Recurrent Preterm Birth

Abstract

Recurrent preterm birth is frequently defined as two or more deliveries before 37 completed weeks of gestation. The recurrence rate varies as a function of the antecedent for preterm birth: spontaneous versus indicated. Spontaneous preterm birth is the result of either preterm labor with intact membranes or preterm prelabor rupture of the membranes. This article reviews the body of literature describing the risk of recurrence of spontaneous and indicated preterm birth. Also discussed are the factors which modify the risk for recurrent spontaneous preterm birth (a short sonographic cervical length and a positive cervicovaginal fetal fibronectin test). Patients with a history of an indicated preterm birth are at risk not only for recurrence of this subtype, but also for spontaneous preterm birth. Individuals of African-American origin have a higher rate of recurrent preterm birth. The potential roles of genetic and environmental factors in recurrent preterm birth are considered.

INTRODUCTION

Preterm birth is the leading cause of perinatal morbidity and mortality worldwide.¹ A preterm delivery is a risk factor for subsequent preterm birth.^2–22 Preterm birth can be the result of three obstetrical circumstances: 1) preterm labor with intact membranes; 2) preterm prelabor rupture of membranes (PROM); and 3) “indicated” preterm birth, which occurs when maternal or fetal indications require delivery before 37 weeks of gestation. The most common indications are preeclampsia and small for gestational age (SGA). Spontaneous preterm parturition is a syndrome caused by multiple etiologies (Figure 1), which are expressed by synchronous or dyssynchronous activation of the common terminal pathway of parturition. The reader is referred to recent reviews for full consideration of this concept.^23,24

Figure 1.

Pathological processes implicated in the preterm parturition syndrome. (Reproduced with permission from Romero et al²⁴, with permission.)

Romero R, Espinoza J, Mazor M, Chaiworapongsa T. The preterm

parturition syndrome. In: Critchely H, Bennett P, Thornton S, editors.

Preterm Birth. London: RCOG Press; 2004. p. 28–60.

Although many studies have focused on the rate of preterm birth,^25–57 an important consideration is whether these deliveries are the result of spontaneous labor (with intact or ruptured membranes) or “indicated” preterm deliveries. The need for this distinction is based on the premise that the risk factors for recurrent preterm PROM, preterm labor with intact membranes, preeclampsia, and/or SGA are different. However, recent observations suggest that there may be overlap among these conditions,^18,19 so that a patient with an “indicated” preterm birth may also be at risk for spontaneous preterm birth.^18,19 The converse may also be true (i.e. that a patient with a spontaneous preterm birth is at risk for an “indicated” preterm birth in a subsequent pregnancy).

This review will present a summary of the literature and aims to clarify the risk of recurrent disease and the biological basis for recurrent preterm birth.

Definition of Preterm Birth

Preterm deliveries are those that occur between fetal viability and 37 completed weeks of gestation (menstrual age). However, the lower limit of gestational age used to define a preterm birth has ranged from 13 to 24 weeks of gestation among various reports.^21,58,59 Our view is that the delivery of a pre-viable fetus should be considered a spontaneous abortion rather than a spontaneous preterm birth. Otherwise, perinatal and infant mortality estimates in a population or country will be biased by the frequency of late spontaneous abortions.

The precise definition of viability, however, is subject to debate given the increased frequency of survival at very low gestational ages. Clearly, some infants can survive around 24 weeks of gestation, but none at 20 weeks. Therefore, we propose the range of 24 to 36 6/7 weeks of gestation for the definition of preterm birth. If and when technological advances allow substantial survival below 24 weeks of gestation, this definition should be revised.

A birthweight of 500 grams has also been used to define the lower limit of viability.^11,60 The popularity of this definition derives from its simplicity. Birthweights can easily be obtained anywhere in the world at a very low cost. The limitations of this approach are that viable neonates born at viable gestational ages and affected by intrauterine growth restriction (IUGR) may have birthweights below 500 grams, and that some pre-viable infants may weigh more than 500 grams. Ideally, gestational age at birth should be used to define viability. There are, however, practical obstacles derived from the uncertainty of gestational age estimation in many countries. This problem can be overcome in areas where ultrasound is performed routinely in early pregnancy, but not elsewhere, including underdeveloped countries. The criteria for the definition of viability have implications for the calculation of vital statistics and comparisons of these among different populations.

RECURRENT PRETERM BIRTH

Recurrent preterm birth is defined as two or more deliveries before 37 completed gestational weeks.^{2,9,12,59,61,62} However, the definition among studies is not uniform. Criteria that have varied and may affect estimation of the rate of recurrent preterm birth include: 1) gestational age thresholds used for defining the upper (i.e. 35 or 36 weeks)^9,12,63,64 and lower (i.e. 13 to 28 weeks)^10,59 limits of preterm birth; 2) inclusion of multiple gestations;^65–67 3) inclusion of spontaneous, as well as indicated preterm births;^11,64 4) the number of preterm births required to fulfill the definition of recurrent preterm birth;⁸ and 5) the requirement that the preterm births be consecutive. ⁸

RECURRENT SPONTANEOUS PRETERM BIRTH

The Frequency of Recurrent Spontaneous Preterm Birth

Recurrent spontaneous preterm birth is defined as more than one preterm birth related to spontaneous onset of labor with intact membranes or preterm PROM.

Several studies have consistently indicated that patients with a previous spontaneous preterm birth are at risk for a subsequent spontaneous preterm delivery.^2–22 Iams et al⁹ reported the results of a secondary analysis of the data from the Preterm Prediction Study, conducted under the leadership of Goldenberg et al.⁶⁸ Among 378 patients with a prior spontaneous preterm birth or spontaneous abortion in the second trimester (gestational age range: 18–36 weeks), the rate of recurrent spontaneous preterm birth (<35 weeks) varied between 14% and 15%, in contrast to the 3% rate of spontaneous preterm birth among 904 parous women with a previous uncomplicated term delivery (Table 1).

Table 1.

Probability and 95% confidence intervals of spontaneous recurrent preterm birth at less than 35 weeks according to the gestational age of the previous spontaneous preterm birth.

Gestational age at delivery in previous spontaneous preterm birth	Probability (95% CI) of spontaneous recurrent preterm birth at <35 weeks
18–26 weeks	0.15 (0.05–0.37)
27–31 weeks	0.15 (0.05–0.38)
32–36 weeks	0.14 (0.05–0.32)
≥ 37 weeks	0.03 (0.03–0.03)

Modified from Iams et al⁹ with permission.

Iams JD, Goldenberg RL, Mercer BM, Moawad A, Thom E, Meis PJ et al. The Preterm Prediction Study: recurrence risk of spontaneous preterm birth. National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. Am. J. Obstet. Gynecol. 1998;178:1035–40.

The rate of recurrent preterm birth was modified according to the sonographic cervical length in the index pregnancy and the presence of a positive test for fetal fibronectin in cervicovaginal fluid at 22–24 weeks of gestation.⁹ Among women with a previous spontaneous preterm birth, the rate of recurrence (<35 weeks) was the highest (64%) among women with a sonographically short cervix (<25 mm) and a positive fetal fibronectin test. The lowest rate of recurrence (7%) occurred among patients with a sonographic cervical length >35 mm and a negative fetal fibronectin test.⁹

Patients with a positive fibronectin test were at higher risk for spontaneous recurrent preterm birth regardless of cervical length: 28% for patients with a positive test compared to only 7% for patients with a negative test. Similarly, a sonographic short cervix contributed to the risk of recurrent spontaneous preterm birth. Among patients with a history of spontaneous preterm birth and a short cervix, the rate of recurrence was 25% if the fetal fibronectin test was negative and 64% if the test was positive. This information can be used to counsel patients about their risk of spontaneous preterm birth. However, further investigation is required in which a mathematical model is generated to predict the individual risk for preterm birth based upon clinical, sonographic, and biochemical parameters in which results are expressed as continuous variables rather than categorical results.

Mercer et al¹⁰ performed a secondary analysis of the same dataset to evaluate the association between prior spontaneous preterm birth and subsequent pregnancy outcome. Women with a history of spontaneous preterm delivery (before 37 weeks of gestation) had a 2.5-fold increased risk (95% confidence interval (CI), 1.9–3.2) of spontaneous preterm delivery in a subsequent pregnancy compared to women with no history of spontaneous preterm delivery (21.7% vs. 8.8%, respectively, p<0.001). This risk was particularly high when the analysis focused on recurrent spontaneous preterm delivery before 28 weeks of gestation (relative risk (RR) 10.6 95% CI, 2.9–38.3).¹⁰ Moreover, the earlier the gestational age of the first preterm delivery, the greater the risk for recurrent spontaneous preterm birth (RR 2.4, 2.7, and 3.1 for prior delivery at 35–36, 28–34, and 23–27 weeks of gestation, respectively).¹⁰

Preterm Prelabor Rupture of Membranes and Recurrent Preterm Birth

PROM is defined as spontaneous rupture of the chorioamniotic membranes before the onset of labor.⁶⁹ Since the consequences of PROM depend on the gestational age at which the episode occurs, this condition has been classified as preterm PROM or term PROM, depending upon whether the rupture of the membranes occurs before or after 37 weeks of gestation, respectively.^70–76 Term PROM occurs in approximately 10% of pregnancies,^{70–72,74,77,78} whereas the frequency of preterm PROM is 2% to 3.5%.^{70–72,74,77,78} Preterm PROM accounts for 30% to 40% of preterm deliveries, ^{70–72,74,77,78} and it is a leading clinically identifiable cause of preterm birth and a major contributor to perinatal morbidity and mortality.^{21,58,70–78}

Analysis of data from the Collaborative Perinatal Project demonstrated that among women with a previous term delivery not complicated by PROM, the frequency of preterm PROM in a subsequent pregnancy is 4%.⁷⁹ In contrast, among patients with two successive singleton pregnancies in the dataset (n=5,230), the frequency of preterm PROM is 21% if the first pregnancy resulted in a preterm delivery due to preterm PROM.⁷⁹

Several investigators have confirmed the high recurrence rate of preterm PROM: 1) Asrat et al⁸⁰ reported a 32% (95% CI: 23.9–40.5) risk of recurrence in 121 patients with a previous episode of preterm PROM; 2) Ekwo et al⁸¹ reported that women with preterm PROM in a previous pregnancy had a 5.5-fold higher risk of preterm PROM in a subsequent pregnancy than those in the control group; and 3) Mercer et al⁵⁸ reported that compared to women with no history of preterm PROM, women with a previous preterm PROM had a higher risk of spontaneous preterm delivery due to preterm PROM in the index pregnancy (13.5% vs. 4.1%, p<0.001; RR: 3.3, 95% CI: 2.1–5.2) as well as a higher risk of preterm PROM at less than 28 weeks (1.8% vs. 0.1%, p<0.01; RR 13.5, 95% CI: 23–80.3).

Mercer et al⁸² used the Preterm Prediction Study population to determine the risk factors for subsequent preterm birth caused by preterm PROM alone. Preterm PROM at less than 35 weeks of gestation occurred in 2% of patients and at less than 37 weeks in 4.5% of patients. Preterm PROM accounted for 32.6% of all preterm deliveries. Clinical characteristics associated with preterm birth caused by preterm PROM, derived from a multivariable analysis and stratified as preterm PROM <37 and <35 weeks are displayed in Table 2. In nulliparous women, the risk factors for preterm PROM were a cervical length ≤25 mm, working during pregnancy, and the presence of medical complications (the odds ratios (OR) ranged between 3 and 3.7).⁸² Among multiparous women, a previous preterm birth caused by preterm PROM was the primary risk factor for preterm PROM in a subsequent pregnancy (OR for preterm PROM at <35 weeks: 4.1; <37 weeks: 3.1). Noteworthy is that a previous preterm birth caused by preterm labor with intact membranes was also a risk factor for preterm PROM, although the odds ratios were lower than if the previous preterm birth was the result of preterm PROM (OR for preterm PROM at <35 weeks: 2.6; <37 weeks: 1.8).⁸²

Table 2.

Risk factors associated with preterm PROM (at less than 35 weeks) stratified by parity.

	Nulliparous		Multiparous
Risk Factor	Odds Ratio	95% Confidence Interval	Odds Ratio	95% Confidence Interval
Cervical length ≤ 25 mm	9.9	3.8–25.9	4.2	2.0–8.9
Previous preterm birth with preterm PROM	--	--	4.1	2.0–8.7
Previous preterm labor with intact membranes	--	--	2.6	1.2–5.3
Working during pregnancy	5.3	1.5–18.7	n.s.	n.s.
Medical complications	4.2	1.1–16.0	n.s.	n.s.
FFN (+)	n.s.	n.s.	n.s.	n.s.
BV	n.s.	n.s.	n.s.	n.s.
FFN (+) and absent BV	n.s.	n.s.	9.0	3.6–22.5
FFN (−) and present BV	n.s.	n.s.	2.8	1.2–6.3

FFN = fetal fibronectin; BV = bacterial vaginosis; n.s. = non-significant.

Modified from Mercer et al,⁸² with permission.

Mercer BM, Goldenberg RL, Meis PJ, Moawad AH, Shellhaas C, Das A et al. The Preterm Prediction Study: prediction of preterm premature rupture of membranes through clinical findings and ancillary testing. The National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. Am. J. Obstet. Gynecol. 2000;183:738–45.

Interestingly, the only risk factor consistently associated with preterm PROM at <37 and <35 weeks in both nulliparous and multiparous women was a short cervical length. Bacterial vaginosis was not found to be a risk factor for recurrent preterm birth.⁸²

Among multiparous women, a short cervix, a positive fetal fibronectin test, and a history of preterm birth following preterm PROM increased the frequency of recurrent preterm PROM at <35 weeks to 25%. If recurrent preterm PROM was defined as <37 weeks, the combination of these three risk factors increased the risk 7.8-fold over the reference group of multiparous women who had none of these risk factors.⁸²

The mechanisms responsible for the association between previous preterm PROM, short cervix, and positive fetal fibronectin and recurrent preterm birth caused by preterm PROM have not been elucidated. It is likely that an insult during gestation (e.g. intrauterine infection) would be resolved by preterm labor with intact membranes or preterm PROM. We have proposed⁸³ that selection of the specific phenotype may be determined by genetic and/or environmental factors. For example, if patients carry DNA variants which predispose to an excess production of matrix-degrading enzymes, such patients will go into premature labor after rupture of membranes. On the other hand, if the genotype is such that the generation of uterotonic agents rather than matrix-degrading enzymes is favored, then preterm labor with intact membranes will be the clinical expression of the preterm parturition syndrome. The genotype may explain the tendency for the same phenotype to occur in subsequent pregnancies (i.e. preterm PROM).

The relationship between a short cervix and preterm PROM could be due to intrauterine infection.^84,85 A long cervix with a well-established mucus plug may serve as an anatomical and biochemical host defense mechanism against ascending intrauterine infection.^86–92 A short cervix may predispose to ascending intrauterine infection by shortening the distance between microorganisms in the lower genital tract and the chorioamniotic membranes.^89,93 In addition, the process of cervical reducing may lead to the loss of the mucus plug. Cervical mucus has been demonstrated to contain antimicrobial properties, attributed at least in part to antimicrobial peptides such as defensins, lactoferrin, calprotectin, and bacterial permeability factor.^86–90

The relationship between a positive cervicovaginal fetal fibronectin test and subsequent preterm PROM has been attributed to the presence of upper genital tract infection.⁶⁸ This interpretation has been based on the association between a positive fetal fibronectin in the midtrimester and the subsequent demonstration of histologic chorioamnionitis at the time of preterm delivery.⁶⁸ However, studies in which amniocentesis was performed in women with positive cervicovaginal fetal fibronectin have found that intra-amniotic infection or inflammation is present in less than 2% of patients with a positive fibronectin.⁹⁴ Because fetal fibronectin is a component of the extracellular matrix located in the chorion leave⁹⁵ and its abundance in cervicovaginal fluid increases prior to both preterm^68,96–109 and term labor^110–119, we propose that detection of this protein is a marker of decidual/membrane activation and, therefore, of the common terminal pathway of parturition. Thus, a history of preterm PROM and a positive test for fibronectin in the midtrimester are likely to reflect activation of the decidua/membrane component of the pathway. We propose a sequence of events that may explain the empirical observations reported by Mercer et al.⁸² A patient with a previous preterm PROM is at risk for subsequent PROM.⁸² If such a patient has a short cervix, the risk of recurrent preterm PROM would increase because of ascending intrauterine infection. If the infection is such that it activates chorion and decidua, then fetal fibronectin will be positive. Of course, it is also possible that a patient with preterm PROM would have activation of the common terminal pathway (and therefore a positive fetal fibronectin) with a long cervix.

Twin Gestation and Recurrent Preterm Birth

There is conflicting data as to whether preterm birth in the context of a multiple gestation is a risk factor for preterm birth in a future singleton pregnancy. Menard et al⁶² were the first to examine the recurrence rate after preterm birth of a twin gestation. The authors reported the outcomes of 144 women who first delivered twins, followed by a subsequent singleton gestation. Preterm delivery (before 37 weeks of gestation), occurred in 59.7% of twin gestations and in 14.6% of the subsequent singleton pregnancies. Among women who delivered preterm twins, 19.6% delivered preterm in the subsequent singleton pregnancy. Preterm birth in twin gestations was associated with a significantly increased risk of preterm delivery in a subsequent singleton pregnancy (RR 2.87, 95% CI 1.02–8.09). Among the subset of patients that delivered twins before 30 weeks, 42% of the subsequent singleton pregnancies delivered preterm (RR 6.11, 95% CI 2.07–18.02). The RR of preterm birth of a singleton after delivery of twins between 30 and 34 weeks of gestation was 3.63 (95% CI 1.02–12.92). However, women who delivered twins between 34 and 37 weeks of gestation did not have an increased risk for recurrent preterm birth.

In contrast, Rydhstrom⁵⁹ reported that a preterm twin delivery, regardless of etiology, did not increase the risk of recurrent preterm birth in a subsequent singleton gestation. However, a prior preterm singleton delivery increased the risk of preterm birth in subsequent singleton and twin pregnancies. Bloom et al¹² reported that women with a singleton gestation that resulted in preterm birth at less than 35 weeks have an increased risk for recurrence (OR 5.6, 95% CI 4.5–7.0). However, those whose first pregnancy resulted in twins delivered at less than 35 weeks did not have a higher risk of recurrent preterm birth (OR 1.9, 95% CI 0.46–8.14).

Cervical Insufficiency as a Cause of Recurrent Midtrimester Abortion/Preterm Birth

The clinical diagnosis of cervical insufficiency is traditionally made in patients with a history of recurrent mid-trimester spontaneous abortions and/or early preterm deliveries in which “the basic process is thought to be the failure of the cervix to remain closed during pregnancy.”¹²⁰ Thus, both by definition and clinical practice, the condition now termed, “cervical insufficiency” is recognized as one that recurs in subsequent pregnancies.

Digital examination of the cervix was the method used to determine cervical status (effacement, dilatation, position, and consistency) before the introduction of ultrasound. Cervical sonography has become an objective and reliable method to assess cervical length, which approximates cervical effacement. The shorter the sonographic cervical length in the mid-trimester, the higher the risk of spontaneous preterm labor/delivery.^121–125 However, there is no agreement concerning what constitutes a sonographic short cervix. For example, Iams et al¹²² proposed that a cervix of 26 mm or shorter at 24 weeks of gestation increases the risk for spontaneous preterm delivery (RR: 6.19, 95% CI: 3.84–9.97). The prevalence of spontaneous preterm delivery (defined as less than 35 weeks) in this study was 4.3%, and the positive predictive value was 17.8% for a cervical length ≤ 25 mm at 24 weeks of gestation.¹²² Other investigators have proposed a cut-off of 15 mm, because a cervical length of 15 mm or less is associated with nearly a 50% rate of spontaneous preterm delivery at 32 weeks of gestation or less, when neonatal morbidity is substantial.^123,125

Sonographic cervical length is not a screening test for spontaneous preterm delivery, because only a small fraction of all patients who will have a spontaneous preterm birth had a short cervix in the mid-trimester. Previous studies conducted at our institution have indicated that only 8% of all patients who will have a preterm delivery at less than 32 weeks of gestation have a cervical length of 15 mm or less in the mid-trimester.¹²⁵ The converse is also true. Among women with a short cervix, some have adverse pregnancy outcomes and others have uncomplicated term deliveries.^{66,121–123,126–140} Only half of women with a cervical length of 15 mm or less deliver before 32 weeks of gestation.¹²⁵ This indicates that cervical length is only one of the factors determining the degree of cervical sufficiency and that a short cervix should not be equated with “cervical insufficiency.”

Sonographic cervical length can modify the a priori risk for preterm delivery. For example, in patients with a history of preterm delivery, a twin gestation, or a triplet gestation, a short cervix confers an increased risk for preterm delivery.^{109,141–149} Indeed, among women with a history of spontaneous preterm birth, the risk of recurrence increases as cervical length shortens.⁹

The hypothesis that cervical competence or sufficiency represents a spectrum was studied by Parikh and Mehta, who used digital examination of the cervix and concluded that degrees of cervical competence do not exist.¹⁵⁰ Iams et al, using sonographic examination of the cervix, suggested that cervical sufficiency/insufficiency is a continuum,⁶⁶ with a strong relationship between cervical length in the index pregnancy and gestational age at delivery in the first pregnancy. This relationship was nearly linear; patients with a typical history of an incompetent cervix (painless dilatation) do not constitute a separate group from those with a history of spontaneous preterm delivery (preterm labor or preterm PROM).⁶⁶ Similar results have been reported by Guzman et al.¹⁵¹ Collectively, these studies suggest a relationship between a history of preterm delivery and the cervical length in a subsequent pregnancy. Inasmuch as patients with a short cervix are at increased risk for a mid-trimester pregnancy loss (clinically referred to as “cervical insufficiency”) or spontaneous preterm delivery with intact or ruptured membranes,^{66,121–123,126–131,133–140,151,152} a short cervix could be considered the expression of a spectrum of cervical diseases or functions.

We have proposed that cervical insufficiency is one of the great “obstetrical syndrome.”¹⁵³ Cervical ripening in the mid-trimester may be the result of: 1) the loss of connective tissue after a cervical operation such as conization^154–156 or LEEP procedure;¹⁵⁶ 2) a congenital disorder such as cervical hypoplasia after diethylstilbestrol exposure;^157–160 3) intrauterine infection;^161,162 and 4) a suspension of progesterone action.¹⁶³ There is experimental evidence that progesterone can reverse cervical compliance induced by the administration of dexamethasone to pregnant sheep.¹⁶⁴ Sherman et al¹⁶⁵ have also generated evidence that the administration of 17 alpha hydroxyprogesterone may be beneficial in patients with clinically diagnosed “cervical insufficiency” and a cervical disorder that manifests itself with the clinical presentation of “cervical insufficiency.” Each of these causes of the syndrome could be affected by genetic or environmental factors. The possibility of novel and yet-to-be-discovered mechanisms of disease playing a role must also be considered.

A proportion of patients presenting with asymptomatic cervical dilatation in the mid-trimester have microbial invasion of the amniotic cavity (MIAC)^161,162 that can be as high as 51.5%.¹⁶² Microbial invasion of the amniotic cavity may be due to premature cervical dilatation with the exposure of the chorioamniotic membranes to the microbial flora of the lower genital tract. Microorganisms may gain access to the amniotic cavity by crossing intact membranes.¹⁶² Under these circumstances, infection would be a secondary phenomenon to primary cervical disease. An alternative explanation is that primary intrauterine infection (ascending, hematogeneous¹⁶⁶), or one caused by activation of microorganisms present within the uterine cavity¹⁶⁷ in the second trimester of pregnancy produces myometrial contractility and cervical ripening. Since uterine contractions are usually clinically silent in the mid-trimester of pregnancy, the clinical picture of an infection-induced spontaneous abortion may be indistinguishable from that of an incompetent cervix.^65,162 Recently, we have established that 9% (5/57) of women with a short endocervix (less than 25 mm) have microbiologically-proven intra-amniotic infection,¹⁶⁸ suggesting that these infections are subclinical and may precede the development of the clinical picture of acute “cervical insufficiency” (dilated and effaced cervix with bulging membranes). The issue of whether subclinical intrauterine infection is a cause of recurrent cervical insufficiency and preterm birth has not been answered.

Women of African-American Origin Have a Greater Risk of Recurrent Preterm Birth than Caucasians

There is a well-established disparity in the rate of preterm birth among ethnic groups in the U.S.^{8,11,169–176} Individuals of African-American origin are at higher risk for recurrent preterm birth.

A large population-based cohort study¹¹ in the state of Georgia found that among women who delivered between 20 and 31 weeks of gestation in their first pregnancy, African-American women had a higher rate of recurrent preterm birth at 20–31 weeks than did Caucasian women (African-American = 13.4% (95% CI: 11.4–15.6) vs Caucasian = 8.2% (95% CI 6.6–10.1)). The same was the case for deliveries between 32 and 36 weeks of gestation.

Of interest was that teenagers whose first preterm delivery occurred between 20 and 31 weeks of gestation had twice the risk of recurrent preterm birth (20–31 weeks) than that of women 20–49 years of age (Table 3). This observation was significant only among African-American women.

Table 3.

Odds Ratios for Recurrence of Preterm Delivery or Low-Birthweight Newborn by Race in Georgia, 1980–1995^*.

	Delivery at 20–31 wk†		Delivery at 32–36 wk†
Maternal Characteristic in Second Pregnancy	White (n=84)	Black (n=145)	White (n=712)	Black (n=1059)
Maternal age (years)
10–17	2.3 (0.9–5.6)	2.0 (1.2–3.5)	1.3 (0.8–2.0)	1.3 (1.1–1.7)
18–19	1.0 (0.5–2.0)	1.2 (0.8–2.0)	1.3 (1.0–1.7)	1.2 (1.0–1.4)
20–49	1.0	1.0	1.0	1.0
Initiation of prenatal care (trimester)
First	1.0	1.0	1.0	1.0
Second, third, or none	1.2 (0.6–2.2)	1.2 (0.8–1.7)	1.1 (0.9–1.4)	1.1 (1.0–1.3)
Interpregnancy interval, months
< 6	1.1 (0.5–2.1)	1.4 (0.9–2.3)	1.0 (0.7–1.3)	1.2 (1.2–1.5)
6–11	1.6 (0.9–2.9)	0.9 (0.5–1.5)	1.2 (0.9–1.5)	1.1 (0.9–1.3)
12–47	1.0	1.0	1.0	1.0
> 47	1.0 (0.4–2.1)	0.8 (0.4–1.5)	0.9 (0.7–1.2)	0.7 (0.6–0.9)
Goodness-of-fit P value§	0.72	0.33	0.29	0.93
Smoking during the pregnancy‡
Yes	0.4 (0.2–1.1)	1.7 (0.2–14.5)	0.8 (0.6–1.2)	0.6 (0.3–1.1)
No	1.0	1.0	1.0	1.0

^*Odds ratio for type of second pregnancy are controlled for all of the other variables in the table except smoking; figures in parenthesis are 95% confidence intervals.

^†Referent group is delivery in second pregnancy at gestation ≥ 37 weeks.

^§Goodness-of-fit for model including all variables except smoking

^‡Analysis restricted to second deliveries occurring from 1989 through 1995. Association adjusted for all other variables in the model.

Modified from Adams et al,¹¹ with permission.

Adams MM, Elam-Evans LD, Wilson HG, Gilbertz DA. Rates of and factors associated with recurrence of preterm delivery. JAMA 2000;283:1591–96.

Kitska et al⁶⁴ used a maternally-linked database from the Missouri Department of Health to study racial disparities and recurrent preterm birth. The study focused on 368,633 mothers who had two or more deliveries between 1978 and 1997. The frequency of recurrent preterm birth was 3.1% among African-Americans and 0.6% among Caucasians (RR, 5.40; 95% CI, 5.06, 5.75). Logistic regression analysis indicated that African-American origin increased the risk for recurrent preterm birth independently of other factors, such as medical complications and low socioeconomic status (adjusted OR, 4.11; 95% CI, 3.78, 4.47). Two additional findings of this study were that: 1) the recurrent preterm birth in women of African-American origin occurred at an earlier median gestational age than in women of Caucasian origin (31 weeks vs. 33 weeks); and 2) the gestational age of the recurrent preterm birth was similar to that of the previous preterm birth and most likely to occur at the same gestational age (Figure 2). This finding was consistent among individuals in both ethnic groups.

Figure 2.

Concordance in timing of preterm (20–34 6/7 weeks’ gestation) birth in Missouri to a mother with previous preterm birth, 1989–1997. The line represents the expected Gaussian curve if concordance in timing is a normally distributed event. The bars represent the timing for each preterm birth after the initial preterm birth for A, all mothers, B, Caucasians, or C, African American in correlation with the expected normal curve. (Reproduced with permission from Kistka et al^x, with permission.)

Kistka ZA, Palomar L, Lee KA, Boslaugh SE, Wangler MF, Cole FS et al. Racial disparity in the frequency of recurrence of preterm birth. Am J Obstet Gynecol 2007;196:131.

Additional Risk Factors for Recurrent Preterm Birth

Several environmental factors have been implicated in recurrent preterm birth. Cnattingius et al⁶¹ studied the association among smoking, previous very early preterm or moderate preterm delivery (before 32 weeks and at 32 to 36 weeks, respectively), and the risk of a subsequent very preterm or moderately preterm delivery in a population-based cohort of 243,858 women in Sweden. The OR for a very early preterm second delivery among the women who smoked 1 to 9 cigarettes per day was 1.4 (95% CI, 1.1, 1.7) and for those who smoked 10 or more cigarettes per day 1.6 (95% CI, 1.3, 2.0), as compared with nonsmokers. Furthermore, women who stopped smoking between pregnancies were not at increased risk for very early or moderate preterm delivery, whereas the women who started to smoke in the second pregnancy had the same risk as those who had continued to smoke.

Merlino et al¹⁷⁷ investigated the association between maternal weight loss and recurrent preterm birth in a cohort of 1,241 patients. Women whose body mass index (BMI) decreased more than 5 kg/m² had more frequent recurrent preterm birth than those whose BMI did not (21.1% vs 9.3%, P ≤ 0.01). For those with a term birth in the first pregnancy, the rate of preterm birth in the subsequent pregnancy was not affected by a decline in BMI. In contrast, women with a preterm birth in the first pregnancy had a higher rate of recurrent preterm birth if BMI decreased more than 5 kg/m² (80.0% vs 28.2%, P = 0.01). Hence, women whose BMI declines between pregnancies are at increased risk for recurrent preterm birth.

The effect of sexual behavior on the risk of recurrent preterm birth was the subject of a secondary analysis of a multicentric observational study of the association between cervical ultrasound at 16–18 weeks and the risk for recurrent preterm birth. Women (n=187) with singleton gestations who were at high risk for preterm birth because of a prior spontaneous preterm birth at less than 32 weeks of gestation were included.¹⁷⁸ A sexual history was obtained by interview at the time of enrollment. Information gathered included the number of sexual partners during the patient’s lifetime, the number of sexual partners during the patient’s pregnancy, and the frequency of sexual intercourse in the preceding month. The greater the number of sexual partners in a woman’s lifetime, the higher the frequency of recurrent preterm birth (one partner 19%, 2 to 3 partners 29%, more than 4 partners 44%, P≤ 0.007). Of interest, neither the frequency of sexual intercourse during early pregnancy nor the number of partners were risk factors for recurrent preterm birth, which is consistent with previous reports. ^179–184

RECURRENT INDICATED PRETERM BIRTH

Indicated preterm births are those that result from delivery of patients before term due to complications that place the mother and/or fetus at risk. Various authors include amont those complications hypertension-related disorders, obstetrical hemorrhage (placenta previa, placental abruption, and other causes of antepartum hemorrhage), and all medically induced preterm deliveries.^12,40,185 Other investigators¹⁸ categorize indicated preterm births into two categories: 1) ischemic placental disease (i.e. preeclampsia, SGA, placental abruption and fetal distress); and 2) miscellaneous (fetal malformations, placenta previa, unexplained vaginal bleeding, chronic hypertension and others).

The incidence of indicated preterm birth has been reported to range from 1% to 5.5% of all deliveries.^{12,176,186,187} However, indicated preterm birth accounts for approximately one-third of all preterm births.^40,185,187

Meis et al¹⁸⁵ reported a study examining the risk factors for indicated preterm birth using the Preterm Prediction Study data set. A history of a previous indicated preterm delivery was associated with an OR of 2.8 (95% CI 1.5–5.4; p=0.002; multivariable analysis by logistic regression including other risk factors for indicated preterm birth).

Bloom et al¹² reported the largest study conducted today in a single unit and concluded that an indicated preterm delivery in singleton gestations is associated with an OR of 5.4 (95% CI 3.1–9.2) for recurrent preterm birth at less than 35 weeks of gestation in comparison to patients who delivered at term in their first pregnancy. Patients who had an indicated preterm birth between 24 and 28 weeks had an OR of 12.5 (95% CI 3.8–40.7) for recurrent preterm delivery and 10 (95% CI 4.8–20.8) if they were delivered between 29 and 32 weeks of gestation. In contrast, patients who were delivered between 33 and 34 weeks did not have a higher risk for recurrent preterm birth.

Patients with one prior preterm birth had an OR of 2.4 (95% CI 1.5–4.1) for indicated preterm delivery compared to women with no history of preterm birth. Moreover, the OR increases to 5.2 (95% CI 2.2–11.9) when the patients had 2 or more previous preterm deliveries.¹⁷⁶ Collectively, these studies suggest that indicated preterm birth is not an isolated event and puts the patient at risk for a subsequent indicated, as well as spontaneous, preterm birth.

The subjects of recurrent preeclampsia and SGA are discussed elsewhere in this issue of the Seminars.

IS THE RECURRENCE RISK FOR PRETERM BIRTH DIFFERENT FOR SPONTANEOUS VERSUS INDICATED PRETERM BIRTH?

Preterm births have been classified as “spontaneous” or “indicated” because of the implicit assumption that preterm labor with intact membranes and preterm PROM share pathophysiologic features and clinical presentation (spontaneous onset of labor). Preterm preeclampsia, fetal distress, and severe IUGR – the most common causes of indicated preterm birth¹⁸⁶ – usually occur in the absence of spontaneous parturition. Thus, the presence or absence of spontaneous parturition has been a sharp dividing line between the phenotypes of indicated and non-indicated preterm birth.

One can also argue that the mechanisms of disease responsible for the phenotypes are shared within the conditions responsible for spontaneous preterm birth and within the conditions responsible for indicated preterm birth. For example, MIAC with bacteria is common in preterm labor^188–204 and preterm PROM,^93,205–211 but rare in preeclampsia and IUGR. In contrast, “failure of physiologic transformation of the spiral arteries” can be observed in all of these four conditions,^212,213 but is more prevalent and severe in preeclampsia and IUGR^214–220 than in preterm labor and preterm PROM.^212,213

Ananth et al¹⁸ provided evidence in support of this view, but also noted that spontaneous preterm birth in the first pregnancy may be followed by an indicated preterm birth in the subsequent pregnancy and vice-versa. The observations are derived from a population-based retrospective cohort study of births in Missouri in which analyses were restricted to women who delivered their first 2 consecutive singleton live births during the study period of 1989–1997.¹⁸ The key observation was that if the first pregnancy resulted in a spontaneous or indicated preterm birth, then women were more likely to have the same type of preterm birth (spontaneous or indicated) in the second pregnancy. Indeed, women with spontaneous preterm birth before 35 weeks of gestation in the first pregnancy had an OR of 3.6 (95% CI 3.2–4.0) for preterm birth before 35 weeks in the second pregnancy. However, the risk for a medically indicated preterm birth was also increased (OR 2.5, 95% CI 2.1–3.0).¹⁸

Similarly, women who delivered at less than 35 weeks because of a medical indication in the first pregnancy were much more likely to have an indicated preterm birth at less than 35 weeks of gestation in their subsequent pregnancy (OR 10.6, 95% CI 9.1–12.4). However, these patients were also at increased risk of having a spontaneous preterm birth (OR 1.6, 95% CI 1.3–2.1), although that risk was lower than the risk of having an indicated preterm birth.¹⁸ Similar findings were evident in pregnancies that ended at less than 32 weeks (Table 4). The greatest risk for recurrence of preterm birth was observed when women delivered their first preterm birth at less than 28 weeks of gestation. The magnitude of the risk for recurrence of preterm birth decreased progressively as gestational age at delivery of the first preterm birth increased.¹⁸

Table 4.

Recurrence of preterm birth at < 37, < 35 and < 32 weeks and subtypes in second pregnancy based on preterm birth at less than 37, less than 35, and less than 32 weeks in the first pregnancy, respectively: Missouri, 1989 to 1997

Preterm birth in the first pregnancy	Preterm birth in second pregnancy, adjusted OR (95% CI)
	All preterm births	Spontaneous births	preterm Medically preterm indicated births
Preterm birth at < 37 wks
Preterm birth at < 37 wks	2.9 (2.8, 3.0)	2.7 (2.5, 2.9)	3.3 (3.1, 3.6)
Spontaneous preterm birth	2.8 (2.7, 3.0)	3.2 (3.1, 3.4)	1.7 (1.5, 1.9)
Medically indicated preterm birth	3.0 (2.8, 3.3)	1.0 (0.9, 1.2)	7.7 (7.0, 8.5)
Preterm birth at < 35 wks
Preterm birth at < 35 wks	3.6 (3.4, 3.9)	3.1 (2.8, 3.4)	4.8 (4.3, 5.4)
Spontaneous preterm birth	3.3 (3.0, 3.6)	3.6 (3.2, 4.0)	2.5 (2.1, 3.0)
Medically indicated preterm birth	4.6 (4.0, 5.2)	1.6 (1.3, 2.1)	10.6 (9.1, 12.4)
Preterm birth at < 32 wks
Preterm birth at < 32 wks	4.9 (4.2, 5.7)	4.1 (3.4, 4.9)	6.5 (5.2, 8.0)
Spontaneous preterm birth	4.5 (3.8, 5.4)	4.6 (3.7, 5.6)	4.3 (3.1, 5.8)
Medically indicated preterm birth	5.8 (4.5, 7.4)	2.7 (1.8, 4.1)	11.3 (8.4, 15.1)

ORs are adjusted for maternal age (second birth), education (second birth), marital status, race/ethnicity, smoking and alcohol use, prepregnancy body mass index, and lack of or late initiation of prenatal care and interpregnancy interval. OR, Odds ratio; CI, confidence interval.

Modified from Ananth et al,¹⁸ with permission.

Ananth CV, Getahun D, Peltier MR, Salihu HM, Vintzileos AM. Recurrence of spontaneous versus medically indicated preterm birth. Am. J. Obstet. Gynecol. 2006;195:643–50.

ISSUES ON THE MANAGEMENT AND PREVENTION OF A PATIENT WITH A HISTORY OF PRETERM BIRTH

Prevention of Recurrent Preterm Birth

Progesterone Administration

Progesterone plays a central role in pregnancy. The proposed functions of progesterone include maintenance of myometrial quiescence, down-regulation of gap-junction formation, and inhibition of cervical ripening.^221–223

Prevention of recurrent preterm birth by progesterone administration has been a matter of debate in the literature for decades.^224–242 This section will review the results of randomized clinical trials and meta-analyses published recently.

da Fonseca et al,²⁴³ reported a randomized, double-blind, placebo-controlled trial of vaginal progesterone versus placebo in decreasing the rate of spontaneous preterm birth. Patients with at least one previous spontaneous preterm birth, a prophylactic cervical cerclage, or a uterine malformation (n=142). Patients were allocated to receive either daily progesterone (100 mg) or placebo by vaginal suppository from 24 to 34 weeks of gestation. The rates of preterm delivery at both less than 37 weeks and less than 34 weeks were lower in the progesterone group than in the placebo group. [For 37 weeks; progesterone: 13.8% (10/72) vs. placebo: 28.5% (20/70); p=0.03 and for 34 weeks; progesterone: 2.8% (2/72) vs. placebo: 18.6% (13/70); p=0.002].

Meis et al²⁴⁴ reported the results of a double-blind placebo-controlled clinical trial comparing the effects of intramuscular 17-OH P versus placebo. Patients with a history of spontaneous preterm delivery (n=463) were enrolled at 16 to 20 weeks of gestation and randomly assigned in a 2:1 ratio to receive weekly injections of 250 mg of 17-OH P or an inert oil placebo until either delivery or 36 weeks of gestation. Treatment with 17-OH P significantly reduced the rate of preterm delivery at less than 37 weeks [17-OH P 36.3% vs. placebo 54.9%, RR 0.66 (95% CI 0.54–0.81)] and less than 35 weeks of gestation [17-OH P 20.6% vs. placebo 30.7%, RR 0.67 (95% CI 0.48–0.93)]. Moreover, neonates born to women treated with 17-P had significantly lower rates of necrotizing enterocolitis [17-P 0% vs. placebo 2.6%, RR could not be calculated], intraventricular hemorrhage [17-OH P 1.3% vs. placebo 5.2%, RR 0.25 (95% CI 0.8–0.82)] and need for supplemental oxygen [17-OH P 14.9% vs. placebo 23.8%, RR 0.62 (95% CI 0.42–0.92)].

Of interest, the effect of 17-OH P in preventing recurrent preterm delivery was demonstrated only in patients whose previous preterm delivery had occurred between 20 and 33.9 weeks of gestation.²⁴⁵ Moreover, it has been estimated that 4.7 women would need to be treated to prevent one recurrent preterm delivery among patients who had delivered between 20–27.9 weeks of gestation in a previous pregnancy. The number needed to treat is similar for women who had delivered at 28–33.9 weeks. Of note, 17-OH P was not associated with a reduction in the rate of recurrent preterm deliveries in patients whose previous preterm delivery had occurred between 34–36.9 weeks of gestation.

The efficacy of 17-OH P in singleton gestations was not demonstrated in twin gestations.²⁴⁶ In this trial, no significant differences were found between the groups in the rates of spontaneous or indicated preterm birth.

In a 2006 Cochrane review, Dodd et al²⁴⁷ reported the results of six randomized trials including 988 patients randomized to receive either 17-OH P or placebo. The administration of 17-OH P was associated with reduced risks for preterm delivery before 37 weeks of gestation (six studies, RR 0.65 95%CI 0.54–0.79) and before 34 weeks gestation (one study, RR 0.15 95% CI 0.04–0.64). Moreover, treatment with progesterone was associated with lower risks for birthweight below 2500 grams (four studies, RR 0.63 95%CI 0.49–0.81) and intraventricular hemorrhage (one study, RR 0.63 95% CI 0.08–0.82). No difference in perinatal death was found between treatment groups (five studies, RR 0.66 95% CI 0.37–1.19). There were no interactions between the dose of progesterone (>500 mg vs. <500 mg 17-OH P weekly) or gestational age at commencing progesterone administration and the reported outcomes (i.e. preterm delivery, neonatal morbidity and mortality). These results were in accord with a previous meta-analysis by this group.²⁴⁸ Additionally, the authors stated that there is currently insufficient information concerning the safety of progesterone supplementation.

Sanchez-Ramos et al²⁴⁹ reported another meta-analysis including ten randomized clinical trails and a total of 1339 patients; eight trials used 17-OH P and two used other progestational agents for the prevention of recurrent preterm birth or recurrent abortion. Patients who were treated with progestational agents had a reduced risk of preterm delivery compared to patients in the placebo group (OR 0.45 95% CI 0.25–0.80). The number needed to treat to prevent a single preterm delivery was 10 (95% CI 6–24). A similar effect was observed when only trials using 17-OH P were included (OR 0.45 95% CI 0.25–0.80); and the number needed to treat was eight (95% CI 5–19).

Odibo et al²⁵⁰ performed a cost-effectiveness analysis of the treatment with 17-OH P for the prevention of preterm birth. The cost savings per quality-adjusted life year gained by using 17-OH P was $3,090 in women with a prior preterm delivery at <32 weeks and $2,963 in women who had delivered at 32–37 weeks of gestation. Moreover, the cost per additional preterm delivery avoided with the use of 17-OH P was $35,319 in women with a previous preterm delivery at <32 weeks and $36,093 in women who had delivered at 32–37 weeks of gestation.

In a recent meeting of the Prematurity Interest Group of the Society for Maternal-Fetal Medicine, da Fonseca et al reported the results of a randomized clinical trial of vaginal progesterone administration to women with a short cervix (<15 mm). A 40% reduction in the rate of preterm birth by was found in women treated with vaginal progesterone (da Fonseca E, Nicolaides K, personal communication). In contrast, the largest randomized clinical trial of vaginal progesterone in women with a history of previous preterm delivery did not demonstrate a beneficial effect of vaginal progesterone (Lewis D et al, personal communication). The FDA has raised questions about a safety signal.²⁵¹ However, this concern was not identified in the trial in twins.²⁴⁶ A review of embryo toxicity in animals has been recently published.²⁵²

In summary, it seems that the administration of progesterone may be an effective intervention to reduce the rate of preterm delivery in a subset of women with a previous preterm delivery. Women with a short cervix may benefit from this intervention.

Treatment of Bacterial Vaginosis

Treatment of patients with bacterial vaginosis and a history of preterm birth is controversial. While some investigators have argued strongly in favor of treatment, ²⁵³ others believe that this intervention is not justifiable.^254–258 Controversy over the choice of antibiotic also exists, with evidence that early treatment with clindamycin is preferable to treatment with metronidazole.^259–262 There is no evidence that treatment of patients with a previous preterm delivery with interconceptional antibiotics will prevent a subsequent preterm birth.^263,264

The Use of Cerclage

The clinical value of cervical cerclage has been subject of many observational and randomized clinical trials^{14,131,133,134,138,265–286} and the studies have been the topic of several systematic reviews.^287–289 The evidence suggests the following conclusions:

1. Cervical cerclage in women with a sonographic short cervix (15 mm or less) and a low risk for preterm delivery (by history) does not reduce the rate of spontaneous preterm birth.²⁸⁶

2. The effectiveness of cervical cerclage in women with a sonographic short cervix and a high risk (by history) for the prevention of preterm delivery is controversial.^{14,136,277,278,290}

3. The role of prophylactic cerclage in high-risk patients without a sonographic short cervix for the prevention of preterm delivery/midtrimester abortion (by history) is unclear.^{269–271,278,285} While the largest trial conducted prior to the introduction of ultrasound evaluation of the cervix suggested a modest beneficial effect,²⁷¹ other trials^269,270 and systematic reviews¹²⁰ prior to the use of ultrasound have indicated that the evidence of effectiveness is either weak or non-existent.

4. In patients at risk for preterm delivery, serial sonographic examination of the cervix followed by cerclage in those who shortened the cervix is a reasonable alternative to prophylactic placement of a cerclage based upon uncontrolled studies.^131,282,291

This evidence indicates that only patients with the clinical presentation of “acute cervical insufficiency” and those with a pregnancy history consistent with “cervical insufficiency” and progressive shortening of the cervix demonstrated with ultrasound may benefit from cerclage placement. However, important to consider is that each conclusion is based on the results of only one randomized clinical trial.^278,283 Sakai et al demonstrated that the inflammatory status of the endocervix may be an additional criterion to distinguish those patients who would benefit from cerclage placement from those in whom this intervention may be ineffective or harmful.²⁹²

Summary

The evaluation of a patient with a previous preterm birth begins with an examination of the obstetrical circumstances responsible for this complication. If the preterm birth was “indicated,” then the risk of recurrence is related to the specific condition, such as preterm preeclampsia, preterm severe IUGR, placenta previa, etc. The reader is referred to the relevant articles in this volume of the Seminars for details about recurrence rate, monitoring of the index pregnancy, and interventions.

If the previous preterm birth was the result of spontaneous labor (with intact or ruptured membranes), the information provided in this article can be used to counsel the patient about the likelihood of recurrence. In general, most patients with a previous spontaneous preterm birth will deliver at term in a subsequent pregnancy.^2–22 However, the earlier the gestational age of the preterm birth, the higher the likelihood of recurrence. It is important to be aware that recurrent preterm births tend to occur at the same gestational age.^10,12,18 Counseling should ideally be conducted prior to conception, and efforts should be made to identify potentially treatable causes such as a Mullerian duct abnormalities. However, the attributable risk of these conditions for preterm birth is extremely low.

The estimates of risk of recurrence for spontaneous preterm birth can be improved by performing a sonographic examination of the uterine cervix and a fetal fibronectin test. A long cervix and a negative fetal fibronectin test reduce the risk just as a short cervix and a positive fetal fibronectin test increase the risk.^9,122

No intervention has been proven effective in reducing the rate of preterm birth in patients with a positive fetal fibronectin test.^293,294 Similarly, the management of patients with a short cervix remains controversial. Evidence suggests that cervical cerclage does not prevent preterm birth in women with a short cervix who do not have a history of previous preterm birth.²⁸⁶ Similarly, a prophylactic cerclage has not been effective in reducing the rate of preterm birth in patients at risk for midtrimester abortion or spontaneous preterm birth.^138,277,282 In contrast, one randomized clinical trial of patients with risk factors or symptoms of cervical insufficiency and a shortened cervix (<25 mm before 27 weeks of gestation) in the index pregnancy found a benefit of “therapeutic cerclage.”²⁷⁸ Though further studies are required to identify effective interventions and the patients who will benefit from them, monitoring cervical length with ultrasound and offering cerclage based on individual risk assessment is a reasonable management strategy.