Treatment modalities for CIN

Historically, the treatment of choice for moderate-to-severe CIN was the cold knife cone (CKC). Its application predated modern colposcopic practice and the widespread availability of electrocautery and laser technology. However, its modern applicability has been limited by high cost, significant intraoperative and postoperative bleeding, substantial perioperative infectious risk, a high level of technical difficulty, and a recognized association with postprocedure cervical stenosis. As a result of these limitations, alternative excisional and ablative procedures were developed that include the laser conization, the loop electrosurgical excisional procedure (LEEP), and various methods of ablation. The laser conization had the advantage of being performed under local anesthesia with less associated bleeding and more accurate tailoring of cone size. However, the thermal damage that was inflicted on the tissue specimen potentially could make pathologic evaluation of the margins impossible. With the advent of widely available electrocautery, the LEEP gradually replaced the CKC as the treatment of choice for CIN. Several studies have documented its advantages over CKC in that it is less expensive, technically easier, less painful, associated with less hemorrhage, and can be performed in an office setting with similar efficacy. Additionally, in contrast to the laser cone, the tissue specimen is more adequate for pathologic evaluation of the surgical margins.

Ablative techniques that are used to treat cervical dysplasia include laser ablation, cryotherapy, electrofulguration, and cold coagulation. Although these ablative techniques provide no tissue specimen for pathologic evaluation and can be applied only to a certain subset of patients, they appear to have similar efficacy with respect to the elimination of CIN and reduction of the risk of progression to cancer. A Cochrane Review about the surgical treatment of CIN reviewed 28 trials that compared the efficacy of both ablative and excisional treatment techniques and concluded that no method was more efficacious than any other. Therefore, the selection of ablative techniques vs excisional techniques should be based on the severity of disease, the adequacy of the colposcopic examination, the histologic findings of the biopsy, the appropriate correspondence of the cytologic and histologic evidence, and the desire for future childbearing.

CKC

It was recognized as early as 1938 that conization may have a negative impact on future pregnancy, with higher incidences of PTD and other complications. Early studies that investigated the association between CKC and obstetric complications were contradictory. Since that time, significant data have been published that has solidified the increased obstetric risk after CKC. A retrospective analysis by Klaritsch et al in 2006 evaluated the risk of PTD and obstetric complications in women with a history of cold knife conization of the cervix relative to the general obstetric population in Austria. The investigators reported PTD in 22.4% of 76 deliveries in the conization group compared with 6.6% of 29,711 deliveries in the general obstetric population (odds ratio [OR], 4.07; 95% confidence interval [CI], 2.22–7.10; P < .001). They further reported nearly an 8-fold increased risk of both preterm premature rupture of membranes (OR, 7.70; 95% CI, 3.87–14.21; P < .001) and cervical tears (OR, 7.53; 95% CI, 2.63–17.57; P < .001) but no significant increase in the risk of cesarean delivery, low birthweight, or duration of labor.

Because many confounding variables such as smoking, sexually transmitted diseases, maternal marital status, and socioeconomic status serve as important risk factors for both CIN and PTB, drawing conclusions from retrospective studies with the use of the general obstetric population as control subjects can be misleading. Kristensen et al attempted to address this issue by including patients with deliveries before conization. The investigators divided the cohort of 14,223 Danish women into 4 groups: those who had their conization before first delivery, those who had it between their first and second deliveries, those who had it after 2 deliveries, and those with no history of conization. The 170 women with a history of CKC, regardless of timing of the procedure, experienced a higher incidence of PTD. Although this risk was higher in women who underwent conization before pregnancy, women who underwent CKC subsequent to both deliveries also had a slightly increased risk of PTD in the precedent pregnancies, when compared with the general population. The authors therefore concluded that CKC was associated with a higher rate of PTB, but that factors other than surgical intervention may contribute to the observed risk. To further elucidate this issue, El-Bastawissi et al retrospectively compared women who had carcinoma in situ of the cervix who were not treated with conization with those women who received the prescribed therapy. Importantly, this study reported no increased risk of preterm birth or cesarean delivery in women with untreated carcinoma in situ over the general population. It did demonstrate, however, an increase in PTD and cesarean delivery among women with a history of conization (OR, 1.6; 95% CI, 1.2–2.0). Bruinsma et al also attempted to clarify the relationship between cervical dysplasia and PTD by reporting on women who underwent treatment for the precancerous changes and women who remained untreated. In contrast to the study just discussed, these authors reported an increased risk among women with untreated cervical dysplasia, compared with the general population (standardized prevalence ratio, 1.5; 95% CI, 1.4–1.7), with an even higher risk among those women who underwent treatment (OR, 2.0; (95% CI, 1.8–2.3). However, once the authors controlled for confounding factors that included marital status, history of multiple induced abortions or miscarriages, maternal age, major maternal medical condition, and illicit drug use, neither group had an increased risk. Despite multiple attempts to clarify this issue, it remains unclear which factors play the greatest role in the risk of PTD and adverse obstetric outcomes in women with cervical dysplasia.

A search of the published literature revealed 2 metaanalyses that addressed obstetric outcome after cervical surgery (ie, CKC, LEEP, laser ablation, laser conization). Both studies reported an increased risk of PTD (relative risk [RR], 2.59; 95% CI, 1.8–3.72; RR, 2.78; 95% CI, 1.72–4.51) and low birthweight (RR, 2.53; 95% CI, 1.19–5.36; RR, 2.86; 95% CI, 1.37–5.97) in patients with a history of CKC. The metaanalysis by Kyrgiou et al also reported a significantly increased rate of cesarean delivery (RR, 3.17; 95% CI, 1.07–9.40). The report published by Arbyn et al in 2008 evaluated the incidence of perinatal death, early PTD (<32-34 weeks of gestation), very early PTD (<28-30 weeks of gestation), and low birthweight (<2000 g) and reported that a history of CKC was associated with increased perinatal mortality rates, severe PTD, extreme PTD, and low birthweight. The results of studies that evaluated CKC and obstetric outcomes are summarized in Table 1 .

Laser conization

Because of the growing concerns regarding CKC and technologic advances, laser conization became an increasingly popular alternative for the treatment of cervical dysplasia. Hagen and Skjeldestad in 1993 were the first investigators to report an increased rate of PTD in patients who underwent laser conization. In this series of 56 women with a history of laser conization who delivered after 22 weeks gestation, the authors demonstrated a 38% rate of PTD among cases, compared with 6% in matched control subjects (OR, 9.0; 95% CI, 3.7–21.7). Other investigators have demonstrated conflicting results. Sadler et al in 2004 reported an increased risk of preterm premature rupture of membranes among women who underwent laser conization with an adjusted RR of 2.7 (95% CI, 1.3–5.6) but failed to demonstrate an increase in spontaneous preterm deliveries (adjusted RR, 1.3; 95% CI, 0.8–2.2). Other studies found similar results including Sagot et al who examined 71 pregnancies in 54 women before laser conization and compared them with 82 pregnancies after the procedure. The authors reported no significant difference in the rate of PTD (13.2% vs 8.5%) or premature rupture of membranes (1.9% vs 0%) before and after treatment but detected a reduced rate of vaginal term deliveries after conization (90% vs 73.6%; P = .025). This study used patients as their own historic control subjects, thus strengthening the results by addressing many confounders that are associated with both cervical dysplasia and preterm birth.

Although studies by Raio et al and Sadler et al reported no overall increase in preterm birth after laser conization, both studies further analyzed outcomes based on cone height and independently detected increased risk for poor obstetric outcome with larger cone size. Raio et al demonstrated an increased risk of PTD in women with a cone height of ≥10 mm, whereas Sadler et al reported a 3-fold increase in risk of preterm premature rupture of membranes and subsequent PTD in women with a cone height of ≥1.7 cm. Finally, one retrospective study assessed the risk of low birthweight in 65 patients with a history of CO ₂ laser conization and reported a 2.2 RR (95% CI, 1.04–4.5) for birthweight <2500 g, a 3.5 RR (95% CI, 1.02–12.0) for birthweight <2000 g, and a 10.0 RR (95% CI, 1.2–85.6) for weight <1500 g, which provides further, albeit different, supporting evidence to indicate poor obstetric outcomes in patients with a history of laser conization for cervical dysplasia.

LEEP

As mentioned earlier, the technical simplicity, decreased blood loss, and outpatient nature of the procedure have all contributed to LEEP becoming the treatment modality of choice for cervical dysplasia. Because of its widespread application, LEEP has the farthest reaching implications for public health impact and therefore should be considered most carefully. Despite the significant volume of data that are available, the effects of LEEP on pregnancy outcomes remain controversial, with evidence supporting both sides of the debate ( Table 2 ). Sjoborg et al published a multiinstitutional retrospective case-control study that evaluated the cases of 742 women with a history of either LEEP or laser conization. In this series, the authors reported the risk of giving birth before 37, 32, and 28 weeks of gestation after treatment with either laser conization or LEEP and compared those rates to control subjects in the general obstetric population. After adjustment for smoking habits, education level, and marital status, the ORs in the treatment group were 3.4 (95% CI, 2.3–5.1), 4.6 (95% CI, 1.7–12.5), and 12.4 (95% CI, 1.6–96.1) for each gestational age, respectively. The authors further reported increased rates of low birthweight and preterm rupture of membranes in women who underwent either excisional procedure relative to control subjects.

Several studies subsequently have focused exclusively on women who were treated with LEEP. Nøhr et al reported an approximately 2-fold increased risk for PTD among women with a history of LEEP, even after controlling for confounders that included smoking, age, parity, obstetric history, and educational status (OR, 1.8; 95% CI, 1.1–2.9). Samson et al examined women who had been treated exclusively with LEEP and reported an increased rate of spontaneous PTD (OR, 3.5; 95% CI, 1.9–6.95) and PTD after premature rupture of membranes (OR, 4.10; 95% CI, 1.48–14.09). Additionally, the authors substantiated previous reports that a history of LEEP conferred a higher risk of low birthweight infants (OR, 3.00; 95% CI, 1.52–6.46).

Over the past 3 years, several European studies that have evaluated the risk of PTD after LEEP have been published. All of them report a significant increase in the risk of PTD after treatment for cervical dysplasia, although many still acknowledge that causation remains theoretic. In a Danish cohort study, 349 women with a LEEP before pregnancy had an OR of 1.8 (95% CI, 1.1–2.9) for PTD at <37 weeks of gestation with an absolute difference of 6.6% vs 3.5% when compared with the untreated contingent of the cohort. There was no adjustment for confounding factors such as marital status, economic status, and smoking. Jakobsson et al reported a similar magnitude of risk in a Finnish cohort of 624 women who delivered after LEEP (RR, 2.61; 95% CI, 2.02–3.2) but added an ideal, internal control with a subgroup of 258 women who had deliveries before and after LEEP. The preterm birth rate was 6.5 % before LEEP and 12% after LEEP, which is a nearly 2-fold increase in the same woman before and after LEEP. In addition, a small retrospective cohort study that was performed in Belgium reported an increase in the frequency of PTD (both at <37 and <34 weeks of gestation), with a mean gestational age of 266 vs 274 days at delivery and a lower birthweight.

One of the largest studies to examine the association of LEEP and preterm birth was performed in a Danish population-based cohort. Ortoft et al reported an increase in preterm birth at <37 weeks of gestation (hazard ratio [HR], 2.4; 95% CI, 1.8–3.1), perinatal death (HR, 4.1; 95% CI, 1.3–13), and preterm premature rupture of membranes at <37 weeks’ gestation (HR, 3.0; 95% CI, 2.2–4.1) in nearly 600 women who were treated with LEEP before pregnancy, compared with control subjects. However, a subgroup analysis of women who gave birth before and after conization procedures did not detect a significant difference in any of these endpoints. This study is one of the only investigations that provided information on the effect of multiple cervical procedures. Although not separated by type of procedure, women who had >1 conization procedure had a nearly 10-fold increase (HR, 9.9; 95% CI, 6–17) in the rate of preterm birth, compared with women who never had an excisional procedure. Although a large recent study in a Norwegian population-based cohort reported a similar increase in the relative risk of preterm birth after a conization when compared with the general population (RR, from 2.4 at 33-36 weeks of gestation to 4.3 at 24-27 weeks of gestation) or to women who subsequently had a conization (RR, from 2.2 at 33-36 weeks of gestation to 3.0 at 24-27 weeks of gestation), the applicability of these findings to the US population or modern practice is limited because information on methods was not available and the accrual of patients spanned >35 years. Despite these limitations, the finding of the highest relative risks for PTD at earlier gestational ages warrants attention because these are the neonates with the greatest risk of perinatal morbidity.

In contradistinction, multiple other studies that used similar designs have failed to demonstrate an increase in poor obstetric outcomes after LEEP. Ferenczy et al reported no difference in PTD or cesarean delivery after LEEP; however, this patient population was restricted to those with cone height <1.5 cm and a mean frontal diameter of 1.8 cm, which raises the question of the effect of larger cone specimens. In another Norwegian study, Acharya et al reported no increase in risk of PTD after LEEP in 79 patients, when compared with 158 matched control subjects. However, there was a 4-fold increase in PTD when cone height was >25 mm (RR, 4.0; 95% CI, 1.0–16.0). In a study from New Zealand, Sadler et al demonstrated no increased risk of PTD in patients who underwent LEEP (adjusted RR, 1.2; 95% CI, 0.8–1.8) but did report an increased risk of preterm premature rupture of membranes with subsequent PTD (adjusted RR, 1.9; 95% CI, 1.0–3.8). Another study differed slightly in that the authors reported no increased risk of PTD but detected a decrease in birthweight among infants who were delivered after LEEP.

A recent study by Werner et al reported on >1300 women who had both a LEEP and singleton pregnancy delivered at a single institution in the United States. Of these women, 511 had the LEEP performed before the examined pregnancy, and another 842 underwent LEEP after the index pregnancy. When compared with the general obstetric population, the rates of PTD in women with LEEP were similar (4% vs 2% vs 4%, respectively; P = .12 and = .22). Further analysis after control for demographic differences showed similar results (4% vs 4% vs 4%; P > .5). This study is of particular importance, given the large US population examined, but it is limited by a lack of information on potential confounders and a lack of information on previous preterm births or LEEP that was performed at other institutions.

A systematic review by Crane in 2003 reviewed 5 articles that compared women who were treated with LEEP to control subjects and concluded that women who had LEEP were more likely to have a PTB in subsequent pregnancies (OR, 1.8; 95% CI, 1.18–2.76) and were more likely to have low birthweight infants (<2500 g; OR, 1.60; 95% CI, 1.01–2.52). When limited to studies that controlled for smoking status, the increased rate of PTD persisted, but the risk of low birthweight equilibrated between groups. A 2006 metaanalysis by Kyrgiou et al included 10 studies (5 of which were included in the aforementioned analysis by Crane ) that compared women who had a LEEP before pregnancy with untreated control subjects and reported pooled RRs that were calculated with a random-effects model. Women with a previous LEEP had an increased risk of PTD (RR, 1.7; 95% CI, 1.24–2.35), premature rupture of membranes (RR, 2.69; 95% CI, 1.62–4.46), and low birthweight (RR, 1.82; 95% CI, 1.09–3.06). This analysis also controlled for confounding factors between study populations with the performance of a subgroup metaanalysis that matched for age, parity, and smoking that resulted in a higher relative risk for PTD (RR, 2.10; 95% CI, 1.34–3.29). A further analysis of the dimensions of tissue excised was also undertaken for those studies that reported these data. There was a significant increase in PTD if cone height was >10 mm (RR, 2.6; 95% CI, 1.3–5.3). If cone height was <10 mm, the data were conflicting between studies, and the risk was not significant (RR, 1.5; 95% CI, 0.6–3.9). Finally, the metaanalysis that was published in 2008 examined data from 7 studies with 3600 women who treated with LEEP before pregnancy. The authors reported no increase in the risk of PTD at <32 or 34 weeks of gestation (RR, 1.2; 95% CI, 0.5–2.9) or perinatal mortality rates (RR, 1.17; 95% CI, 0.74–1.87) but did note an increase in women who were treated with CKC. Despite the lack of a demonstrable increase in the risk of PTD at any gestational age, the authors still cautioned that LEEP cannot be considered completely free of adverse outcomes.

Ablative procedures

After the potential complications associated with conization are highlighted, it would be remiss to avoid discussion of the potential obstetric complications after ablative procedures as alternatives to conization. A handful of retrospective case-control studies have examined ablative procedures, and, much like the data in the preceding sections, the results are contradictory. Of the 4 retrospective studies that were examined, 3 concluded no increase in PTB after laser ablation of the cervix, although the other reported an RR of 1.39 (95% CI, 1.18–1.63) El-Bastawissi et al and Crane examined several different types of treatment for CIN that included laser vaporization and cryotherapy and reported no increase in poor obstetric outcomes after ablative procedures. Neither metaanalysis that evaluated this issue detected increased obstetric risk after laser ablation or cryotherapy. Arbyn et al reported an overall relative risk for PTD of 0.87 (95% CI, 0.53–1.45) when they analyzed all ablative techniques collectively, which is a result similar to that reported by Kyrgiou et al for only laser ablation (RR, 0.87; 95% CI, 0.63–1.20). Neither metaanalysis reported an increase in incidence of low birthweight infants of patients who were treated with ablative techniques. The only article that reported an increase in complications after ablation for CIN was by Jakobsson et al, who reported an RR 1.47 (95% CI, 1.29–1.67) among 9000 patients who underwent ablation after adjustment for smoking, age, and parity. Although the data regarding such procedures are mixed, far fewer studies indicate a risk that is associated with ablative procedures, which is an important consideration for clinicians when treating reproductive-aged women for cervical dysplasia.