Cervical cancer is the second most common cancer in female worldwide, and it is the leading cause of gynecological cancer-related morbidity and mortality in developing countries.1 However, well-organized cervical screening programs and the appropriate management of screen-detected intraepithelial lesions has reduced the incidence of cervical cancer by up to 80%.2

Among the treatment options of excisional procedures, large loop excision of the transformation zone, also known as loop electrosurgical excision procedure (LEEP) has become a standard treatment for women with cervical intraepithelial neoplasia (CIN) in the industrialized world.3,4 However, the difference in the excisional procedure may potentially affect the outcomes of subsequent pregnancies, especially the risk of preterm delivery in women with a history of cervical conization.

Cerclage of the uterine cervix has been tried as an effort to reduce the incidence of preterm delivery in patients with a history of conization, but its efficacy still remains unclear. With this background, this study was conducted to evaluate the influence of conization on the following pregnancy, and whether prophylactic cervical cerclage can change the outcome of pregnancy in women with a history of electrosurgical cervical conization.

We retrospectively reviewed the medical records to evaluate the obstetric and neonatal outcomes of all patients with a singleton pregnancy and a history of conization due to CIN or carcinoma in situ (CIS), who were treated and delivered at the Samsung Medical Center between January 2001 and December 2008. In all patients, thorough history including prior preterm delivery as well as complete clinical and physical examinations were done antenatally. Cervical length was measured by transvaginal ultrasonography before cerclage. Although a cervical length of less than 25 mm was considered to be short, the decision of cerclage was made by patients after obtaining explanation of her cervical status and the controversial role of cerclage in preventing preterm delivery after conization. Of the 56 cases, 25 women underwent cerclage (cerclage group) and 31 pregnant women did not want cerclage (expectant group). Pregnancies complicated by fetal death, multiple pregnancy, fetal chromosomal or non-chromosomal anomalies, diabetes, and serious maternal medical diseases were excluded from this study.

The conization procedure is noted as follows. Firstly, a colposcopic examination was taken to identify the margins of the lesion. Then, the electrosurgical conization was performed by using a right-angled triangular loop, carrying a high frequency current which can penetrate tissue more deeply toward the uterine endocervix than the conventional round loop.5 Thereafter, a cold coagulator (120℃) was applied to the cone bed for 30-60 seconds after excision for the purpose of hemostasis and destruction of any residual lesion after electrosurgical excision. Fig. 1 shows an example of a sagittal T2-weighted magnetic resonance image of a large iatrogenic cervical defect caused by electrosurgical conization.

Cervical cerclage was performed prophylactically between 14 and 19 gestational weeks by two maternal-fetal medicine specialists. The McDonald technique was used in all patients in the cerclage group. All of them were given perioperative intravenous antibiotics. The knots of the cerclage were removed regardless of membrane status if a patient went into spontaneous preterm labor. Otherwise, the knots were removed electively at 36-37 gestational weeks and patients were allowed to await spontaneous onset of labor, or induction of labor was proceeded when indicated. Two consecutive doses of corticosteroid (betamethasone) were administered intramuscularly for fetal lung maturation if there was any sign of preterm delivery (i.e., regular uterine contractions and cervical dilatation despite the cerclage, preterm premature rupture of membranes [PPROM]) in patients with gestational age 24-34 weeks. The use of tocolytics was at the discretion of the attending physician when regular uterine contractions were developed. Tocolytics were used after removal of the cerclage when spontaneous preterm labor was noticed regardless of cervical status.

The antenatal parameters analyzed were maternal demographic characteristics, the indication and method of conization, depth of cone, interval between conization and pregnancy, and cervical length before cerclage. Obstetric outcomes analyzed were antepartum bleeding episodes, abortion, frequency of admission due to preterm labor per patient, admission duration for preterm labor per patient, tocolytics and antenatal corticosteroid use, preterm delivery, PPROM, gestational age at delivery, and mode of delivery. Neonatal outcomes were analyzed with respect to birth weight, Apgar scores, necessity and duration of ventilator therapy, and duration of neonatal intensive care unit (NICU) stay. In addition, neonatal morbidities including respiratory distress syndrome (RDS), bronchopulmonary dysplasia (BPD), periventricular leukomalacia (PVL), grade 3-4 intraventricular hemorrhage (IVH), grade 3-4 retinopathy of prematurity (ROP), stage 2-3 necrotizing enterocolitis (NEC), suspected or proven early and late neonatal sepsis, and mortality were analyzed. RDS was diagnosed in the presence of respiratory grunting and retracting, an increased oxygen requirement (FiO₂>0.4) and diagnostic radiographic chest findings. BPD was defined as a need for supplementary oxygen for ≥28 days or by diagnostic radiographic or histological findings. IVH and PVL were diagnosed and graded by ultrasonographic examination of the neonatal brain. IVH was defined as intraventricular bleeding with ventricular dilatation (grade 3) or with parenchymal involvement (grade 4). PVL was defined as the presence of an obvious hypoechoic cyst in the periventricular white matter. ROP was diagnosed by ophthalmologists and its grading was based on the International Classification of Retinopathy of Prematurity.6 NEC was diagnosed on the basis of various clinical and radiological parameters and grouped in 3 stages as per modified Bell's classification.7 The diagnosis of neonatal sepsis was based on the presence of a positive blood culture (proven sepsis) or positive laboratory evidence in clinically suspected neonates (suspected sepsis). Composite morbidity was defined as having more than one of the followings: fetal death, RDS, BPD, PVL, IVH (≥ grade 3), ROP (≥ grade 3), NEC (≥ stage 2), and neonatal sepsis.

Data were presented as percent for categorical variables and median (range) for continuous variables. Fisher exact or chi-square tests were used for the statistical analysis of categorical variables, as appropriate, and Mann-Whitney U tests were used for continuous variables. The results were considered statistically significant when p-values were <0.05.

During the 8-year period of review, 56 cases of patients who met the inclusion criteria of this study were identified among 22,126 deliveries. The rate of preterm delivery was significantly higher in pregnant women having a history of conization than the remainder of the study population (32.1% [18/56] vs. 15.2% [3,355/22,070], p<0.001).

Of the 56 cases, 31 pregnant women were included in the expectant group and 25 women were included in the cerclage group. The demographic characteristics of the patients, indication for conization, cone depth and the interval between conization and pregnancy were similar in the two groups (Table 1). Cervical length of the patients in the conization group was significantly shorter than those in the expectant group. In addition, cervical cerclage rate was higher in patients with cervical length shorter than 25 mm (12/19, 63.2%) compared to those with normal cervical length (13/37, 35.1%), but the difference did not reach statistical significance (p=0.055). The mean interval from admission due to preterm labor (cerclage removal) to delivery in patients who received tocolytics was 3.00±3.69 days in the cerclage group and 6.13±5.94 days in the expectant group (p=0.281).

Obstetric outcomes including frequency of antepartum bleeding episode, admission due to preterm labor, use of tocolytics or antenatal corticosteroid were not significantly different between the two groups (Table 2). The rate of preterm delivery and PPROM and median gestational age at delivery were similar in the two groups. The mode of delivery and indication of cesarean delivery of the two groups were also not significantly different. In accordance to this result, the neonatal outcomes were similar in the two groups (Table 3). Composite neonatal morbidity also did not show statistical difference between the cerclage and the expectant group.

Since short cervical length is a well known risk factor for preterm delivery, and the cervical length of the study subjects were different between the two groups, we re-analyzed the data of 19 patients with short cervical length (less than 25 mm). Seven women were managed expectantly and the other 12 women underwent cerclage operation. Maternal age (expectant group vs. cerclage group, 31.0 years [29 to 35] vs. 30.5 years [27 to 40], p=0.910), depth of cone (2.2 cm [1.5 to 2.5] vs. 2.2 cm [2.0 to 3.2], p=0.321), cervical length (2.0 cm [1.5 to 2.4] vs. 2.0 cm [1.1 to 3.1], p=0.837), interval between conization and pregnancy (24 months [7 to 36] vs. 14 months [4 to 96], p=0.650), gestational age at delivery (37.0 weeks [27.3 to 39.0] vs. 38.3 weeks [25.0 to 39.6], p=0.343), rate of preterm delivery (3/7 [42.9%] vs. 3/12 [25.0%], p=0.617) was not significantly different between patients who underwent cerclage and those managed expectantly. All neonatal outcomes also were not different significantly between the two groups (data not shown).

Because cerclage was not preventive for preterm delivery in patients with a history of electrosurgical conization, we tried to find out other risk factors for preterm delivery in these patients. Of the 56 cases, 18 pregnant women delivered preterm (preterm group) and 38 women delivered at term (full term group). We compared the demographic characteristics and potential antenatal risk factors for preterm delivery between the two groups (Table 4). However, maternal characteristics, cone depth, cervical length, and interval between conization and pregnancy were not significant different between the two groups.

This study evaluated the outcomes of pregnancy after electrosurgical conization: preterm delivery rate and the effect of the prophylactic cervical cerclage. Our results showed that the risk of preterm delivery was significantly increased in women with a history of electrosurgical cervical conization before pregnancy. However, it is questionable whether prophylactic cervical cerclage is beneficial for these high-risk patients in preventing preterm delivery.

Any surgical procedure of the cervix including conization has a potential of causing cervical incompetence which subsequently leads to abortion or preterm delivery during the following pregnancy. Early study results show that conization does not constitute a hazard to the outcome of future pregnancies.8-10 However, recent studies have reported that conization can lead to impaired outcomes of pregnancy.11-13 The incidence of preterm deliveries following conization varies in recent publications between 14 and 25%.11,12 These conflicting findings might be explained by differences in the study period population, and operative techniques such as instruments used for conization, depth, width and volume of the removed cones. Leiman et al.14 found that the risk of premature delivery and late spontaneous abortion increased in direct proportion to the cone size. In our study, cone depth was not associated with risk of preterm delivery in pregnancies after conization. However, the volume of removed cones was not evaluated, and it might be presumed the electrosurgical conization method using a right-angled, triangular loop with cold coagulation would have removed large volume of cone tissues as shown in Fig. 1.

The overall obstetric outcomes after electrosurgical excision in our study showed different results from earlier reports by others.15-17 Common conclusions of earlier studies were that the diathermy loop excision does not affect fertility and pregnancy outcome. However, none of these studies seem to have analyzed the overall rates of pregnancy and preterm delivery after conization, even in a single institute. Because conization itself most often cures the cervical lesion, the patients are not usually required for long term follow-up at a gynecologic oncologic clinic, and a considerable number of patients is lost to follow-up. In our study, we recruited almost all patients who became pregnant after conization in our institute. All excisional procedures used to treat cervical intraepithelial neoplasia have a potential to bring about adverse obstetric morbidity, but only cold-knife conization is proven to significantly increase adverse outcomes that is similar to our results.18

An increasing number of reproductive-age women with cervical lesions are being treated with local excisional procedures. In particular, the right-angled triangular shape cone biopsy excisors with a ball diathermy is known to be more effective than the U-shaped loop in women who want to preserve fertility because of its higher rate of removing the transformation zone at once, or with fewer times and with a lower rate of margin positive and recurrent rate, and a better maintenance of shape of uterine cervix after the procedure.19,20 Despite the advantages, these procedures may alter the cervical environment by damaging the integrity of tissue, and thus might affect cervical competence. These women might also have an increased risk of preterm delivery because cervical conization removes a sizable part of the connective tissue which can possibly lead to a weakening of cervical tissue, and inability to support the pregnant uterus.21 Therefore, cervical cerclage has been performed as an effort to reduce the risk of preterm delivery. However, the cerclage may itself be a risk factor for preterm delivery. Sutures can act as a foreign body which may cause uterine irritability and lead to contractions after a cerclage procedure.22 Moreover, some authors reported a significant increase of pathogenic flora in the vagina and cervix after cerclage.23 Along with previous reports in which cerclage after conization had no benefit on pregnancy outcomes,24-26 our study also showed that cervical cerclage after electrosurgical conization did not convincingly reduce the rate of spontaneous preterm delivery. We performed the McDonald procedure for cerclage in this study, but we are not sure that the Shirodkar procedure would have made a difference in the rate of preterm delivery, because preterm delivery rates were similar whether the cervical length were more or less than 25 mm before cerclage. Therefore, we suggest that using a colposcope-guided conization may be an option to reduce the volume of removed tissue so as to minimize the adverse effects on future pregnancies. We also suggest a more sparing use of cerclage for women with a history of conization because the efficacy and safety of the prophylactic cervical cerclage is still controversial.

To date, there is no randomized trial to evaluate the effect of prophylactic cervical cerclage after conization. The main strength of our study is that almost all patients were followed up for the antenatal care and delivery after conization, and the procedure of each conization and cerclage was performed with the same method in a single institute. However, the main limitation of our study is its non-randomized retrospective nature and the small number of subjects included in the study. Therefore a larger number of prospective randomized controlled trials are needed to confirm the effect of prophylactic cervical cerclage after conization.