During pregnancy, several conditions may necessitate operative procedures on the vulva, vagina, or cervix. Of surgeries, cervical cerclage is one of the more common. Other procedures are used during delivery and include Dührssens incisions, division of a vaginal septum, and release of female genital mutilation scarring. A brief review of procedures relevant to cervical dysplasia and cervical polyps concludes the chapter.
The primary function of the cervix during pregnancy is to keep the uterus and its contents sequestered until controlled cervical dilatation and delivery ensues at term. Failure of this function may result in preterm birth. Thus, to reinforce an insufficient cervix, cerclage procedures are often performed.
When the cervix fails because of an intrinsic weakness during the midtrimester, it has been historically referred to as cervical incompetence. Today, the term cervical insufficiency is preferred to avoid negative connotations. The intrinsic cervical defect classically results in painless dilatation of the cervix with pregnancy loss during the midtrimester. Easterday and Reid (1959) described this process: “The cervix in these patients usually dilates without discomfort, over a period of days or possibly weeks, to the point where the membranes are plainly visible on speculum examination. Unless this is recognized early, the membranes will rupture, and the pregnancy will terminate prematurely.” However, overreliance on this classic history may impede the diagnosis of cervical insufficiency. In fact, early symptoms frequently develop and include urinary frequency and urgency, lower abdominal pressure, or watery discharge (Toaff, 1974). After rupture of membranes, this process may become overtly painful due to contractions, further distention of the cervix, and passage of the uterine contents. Such devastating early losses often recur in subsequent pregnancies, and this supports the concept of intrinsic cervical deficiency.
The etiology of this cervical deficiency has been debated and may stem from either congenital or acquired defects. Given that 25 percent of women with a history of cervical insufficiency have a first-degree relative with the condition, a genetic factor seems very plausible (Warren, 2009). As putative elements, extracellular matrix components and several genes have been studied in affected women. Notably, women with prior cervical insufficiency do not have intrinsically low collagen levels within the extracellular matrix, nor do they appear to have an inferior quality of collagen or an excessive number of smooth muscle cells (Oxlund, 2010). Although polymorphisms in certain genes associated with inflammation and collagen metabolism have been identified in women with cervical insufficiency, their role in intrinsic cervical deficiency remains unclear (Warren, 2009).
During pregnancy, the biochemistry and structure of the cervix undergo important changes. These alterations include significantly decreased stromal stiffness, greater water content, increased sulfated glycosaminoglycan content, increased collagen solubility, and decreased collagen organization (Myers, 2008, 2009). These changes occur early, typically within the first 4 to 6 weeks of pregnancy.
In preparation for labor, further changes develop in the cervix, and these may differ in women with preterm labor. With cervical ripening in both term and preterm cervices, there is increased transition from high-molecular weight glycosaminoglycans to low-molecular weight hyaluronan. Endocervical levels of hyaluronan at the time of ultrasound-indicated cerclage performed between 15 and 25 weeks are higher in women who delivered preterm as opposed to those delivering at term (Eglinton, 2011). In preterm cervices, this transition to lower-molecular-weight hyaluronan is associated with increased Has2 (hyaluronan synthase) gene activity. This contrasts with term cervices, in which Has1 dominates (Akgul, 2012).
Since 1978, numerous cases of cervical insufficiency associated with in utero diethylstilbestrol (DES) exposure have been reported (Goldstein, 1978; Mangan, 1982; Sandberg, 1981; Singer, 1978). Jefferies and colleagues (1984) studied 367 women with complete records regarding DES exposure. They noted that anomalies were strongly tied to gestational week at first dose and total dose. In addition to cervical abnormalities, uterine malformations are also common among women with DES exposure in utero. However, outcomes are not improved by cervical cerclage placement in women with DES exposure in utero (Kaufman, 1984). The Food and Drug Administration made pregnancy a contraindication to use of DES in 1971, and greatly restricted its use as a postcoital contraceptive in 1975. Despite this, DES use continued in several countries, and thus providers should remain vigilant for DES-exposed patients.
Among women with uterine malformations, cervical length has been reported to be shorter. For example, Crane and coworkers (2012) reported that cervical length was significantly shorter in women with bicornuate (3.46 cm) or unicornuate (2.20 cm) uteri compared with low-risk controls (4.32 cm). Women with a uterine malformation were more likely to deliver preterm. Namely, 14 percent delivered before 35 weeks, and 27 percent delivered before 37 weeks compared with 3.3 percent of controls. However, the cause of the increased preterm birth rate is unclear. Distention of a malformed uterus or an intrinsic cervical deficiency associated with uterine malformations has been implicated.
Given this increased risk of preterm birth, a handful of studies have sought to determine the utility of cerclage among women with uterine malformations. In one retrospective study, 30 of 88 women with a prior second-trimester loss had müllerian anomalies of the upper genital tract (Ayers, 1988). These included 12 arcuate, 10 septate, and eight bicornuate uteri. Of these 30 women, 24 (80 percent) had cervical shortening seen sonographically and underwent cerclage. The remaining six women did not receive a cerclage. Good outcomes were reported in both cohorts. Moreover, among the total population of 88 women, 68 of the 70 women who underwent cerclage delivered after 35 weeks, and of those 18 who did not have cerclage, 95 percent delivered after 35 weeks.
Of specific müllerian defects, Yassaee and Mostafaee (2011) reported outcomes in 40 women with bicornuate and arcuate uteri. Women who had a bicornuate uterus and who underwent cerclage had higher term delivery rates than those with a bicornuate uterus and no cerclage—76 percent versus 27 percent, respectively. In contrast, such rates were not significantly different among women with an arcuate uterus and with or without a cerclage. In light of the heterogeneity of uterine malformations and the scant data regarding cerclage in affected women, it remains unclear whether this surgery improves outcomes for women with uterine anomalies in general.
For decades, published data have been conflicted with respect to rates of second-trimester losses and preterm birth among women who have undergone cervical conization. Recent literature indicates that this controversy continues. For example, Fischer and coworkers (2010) compared 85 women with a history of cold-knife conization or loop electrosurgical excision procedure (LEEP) and 85 matched controls. Although the average cervical length of women with a history of either procedure was shorter than that of controls—3.3 versus 3.9 cm—the fraction of women with cervical lengths <2.5 cm did not significantly differ between the groups. No woman with a cervical length <2.5 cm delivered prior to 34 weeks, and there were no second-trimester losses consistent with cervical insufficiency. In contrast, Armarnik and associates (2011) analyzed 53 cases of conization—mostly LEEP—among 104,670 deliveries. They found a significantly increased risk of preterm birth before <34 weeks’ gestation among the women who had this surgery.
When limited to only include LEEP, controversy exists with respect to the subsequent preterm birth risk. Heinonen and colleagues (2013) found that LEEP—especially repeat procedures—increased the risk of preterm birth. The interval between procedures did not affect this risk. Other studies suggest that short intervals may adversely affect outcomes (Ciavattini, 2015; Conner, 2013). In a study of 241,701 women who were delivered of singleton pregnancies at Parkland Hospital, 511 women had previously undergone LEEP and another 842 had undergone LEEP after the index pregnancy (Werner, 2010). In this group of treated gravidas, the risk of preterm birth was not increased compared with that of the general obstetric population.
Given the lack of agreement as to whether conization increases preterm birth risk, it is no surprise that cerclage has not been shown to benefit such women, and it may cause harm. Zeisler and associates (1997) compared 30 women with prior conization who underwent cerclage with 39 control women who did not receive cerclage. They reported no significant difference in the preterm birth rates (23 and 21 percent, respectively). However, women who underwent cerclage were twice as likely to be hospitalized for preterm labor—67 percent versus 33 percent, respectively. These authors recommended that prophylactic cerclage should be used more sparingly in women with a history of conization. We agree.
Several recent studies indicate that pregnancy termination confers a small but statistically significant risk of cervical incompetence and preterm birth. Scholten and coworkers (2013) studied a perinatal database encompassing women with singleton pregnancies that reached at least 20 weeks. Within this database, 16,000 women (1.2 percent) had undergone pregnancy termination, and 90 to 95 percent of these cases involved a surgical method of evacuation. Among this subgroup of women, rates of cervical incompetence treated by cerclage were significantly higher, with an adjusted OR of 1.11. In their review of 37 studies, Shah and associates (2009) found that a single induced pregnancy termination confers a small increased risk for low birthweight and preterm birth but not delivery of a small-for-gestational-age neonate. More than one induced pregnancy termination further increases the risk of low birthweight. Finally, Kilpatrick and colleagues (2006) reported that prior termination before <20 weeks’ gestation was associated with previable ruptured membranes and cervical dilatation. Despite this apparently increased risk of preterm birth among women with a prior pregnancy termination, no study has shown a benefit from cerclage. Similar to women with previous conization, expectant management is reasonable for women with a prior pregnancy termination.
The main surgery for cervical insufficiency involves placement of a cerclage. The timing of this procedure results in three categories of cerclage: ultrasound-indicated, prophylactic (history-indicated), or rescue cerclage. In many cases, these overlap.
Of these, an ultrasound-indicated cerclage is currently favored for women who have a prior loss suggesting cervical insufficiency and who develop cervical shortening during sonographic surveillance. Prophylactic cerclage is generally restricted to women with histories strongly suggesting cervical insufficiency. At our institution, we generally offer a prophylactic cerclage to women who have had a prior cerclage and to women with a history of at least one midtrimester loss that appears attributable to cervical insufficiency. Last, the concept of rescue cerclage, alternatively termed an emergency cerclage, is not well defined. The term is generally used when the cervix is dilated and membranes are visible or prolapsing during the mid-second trimester. Importantly, inevitable abortion or preterm labor is excluded in such cases prior to cerclage placement. If not, the force of uterine contents against cerclage stitches during contractions can ultimately tear the cervix and lead to significant cervical trauma and bleeding.
With the exception of some prophylactic procedures, most cerclages are placed transvaginally. Generally speaking, a transvaginal cerclage is not placed until the second trimester. This is because 15 to 20 percent of pregnancies will abort spontaneously and render a cerclage unnecessary and potentially dangerous from lacerations just described. Prior to cerclage placement, most practitioners sonographically document fetal cardiac activity and exclude obvious severe or lethal fetal anomalies. In particular, many women undergo first-trimester screening with serum analytes, nuchal translucency, and/or free fetal DNA evaluation.
Most authorities do not recommend cerclage after 24 weeks, given the risks for prematurely ruptured membranes and pregnancy loss. At Parkland Hospital, cerclage procedures are generally not done once putative fetal viability is reached after 23 to 24 weeks. Cerclage is not performed in women with a suspected indication for delivery, such as preeclampsia or chorioamnionitis. Preterm labor, active bleeding, or ruptured membranes are other contraindications. Evidence best supports the use of cerclage in singleton pregnancies, and at Parkland Hospital, we currently do not place a cerclage in women with a multifetal gestation. The American College of Obstetricians and Gynecologists (2014) does not recommend the use of cerclage in twin pregnancies.
Several techniques for transvaginal cervical cerclage have been described, including the Lash procedure (1950), Shirodkar cerclage (1955), McDonald cerclage (1957), and the Wurm procedure (Hefner, 1951). In 1965, Benson and Durfee described the transabdominal cerclage, although laparoscopic or robotic placement is increasingly used now. Today, the most commonly used transvaginal procedures are variations of the Shirodkar and McDonald cerclages. Neither appears more effective than the other (Odibo, 2007).
For transvaginal cerclage, no studies guide selection of optimal suture material. In a secondary analysis of a randomized trial evaluating McDonald versus no cerclage, 84 women received polyester braided suture, 46 women received polyester braided tape (Mersilene), and eight women received monofilament suture (Berghella, 2012; Owen, 2009). The women with monofilament suture were excluded from analysis due to the small number. Outcomes did not differ according to suture material.
In his original manuscript, Dr. Vithal Naresh Shirodkar (1955) described a technique using fascia lata. This has subsequently been replaced by polyester braided surgical tape such as Mersilene. The surgical technique remains similar today and is typically performed as outlined in the subsequent steps.
Regional analgesia is suitable and preferred. The woman is then placed in the standard dorsal lithotomy position, typically using candy-cane stirrups. The vagina and perineum are surgically prepared, drapes are positioned, and the bladder is drained. Some authors do not use antiseptic solution on the exposed amnionic membranes because of the irritative effect and instead use warm saline solution (Pelosi, 1990). That said, no evidence indicates that either method yields superior outcomes.
To begin, a weighted Auvard speculum is placed along the posterior vaginal wall, and the cervix is grasped with DeLee ovum or ring forceps anteriorly, posteriorly, and laterally. Right-angle retractors held by an assistant are used to ensure additional exposure. With traction on the ring forceps, the intersection of the cardinal ligament at the lateral aspect of the cervix is palpated and serves to approximate the level of the internal os.
The cervix is then pulled down and a transverse incision is made near the reflection of the bladder (Fig. 11-1). The bladder can then be mobilized cephalad. This is often completed with blunt dissection to the level of the internal os and near where the vesicouterine peritoneal fold is encountered (Figs. 11-2 and 11-3). The cervix is then pulled upward, and a transverse incision is made through the vaginal epithelium across the posterior aspect of the cervix. Similar mobilization to the level of the internal os can then be performed posteriorly.
Allis clamps are placed bilaterally at the sides of the cervix and within the dissected span of cervix (Fig. 11-4). The Allis clamps are closed with the objective of bunching the paracervical vessels within the clamp and pulling them laterally. This allows the needle and suture to pass under the clamp tip while avoiding paracervical vessel puncture (Rust, 1967). A 5-mm Mersilene tape that is armed on each end with a needle is lubricated with sterile gel lubricant. One needle is then advanced from posterior to anterior on one side (Harger, 1980). Importantly, passing the needle too deep risks membrane rupture. The tape ideally lies flat and without twists across the posterior aspect of the cervix. The other needle is then advanced on the opposite side in a similar fashion (Fig. 11-5).
The tape is tied anteriorly to aid later removal, and the free ends are left exposed. As originally described, the anterior and posterior incisions were closed with chromic catgut suture, and the patient was anticipated to deliver by cesarean. To aid later tape removal, the previous edition of this book did not advocate closure of the incisions unless there was bleeding. For those who prefer to close the incisions, a tail of the cerclage suture can be left to protrude through the incision to aid later removal (Fig. 11-6) (Curet, 1980).
In 1957, Dr. Ian McDonald described a simpler technique that placed a purse-string suture around the cervix and did not require bladder mobilization. Of the 70 women in his case series, 33 women were delivered of surviving infants, and another 16 women had their pregnancies extended by at least 4 weeks. The surgical procedure used today is similar to McDonald’s original description, but a nonabsorbable suture is typically used. Options include a no. 1 or 2 monofilament nylon or polypropylene suture, a polyester braided suture, or Mersilene tape.
As with the Shirodkar cerclage, spinal analgesia is recommended by most. The woman is placed in standard dorsal lithotomy position and prepared as described in the previous section. The cervix is grasped with DeLee ovum or ring forceps, and the cervicovesical junction is identified. One marker of this junction is the transition from the rugose epithelium overlying the bladder to smooth epithelium of the portio vaginalis. Identification is aided by gently moving the cervix back and forth.
Four to six bites of a purse-string suture are circumferentially placed as high as possible, beginning at the cervicovesical junction (Figs. 11-7 and 11-8). At the 3 o’clock and 9 o’clock positions, the needle is maneuvered under the lateral cervical vasculature to avoid bleeding. McDonald cautioned against taking bites that are too shallow posteriorly, and these should be deep enough to include the stroma without entering the endocervical canal. The knot is tied anteriorly, typically without completely occluding the endocervical canal. The free ends are left long enough to more easily identify them for later removal.
FIGURE 11-7
McDonald cerclage. The cervix is grasped with DeLee ovum forceps, and the cervicovesical junction is identified at the intersection of the smooth epithelium of the portio vaginalis and the rugose epithelium overlying the urinary bladder. A nonabsorbable suture with attached needle is used to begin a purse-string suture at the 12 o’clock position.
FIGURE 11-8
The needle is maneuvered under the lateral cervical vasculature at the 9 o’clock position to avoid bleeding. One to 2 bites are taken posteriorly and deep enough to purchase adequate cervical stroma without traversing the endocervical canal. This is repeated at the 3 o’clock position. The knot is tied anteriorly with multiple (typically eight) throws when using nonabsorbable suture. The free ends are left long enough to aid their later identification during cerclage removal.
Benson and Durfee (1965) described placement of a transabdominal cerclage for women who are not suitable candidates for transvaginal cerclage. One of the following serves as possible criterion: “(1) an obvious congenitally short or extensively amputated cervix; (2) marked scarring of the cervix—as after previous unsuccessful cervical cerclage; (3) deeply notched multiple cervical defects; (4) unhealed, penetrating, forniceal lacerations; and (5) subacute cervicitis.” Due to the significantly increased risk of bleeding and complications associated with transabdominal cerclage, Rand and Norwitz (2003) recommend that the transabdominal approach be reserved for selected patients. Examples include women with failed transvaginal cerclages or those in whom a transvaginal cerclage is technically impossible to perform because of extreme cervical shortening, scarring, or laceration. The recommendations of Rust and Roberts (2005) are similar.
Placement of a cervicoisthmic cerclage was originally described using laparotomy, but several reports additionally detail laparoscopic or robotically assisted cervicoisthmic cerclages (Foster, 2013; Menderes, 2015; Moore, 2012; Riiskjaer, 2012; Tulandi, 2014). Although transvaginally placed cervicoisthmic cerclages have been described, these are essentially modifications of the Shirodkar technique that seek higher placement (Golfier, 2001; Katz, 2005).
Transabdominal cerclage placement becomes increasingly difficult with an enlarging uterus. Therefore, women ideally undergo cerclage before pregnancy or at 11 to 14 weeks’ gestation, but after initial screening for aneuploidy and obvious malformation is completed (Rand, 2003). Tulandi and colleagues (2014) evaluated 16 studies involving 678 pregnancies. They found that results of placement before pregnancy and during pregnancy were similar, whether performed laparoscopically or by laparotomy. Placement before conception is more practical.
Following induction of general anesthesia, the patient is positioned supine, the abdomen and vagina are surgically prepared, and a Foley catheter is inserted. A Pfannenstiel incision provides suitable access. After entry into the abdomen, the vesicouterine peritoneal reflection is incised, and the vesicouterine space is entered. The loose connective tissue in this space is easily disrupted bluntly or sharply to permit caudal mobilization of the bladder (Fig. 11-9). The bladder is moved to expose the uterine isthmus.
The uterine artery is identified bilaterally, and this is aided by gentle upward traction of the uterine fundus. Invaluable comments by Benson and Durfee (1965) describe the uterine vessels and a potential “free space” that lies between these vessels and the lateral aspect of the uterine isthmus at the level of the internal os. They encourage careful dissection to define this space, while avoiding the ureter, which lies posterolateral (Fig. 11-10). Subsequently, the broad ligament’s posterior leaf is punctured in an avascular area on each side of the uterus. This allows the Mersilene tape to be threaded through the broad ligament and the avascular space at a level just above the cervicoisthmic junction. Passing the tape medially to the uterine vessels avoids their occlusion when the cerclage is tied.
FIGURE 11-10
The uterine vessels will be prominent near the level of the internal os, which is typically approximately 1 cm superior to the insertion of the uterosacral ligaments. In the area of the internal os, a window is made in free space medial to the vessels. This avoids vessel compression by the tightened cerclage. Care is also taken to avoid the ureter, which is lateral and posterior.
The Mersilene strand is then passed around the isthmus and tied. Suture may be passed anterior to posterior, and the knot tied behind the uterus (Herron, 1988). Alternatively, suture ends may be threaded so that they are knotted at the front of the uterus.
With their modification, Debbs and associates (2007) grasp the uterine vessels laterally at the cervicoisthmic junction and create a window with a right-angle clamp at a point 1 cm superior to the uterosacral ligaments. An assistant then passes the 5-mm Mersilene tape on a long Kelly clamp to the open right-angle clamp.
These needleless approaches can minimize the risk of vessel puncture and hemorrhage, particularly during laparoscopic approaches (Menderes, 2015). Alternatively, for laparoscopy, Mersilene tape attached to blunt needles that have been straightened, to enable passage through a trocar, can be used (see Fig. 11-10) (Tusheva, 2012). These authors described their laparoscopic approach to placing the cerclage at the level of the internal os as follows: “The landmarks for this placement include the uterosacral ligaments; a distance of 1.5 cm superior and 1 cm lateral to the insertion of the uterosacral ligament on the posterior uterus is a good initial guide for needle placement. The needles are then cut off and removed, and the Mersilene suture is then tied tightly around the cervix with six knots using intracorporeal knot tying.” The bladder flap is typically replaced over the cerclage with absorbable suture for both open and laparoscopic approaches (Fig. 11-11).
The chief difficulty that distinguishes the emergent or rescue cerclage from those described previously is that the cervix has begun to dilate and efface, and typically the membranes are exposed. Frequently, the membranes are protruding from an open cervix, and numerous techniques for reducing the membranes have been described (Table 11-1). Of these, use of a moist swab or sponge stick is shown in Figure 11-12.
| Method | Investigator |
|---|---|
| Traction on cervical edges | Daskalakis, 2006; Yip, 1998 |
| Moist swab to reduce | Olatunbosun, 1981; Daskalakis, 2006 |
| General anesthesia | MacDougall, 1991; Olatunbosun, 1981; Yip, 1998 |
| Multiple cervical stay sutures | Olatunbosun, 1981 |
| Overfilling the bladder | Scheerer, 1989 |
| Amnioreduction | Locatelli, 1999; MacDougall, 1991 |
| Foley catheter | MacDougall, 1991; Yip, 1998 |
| Inflated balloon | Tsatsaris, 2001 |
| Steep Trendelenburg position | MacDougall, 1991; Olatunbosun, 1981; Tsatsaris, 2001; Yip, 1998 |
| Knee-chest position | Ogawa, 1999 |
FIGURE 11-12
At times, the membranes may be exposed or prolapsed, creating interference and the potential for their inadvertent rupture. Traction with DeLee ovum forceps on the cervix combined with gentle elevation of a moistened sponge swab can push the membranes cephalad. Several other techniques to reduce exposed membranes are listed in Table 11-1.
Several randomized trials have evaluated transvaginal cerclage efficacy. Three of the earliest focused on placement of a prophylactic cerclage based on obstetric history. These are summarized in Table 11-2.
| Study | n | Obstetric History | Outcomes |
|---|---|---|---|
| Rush, 1984 | 194 | History of 2 to 4 pregnancies that ended spontaneously before 37 weeks, with at least 1 pregnancy that ended spontaneously between 14 and 36 weeks | No significant differences for rates of preterm birth before 37 weeks, low birthweight, neonatal mortality or morbidity. Cerclage patients had significantly longer hospitalizations. 18 of 96 women with cerclage and 12 of 98 women without cerclage had ruptured membranes (p = NS) |
| Lazar, 1984 | 506 | Complex multifactor scoring chart: women with a moderate risk of preterm birth (scores ≥9 but <20) were eligible | No significant differences in the rates of preterm birth or perinatal deaths. Women with cerclage were twofold more likely to be hospitalized (p < .001) |
| MRC/RCOG, 1993 | 1292 | Women were eligible if their obstetricians were uncertain whether to advise them to have a cerclage based on prior history | Fewer deliveries before 33 weeks (13 vs 17%, p = .03) and fewer very-low-birthweight neonates (10 vs 13%, p = .05) in the cerclage group, but the overall preterm birth rates were not significantly different (26 vs 31%, p = .07). Rates of miscarriage, stillbirth, and neonatal death did not differ. Hospital admission and tocolysis was more common in the cerclage group, among which there was a twofold increased risk of puerperal fever (p = .03) |
More recent randomized trials have studied sonographic screening of cervical length to guide cerclage placement. As shown in Table 11-3, screened cohorts were diverse and included women at low risk for pregnancy complication and women at risk for preterm birth. Of six prospective trials that investigated primarily singleton pregnancies, the results have been mixed. Four studies reported no benefit, and two showed more favorable outcomes.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
