Cervical Insufficiency




Key Abbreviations


17-α-hydroxyprogesterone 17-OH-P


American College of Obstetricians and Gynecologists ACOG


Cervical insufficiency CI


Cervical intraepithelial neoplasia CIN


Cervical length CL


Diethylstilbestrol DES


Intramuscular IM


Interquartile range IQR


Intraamniotic infection IAI


Large loop excision of the transformation zone LLETZ


Loop electrosurgical excision procedure LEEP


National Institutes of Child Health and Human Development NICHD


Preterm birth PTB


Preterm premature rupture of the membranes PPROM


Randomized controlled trial RCT


Receiver operating characteristic curve ROC


Royal College of Obstetricians and Gynaecologists RCOG


Society for Maternal-Fetal Medicine SMFM


Spontaneous preterm birth sPTB


Transabdominal ultrasound TAU


Translabial ultrasound TLU


Transvaginal ultrasound TVU




Overview


Since the initial description in 1658 of the cervix being “so slack that it cannot keep in the seed” by Cole and Culpepper, few subjects in obstetrics have generated as much controversy as the term cervical incompetence, now more correctly termed cervical insufficiency . The competent or “sufficient” human uterine cervix is a complex organ that undergoes extensive changes throughout gestation and parturition. It is a key structure responsible for keeping the fetus and membranes inside the uterus until the end of gestation and for undergoing significant changes to allow the delivery of the baby during spontaneous or induced labor.


The cervix is primarily fibrous connective tissue composed of an extracellular matrix that consists of collagen types I and II, elastin, proteoglycans, and a cellular portion that consists of smooth muscle and blood vessels. A complex remodeling process of the cervix occurs during gestation that involves timed biochemical cascades, interactions between the extracellular and cellular compartments, and cervical stromal infiltration by inflammatory cells. Any disarray in this timed interaction could result in early cervical ripening, cervical insufficiency, and preterm birth (PTB) or miscarriage.


The incidence of cervical insufficiency in the general obstetric population is reported to vary between approximately 1/100 and 1/2000. If women with a singleton gestation, a prior spontaneous preterm birth (sPTB), and a current transvaginal ultrasound (TVU) cervical length (CL) less than 25 mm are labeled as having cervical insufficiency, the incidence is about 3% to 4%. Wide variations in estimating the incidence are likely due to real biologic differences among study populations, the criteria used to establish the diagnosis, and reporting bias between general practitioners and referral centers. Once considered a discrete entity reserved for patients with a history of midtrimester loss associated with painless cervical dilation, in the last few years, this diagnosis has become a component of the spontaneous preterm birth syndrome (see Chapter 29 ). Strategies that include the identification of a short cervix in the index gestation have been developed with different management protocols utilized to reduce the rate of PTB.




Cervical Insufficiency: a Distinct Entity or Evidence of Preterm Parturition?


In 1962, Danforth and Buckingham suggested that cervical incompetence was not an all-or-none phenomenon; rather it comprised degrees of insufficiency, and combinations of factors could cause “cervical failure.” These classic investigations demonstrated that unlike the uterine corpus, the normal cervix is composed predominantly of connective tissue. This fibrous band is the chief mechanical barrier against the loss of the enlarging products of conception. The cervix and mucous glands also play an important immunologic role in preventing vaginal flora from ascending into the normally sterile intrauterine environment.


In a subsequent report, these investigators analyzed cervical biopsies taken from postpartum women and compared them with hysterectomy specimens from nonpregnant patients. Pregnancy was associated with increased water content, a marked decline in collagen and glycoprotein, and increased glycosaminoglycans. The cellular and biochemical changes suggested that cervical dilation in pregnancy is a dynamic process, and this might explain why a woman could have a pregnancy outcome consistent with cervical insufficiency in one pregnancy but then, without treatment, have a subsequent term birth. Presumably, the factors that incite the pathologic cervical changes might vary among pregnancies. Women with a more muscular cervix might have heightened susceptibility to factors that lead to cervical changes and PTB.


These earlier observations were enlarged by Leppert and colleagues, who reported an absence of elastic fibers in the cervix of women with clinically well-characterized cervical insufficiency on the basis of their reproductive history. Conversely, cervical biopsy specimens from women with normal pregnancies showed normal amounts and orientation of these elastic fibers. It is unknown whether these microstructural and biochemical phenomena were congenital, acquired from previous trauma, or the result of other pregnancy-associated pathologic conditions. Collectively, these biochemical and ultrastructural findings support the variable, and often unpredictable, clinical course of women with a history of cervical insufficiency.


Whereas the traditional paradigm depicted the cervix as either competent or incompetent, both clinical data and interpre­tative reviews suggest that, as with most other biologic processes, cervical “competence” is rarely an all-or-none phenomenon. The term cervical insufficiency has been historically thought to represent a defect inherent in the cervix itself that leads to an inability to retain a pregnancy. Initial treatment, as we will discuss later, was aimed at preventing pregnancy loss by reinforcing the structural integrity of the cervix and repairing the defect in the cervical stroma. However, as we have gained more insight into the various surgical, mechanical, and biochemical modalities used to treat cervical pathology, inves­tigators have questioned whether abnormal cervical anatomy is the dominant cause or whether, at times, other factors play a major role.


Cervical insufficiency, as classically defined, is recurrent painless cervical dilation that leads to three or more midtrimester births. However, painful contractions are not a prominent feature of early cervical change, with the magnitude of contraction-related pain inversely proportional to the extent of cervical ripening prior to their onset. In contrast to a proposed mechanism of defective cervical tissue, although some patients may have tangible anatomic evidence of poor cervical integrity, most women with a clinical diagnosis of cervical insufficiency have ostensibly normal cervical anatomy. Knowing whether cervical integrity was compromised by a primary mechanical deficiency, versus other local or systemic factors, would help to define the optimal therapy. In a proposed model of cervical competence as a continuum, a poor obstetric history attributed to cervical insufficiency likely results from a process of premature cervical ripening induced by myriad underlying processes that include infection, inflammation, local or systemic hormonal effects, or even genetic predisposition. If and when cervical integrity is compromised, other processes may be stimulated (e.g., premature membrane rupture, preterm labor) and appear clinically as the spontaneous preterm parturition syndrome, which may also contain features related to the uterus and chorioamnion. Thus the term cervical insufficiency may be increasingly considered a clinically convenient label to describe a complex and poorly understood process of premature cervical ripening and the midtrimester onset of parturition. Given this, we recognize that newer approaches to the management of cervical insufficiency have been included within the broader topic of PTB prevention.




Short Cervix


Traditionally, cervical insufficiency (CI) has been defined by criteria based solely on obstetric history: painless cervical dilation that leads to recurrent second-trimester births in the absence of other causes. The modern approach allows the diagnosis of CI to be made in primigravidas or in multigravidas without multiple prior pregnancy losses. This new description defines CI by the presence of both (1) TVU cervical length less than 25 mm and/or cervical changes detected on physical examination before 24 weeks of gestation and (2) prior sPTB at less than 37 weeks. Regarding this second criterion, some limit it to sPTB up to 34 weeks. TVU screening is the gold standard for evaluation of the CL for prediction of PTB. CL can also be measured with transabdominal ultrasound (TAU) or translabial ultrasound (TLU), but these techniques should not be used clinically for prediction of PTB because TVU is superior in many ways. The TAU CL measurement is less sensitive in detecting a short CL, is noted to overestimate CL and underdiagnose short CL, and has several additional limitations: (1) a full bladder is needed; (2) the cervix maybe obscured by the fetal parts; and (3) the image is of poorer quality because of the distance from the abdominal probe to the cervix. TVU CL screening has been shown to have superior cost-effectiveness compared with TAU because TVU is associated with better prevention of PTB. All major guidelines that have described CL screening have clearly recommended TVU, including the Society for Maternal-Fetal Medicine (SMFM), American College of Obstetricians and Gynecologists (ACOG), and the Royal College of Obstetricians and Gynaecologists (RCOG) . TLU is also less sensitive and less predictive than TVU. Given this evidence, TAU or TLU should not be used for CL screening. Moreover, TVU has been shown to be more predictive of sPTB compared with digital manual examination of the cervix.


TVU CL screening should be done using the technique described by the Cervical Length Education and Review (CLEAR) program and the SMFM through its Perinatal Quality Foundation . The CLEAR program provides an official education program on TVU CL screening with examination and continuing image review to certify perinatologists, obstetricians, residents/fellows in training, and sonographers. TVU CL screening should be performed in clinical practice only after this program has been completed.


TVU CL measurement is a safe, acceptable, and reliable screening test that is now widely used to screen women for their risk for PTB. TVU CL screening is done after the patient has emptied her bladder. Then, a sterile transvaginal probe is inserted into the anterior vaginal fornix. Initially the probe is withdrawn until the image blurs to reduce compression from the transducer, then just enough pressure is reapplied to create the best image. The TVU image of the cervix should occupy 75% of the screen, and the lower tip of the bladder should also be visible. The anterior width of cervical thickness should be equal to the posterior cervical thickness, and there should be no increased echogenicity in the cervix because of excessive pressure. Prior to the measurement, the internal and external os and the entire endocervical canal should be identified. The calipers should be correctly placed to measure the distance from the internal to the external os ( Fig. 28-1 ). Three measurements of CL are usually obtained, then mild fundal pressure is applied for about 15 seconds to watch for funneling and/or cervical shortening. It is advisable to reduce probe pressure while fundal or suprapubic pressure is applied. The total time for TVU CL screening should be no less than approximately 5 minutes. With proper technique, the intraobserver and interobserver variabilities are both less than 10%. The population studied can affect the performance of CL screening. These include singleton versus multiple gestations, symptomatic versus asymptomatic women, intact membranes versus ruptured membranes, and prior PTB versus no prior PTB, as well as many others. Because TVU CL has different predictive characteristics in these different patient populations, they should be evaluated separately. Currently, the most appropriate groups to be considered for TVU CL screening are :




  • Asymptomatic singleton pregnancies without prior PTB



  • Asymptomatic singleton pregnancies with prior PTB



  • Asymptomatic multiple pregnancies



  • Symptomatic singleton pregnancies




FIG 28-1


Normal cervical length as measured by transvaginal ultrasound.


The sensitivity and positive predictive value (PPV) of a short CL is very different, for example, in women with or without a prior PTB. In singleton gestations with no history of sPTB, the sensitivity of a short CL for subsequent PTB is approximately 35% to 45% and the PPV is approximately 20% to 30%, which means that the majority of women with a short cervical length will deliver at 35 weeks or later. The sensitivity of TVU CL is also approximately 35% in twins. However, the sensitivity of a short CL is much higher, about 70%, in a singleton pregnancy with a prior sPTB.




Risk Factors for Cervical Insufficiency


Based largely on the epidemiologic associations between the clinical diagnosis of cervical insufficiency and antecedent historic factors, numerous risk factors for CI have been suggested. Risk factors include prior cervical destructive surgery (i.e., trachelectomy, loop electrosurgical excision procedure [LEEP], large loop excision of the transformation zone [LLETZ], laser conization, or cold-knife cone biopsy), in utero diethylstilbestrol (DES) exposure, prior induced or spontaneous first- and second-trimester abortions, uterine anomalies, multiple gestations, or even prior sPTBs that did not meet typical clinical criteria for cervical insufficiency. Because DES usage was effectively abandoned in the early 1970s, this congenital risk factor is now only of historic interest.


Conversely, forced cervical dilation, which generally accompanies midtrimester dilation and evacuation, has been associated with midtrimester spontaneous loss in older studies. A review by Atrash and colleagues of literature published prior to 1987 indicated a nominal relative risk of 3. The authors appropriately recognized the limitations of older investigations that did not include proper control groups or consider important confounders. Particularly unclear was whether and to what extent the recorded spontaneous abortions were characterized by clinical circumstances consistent with CI. Women whose first pregnancy was terminated by an induced abortion, generally by vacuum aspiration, do not appear to have a significant risk for CI but do have a risk of PTB. A recent series of midtrimester surgical abortion after cervical dilation with osmotic dilators up to 24 weeks’ gestation did not confirm the risks observed in the older literature ; however, the investigators speculate that their primary use of osmotic dilators, as opposed to mechanical dilators, may have diminished the associated risks.


Because trachelectomy is now rarely performed in reproductive-age women, the associated risks in contemporary practice are difficult to ascertain; however, older series confirm an appreciable risk of early PTB. Whether LEEP, LLETZ, and cervical conization (cold knife or laser) are significant risk factors for CI is still debated because the relevant outcome data are both contradictory for a significant pregnancy effect and investigators have generally focused on the risk of PTB and not CI per se. These associations are further confounded by the presence of and risk factors for intraepithelial neoplasia in addition to the type of procedure used; the amount of cervical tissue removed, especially after repeat procedures; and the residual CL. Moreover, because almost all studies are retrospective and subject to considerable bias, existing evidence must be cautiously interpreted. Given the increasingly widespread use of LEEP/LLETZ and the obvious public health concerns, defining its associated risk is of particular importance.


Most but not all investigators have reported an increased risk of PTB before 37 weeks after LEEP or conization. In a meta-analysis that incorporated data from 27 studies, cold knife conization was associated with a 2.6-fold increased risk of PTB before 37 weeks (95% confidence interval [CI], 1.80 to 3.72). Women with prior LEEP were found to have a 1.7-fold increased risk of PTB (95% CI, 1.24 to 2.35) and a nearly threefold increased risk of preterm premature rupture of membranes (PPROM; 95% CI, 1.62 to 4.46); no increased risk was observed after laser ablation. A subgroup analysis that controlled for age, parity, and smoking also revealed an increased relative risk for preterm delivery after LEEP (relative risk [RR], 2.10; 95% CI, 1.34 to 2.69). Two studies that utilized large Finnish registries reported an increased risk of PTB after LEEP or conization. The first one reported a nearly threefold increased risk for PTB after repeat LEEP, consistent with other literature. The second noted a 2.9 relative risk (95% CI, 2.2 to 3.7) of very preterm delivery (28 to 31 weeks) and a 2.1 relative risk (95% CI, 1.47 to 2.99) of extremely preterm delivery (<28 weeks) following cervical conization in a population of over 25,000 women who had undergone surgical treatment of cervical intraepithelial neoplasia (CIN). Possibly the most compelling association between surgical treatment of intraepithelial neoplasia and midtrimester birth comes from a recent systematic review and meta-analysis. These investigators identified 14 studies that reported early pregnancy outcomes, eight of which included midtrimester births. Although seven of these did not individually report a significant association, the inclusion of a large population-based study from Norway confirmed a highly significant association with a pooled risk ratio of 2.6 (95% CI, 1.5 to 4.7). The authors suggested that this increased risk was the result of cervical insufficiency after proportionally large excisions but could not confirm that other components of the sPTB syndrome were excluded.


Analogous to the findings of Sadler and colleagues, several investigators have studied the effect of actual specimen size from LEEP or conization as a better predictor of PTB. Leiman concluded that the risk of PTB was increased only when the maximum cone height was greater than 2 cm or the volume was greater than 4 mL. Raio and colleagues performed a matched cohort study of 64 women who had undergone prior laser conization and observed no difference in the incidence of PTB compared with their controls (9.4% vs. 4.7%). They further reported statistically similar birthweights and gestational ages at delivery. However, in a secondary analysis, a laser cone height greater than 10 mm was a significant independent risk factor for PTB. In a more recent retrospective study of 321 pregnancies following LEEP, a threefold increase in PTB at less than 37 weeks was observed if the excision volume exceeded 6 mL (95% CI, 1.45 to 5.92) or if the thickness of the excised tissue was greater than 12 mm (95% CI, 1.27 to 7.01).


Some investigators have proposed that PTB risk increases merely in the presence of cervical dysplasia, suggesting that the indication for the procedure—rather than the surgical procedure itself—is the more important biologic risk factor. In a recent meta-analysis that compared 6589 patients with a history of LEEP to over 1 million patients without a history of excision, Conner and associates reported a modestly increased risk of PTB before 37 weeks following LEEP (RR, 1.61; 95% CI, 1.35 to 1.92). However, on further analysis, those with prior LEEP had a risk of PTB similar to those with known dysplasia who had never undergone surgical excision. In a recent retrospective study that utilized a Finnish registry of approximately 450,000 patients, prior LEEP was associated with a 1.61-fold increased risk for preterm delivery before 37 weeks, but the severity of CIN did not increase this risk. Instead, the investigators reported a twofold increased risk (95% CI, 1.5 to 2.9) of PTB in those who underwent LEEP for non-CIN lesions confirmed on histopathology (e.g., condyloma acuminatum); these data suggest that surgery does pose an independent risk.


Poon and colleagues performed a secondary analysis of data from a clinical trial of PTB prevention and suggested that the risk of PTB from prior LEEP may be better reflected by a midgestation CL. In the study, 26,867 women undergoing routine prenatal care had transvaginal sonographic CL screening performed between 20 and 24 6/7 weeks of gestation. In 473 women who had a previous LEEP, the rate of preterm delivery before 34 weeks was increased (3.4% vs. 1.3%, P = .0002), and the median CL was significantly shorter (32 mm; interquartile range [IQR], 27 to 38 mm) versus 34 mm (IQR, 30 to 39 mm; P < .0001). After controlling for cervical length, prior LEEP was not a significant predictor of PTB.


Regarding history-indicated “prophylactic” treatment of presumptive insufficiency following cervical surgery, Kuoppala and Saarikoski retrospectively reviewed 62 women who had undergone cone biopsy and an equal number of matched control patients. The pregnancy outcomes of the 22 who underwent history-indicated cerclage were similar to those managed without cerclage, with fetal salvage rates of 97% and 100%, respectively. On the basis of their findings and a review of seven other published reports, they concluded that history-indicated cerclage should not be routinely recommended for this possible risk factor. Of note, in the largest published randomized trial of history-indicated cerclage for risk factors, summarized later, women with one or more cone biopsies or cervical amputations had an overall PTB rate before 33 weeks of 35%; however, in this population, no benefit was reported from history-indicated cerclage placement.


In summary, the weight of available evidence confirms an association between the surgical treatment of CIN and spontaneous PTB. However, it should be noted that these data do not confirm a disproportionate incidence of midtrimester loss with a clinical presentation consistent with the clinical diagnosis of cervical insufficiency. Women who have had a large cone specimen removed that included cervical amputations and those who have undergone multiple LEEP or conization procedures are the ones most at an increased risk of sPTB and presumably earlier deliveries. Whether history-indicated cerclage would be an effective preventive strategy in these at-risk women remains speculative. Because the available clinical trial data do not suggest a benefit from history-indicated cerclage, these women may be followed for evidence of premature cervical changes suggestive of insufficiency. Women with a history of prior cervical surgery and recurrent spontaneous midtrimester loss that suggests a clinical diagnosis of insufficiency should be considered for a history-indicated cerclage in future pregnancies. Those with a history of LEEP or conization without a history of prior spontaneous midtrimester loss may be considered as candidates for clinical surveillance.




Tests for Cervical Insufficiency


Because cervical insufficiency is likely part of a broader preterm parturition syndrome, there are few proven objective criteria other than perhaps a rare, gross cervical defect. For this reason, along with the greater research initiatives into PTB prevention, considerable effort has been devoted to developing objective and reproducible criteria for cervical insufficiency.


Most of the earlier reported tests for cervical insufficiency were based on the physiology of the internal os in the nonpregnant state and are of historic interest. Attempts at objective assessments include passage of a #8 Hegar dilator into the nonpregnant cervical canal without resistance, diminished force required to dislodge a Foley catheter whose balloon was placed above the internal os and filled with 1 mL of saline solution, and assessment of cervical elastic properties.


All such attempts at providing an objective diagnosis of cervical insufficiency failed because the tests were not evaluated with regard to standard characteristics (e.g., sensitivity, specificity) against some reference standard for the diagnosis. Moreover, none of these tests could reasonably predict pregnancy-associated conditions that would lead to premature ripening and cervical dilation (i.e., functional insufficiency). Finally, because no universally applicable standard exists for the diagnosis of cervical insufficiency, and because the results of such tests were never evaluated and linked to a proven effective treatment, their clinical utility was at best theoretic. Because no test for cervical insufficiency in the nonpregnant patient has been validated, none of these tests are in common use today.


Although the literature devoted to screening for cervical insufficiency has been dominated by cervical length measurements, there is far more of importance beyond length in assessing the pregnant cervix. Multiple approaches have been described to assess the cervical microstructure, encompassing techniques to evaluate tissue hydration, collagen structure, and tissue elasticity. In a recent review by Feltovich and colleagues, several of these techniques were designated as having high clinical promise in assessment of the cervix, including Raman spectroscopy, backscattered power loss, and shear wave speed. Currently, these innovative modalities are still in the early stages of development. However, they may ultimately improve our understanding of the physiology of the cervix and may have a role in the clinical prediction of premature cervical ripening.




Clinical Diagnosis of Cervical Insufficiency


Patient History


Cervical insufficiency has primarily been a clinical diagnosis characterized by a history of recurrent painless dilation and spontaneous midtrimester (16 to 24 weeks) birth of a nonanomalous living fetus that usually suffers neonatal death or serious long-term morbidities from extreme prematurity. The diagnosis is usually retrospective and made only after poor obstetric outcomes have occurred. On occasion, a patient will be seen who has experienced painless cervical dilation before delivery. In such cases, careful documentation of what has happened is essential. In most cases, however, a careful history and review of the past obstetric records are crucial to the diagnosis. However, in many instances, the records are incomplete or unavailable, and many women cannot provide a reliable history. Even with excellent records and accurate history, clinicians might disagree except in the most classic cases. Confounding factors in the history, records, or physical examination might be used to either support or refute the diagnosis based on their perceived importance.


As noted above, the physician managing a patient who experiences a spontaneous midtrimester birth is in the optimal position to assess and document whether the clinical criteria for cervical insufficiency were met (e.g., hourglassing membranes without painful regular uterine contractions) and to rule out other causes of midtrimester birth (e.g., placental abruption, antecedent fetal death, or fetal anomaly). However, because the preterm parturition syndrome includes other anatomic components, it is possible that some cases of cervical insufficiency are preceded by premature chorioamnion rupture or develop clinically apparent uterine activity. Although intrauterine infection might conceivably exclude the diagnosis, if cervical ripening and occult dilation have caused loss of the mucus plug, the normal barrier between the vaginal flora and chorioamnion is disrupted, and infection might be the presenting clinical event. Although insufficiency is generally considered a diagnosis of exclusion, precisely how other causes of PTB should be excluded has never been explicitly defined.


Because cervical insufficiency is generally a retrospective diagnosis that depends on a history of poor outcomes, clinicians have sought criteria that might lead to a prospective and more objective diagnosis. In women considered to be at risk for cervical insufficiency based on an atypical history, prior cerclage for questionable historic indications, or because of other identified risk factors as mentioned above, serial examinations may be performed to detect progressive shortening and dilation, leading to a presumptive diagnosis of insufficiency, which may then be amenable to therapeutic intervention ( Box 28-1 ). Use of cerclage merely for the presence of various risk factors, as described above, is not recommended.



Box 28-1

Diagnostic Criteria for the Clinical Diagnosis of Cervical Insufficiency





  • History of (recurrent) painless cervical dilation and midtrimester birth



  • Painless midtrimester cervical shortening and dilation detected by serial digital evaluations






Sonographic Diagnosis of Cervical Insufficiency


Over the past several decades, numerous investigators have suggested that cervical insufficiency can be diagnosed by midtrimester sonographic evaluation of the cervix. Various sonographic findings that include CL, funneling at the internal os, and dynamic response to provocative maneuvers (e.g., fundal pressure) have been proposed ( Fig. 28-2 ). In these earlier reports, the sonographic evaluations were not blinded, which led to uncontrolled interventions and difficulty determining their value. Diagnostic criteria were disparate and, in some cases, were not described in a quantitative or reproducible manner.




FIG 28-2


Short cervix (calipers) by transvaginal sonogram at 20 5/7 weeks’ gestation. Note the presence of a biofilm (sludge) that has been associated as a potential marker for subclinical infection.

(From Romero R, Kusamovic JP, Espinoza J, et al. What is amniotic fluid “sludge”? Ultrasound Obstet Gynecol. 2007;30:793-798.)


Later, large, blinded observational studies that used reproducible methods were published. These investigators reported the relationship between midtrimester cervical sonographic findings and PTB and sought to identify cervical changes that indicated a dominant cervical etiology and the possibility for intervention. The Eunice Kennedy Shriver National Institutes of Child Health and Human Development (NICHD) Maternal-Fetal Medicine Units (MFMU) Network completed a study of 2915 unselected women with a singleton pregnancy who underwent a blinded cervical sonographic evaluation at 22 to 24 weeks’ gestation: the relative risk of sPTB increased steadily as the measured CL shortened. In spite of this highly significant relationship, as a test for predicting spontaneous PTB at less than 35 weeks, a CL cutoff of less than 26 mm (the population 10th percentile) had low sensitivity (37%) and poor PPV (18%). Midtrimester birth consistent with cervical insufficiency was not specifically reported, but examination of the survival curve suggests that less than 5% of women with a CL of 25 mm or less at 22 to 24 weeks delivered prior to 28 weeks. Moreover, the circumstances of these early births were not reported, so clinical criteria for insufficiency could not be assessed. Thus in an unselected population, shortened CL at 22 to 24 weeks appeared to be a very ineffective tool for identifying women with cervical insufficiency.


In a subsequent study, the NICHD MFMU Network examined the utility of cervical ultrasound as a predictor of sPTB at less than 35 weeks in high-risk women, defined as those having at least one prior sPTB at less than 32 weeks. Women with a clinical diagnosis of cervical insufficiency were excluded. Beginning at 16 to 18 weeks of gestation, 183 gravidas underwent serial, biweekly sonographic evaluations until the 23rd week of gestation. The study design permitted analysis of the shortest observed CL over time, which also included any fundal pressure–induced or spontaneously occurring CL shortening. As in the previous study, a highly significant inverse relationship was found between CL and spontaneous PTB. However, in this high-risk population, at a CL cutoff of less than 25 mm, the sensitivity increased to 69%, and the PPV increased to 55%. Importantly, a secondary analysis of the data suggested that these high-risk women with a shortened CL might have a clinically significant component of diminished cervical competence because a preponderance of midtrimester births occurred at less than 27 weeks in this group.


These reports support the concept of shortened CL as a surrogate measure of cervical competence, but the actual identification of an appropriate CL cutoff that should lead to clinical intervention and confirmation of the potential contribution of related cervical sonographic findings (e.g., funneling at the internal os) remains problematic. Clearly, cervical sonography has a lower sensitivity (about 30% to 40%) as a screening test in low-risk women (i.e., singleton gestations without prior sPTB), but it appears to have significant utility for identifying clinically significant cervical pathology in women with a prior early sPTB.


Although some investigators have included women with various risk factors in their study populations composed primarily of women with previous sPTB, the results could not be subcategorized because of small sample sizes. However, in a recent series of 64 women with various uterine anomalies, the authors observed an overall preterm delivery rate at less than 35 weeks of 11% and found a significant relationship between CLs of less than 25 mm and PTB, with summary predictive values similar to those of other high-risk populations.


Use of cervical ultrasound in twin gestations has also been reported. However, the screening test characteristics, especially sensitivity and PPV (<40%), appear to be generally lower than for women with a prior early sPTB. A systematic review summarized the predictive value of vaginal sonography for PTB in 46 published series of both asymptomatic and symptomatic gravidas carrying singleton or twin gestations. Eleven series of asymptomatic women with twins were included. The relationship between cervical shortening and PTB was much stronger in the singleton, as opposed to the twin, populations in both receiver operator curve and meta-analytic assessment of likelihood ratios. Because the management of twins with shortened CL has been particularly problematic, it is unclear how we should be using CL measurements in multiple gestations outside of clinical trials ( Box 28-2 ).



Box 28-2

Criterion for the Sonographic Diagnosis of Cervical Insufficiency





  • Shortened cervical length (<25 mm) detected by midtrimester sonography in women with singletons and a history of spontaneous preterm birth at <34-37 weeks






Diagnosis of Cervical Insufficiency on Physical Examination


Uncommonly, a patient in the midtrimester will present with vague pelvic symptoms such as increased pressure and vaginal discharge with increased urinary frequency but without other symptoms of urinary tract infection. Physical exam with a speculum reveals a relatively uneffaced but dilated cervix (at least 1 to 2 cm but generally less than 5 cm) and membranes visible at or beyond the level of the external os. Fetal parts or the umbilical cord might be seen behind the membranes or even contained in the prolapsing sac. Significant uterine contractions and clinical evidence of intrauterine infection (e.g., fever, uterine tenderness) are absent, and a period of observation and monitoring for overt infection and labor are generally performed to establish a diagnosis. This is termed acute cervical insufficiency . Presumably, these findings were antecedent events in most if not all cases of midtrimester birth later attributed to cervical insufficiency based on historic criteria. However, this presentation provides a unique clinical opportunity to witness the natural history of cervical insufficiency, explore possible etiologies, and consider the effectiveness of different interventions ( Box 28-3 ).



Box 28-3

Criteria for the Diagnosis of Cervical Insufficiency on Physical Examination





  • Midtrimester cervical dilation and membranes visible at or beyond the external os in the absence of clinically defined antecedent labor or overt intrauterine infection



  • Significant (serial) asymptomatic cervical dilation detected by palpation at midtrimester






Treatment: Cerclage


The contemporary mainstay of treatment has been a surgical approach using one of the classic cerclage procedures, although both medical treatments and other mechanical supportive therapies have been used.


Cerclage Technique


The proper placement of a suture in the uterine cervical stroma and/or around its perimeter to treat cervical insufficiency or prevent preterm delivery has been the subject of debate because of the lack of proper randomized studies to assess different perioperative strategies. Techniques based on indications are described below.


History-Indicated Cerclage


Women with a history of recurrent midtrimester losses characterized by painless cervical dilation in the absence of overt labor or placental abruption are considered candidates for history-indicated cerclage. This may also be utilized in women with a prior history of cerclage placement secondary to painless cervical dilation in the midtrimester; that is, those with a prior cerclage based on findings on physical examination. History-indicated cerclage placement is generally performed at 12 to 15 weeks of gestation to avoid the complications of early spontaneous loss often attributable to chromosomal abnormalities.


In 1950, Lash and Lash described repair of the cervix in the nonpregnant state that involved partial excision of the cervix to remove the area of presumed weakness. Unfortunately, this technique was associated with a high incidence of subsequent infertility. In 1955, Shirodkar reported successful management of cervical insufficiency with the use of a submucosal band. Initially, he used catgut as suture material, and later he used Mersilene (Ethicon, Somerville, NJ) placed at the level of the internal cervical os. The procedure required anterior displacement of the bladder in an attempt to place the suture as high as possible at the level of cervical internal os. This type of procedure resulted in a greater number of patients being delivered by cesarean delivery because of the difficulty in removing the suture buried under the cervical surface, and it often required leaving the suture in place postpartum. Several years later, McDonald described a suture technique in the form of a purse string that did not require cervical dissection and that was easily placed during pregnancy. This technique involves taking four or five bites as high as possible in the cervix, trying to avoid injury to the bladder or the rectum, with placement of a knot anteriorly to facilitate removal ( Fig. 28-3 ). Most randomized studies of cerclage have used the McDonald technique. A randomized controlled trial (RCT) based in Europe utilizing the Shirodkar technique did not demonstrate benefit. Therefore, owing to its simplicity and effectiveness, a McDonald technique is recommended as the first-line procedure.




FIG 28-3


Placement of sutures for McDonald cerclage. A, Double-headed Mersilene band with four bites in the cervix, avoiding the vessels. B, The suture is placed high up in the cervix, close to the cervicovaginal junction, approximately at the level of the internal os.


Several types of suture material have been used, and no randomized studies had been done to compare the different types of cerclage suture and/or the preferable type of needle to use. We have been successful in using a Mersilene tape. However, the use of thinner suture material, such as Prolene (Ethicon) or other synthetic nonabsorbable sutures like Ethibond (Ethicon), is advocated by others with the argument that the width of the Mersilene tape places the patient at greater risk for infection. Currently, no evidence suggests that placing two sutures results in better outcomes than placing one. A recent large retrospective study did not show a difference in PTB rates between the two- versus one-stitch groups. Furthermore, placement of a second occlusive suture at the level of the external os to keep the mucous plug in place has not shown benefit in a recent RCT.


Preoperative patient preparations that include the use of prophylactic antibiotics or tocolytics have not been proven to be of benefit, and evidence is insufficient to recommend their use. The value of performing cultures before the cerclage procedure has not been properly studied, and this practice—as well as giving perioperative antibiotics—should be individualized. Similarly, the use of amniocentesis prior to cerclage has not been subjected to RCTs. It appears that for a history-indicated cerclage, the rate of subclinical intraamniotic infection (IAI) is very low, and amniocentesis does not seem justifiable. However, for ultrasound-indicated cerclage, the rate of subclinical IAI could be as high as 1% to 2%. The value of amniocentesis in this setting needs to be evaluated by an RCT.


The choice of anesthesia for cerclage varies. Chen and colleagues did not observe a difference in outcome between general versus regional anesthesia. In our experience, a short-acting regional anesthetic is sufficient, and spinal anesthesia appears to be the preferred anesthesia for cerclage.


The value of bed rest postoperatively has been questioned and even criticized, and its value in the setting of cerclage placement has not been studied. Decisions regarding physical activity and intercourse are individualized and are based on the status of the cervix as determined by outpatient digital evaluation or sonographic findings ( Fig. 28-4 ).




FIG 28-4


Transvaginal sonogram of the cervix after cerclage placement. The internal os is closed, and no funneling is evident. Echogenic spots in the cervix correspond to cerclage ( arrows ).


The approach to the patient with preterm contractions and a cerclage in place has not been sufficiently evaluated. Most authorities will only remove the cerclage if there is tension in the suture because of the concern for cervical lacerations if the suture is left behind. The suture is usually removed electively at 36 to 37 weeks. Systematic review of the literature showed that planned removal of cervical cerclage before the onset of labor at 36 to 37 weeks is associated with a 28% nonsignificant reduction in the incidence of cervical lacerations compared with removal in labor (6.4% vs. 11.4%; RR, 0.72; 95% CI, 0.35 to 1.49). The mean interval between cerclage removal and spontaneous delivery is 14 days.


In patients with a prior failed vaginal cerclage, an abdominal cerclage has been recommended. A prior history-indicated vaginal cerclage resulting in an sPTB before 33 weeks (i.e., a failed cerclage) is currently the only indication proven by controlled studies to show benefit from a transabdominal cerclage compared with a repeat transvaginal cerclage. This procedure is usually done either in pregnancy between 11 and 12 weeks gestation or prior to conception. The procedure can be done either with a laparotomy or using a laparoscopic or robotic approach. A bladder flap may or may not be created. The procedure requires the placement of a Mersilene tape at the level of the junction between the lower uterine segment and the cervix lateral to the uterus and medial to the uterine vessels ( Fig. 28-5 ). Greater morbidity, including injury to the uterine vessels, requires expertise when performing this procedure. In our experience, we have found it helpful to have an assistant provide fundal traction, while the surgeon grasps the uterine vessels and retracts them laterally to expose an avascular space between the artery and the cervix. A right-angle clamp is passed anterior to posterior through this avascular space, tenting and incising the posterior leaf of the broad ligament and grasping a Mersilene tape brought back through the space. The same procedure is repeated on the opposite side, and the tape is tied anteriorly ( Figs. 28-6 through 28-9 ). Several investigators have reported extensive experience with this procedure with low morbidity and favorable outcomes. Cesarean delivery is necessary, and the suture is left in place if future fertility is desired.


Mar 31, 2019 | Posted by in OBSTETRICS | Comments Off on Cervical Insufficiency

Full access? Get Clinical Tree

Get Clinical Tree app for offline access