement may be difficult), or women with a history of a failed vaginally-placed cervical suture. Abdominal cerclage is associated with a higher incidence of serious maternal operative complications than vaginal cerclage (3.4 vs 0%) and their use should be reserved for those at highest risk with a previously failed vaginal cerclage. A systematic review of 13 case series and one controlled non-randomised study of transabdominal cervical cerclage in women with a history of failed transvaginally-placed cervical cerclage found a lower risk of perinatal death or delivery before 24 weeks in women who had a
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
Fetal fibronectin Fetal fibronectin is a glycoprotein found in the amniotic fluid, placenta, and between the chorion and decidua. It is described as a “glue” between chorion and decidua and is not normally detectable in vaginal secretions after fusion of decidua and fetal membranes. The presence of fFN in vaginal secretions after 18 weeks occurs following disruption of the choriodecidual interface due to inflammation or mechanical forces and is associated with an increased risk of preterm delivery. A swab taken from the posterior fornix can be used to detect the presence of fFN in a bedside test, with results given qualitatively (a positive corresponding to a concentration of >50 ng/mL, or negative <50 ng/ mL), or quantitatively. Compared to women with threatened preterm labour, fFN has a lower sensitivity for the prediction of preterm birth in asymptomatic women (78e89% compared to 68 e76%), although the specificity appears similar for both asymptomatic (88e89%), and symptomatic women (86%). Currently, the greatest utility for fetal fibronectin is in its negative predictive value for women symptomatic of preterm labour, and the reassurance that imminent delivery is highly unlikely. It is not clinically useful as a stand-alone test for prediction of preterm birth in asymptomatic women with or without other risk factors for preterm birth. Quantitative fFN shows an improved prediction of preterm birth compared with qualitative testing. The risk of birth before 35 weeks increases with increasing levels of fFN from >20 ng/ml to 300 ng/ml. Using a threshold of
355
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
1st centile) had a significantly increased risk of delivery prior to 35 weeks (relative risk 2.4, 3.8, 9.5 and 14 respectively). It should be noted however, that less than half the women with the shortest cervices (less than 15 mm), delivered prior to 35 weeks. These findings have since been replicated in other large prospective studies.
10 ng/mL results in a higher sensitivity for prediction of preterm labour and when combined with cervical length measurement, is a more cost-effective approach to identifying the symptomatic women at greatest risk of preterm birth (see below). Phosphorylated insulin-like growth factor binding protein-1 Phosphorylated insulin-like growth factor binding protein-1 is a protein secreted by decidual cells. Detachment of fetal membranes from the decidua causes leakage of phIGFBP-1 into vaginal secretions. A qualitative bedside test (Actim PartisTM) is commercially available. As for fFN, this test is measured from sections obtained on a vaginal swab. Like fFN, phIGFBP-1 has better negative predictive value than positive predictive value for preterm birth in symptomatic women. Its predictive accuracy in asymptomatic women is limited, with a recent meta-analysis suggesting a pooled positive likelihood ratio of 14e47%, and negative likelihood ratio of 76e93% for preterm delivery.
Combined fFN and cervical length measurements The combination of transvaginal cervical length measurement and fFN testing is better at predicting spontaneous preterm delivery compared to fFN alone. A subsequent analysis of NICHD data looked at the effect of cervical length and fFN at 24 weeks on the risk of preterm delivery in women with a history of previous preterm birth. In women with a previous preterm birth, both cervical length and fetal fibronectin test results at 24 weeks modified the risk of preterm delivery. The highest risk of preterm delivery was in women with a positive fFN and a cervical length <25 mm (risk of delivery <35 weeks 64%). The risk was lowest for women with a negative fetal fibronectin and a cervical length >35 mm or more (7% risk of delivery prior to 35 weeks). Quantitative fFN can also be used in conjunction with cervical length measurement. In a study of asymptomatic women with risk factors for preterm delivery, women with a mid-trimester cervix length <25 mm had a 9.5% risk of delivery <34 weeks when qfFN <10 ng/mL. The risk of delivery <34 weeks increased to 55% for those with levels >200 ng/mL. A risk prediction tool for both symptomatic and high-risk asymptomatic women has been developed, combining quantitative fFN, cervical length, and past obstetric history. This tool has the advantage that it provides an individualised estimate of preterm delivery, rather than just a summary estimate of risk. The algorithm is available to clinicians as a smart phone app; further information is available at www.quipp.org.
Placental alpha microglobulin-1 Placental alpha microglobulin-1 (PAMG-1) is a large glycoprotein synthesised by decidual cells. PAMG-1 is found in amniotic fluid. Its detection in the vaginal secretions is normally low but has been used to detect ruptured membranes in a bedside test marketed as AmnisureTM. More recently, studies have assessed the role of PAMG-1 in detecting preterm labour independently from rupture of membranes in women with signs or symptoms of preterm labour. The test appears to perform with high sensitivity and specificity for symptomatic women. However, as these studies were of relatively small sample size, further research is required to confirm the test’s accuracy, and to clarify its role in the assessment of women who are asymptomatic. Cervical length measurement Large prospective studies have consistently observed an increased risk of preterm birth for women with a shortened cervix during pregnancy, and that the risk of preterm delivery increases with decreasing cervical length. Transvaginal ultrasound is the gold standard method for visualising and measuring cervical length, with high likelihood of obtaining measurements and low interobserver variability. The alternative transabdominal approach may be affected by maternal bladder filling, and is associated with a lower likelihood of obtaining adequate measurements. Cervical length measurements are performed over a 5-min period with an empty maternal bladder and in the absence of undue probe pressure. The cervical length is taken as the shortest measurement that displays the landmarks of internal, external os, cervical canal. Assessment of cervical length in women at high risk of preterm birth is usually performed between 14 and 24 weeks; risk of spontaneous preterm birth increases as cervical length decreases and gestational age decreases. The National Institute of Child Health and Development (NICHD) Preterm Prediction study was the first large prospective study to evaluate cervical length as a predictor for preterm birth in asymptomatic women. Cervical length was measured in 3000 women at 22e24þ6 weeks. Compared to women with a cervical length of more than 40 mm (the 75th centile) women with a cervical length less than 40, 35, 26 or 13 mm (75th, 50th, 10th or
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
Antenatal strategies to reduce the risk of preterm birth What antenatal strategies are proven to be effective at reducing the risk of preterm birth in women with a prior history of preterm birth? Cervical cerclage A cervical cerclage is typically performed in the second trimester for one of three indications: a history of recurrent preterm losses (history indicated), cervical shortening detectable on ultrasound (ultrasound indicated), or in women presenting with signs of cervical dilation or visible fetal membranes (rescue cerclage). The most common techniques for cerclage placement are the Shirodkar and McDonald techniques, both of which result in a purse-string suture around the cervix. The Shirodkar technique describes dissection of the vaginal mucosa anteriorly and posteriorly to expose the cervix as close to the level of the internal os as possible, whereas the McDonald technique does not conventionally involve dissection. Many studies of cervical cerclage do not stipulate which technique has been used and there are no prospective studies evaluating pregnancy outcomes by comparison of the two techniques, so choice of technique is generally governed by surgeon preference. There is also little evidence to guide choice of suture material. Traditionally, a braided filament has been the suture of choice, due to the perceived superior
356
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
strength, and ease of removal compared to monofilament sutures. However, the multifilament braided suture favours bacterial colonisation, and has been associated with increased infectious morbidity and poorer tissue healing in non-obstetric settings. There is concern that use of braided filaments for cerclage may introduce a route through which preterm birth is stimulated. The C-STITCH study is a randomised trial that is currently recruiting, and aims to assess the effect on pregnancy and neonatal outcomes of monofilament versus braided suture material in women undergoing cerclage.
Cerclage or progesterone? Cerclage requires a skilled operator and is associated with maternal risks relating to anaesthesia and surgery (cervical laceration, fever, bleeding, chorioamnionitis and ruptured membranes). Therefore, a question of considerable interest is whether cerclage is better at preventing preterm birth than other treatments. There is a paucity of data comparing progesterone to cervical cerclage (or pessary). A recently published indirect meta-analysis (comparing interventions head-to-head through a common comparison group) has suggested progesterone is more effective at preventing preterm delivery (<34, <37 weeks) than cerclage. The SuPPoRT trial is currently randomising women with a cervix <25 mm on mid trimester ultrasound to cervical cerclage, progesterone or pessary to determine the relative efficacy of these interventions on pregnancy duration and neonatal outcomes. There is currently insufficient evidence to support combined use of multiple (cerclage and progesterone, progesterone and pessary) vs single interventions.
History indicated cerclage In an international multi-centre study, 1292 women were randomised to cerclage or no cerclage when there was clinical uncertainty as to whether or not to perform a cervical cerclage. The cerclage group had significantly fewer deliveries prior to 33 weeks (number needed to treat to prevent one delivery <33 weeks ¼ 25), however subgroup analysis showed that only women with at least three previous pregnancies ending prior to 37 weeks benefited from cerclage. In this group, the number of preterm deliveries prior to 33 weeks was halved; however, no differences in neonatal outcomes were detected. It is possible that this study underestimated the magnitude of benefit of cerclage, as it excluded women who were potentially the most likely to benefit from cerclage e those that clinicians were not uncertain as to whether cerclage would provide benefit.
Progesterone Progesterone is a key component in the maintenance of pregnancy. It has anti-inflammatory properties and plays a role in the prevention of cervical ripening and maintenance of myometrial quiescence. There have now been over 40 published randomised trials studying the efficacy of progesterone in reducing preterm birth. The two main types of progesterone studied are natural progesterone (rapidly absorbed via vaginal mucosa) and 17 alpha hydroxyprogesterone capronate (which has a longer halflife and is administered as weekly intramuscular injections). Of the two, vaginal progesterone is cheaper, and appears to be more effective. Meta-analyses of earlier randomised trials suggested a benefit of progesterone in the prevention of preterm birth in women with a shortened cervix, or history of a previous preterm birth. The use of progesterone for preterm birth prevention has been endorsed by clinical guidelines groups, including the National Institute for Health and Care Excellence. More recently, two large randomised studies (OPPTIMUM and PROGRESS) have shown no benefit of vaginal progesterone in reducing preterm birth, composite neonatal morbidity and mortality, or childhood cognitive and neurosensory outcomes at age 2 years. A subsequent metanalysis which included data from the OPPTIMUM study concluded that progesterone reduced preterm birth, neonatal morbidity and mortality when given antenatally to women with a shortened cervix. However, there remains debate regarding the specific conditions where progesterone may reduce risk of preterm birth, and the optimal formulation, dose, route also remain unclear. The Evaluating Progestogens for Prevention of Preterm birth International Collaborative (EPPPIC) is an independent meta-analysis of individual participant data from the totality of RCTs of progesterone in the prevention of preterm birth. Publication is expected in late 2018 and will hopefully clarify whether there are particular subgroups of pregnant women who may benefit from progesterone therapy.
Ultrasound indicated cerclage in women with a history of preterm birth A meta-analysis using individual patient level data compared cerclage with expectant management in women with a shortened cervix (less than 25 mm) and found that cerclage was associated with a reduction in preterm birth before 35 weeks in those with singleton pregnancies and a history of prior preterm birth or second trimester loss (RR 0.63). A further randomised study of women with a history of preterm birth and a cervical length of <25 mm showed a reduction in pre-viable delivery and perinatal mortality, although there was no reduction in delivery in less than 35 weeks unless cervical length was below 15 mm. Rescue cerclage Ideally, women with a previous preterm birth should be offered elective cervical cerclage or serial surveillance of cervical length (depending on prior history), thereby avoiding a presentation with cervical dilatation. However, in the event that cervical dilatation occurs, a rescue cerclage should be considered. A small randomised study, comparing rescue cerclage with expectant management (one week of broad spectrum antibiotic therapy and bed rest until 30 weeks) found cerclage delayed delivery by an average of 4 weeks, and had a reduction in delivery prior to 34 weeks, perinatal morbidity and a trend towards reduced perinatal mortality. Similar delays have been seen in small nonrandomised studies. Cerclage may be less likely to be effective with significant membrane prolapse (beyond the external os), cervical dilatation of more than 4 cm, and should never be performed when there are signs or symptoms of chorioamnionitis, fetal compromise, death or fetal anomaly not compatible with life, PPROM or ongoing vaginal bleeding.
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
Cervical pessary Cervical pessaries are proposed to prevent preterm labour via the alteration of the inclination of the cervical canal relative to the uterus (resulting in a more acute utero-cervical angle),
357
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
controlled trials have not found a reduction in the rate of preterm birth when women are treated for this condition.
preventing direct pressure on the cervix and fetal membranes at the level of the internal os. Pessaries are a “one-off” treatment, well tolerated with minimal side effects and are of relatively low cost, they are an attractive alternative to cerclage and progesterone treatments, particularly for women from countries where resources are limited. The first randomised controlled trial using Arabin pessaries for the prevention of preterm birth in women with a shortened cervix detected on a mid-trimester ultrasound scan demonstrated reduced rates of preterm delivery and a reduction in composite neonatal adverse outcomes. Subsequent randomised studies have shown conflicting results, with a recent meta-analysis suggesting that pessary use does not reduce rates of preterm birth or improve perinatal outcomes in women with a shortened cervix. Interestingly, there appears to be an effect of clinician-learning curve on the effectiveness of pessary treatment, with significant differences in outcomes between those recruited early and those recruited later to trials of pessary. Further randomised trials are underway, as strong evidence of efficacy is lacking, and adequate training in operation and insertion is needed to ascertain whether cervical pessaries can prolong gestation and improve perinatal outcomes for women at risk of preterm birth.
Bed-rest, relaxation, or stress reduction Although many studies have demonstrated an association between psychological stress and preterm birth, a limited number of studies have assessed the effect of relaxation techniques on preterm birth rates. The studies published use different relaxation techniques, so are difficult to compare directly; however while studies show evidence of reduced stress and anxiety scores, they do not show evidence of reduced risk of preterm birth. Similarly, there is no evidence that reducing work or reducing sexual activity reduces the likelihood of preterm birth. Management of women presenting with preterm labour The management of women who present with threatened preterm labour does not differ between those who have and have not had a previous preterm birth. In brief, management begins with a careful history, examination, and investigations aimed at determining the likelihood of preterm labour, and excluding the presence of serious conditions in mother and baby that require imminent delivery (for example, fetal distress, placental abruption or chorioamnionitis). Once conditions requiring expedited delivery have been excluded, adjunct tests such as ultrasound assessment of cervical length (as well as determining fetal presentation and growth), fFN, phIGFBP-1 and PAMG can be considered. This allows targeted use of therapies most likely to improve neonatal outcomes should preterm delivery ensue: maternal administration of corticosteroids (if gestational age is <35 weeks); maternal administration of magnesium sulphate where delivery is imminent (if gestational age is <30 weeks); transfer to a unit able to provide appropriate neonatal care and the considered treatment of any precipitating causes of labour.
Preterm birth clinic Despite a growing body of evidence supporting the use of cervical cerclage and progesterone for the prevention of preterm birth, the counselling, availability and clinical use of these interventions is inconsistent. In obstetrics (and many other areas of medicine), there is a trend toward developing specialised clinics which focus on prevention and provide a means for consistent application of the most up-to-date evidence-based practice. There is limited evidence that attendance at preterm birth prevention clinics reduce the rates of preterm birth, and such clinics are now an accepted part of service provision in many centres and offer other benefits such as an opportunity for teaching of both clinicians and patients, and an ideal setting for further research into the aetiology and prevention of preterm birth.
Summary and conclusions A history of a preterm birth is an important risk factor for spontaneous preterm birth in a subsequent pregnancy, although the majority of women with a previous preterm birth will deliver at term. Risk factors for spontaneous preterm birth have been identified, however many of these are non-modifiable; and those that can be changed e such as smoking cessation and weight optimisation - may result in a modest reduction in risk of preterm delivery at best. There is evidence that interventions such as progesterone and cervical cerclage may be beneficial in certain women at high risk of preterm birth, however questions still remain as to how best to screen these women, the optimal treatment regimen, and whether these treatments improve perinatal outcomes. In asymptomatic women, the best biomarkers for the prediction of preterm delivery are cervical length measurement, and detection of fFN in the cervicovaginal secretions. However, without effective strategies that diminish both the rate of preterm birth and improve perinatal outcomes, the role of these tests in the management of these women is less well defined. These biomarkers should be used in future research to direct trials of new therapies to those at the highest risk of
Antenatal therapies with insufficient evidence of benefit Screening and treatment of asymptomatic women for lower genital tract infection Although genital tract infection is associated with preterm birth, there is not consistent evidence that the screening and treatment of asymptomatic women improves outcomes. A recent systematic review found screening women in the early second trimester and treating for candidiasis, bacterial vaginosis (BV) and trichomoniasis halved the rate of delivery before 37 weeks. In contrast, studies assessing the effect of screening and treatment of BV or trichomoniasis independently have found that although treatments are effective at eradicating infection, rates of preterm birth were either no different, or increased compared to untreated women. Treatment of periodontal disease Increasing severity of periodontal disease is associated with increasing likelihood of preterm birth. However, randomised
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
358
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
preterm birth. Out of this mass of evidence have arisen a variety of guidelines worldwide for the prediction and prevention of preterm labour. An example is the NICE (National Institute of Clinical Excellence) guideline ‘Preterm Labour and Birth’, published in the UK in 2015, but there are others. A
Practice points C
C
FURTHER READING Abbott DS, Hezelgrave NL, Seed PT, et al. Quantitative fetal fibronectin to predict preterm birth in asymptomatic women at high risk. Obstet Gynecol 2015; 125: 1168e76. https://doi.org/10.1097/AOG. 0000000000000754. Althuisius SM, Geijn HP. Strategies for preventiondcervical cerclage. BJOG An Int J Obstet Gynaecol 2005; 112: 51e6. https://doi.org/ 10.1111/j.1471-0528.2005.00585.x. Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology and causes of preterm birth. Lancet 2008; 371: 75e84. https://doi.org/ 10.1016/S0140-6736(08)60074-4. ‘Preterm labour and birth’ NICE guideline 25. 2015, https://www.nice. org.uk/guidance/ng25. Shennan AH. Prediction and prevention of preterm birth: a quagmire of evidence. Ultrasound Obstet Gynecol 2018; 51: 569e70. https:// doi.org/10.1002/uog.19063.
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
C
C
C
359
Preterm birth may be considered as either spontaneous or indicated. For women with a previous indicated preterm birth, risk of recurrence depends on the specific condition necessitating early delivery. However, these women are also at slightly higher risk of spontaneous preterm birth. For women with a previous spontaneous preterm birth, risk of early delivery increases with the number of previous preterm deliveries and is more likely in those who previously delivered at earlier gestations. However, most women with a previous preterm delivery will not deliver preterm in subsequent pregnancies. In a general population, risk estimates for preterm delivery can be improved by transvaginal ultrasound measurement of cervical length, and fetal fibronectin testing. However, the role of these investigations in women with a previous preterm birth is not well defined, and further research is needed to clarify what optimal cut-offs are whether available interventions provide significant benefit for those who “screen positive”. Progesterone and cervical cerclage are interventions which may reduce the risk of preterm delivery, in certain groups of high risk women.
Ó 2018 Elsevier Ltd. All rights reserved.
Management of a woman with a previous spontaneous preterm birth
Preterm birth may be considered as indicated (iatrogenic) or spontaneous. Indicated birth accounts for at least one-third of preterm deliveries and occurs when complications arise that put mother or fetus at risk unless delivery is expedited, for example severe fetal growth restriction, massive antepartum haemorrhage, or severe pre-eclampsia. The remainder of preterm births are considered spontaneous (including those preceded by preterm pre-labour rupture of membranes). Spontaneous preterm labour can be initiated by a range of pathology, including infection, uterine stretch, placental ischaemia, haemorrhage, and other inflammatory causes. This is likely to explain why, despite considerable healthcare advances, successful prediction and prevention of spontaneous preterm birth has remained so elusive: as preterm birth represents a final common pathway of a heterogeneous group of conditions, each predictor or intervention is likely to only identify or improve outcomes for a subset of cases. There are many risk factors for preterm birth; however a prior history of spontaneous preterm birth is the most significant and consistently identified clinical risk factor, and the identification of women with a history of preterm birth gives an opportunity to optimise care for subsequent pregnancies. This article will review and discuss risk assessment and management of singleton pregnancies where there is a history of a prior preterm birth.
Charlotte Oyston Katie Groom
Abstract Preterm birth is an important cause of neonatal morbidity and mortality and has long term adverse health consequences. Worldwide, close to 15 million babies are born preterm each year, and there is no sign that the rate of preterm birth is slowing. A history of a previous spontaneous preterm birth is a significant risk factor for a subsequent spontaneous preterm birth; identifying these women provides an opportunity to optimise care in future pregnancies. Antenatal identification of women at the highest risk of preterm birth is challenging, as tests that accurately identify asymptomatic women who go on to deliver preterm are lacking. Furthermore, the short- and long-term benefits of interventions such as cerclage and progesterone remain unclear. Research is underway to develop biomarkers that can accurately predict women who will deliver preterm. However, without effective strategies that diminish rates of preterm birth and improve perinatal outcomes, the clinical role of these tests is less well defined.
Case 1 Pre-conception risk prediction for women with a previous preterm birth Mrs PT is a 32-year-old G3P1 who attends your clinic prepregnancy. Her notes document an elective early surgical termination in her first pregnancy, and a previous delivery at 27 weeks, with her baby spending 9 weeks in the neonatal unit. She is considering becoming pregnant again and is extremely anxious. She wants to know the likelihood of this baby being born preterm. Risk scoring tools based on clinical history alone perform poorly. However, details of the circumstances leading to the previous preterm birth may aid prediction and so a full history and case note review of the previous births is important to assess future risk. There may be areas where there is opportunity to reduce or eliminate risk factors for recurrent preterm birth as well as other pregnancy complications.
Keywords cervical cerclage; cervical length; fetal fibronectin; preterm birth; prevention; progesterone
Introduction Preterm birth is defined as birth occurring prior to 37 completed weeks of gestation. Preterm birth is a leading cause of perinatal morbidity and mortality: worldwide almost 15 million babies are born preterm, and over one million die from complications of prematurity each year. Those that survive the neonatal period have greater rates of respiratory disease, visual impairment, deafness, and neurodevelopmental disability compared to their term peers. Those born preterm are also more likely to develop non-communicable diseases such as obesity, type II diabetes, hypertension and cardiovascular disease in adulthood. Despite extensive ongoing research into causes, predictors and treatments for preterm birth, the rate of preterm birth has not fallen, even increasing in some countries.
Number and timing of previous deliveries The risk of recurrent spontaneous preterm birth is dependent on the aetiology of the previous preterm birth, the number of previous preterm births, and the severity of prematurity. The likelihood of having a preterm birth in a given pregnancy increases with the number of previous preterm births, with a reducing chance of delivery after 33 weeks of 83, 86 and 68% after 1, 2 or 3 prior preterm births. Risk of preterm birth also increases as the gestational age of previous preterm births declines, with the greatest risk for those with a previous delivery at less than 28 weeks’ gestation. If the previous preterm delivery was an indicated delivery, the risk of recurrence will be related to the recurrence risk of the condition leading to the indicated delivery. A history of previous indicated preterm birth is also associated with an increased risk of spontaneous preterm birth in subsequent pregnancy. Presumably this is due to a shared aetiology
Charlotte Oyston MB ChB BMedSci (Hons) Dip OMG PhD Advanced Trainee in Obstetrics and Gynaecology, Middlemore Hospital, Auckland, New Zealand. Conflicts of interest: none declared. Katie Groom MB BS BSc FRANZCOG PhD CMFM Associate Professor of Maternal and Perinatal Health, The Liggins Institute, University of Auckland; Maternal Fetal Medicine Subspecialist, Department of Obstetrics and Gynaecology, National Women’s Health, Auckland City Hospital, Auckland, New Zealand. Conflicts of interest: none declared.
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
353
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
women with uterine anomalies may also have a cervical anomaly which contributes to cervical insufficiency in pregnancy.
(such as placental ischaemia) between pregnancy complications such as pre-eclampsia and preterm birth. The timing of the last pregnancy should also be reviewed, as risk of early delivery increases with reducing interpregnancy interval. An analysis of 6181 women with a previous preterm birth demonstrated that after adjusting for confounding factors, those with an interpregnancy interval of less than 6 months had the highest increase in risk of preterm delivery (44% increase in risk of recurrent preterm birth), those with an interval of 6e12 months had a modest increase, and there was no increase in risk of recurrent preterm birth with an interpregnancy interval of 12e18 months, compared to the risk of those with an interval of 18 months or more.
Use of assisted reproductive technologies Several systematic reviews have shown an increase in risk of preterm delivery in women undergoing IVF treatment compared to those who conceive naturally, with the risk of preterm delivery doubling for delivery prior to 37 weeks and tripling for delivery prior to 33 weeks for singleton pregnancies. It is unclear whether this increase in risk results from the fertility treatment itself, underlying maternal factors associated with infertility, or iatrogenic bias in the care of these pregnancies. Demographic, socioeconomic, psychosocial and lifestyle factors Women who are obese or significantly under weight prepregnancy are at increased risk of preterm delivery, as well as other pregnancy complications. A large meta-analysis of data from over one million individuals found that women from developed countries with a pre-pregnancy body mass index (BMI) of less than 18.5 kg m2 are more likely to have a preterm birth e either spontaneous or indicated d compared to women with a BMI in the normal range (RR 1.22). Women with a BMI of 35 kgm2 also appear to be at increased risk of indicated preterm birth (OR 1.5e1.8), which may be partially explained by an increased prevalence of chronic disease (such as hypertension or diabetes) in this population. Race and ethnicity are important risk factors for preterm birth. After adjusting for risk factors such as age, education and parity, black women have a two- to three-fold higher risk of preterm birth compared to non-black women. The aetiology of this association is unclear but is possibly interlinked with other lifestyle factors and/or social disparities; rates of preterm birth are increased amongst socioeconomically deprived women, women who have high levels of anxiety or perceived stress or low levels of support, are at the extremes of maternal age, and those who smoke, consume alcohol, and use illicit drugs during pregnancy.
History of cervical surgery or trauma/uterine instrumentation Women undergoing excisional treatment for cervical dysplasia have an increased risk of preterm birth, although the mechanism(s) mediating this increase in risk are unclear. One hypothesis is that these procedures reduce the mechanical support of the cervix which may then increase susceptibility to cervical insufficiency, or spontaneous loss of the cervical mucous plug leading to ascending infection. Cervical conisation has the most consistent and strongest association with preterm birth. Women with a history of conisation are 2e3 times more likely to have a preterm birth than healthy controls, and those conceiving within 2e3 months of conisation or large conisation (>1 cm depth) may have the greatest risk. Meta-analyses also suggest an increased risk of preterm birth with other excisional therapies e such as large loop excision of the transformation zone (LLETZ) (RR 1.70 compared to healthy controls). Although many studies within the meta-analyses use healthy controls as the comparison group, an increased risk of preterm birth with all excisional therapies remains d albeit attenuated d when women undergoing cervical treatment are compared to those who had been diagnosed with precancerous lesions, but not treated. This suggests that the increase in risk may be in part due to the underlying pathology of the precancerous lesion, as well as the cervical procedure itself. Cervical dilation and curettage for miscarriage or termination of pregnancy also increase the risk of spontaneous preterm birth in subsequent pregnancies. While some large cohort studies have found no significant difference in rates, more recent good quality studies have demonstrated a small but significant increase in risk of spontaneous preterm birth after these procedures, with odds ratios of up to 1.8 and evidence that the risk of preterm birth in subsequent pregnancies rises with the number of procedures performed.
Pre-conception strategies to reduce risk of preterm birth What advice should be given and what medical or surgical interventions are available pre-conception that could reduce the risk of preterm delivery? There are few pre-conception interventions that improve outcomes in preterm birth. Many risk factors for preterm birth are non-modifiable, and those that can be modified only minimally increase absolute risk and are poorly predictive of subsequent preterm birth. When assessing the impact of strategies to prevent preterm birth, it is important to assess impact on perinatal morbidity and mortality, as in some instances, preterm birth may represent a mechanism through which the fetus escapes an adverse intrauterine environment; prolonging these pregnancies may be associated with neonatal harm rather than benefit.
Presence of congenital uterine anomalies Congenital uterine anomalies are found more commonly in women with a history of second trimester loss or preterm birth, with a prevalence of up to 25% (compared to 6% in women from an unselected population). The risk of preterm birth varies with the type of uterine anomaly; women with a uterine didelphys or a septate uterus have the highest risk with up to 33% of pregnancies ending in preterm delivery. It has long been thought that the increased rate of preterm birth results from a reduced volume and distensibility of the uterine cavity; however it is most likely that
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
Pre-conception advice and medical interventions Ideally, counselling and assistance regarding optimal pregnancy spacing, the benefits of maintaining a healthy weight, and abstinence from smoking, alcohol and drugs should be provided
354
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
in the preconception period. Smoking cessation (and limiting passive exposure) is of particular importance, as this is the largest modifiable risk factor for preterm birth. At an individual level, providing behavioural counselling with nicotine replacement therapy (safer than continuing to smoke, and associated with increased rates of cessation) is the most effective strategy for smoking cessation. Incentive schemes (where incentives such as cash or vouchers are used to encourage cessation), also improve cessation rates. At a population level, clinicians should support and promote smoke-free environments: rates of preterm birth reduce by an estimated 10% after introduction of smokefree legislation. Medical management of chronic medical conditions such as diabetes and hypertension should be optimised; although this may not be associated with reductions in rate of spontaneous preterm birth, optimal control of chronic conditions may lower the risk of additional pregnancy complications, and therefore the rates of indicated preterm birth, as well as potentially leading to better long-term outcomes for mother and baby.
transabdominal cerclage, compared to those who had a repeat vaginally-placed cerclage (6 vs 12.5%). The MAVRIC trial is the first randomised trial comparing abdominal with vaginal cerclage. Preliminary data suggests that for women with a previously unsuccessful vaginally-placed cervical cerclage, a transabdominally-placed cerclage is superior to repeat vaginallyplaced cerclage in reducing risk of delivery prior to 32 weeks and fetal death.
Case 2 Antenatal tests for prediction of preterm birth Ms S is a 24-year-old G3P2 with a history of two previous preterm deliveries at 32 weeks. You see her in clinic at 12 weeks and she asks if there are any tests she can have that will detect whether she will go into labour early. What tests can be used in pregnancy to help stratify the risk of preterm delivery in asymptomatic women? Accurately identifying women who are at high risk of delivering preterm provides an opportunity to optimise the timing and administration of interventions such as antenatal corticosteroids, MgSO4 and arrange for transfer to a centre with tertiary neonatal facilities. A negative test also provides reassurance for those who are unlikely to deliver preterm. Tests that are currently available in clinical practice for the prediction of spontaneous preterm birth include sonographic cervical length measurement, and the vaginal biomarkers, fetal fibronectin (fFN), placental alpha macroglobulin-1 (PAMG-1) and cervical phosphorylated insulinlike growth factor binding protein-1 (phIGFBP-1). Of the biomarkers tests, only fetal fibronectin has published data to support use in asymptomatic women.
Pre-conception surgical interventions As uterine anomalies are associated with increased risk of preterm birth, it seems logical that normalising uterine anatomy would be associated with a lowering of preterm delivery. For women with a septate uterus, there is evidence from observational studies that pregnancy outcomes (including preterm birth rates) may be improved with hysteroscopic resection of the septum. However, given the paucity of adequately powered randomised studies comparing outcomes of hysteroscopic resection with no treatment, clear recommendations as to which women are appropriate candidates for this procedure cannot be made. Similarly, other surgeries aimed at normalising anatomy in women with different congenital uterine malformations have been described, but evidence is limited to case reports and small observational studies. Where surgery is considered for women with a uterine anomaly, the benefits and risks should be carefully considered on a case by case basis. Cervical cerclage is a procedure where a non-absorbable suture is placed around the cervix to reduce the risk or preterm delivery. The mechanism of action is thought to be through increasing mechanical support, or through reducing the risk of ascending infection through cervical lengthening or retention of the mucous plug. Cerclage is typically performed vaginally in the second trimester for at-risk pregnancies (the indications for cerclage placement during pregnancy are discussed in more detail in the section ‘Antenatal strategies to reduce the risk of preterm birth’ below). Abdominal placement of a cervical suture via laparotomy or laparoscopy is sometimes carried out preconception in women with a very short or scarred cervix (where transvaginal placement may be difficult), or women with a history of a failed vaginally-placed cervical suture. Abdominal cerclage is associated with a higher incidence of serious maternal operative complications than vaginal cerclage (3.4 vs 0%) and their use should be reserved for those at highest risk with a previously failed vaginal cerclage. A systematic review of 13 case series and one controlled non-randomised study of transabdominal cervical cerclage in women with a history of failed transvaginally-placed cervical cerclage found a lower risk of perinatal death or delivery before 24 weeks in women who had a
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
Fetal fibronectin Fetal fibronectin is a glycoprotein found in the amniotic fluid, placenta, and between the chorion and decidua. It is described as a “glue” between chorion and decidua and is not normally detectable in vaginal secretions after fusion of decidua and fetal membranes. The presence of fFN in vaginal secretions after 18 weeks occurs following disruption of the choriodecidual interface due to inflammation or mechanical forces and is associated with an increased risk of preterm delivery. A swab taken from the posterior fornix can be used to detect the presence of fFN in a bedside test, with results given qualitatively (a positive corresponding to a concentration of >50 ng/mL, or negative <50 ng/ mL), or quantitatively. Compared to women with threatened preterm labour, fFN has a lower sensitivity for the prediction of preterm birth in asymptomatic women (78e89% compared to 68 e76%), although the specificity appears similar for both asymptomatic (88e89%), and symptomatic women (86%). Currently, the greatest utility for fetal fibronectin is in its negative predictive value for women symptomatic of preterm labour, and the reassurance that imminent delivery is highly unlikely. It is not clinically useful as a stand-alone test for prediction of preterm birth in asymptomatic women with or without other risk factors for preterm birth. Quantitative fFN shows an improved prediction of preterm birth compared with qualitative testing. The risk of birth before 35 weeks increases with increasing levels of fFN from >20 ng/ml to 300 ng/ml. Using a threshold of
355
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
1st centile) had a significantly increased risk of delivery prior to 35 weeks (relative risk 2.4, 3.8, 9.5 and 14 respectively). It should be noted however, that less than half the women with the shortest cervices (less than 15 mm), delivered prior to 35 weeks. These findings have since been replicated in other large prospective studies.
10 ng/mL results in a higher sensitivity for prediction of preterm labour and when combined with cervical length measurement, is a more cost-effective approach to identifying the symptomatic women at greatest risk of preterm birth (see below). Phosphorylated insulin-like growth factor binding protein-1 Phosphorylated insulin-like growth factor binding protein-1 is a protein secreted by decidual cells. Detachment of fetal membranes from the decidua causes leakage of phIGFBP-1 into vaginal secretions. A qualitative bedside test (Actim PartisTM) is commercially available. As for fFN, this test is measured from sections obtained on a vaginal swab. Like fFN, phIGFBP-1 has better negative predictive value than positive predictive value for preterm birth in symptomatic women. Its predictive accuracy in asymptomatic women is limited, with a recent meta-analysis suggesting a pooled positive likelihood ratio of 14e47%, and negative likelihood ratio of 76e93% for preterm delivery.
Combined fFN and cervical length measurements The combination of transvaginal cervical length measurement and fFN testing is better at predicting spontaneous preterm delivery compared to fFN alone. A subsequent analysis of NICHD data looked at the effect of cervical length and fFN at 24 weeks on the risk of preterm delivery in women with a history of previous preterm birth. In women with a previous preterm birth, both cervical length and fetal fibronectin test results at 24 weeks modified the risk of preterm delivery. The highest risk of preterm delivery was in women with a positive fFN and a cervical length <25 mm (risk of delivery <35 weeks 64%). The risk was lowest for women with a negative fetal fibronectin and a cervical length >35 mm or more (7% risk of delivery prior to 35 weeks). Quantitative fFN can also be used in conjunction with cervical length measurement. In a study of asymptomatic women with risk factors for preterm delivery, women with a mid-trimester cervix length <25 mm had a 9.5% risk of delivery <34 weeks when qfFN <10 ng/mL. The risk of delivery <34 weeks increased to 55% for those with levels >200 ng/mL. A risk prediction tool for both symptomatic and high-risk asymptomatic women has been developed, combining quantitative fFN, cervical length, and past obstetric history. This tool has the advantage that it provides an individualised estimate of preterm delivery, rather than just a summary estimate of risk. The algorithm is available to clinicians as a smart phone app; further information is available at www.quipp.org.
Placental alpha microglobulin-1 Placental alpha microglobulin-1 (PAMG-1) is a large glycoprotein synthesised by decidual cells. PAMG-1 is found in amniotic fluid. Its detection in the vaginal secretions is normally low but has been used to detect ruptured membranes in a bedside test marketed as AmnisureTM. More recently, studies have assessed the role of PAMG-1 in detecting preterm labour independently from rupture of membranes in women with signs or symptoms of preterm labour. The test appears to perform with high sensitivity and specificity for symptomatic women. However, as these studies were of relatively small sample size, further research is required to confirm the test’s accuracy, and to clarify its role in the assessment of women who are asymptomatic. Cervical length measurement Large prospective studies have consistently observed an increased risk of preterm birth for women with a shortened cervix during pregnancy, and that the risk of preterm delivery increases with decreasing cervical length. Transvaginal ultrasound is the gold standard method for visualising and measuring cervical length, with high likelihood of obtaining measurements and low interobserver variability. The alternative transabdominal approach may be affected by maternal bladder filling, and is associated with a lower likelihood of obtaining adequate measurements. Cervical length measurements are performed over a 5-min period with an empty maternal bladder and in the absence of undue probe pressure. The cervical length is taken as the shortest measurement that displays the landmarks of internal, external os, cervical canal. Assessment of cervical length in women at high risk of preterm birth is usually performed between 14 and 24 weeks; risk of spontaneous preterm birth increases as cervical length decreases and gestational age decreases. The National Institute of Child Health and Development (NICHD) Preterm Prediction study was the first large prospective study to evaluate cervical length as a predictor for preterm birth in asymptomatic women. Cervical length was measured in 3000 women at 22e24þ6 weeks. Compared to women with a cervical length of more than 40 mm (the 75th centile) women with a cervical length less than 40, 35, 26 or 13 mm (75th, 50th, 10th or
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
Antenatal strategies to reduce the risk of preterm birth What antenatal strategies are proven to be effective at reducing the risk of preterm birth in women with a prior history of preterm birth? Cervical cerclage A cervical cerclage is typically performed in the second trimester for one of three indications: a history of recurrent preterm losses (history indicated), cervical shortening detectable on ultrasound (ultrasound indicated), or in women presenting with signs of cervical dilation or visible fetal membranes (rescue cerclage). The most common techniques for cerclage placement are the Shirodkar and McDonald techniques, both of which result in a purse-string suture around the cervix. The Shirodkar technique describes dissection of the vaginal mucosa anteriorly and posteriorly to expose the cervix as close to the level of the internal os as possible, whereas the McDonald technique does not conventionally involve dissection. Many studies of cervical cerclage do not stipulate which technique has been used and there are no prospective studies evaluating pregnancy outcomes by comparison of the two techniques, so choice of technique is generally governed by surgeon preference. There is also little evidence to guide choice of suture material. Traditionally, a braided filament has been the suture of choice, due to the perceived superior
356
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
strength, and ease of removal compared to monofilament sutures. However, the multifilament braided suture favours bacterial colonisation, and has been associated with increased infectious morbidity and poorer tissue healing in non-obstetric settings. There is concern that use of braided filaments for cerclage may introduce a route through which preterm birth is stimulated. The C-STITCH study is a randomised trial that is currently recruiting, and aims to assess the effect on pregnancy and neonatal outcomes of monofilament versus braided suture material in women undergoing cerclage.
Cerclage or progesterone? Cerclage requires a skilled operator and is associated with maternal risks relating to anaesthesia and surgery (cervical laceration, fever, bleeding, chorioamnionitis and ruptured membranes). Therefore, a question of considerable interest is whether cerclage is better at preventing preterm birth than other treatments. There is a paucity of data comparing progesterone to cervical cerclage (or pessary). A recently published indirect meta-analysis (comparing interventions head-to-head through a common comparison group) has suggested progesterone is more effective at preventing preterm delivery (<34, <37 weeks) than cerclage. The SuPPoRT trial is currently randomising women with a cervix <25 mm on mid trimester ultrasound to cervical cerclage, progesterone or pessary to determine the relative efficacy of these interventions on pregnancy duration and neonatal outcomes. There is currently insufficient evidence to support combined use of multiple (cerclage and progesterone, progesterone and pessary) vs single interventions.
History indicated cerclage In an international multi-centre study, 1292 women were randomised to cerclage or no cerclage when there was clinical uncertainty as to whether or not to perform a cervical cerclage. The cerclage group had significantly fewer deliveries prior to 33 weeks (number needed to treat to prevent one delivery <33 weeks ¼ 25), however subgroup analysis showed that only women with at least three previous pregnancies ending prior to 37 weeks benefited from cerclage. In this group, the number of preterm deliveries prior to 33 weeks was halved; however, no differences in neonatal outcomes were detected. It is possible that this study underestimated the magnitude of benefit of cerclage, as it excluded women who were potentially the most likely to benefit from cerclage e those that clinicians were not uncertain as to whether cerclage would provide benefit.
Progesterone Progesterone is a key component in the maintenance of pregnancy. It has anti-inflammatory properties and plays a role in the prevention of cervical ripening and maintenance of myometrial quiescence. There have now been over 40 published randomised trials studying the efficacy of progesterone in reducing preterm birth. The two main types of progesterone studied are natural progesterone (rapidly absorbed via vaginal mucosa) and 17 alpha hydroxyprogesterone capronate (which has a longer halflife and is administered as weekly intramuscular injections). Of the two, vaginal progesterone is cheaper, and appears to be more effective. Meta-analyses of earlier randomised trials suggested a benefit of progesterone in the prevention of preterm birth in women with a shortened cervix, or history of a previous preterm birth. The use of progesterone for preterm birth prevention has been endorsed by clinical guidelines groups, including the National Institute for Health and Care Excellence. More recently, two large randomised studies (OPPTIMUM and PROGRESS) have shown no benefit of vaginal progesterone in reducing preterm birth, composite neonatal morbidity and mortality, or childhood cognitive and neurosensory outcomes at age 2 years. A subsequent metanalysis which included data from the OPPTIMUM study concluded that progesterone reduced preterm birth, neonatal morbidity and mortality when given antenatally to women with a shortened cervix. However, there remains debate regarding the specific conditions where progesterone may reduce risk of preterm birth, and the optimal formulation, dose, route also remain unclear. The Evaluating Progestogens for Prevention of Preterm birth International Collaborative (EPPPIC) is an independent meta-analysis of individual participant data from the totality of RCTs of progesterone in the prevention of preterm birth. Publication is expected in late 2018 and will hopefully clarify whether there are particular subgroups of pregnant women who may benefit from progesterone therapy.
Ultrasound indicated cerclage in women with a history of preterm birth A meta-analysis using individual patient level data compared cerclage with expectant management in women with a shortened cervix (less than 25 mm) and found that cerclage was associated with a reduction in preterm birth before 35 weeks in those with singleton pregnancies and a history of prior preterm birth or second trimester loss (RR 0.63). A further randomised study of women with a history of preterm birth and a cervical length of <25 mm showed a reduction in pre-viable delivery and perinatal mortality, although there was no reduction in delivery in less than 35 weeks unless cervical length was below 15 mm. Rescue cerclage Ideally, women with a previous preterm birth should be offered elective cervical cerclage or serial surveillance of cervical length (depending on prior history), thereby avoiding a presentation with cervical dilatation. However, in the event that cervical dilatation occurs, a rescue cerclage should be considered. A small randomised study, comparing rescue cerclage with expectant management (one week of broad spectrum antibiotic therapy and bed rest until 30 weeks) found cerclage delayed delivery by an average of 4 weeks, and had a reduction in delivery prior to 34 weeks, perinatal morbidity and a trend towards reduced perinatal mortality. Similar delays have been seen in small nonrandomised studies. Cerclage may be less likely to be effective with significant membrane prolapse (beyond the external os), cervical dilatation of more than 4 cm, and should never be performed when there are signs or symptoms of chorioamnionitis, fetal compromise, death or fetal anomaly not compatible with life, PPROM or ongoing vaginal bleeding.
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
Cervical pessary Cervical pessaries are proposed to prevent preterm labour via the alteration of the inclination of the cervical canal relative to the uterus (resulting in a more acute utero-cervical angle),
357
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
controlled trials have not found a reduction in the rate of preterm birth when women are treated for this condition.
preventing direct pressure on the cervix and fetal membranes at the level of the internal os. Pessaries are a “one-off” treatment, well tolerated with minimal side effects and are of relatively low cost, they are an attractive alternative to cerclage and progesterone treatments, particularly for women from countries where resources are limited. The first randomised controlled trial using Arabin pessaries for the prevention of preterm birth in women with a shortened cervix detected on a mid-trimester ultrasound scan demonstrated reduced rates of preterm delivery and a reduction in composite neonatal adverse outcomes. Subsequent randomised studies have shown conflicting results, with a recent meta-analysis suggesting that pessary use does not reduce rates of preterm birth or improve perinatal outcomes in women with a shortened cervix. Interestingly, there appears to be an effect of clinician-learning curve on the effectiveness of pessary treatment, with significant differences in outcomes between those recruited early and those recruited later to trials of pessary. Further randomised trials are underway, as strong evidence of efficacy is lacking, and adequate training in operation and insertion is needed to ascertain whether cervical pessaries can prolong gestation and improve perinatal outcomes for women at risk of preterm birth.
Bed-rest, relaxation, or stress reduction Although many studies have demonstrated an association between psychological stress and preterm birth, a limited number of studies have assessed the effect of relaxation techniques on preterm birth rates. The studies published use different relaxation techniques, so are difficult to compare directly; however while studies show evidence of reduced stress and anxiety scores, they do not show evidence of reduced risk of preterm birth. Similarly, there is no evidence that reducing work or reducing sexual activity reduces the likelihood of preterm birth. Management of women presenting with preterm labour The management of women who present with threatened preterm labour does not differ between those who have and have not had a previous preterm birth. In brief, management begins with a careful history, examination, and investigations aimed at determining the likelihood of preterm labour, and excluding the presence of serious conditions in mother and baby that require imminent delivery (for example, fetal distress, placental abruption or chorioamnionitis). Once conditions requiring expedited delivery have been excluded, adjunct tests such as ultrasound assessment of cervical length (as well as determining fetal presentation and growth), fFN, phIGFBP-1 and PAMG can be considered. This allows targeted use of therapies most likely to improve neonatal outcomes should preterm delivery ensue: maternal administration of corticosteroids (if gestational age is <35 weeks); maternal administration of magnesium sulphate where delivery is imminent (if gestational age is <30 weeks); transfer to a unit able to provide appropriate neonatal care and the considered treatment of any precipitating causes of labour.
Preterm birth clinic Despite a growing body of evidence supporting the use of cervical cerclage and progesterone for the prevention of preterm birth, the counselling, availability and clinical use of these interventions is inconsistent. In obstetrics (and many other areas of medicine), there is a trend toward developing specialised clinics which focus on prevention and provide a means for consistent application of the most up-to-date evidence-based practice. There is limited evidence that attendance at preterm birth prevention clinics reduce the rates of preterm birth, and such clinics are now an accepted part of service provision in many centres and offer other benefits such as an opportunity for teaching of both clinicians and patients, and an ideal setting for further research into the aetiology and prevention of preterm birth.
Summary and conclusions A history of a preterm birth is an important risk factor for spontaneous preterm birth in a subsequent pregnancy, although the majority of women with a previous preterm birth will deliver at term. Risk factors for spontaneous preterm birth have been identified, however many of these are non-modifiable; and those that can be changed e such as smoking cessation and weight optimisation - may result in a modest reduction in risk of preterm delivery at best. There is evidence that interventions such as progesterone and cervical cerclage may be beneficial in certain women at high risk of preterm birth, however questions still remain as to how best to screen these women, the optimal treatment regimen, and whether these treatments improve perinatal outcomes. In asymptomatic women, the best biomarkers for the prediction of preterm delivery are cervical length measurement, and detection of fFN in the cervicovaginal secretions. However, without effective strategies that diminish both the rate of preterm birth and improve perinatal outcomes, the role of these tests in the management of these women is less well defined. These biomarkers should be used in future research to direct trials of new therapies to those at the highest risk of
Antenatal therapies with insufficient evidence of benefit Screening and treatment of asymptomatic women for lower genital tract infection Although genital tract infection is associated with preterm birth, there is not consistent evidence that the screening and treatment of asymptomatic women improves outcomes. A recent systematic review found screening women in the early second trimester and treating for candidiasis, bacterial vaginosis (BV) and trichomoniasis halved the rate of delivery before 37 weeks. In contrast, studies assessing the effect of screening and treatment of BV or trichomoniasis independently have found that although treatments are effective at eradicating infection, rates of preterm birth were either no different, or increased compared to untreated women. Treatment of periodontal disease Increasing severity of periodontal disease is associated with increasing likelihood of preterm birth. However, randomised
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
358
Ó 2018 Elsevier Ltd. All rights reserved.
CASE-BASED LEARNING
preterm birth. Out of this mass of evidence have arisen a variety of guidelines worldwide for the prediction and prevention of preterm labour. An example is the NICE (National Institute of Clinical Excellence) guideline ‘Preterm Labour and Birth’, published in the UK in 2015, but there are others. A
Practice points C
C
FURTHER READING Abbott DS, Hezelgrave NL, Seed PT, et al. Quantitative fetal fibronectin to predict preterm birth in asymptomatic women at high risk. Obstet Gynecol 2015; 125: 1168e76. https://doi.org/10.1097/AOG. 0000000000000754. Althuisius SM, Geijn HP. Strategies for preventiondcervical cerclage. BJOG An Int J Obstet Gynaecol 2005; 112: 51e6. https://doi.org/ 10.1111/j.1471-0528.2005.00585.x. Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology and causes of preterm birth. Lancet 2008; 371: 75e84. https://doi.org/ 10.1016/S0140-6736(08)60074-4. ‘Preterm labour and birth’ NICE guideline 25. 2015, https://www.nice. org.uk/guidance/ng25. Shennan AH. Prediction and prevention of preterm birth: a quagmire of evidence. Ultrasound Obstet Gynecol 2018; 51: 569e70. https:// doi.org/10.1002/uog.19063.
OBSTETRICS, GYNAECOLOGY AND REPRODUCTIVE MEDICINE 28:11-12
C
C
C
359
Preterm birth may be considered as either spontaneous or indicated. For women with a previous indicated preterm birth, risk of recurrence depends on the specific condition necessitating early delivery. However, these women are also at slightly higher risk of spontaneous preterm birth. For women with a previous spontaneous preterm birth, risk of early delivery increases with the number of previous preterm deliveries and is more likely in those who previously delivered at earlier gestations. However, most women with a previous preterm delivery will not deliver preterm in subsequent pregnancies. In a general population, risk estimates for preterm delivery can be improved by transvaginal ultrasound measurement of cervical length, and fetal fibronectin testing. However, the role of these investigations in women with a previous preterm birth is not well defined, and further research is needed to clarify what optimal cut-offs are whether available interventions provide significant benefit for those who “screen positive”. Progesterone and cervical cerclage are interventions which may reduce the risk of preterm delivery, in certain groups of high risk women.
Ó 2018 Elsevier Ltd. All rights reserved.