Preterm birth affects 15 million babies per year and leads to significant newborn morbidity and mortality. A short cervix is the strongest clinical risk factor for predicting preterm birth. The definition of a short cervix depends on gestational age and fetal number; the most common definition is less than 25 mm as measured by transvaginal ultrasound. The predictive accuracy of a short cervix operates on a continuum with a shorter cervix and earlier gestational age at diagnosis conferring the highest risk. Though preterm birth is a heterogeneous disorder, cervical shortening is final common pathway of several disorders, including cervical insufficiency, intrauterine infection, and uterine overdistention. Clinical management of a patient with a short cervix depends on whether there is a history of prior preterm birth. For women with a prior preterm birth, management involves supplemental intramuscular progesterone with consideration of cervical cerclage. For women with no prior preterm birth, vaginal progesterone is most commonly offered. Whether there is benefit of a cervical pessary and whether cervical length screening should be offered to all women are areas of active investigation.
Keywordscervix, preterm birth, short cervix, cerclage
Preterm birth (PTB) touches communities everywhere. Around the world, one in 10 babies is born too early. The toll is more than 15,000,000 babies per year. The deep impact of this problem has fostered widespread efforts to use the best available medical practices to address it. Accordingly, the U.S. PTB rate fell from 10.44% to 9.57% from 2007–2014, although, unfortunately, preliminary 2015 birth data shows an increase (to 9.62%). Risk of PTB is inversely associated with cervical length (CL), and midtrimester CL is currently the strongest clinical risk factor for PTB. This chapter addresses the (1) clinical significance of a short cervix, (2) pathophysiology of cervical shortening, (3) ultrasound (US) technique for cervical measurement, and (4) management options for the short cervix.
The cervix is a cylindric anatomic structure with a central canal that forms the lower part of the uterus. It has a dual role: first, it must remain closed throughout pregnancy to allow fetal growth and development during pregnancy, and then open completely at term to allow fetal passage. The responsible molecular and biomechanic mechanisms are incompletely elucidated.
The inverse relationship between CL and PTB risk was identified 20 years ago; specifically, the shorter the cervix, the greater the risk. The definition of a short cervix depends on gestational age and fetal number, but the cutoff is usually 20–30 mm. The most common definition is less than 25 mm, but significance varies with gestational age; 25 mm corresponds to the first percentile at 18 weeks, the third percentile at 22 weeks, the fifth percentile at 24 weeks, and the 10th percentile at 28 weeks of gestation. The definition of 25 mm is also widely applied to twin gestation, and short cervix is more common in multiples; in the National Institute of Child Health and Human Development Maternal Fetal Medicine Unit (NICHD MFMU) network’s Preterm Prediction Study, 18% of women with twins had a short cervix in the midtrimester, compared to 9% in singletons. Another definition of a short cervix is 10–20 mm; this definition was used in a randomized, double-blinded, placebo-controlled trial of vaginal progesterone for PTB prevention. Another common definition is less than 15 mm because the risk of PTB increases exponentially below that value.
Prevalence and Epidemiology
The predictive accuracy of a short cervix for spontaneous PTB depends on the length of the cervix, gestational age at testing, fetal number, and risk factors for PTB ( Table 124.1 ).
|Incidence Spb <35 Weeks’ Gestation||Risk||Characteristics|
|>50%||Highest risk||Singleton, candidate for a physical examination–indicated cerclage; or cervix <15 mm, history of SPB, no cerclage; twin, cervix <15 mm|
|20%–50%||High risk||Singleton, cervix <15 mm, no history of SPB, no therapy; or cervix 15–25 mm, history of SPB, no therapy; or cervix <25 mm + progesterone ± cerclage; Twin, cervix 15–25 mm|
|10%–20%||Medium risk||Singleton, cervix >25 mm, history of SPB or other risk factor; or cervix 15–25 mm, no history of SPB; Twin, cervix >25 mm|
|<10%||Low risk||Singleton; no history of SPB or other risk factor|
Length of Cervix
As noted earlier, the risk of PTB increases as the cervix shortens. In an NICHD MFMU network trial, transvaginal CL was measured in asymptomatic women at 24 weeks of gestation to predict PTB before 35 weeks. The relative risk (RR) increased steadily as CL shortened, and reached its highest (RR = 14) when the cervix was less than or equal to 13 mm. A systematic review of 17 studies of asymptomatic singletons reported a significantly increased risk of PTB with CL <25 mm before 34 weeks of gestation; the likelihood ratio was 6.29 (95% confidence interval [CI], 3.29–12.02).
Gestational Age at Testing
A short cervix before 20 weeks of gestation has a worse prognosis than a short cervix later in pregnancy. In a systematic review of 17 pooled studies of asymptomatic singletons, the likelihood ratio for PTB was increased if the gestational age at testing was less than 20 weeks compared with 20 to 24 weeks. This finding is not surprising; 25 mm corresponds to the first percentile at 18 weeks but the 10th percentile at 28 weeks, so a 25-mm cervix at 18 weeks is more concerning.
Risk Factors for Preterm Birth
The sensitivity of CL for predicting PTB depends on the patient population. For instance, multiple gestation increases the PTB risk, but a short cervix increases it further. In the Preterm Prediction study, the risk of PTB in twins with CL 25 mm or less was similar to that for singletons with CL 15 mm or less. A systematic review of 11 pooled studies of asymptomatic twins found an 18.5% baseline risk of PTB before 34 weeks’ gestation compared with 29%–48% when the cervix was less than 25 mm. In singleton pregnancy, women with a history of spontaneous PTB and a short cervix are at highest risk; in an unselected, low-risk patient population, the sensitivity of a CL less than 25 mm for predicting PTB before 35 weeks of gestation was 37.3%, but in a population of women at high risk based on a prior history of PTB, the sensitivity increased to 69%.
Etiology and Pathophysiology
Although a short cervix is clearly associated with PTB, the question of whether a short cervix is a “cause” of PTB (i.e., cervical insufficiency or “weakness”) or a “consequence” of a pathophysiology unrelated to cervical “weakness” is unclear.
In a study of 52 women with cervical dilation at least 1.5 cm, without active labor (contractions or bleeding), 80% had evidence of infection/inflammation. Positive amniotic fluid culture was present in half of cases and, even when this was negative, evidence of inflammation (higher intraamniotic interleukin-6 levels) was often present.
Similarly, infection/inflammation is common in the setting of a short cervix (less than 25 mm), particularly with a very short cervix. In a study of 401 patients admitted for preterm labor at 22 to 35 weeks of gestation, microbial invasion was more prevalent when the cervix was less than 15 mm (26.3%) compared with 15 to 25 mm (3.8%). In another study in which a panel of inflammatory cytokines was studied as a function of CL, most inflammatory cytokines became elevated at a CL less than or equal to 5 mm.
These studies confirm that intraamniotic infection and inflammation are common when the cervix is short.
Although it seems intuitive that uterine contractility would cause a short cervix, this association is weak at best. In a secondary analysis of a NICHD MFMU study in which subjects had both home uterine activity monitoring and CL measurement, there was no association between CL (less than 25 mm versus greater than or equal to 25 mm) and frequency of uterine contractions.
Uterine overdistention from multiple gestation or polyhydramnios is associated with PTB and cervical shortening. Although the risk of PTB increases rapidly when CL is less than 15 mm in a singleton gestation, in a twin gestation, a similar increase in risk is seen at a higher threshold, with a CL of less than 25 mm. The pathophysiology of PTB is clearly different for multiple versus singleton gestations.
Cervical “Insufficiency” or “Incompetence”
Cervical tissue gets its “strength” from its collagen-rich connective tissue, and abnormalities of cervical tissue have been associated with adverse outcomes. However, the definition of cervical “insufficiency” or “incompetence” has recently come under scrutiny. The term has traditionally been applied to pre- or periviable midtrimester pregnancy loss, and presumed to be caused by “weak” tissue. However, studies that have tried to identify signs of acquired “weakness” (cone biopsy, obstetric laceration) or inherent “weakness” (uterine anomalies, genetic disorders involving collagen) have been inconclusive. A significant limitation of studies linking “weak” cervical tissue with cervical shortening is the lack of a clinically acceptable method to measure tissue properties quantitatively and objectively. Noninvasive quantitative US methods such as shear wave elasticity imaging are under active investigation ( Chapter 170 ).
It is likely that cervical shortening is associated with several different pathophysiologic pathways. Research aimed at clarifying the role of a short cervix in PTB will aid efforts to tailor suitable preventive therapies.
Manifestations of Disease
Clinical Presentation and Screening
Most pregnant women with a short cervix are asymptomatic, regardless of whether the discovery is incidental or the result of screening. The 2016 Society of Maternal Fetal Medicine (SMFM) recommendations for CL screening, which closely approximate those of the American College of Obstetricians and Gynecologists, and the International Society of Ultrasound in Obstetrics and Gynecology, support routine screening in high-risk women (i.e., those with a history of spontaneous PTB) and are noncommittal on screening in low-risk women. Specifically, the latter practice is considered reasonable but not mandatory. Serial transvaginal CL screening of high-risk women should occur at least every 2 weeks between 16 and 24 weeks (grade 1A evidence). If the cervix is 25–29 mm, it is reasonable to perform measurements weekly. Further, SMFM guidelines state that appropriate training should be undertaken in centers that perform CL evaluation because of the potential for over- and underdiagnoses and treatment with erroneous measurements.
Imaging Technique and Findings
The gold standard for CL measurement is transvaginal US, via a technique well described by the NICHD MFMU network trials. This approach is more reproducible and sensitive than transabdominal assessment, and is also unaffected by maternal obesity or shadowing from fetal parts. However, the practicality of transabdominal US may make it useful as an initial screening step.
For transvaginal measurement of CL, the bladder is emptied, and the patient positioned in the dorsal lithotomy position. The transducer is introduced into the vagina and gently applied to the cervix in the anterior fornix. Pressure is applied until a satisfactory image is seen and then withdrawn slightly until the image starts to degrade. The anterior and posterior width of the cervix should be equal; excessive pressure from the transducer may falsely lengthen the cervix, and too little pressure may result in poor image quality. CL is determined by measuring the linear distance between the internal and external cervical os ( Fig. 124.1A ). If the cervix is curved, a linear distance may underestimate the CL, making it best to measure with two or three sequential lines ( Fig. 124.1B ). Out of three images, the shortest, best measurement should be recorded.