Introduction

Cervical intraepithelial neoplasia (CIN) is most often diagnosed in women of reproductive age [1]. Despite the progression of the cervical cytology screening system and the development of other screening tools such as human papilloma virus (HPV) testing, the incidence of this condition in reproductive age is still rising in Japan [2, 3]. Considering age, various fertility-preserving procedures such as conization, cryotherapy, laser treatment, and the loop electrosurgical excision procedure (LEEP) have been used for patients with this condition. The 5-year survival rates of patients who undergo these operations are very high and in most of them the disease does not recur. On the other hand, conization or LEEP of the cervix is reported to be related to adverse pregnancy outcomes in subsequent pregnancies, including preterm delivery, low-birth-weight infants, incompetent cervix, and cervical stenosis [4, 5]. The decrease of mechanical support of the residual cervix and chorioamnionitis (CAM) caused by disruption of the endocervical glands and reduced secretion of mucus are causes of such adverse effects [6]. Therefore, complete cure with a better obstetrical prognosis is the best policy for the treatment of CIN. As a higher medical institution, our university hospital mainly deals with patients having high risks and those with complicated pathological diagnoses. Though we perform CO2 laser conization to obtain an accurate pathological diagnosis, laser conization is a more aggressive treatment modality than LEEP or laser ablation, which is mainly used for patients with lower risks. However, it is still unclear whether more conservative conization improves obstetrical outcomes without increasing adverse oncological effects.

Therefore, we have recently changed our laser conization method from the standard cone-shaped one to a more conservative and shallower, coin-shaped resection for young patients with CIN in whom we could confirm the entire squamocolumnar junction (SCJ) by colposcopy. This study reports both the obstetrical and gynecological prognoses of patients who underwent coin-shaped resection in comparison with those who underwent the standard cone-shaped resection.

Patients and methods

A retrospective study was performed for patients of reproductive age with cervical intraepithelial neoplasia (CIN). Laser conization was performed for patients with CIN 3, high-risk patients with CIN 2, and patients with high suspicion of CIN 3 who had discrepancies between the cytology and colposcopic-directed biopsies such as CIN 3 cytology with CIN 2 or less colposcopic biopsies. We explained the advantages and disadvantages of conization, LEEP and other conservative treatments to them in detail. Then we obtained written informed consent from all the patients who agreed to undergo laser conization. We introduced those patients who desired LEEP or other conservative treatments and those who did not appear to need surgery to other clinics for further treatments.

Altogether, 401 women 44 years old or younger agreed to undergo CO2 laser conization and were treated in our university hospital between 2003 and 2012. None of them were pregnant at the time of the operation. Subsequently 50 of them went on to become pregnant and were followed-up in our university hospital. The patients were divided into the following two groups: (1) a standard cone-shaped conization group and (2) a shallow coin-shaped conization group. In both groups, CO2 laser conization was carried out to perform complete resection of the lesion. In principle, we performed standard cone-shaped conization for patients with CIN until 2008, and then we changed our operative procedure to the more conservative coin-shaped conization for them. Operations were performed according to the following criteria. The width of the conization at the ectocervix was determined by the size of the transformation zone and the location of abnormal ectocervical lesions. The depth of the cone in the endocervical portion was determined by the location of the SCJ, the presence or absence of endocervical disease, and the status of the glandular lesion. Patients for whom invasive cancer appeared to be a possibility and those with glandular lesions were excluded from this study.

For group (1), we performed standard high cone-shaped conization including a sufficient cervical canal for 160 patients between 2003 and 2008 (Fig. 1a), and 25 of these patients later became pregnant. For group (2), shallow coin-shaped conization intended to include the entire pathologic lesion (Fig. 1b) was performed for the 241 patients since 2008 after explanation to the patients in detail and obtaining written informed consent, and 25 of them later became pregnant. The pregnancy courses and oncological prognoses of these two groups were studied.

Fig. 1
figure 1

Representative schema of conization for patients with CIN3. a Schema of standard cone-shaped conization. b Schema of coin (or sombrero)-shaped conization

Full size image

Patients were followed-up depending on their pathological grades. For patients with CIN 3, cytological tests were performed 1, 3, 6 months, and 1 year after the operation. After that, yearly cytological examinations were performed for at least 10 years. For patients with CIN 2 or less, cytological examinations were performed 6 months and 1 year after the operation at our hospital. Once dysplasia was suspected by cytology, colposcopic examination and biopsies were performed. Diagnosis of CIN 3 was regarded as recurrence.

Statistical analysis

Student’s t test was employed for the analysis of all patients who underwent conization, and the nonparametric Mann–Whitney U-test was used for the analysis of subgroups of the standard cone-shaped conization group and coin-shaped conization group who might have had a skewed distribution, with the use of the Microsft Excel. p < 0.05 was considered significant.

Results

Differences of clinical characteristics, gynecological prognoses, and cutting depths between the patients of the two groups

Table 1 shows the differences of various clinical characteristics and gynecological prognoses between the patients of the two groups. There were almost no significant differences between the two groups with regard to the clinical characteristics, gynecologic prognoses, recurrence rates, and residual disease rates. However the reoperation rate was higher in the standard cone-shaped conization group. None of the patients died of the disease. Cases of CIN 1, although they received laser conization under the diagnosis of “high suspicion of CIN 3”, which means repeated abnormal cytological findings corresponding to CIN 3 with CIN 2 or less in colposcopic biopsies, the final pathological diagnosis of the conization specimen was CIN 1.

Table 1 Clinical characteristics and gynecologic prognoses of all the patients who underwent conization for CIN
Full size table

Conization height in the standard conization for all the patients was 1.40 cm, and that for pregnant patients was also 1.40 cm. The conization height of the coin-shaped conization was 1.28 cm, and that for pregnant patients was 1.12 cm. Thus, cone height reduction of about 2–3 mm was achieved by this more conservative operative procedure. However, the reduction of the cone height was not associated with reduction of the free margin length. This was probably because the coin-shaped resection was not really “coin”-shaped, but was rather a “sombrero”-like shape, which meant wide resection of the ectocervix and narrow and sufficient cutting of the endocervix (Fig. 1b).

Difference between obstetrical prognoses of the two groups

Table 2 shows the differences of clinical characteristics of pregnant patients of the two groups. As was the case for all patients, there was no significant difference between the groups except for the number of patients who underwent IVF-ET. No one underwent prophylactic cervical cerclage in either group.

Table 2 Clinical characteristics of patients who became pregnant
Full size table

Table 3 shows the differences of obstetrical characteristics of pregnant patients of the two groups. In the original standard conization group, 7 patients (28 %) delivered before term and 18 patients (72 %) delivered at term. The overall mean gestational age at delivery was 37 weeks 3 days ± 2 weeks 2 days. The overall mean interval between conization and delivery was 47 months. On the other hand, in the shallow coin-shaped conization group, 5 patients (20 %) delivered before term and 20 patients (80 %) delivered at term. The overall mean gestational age at delivery was 38 weeks 1 day ± 2 weeks 1 day. The overall mean interval between conization and delivery was 28.5 months. There were no significant differences in preterm delivery rates, the occurrences of pPROM and CAM, or the neonatal prognoses between the two groups. However, as shown in Fig. 2, changes in the uterine cervical length during the pregnancy showed a clear difference. The rate of reduction of cervical length over the pregnancy was smaller in the coin-shaped group, and the number of patients with a cervix shorter than 2 cm was smaller in the coin-shaped group.

Table 3 Obstetrical prognoses of the patients after conization
Full size table
Fig. 2
figure 2

Changes of cervical length measured by transvaginal ultrasonography. The rate of reduction of cervical length over the pregnancy was smaller in the coin-shaped group, and the number of patients with a cervix shorter than 2 cm was smaller in the coin-shaped group especially after 25 weeks of pregnancy

Full size image

Discussion

Although conservative coin-shaped conization did not markedly improve the obstetrical prognosis, this operative procedure improved the reduction rate of uterine cervical length over the pregnancy without any increase in oncological adverse effects or worsening in the gynecologic prognosis. In Japan, we tend to intervene against threatened abortion in patients who have received conization earlier and more aggressively than in other countries [7]. For example, we commonly perform oral administration of the beta-adrenergic agent ritodrine hydrochloride for patients who feel abdominal tension and for those who show shortening of the uterine cervix to less than 3 cm at outpatient clinics. The effect of such prophylactic administration of ritodrine hydrochloride is still controversial or considered to be negative in the USA and European countries [8]. However, such oral administration as well as continuous low-dose intravenous administration of ritodrine hydrochloride under hospitalization is still widely used in Japan, including in our hospital. This might have contributed to the lack of a significant difference in adverse obstetrical effects between the two groups. In any case, conservative treatment would have a positive impact on the pregnancy course.

Conization of the uterine cervix has been widely used to diagnose and treat CIN, and even to treat microinvasive carcinoma of the uterine cervix. Although less invasive techniques such as LEEP have reduced the need for diagnostic conization dramatically, conization, mainly by laser, still remains a useful operative method for these conditions because it provides a more accurate pathological diagnosis and more complete resection than LEEP and ablation [5, 9].

However, any surgical procedure involving the cervix, including conization, has the potential to cause cervical incompetence, which subsequently leads to abortion or preterm delivery during the following pregnancy. Many recent studies have reported that conization can lead to impaired outcomes of pregnancy [5, 10, 11]. Cone height in particular seems to affect the obstetrical outcomes. Leiman et al. reported that the risk of preterm delivery and late spontaneous abortion increased in direct proportion to the cone size [12].

Therefore, more conservative conization, like that we performed, and prophylactic cerclage are promising candidates for improvement of the obstetrical prognosis after conization.

The benefits of prophylactic cerclage for women with a history of conization are not clear. Many studies could not detect significant merits of prophylactic cerclage and reported that cerclage itself could be a risk factor [13–15]. Sutures can act as a foreign body and lead to uterine irritability and contractions after cerclage [16]. Moreover, it is also reported that such cerclage can cause a significant increase in pathogenic flora in the vagina and cervix [17]. Our hospital usually does not perform prophylactic cerclage for these patients as long as they do not show symptoms of cervical atony by transvaginal ultrasonography. However, as we recently reported, prophylactic cerclage seems to be inevitable for patients who undergo radical trachelectomy[18, 19]. This procedure also seems to be necessary for patients who undergo conization twice or more. We believe that the remaining cervical length, in other words, the size of the specimen removed, is an important factor to determine the necessity for prophylactic cerclage. Of course, loop excision removes a smaller cone volume than conization as Grimm et al. reported [20]. Therefore loop excision and cervical ablation might be good treatment modalities for patients with lower risks. However, even in cases where conization is needed, we believe we should consider more conservative conization, such as shallow coin-shaped conization.

Conization can certainly lead to impaired outcomes of pregnancy. The incidence of preterm deliveries following conization varies between 14 and 25 % in recent reports [21, 22] depending on differences in the study population and operative techniques such as the depth, width and volume of the removed cone size. However, as we mentioned above, we tend to intervene against abdominal tension and changes of cervical length earlier and more thoroughly than in other countries. Therefore the preterm delivery rates after conization might be lower than in reports from other countries. Our standard cone-shaped conization also showed good results, and our conservative conization seemed to have a minimal impact on improving the obstetrical prognosis. However, considering the changes of cervical length over the pregnancy, this operative procedure actually had good effects on pregnancy courses. No increase of recurrent disease or residual disease was detected and the oncological prognosis was also good. In this sense, this conservative coin-shaped conization is useful.

We believe the most important thing to obtain the best clinical results via this conservative conization is to perform accurate colposcopic examination [23]. If the SCJ is completely visible, and even if only a small part of the CIN lesion has gone into the cervical canal, a shallow cone biopsy may suffice. In the case of recurrence, it leaves enough of the cervix to perform a second conization. On the other hand, if the SCJ is not visible at all in the presence of a high-grade smear, then a deep cone biopsy including the entire cervical canal must be performed. In addition, when invasive cancer or glandular disease such as adenocarcinoma in situ (AIS) is suspected, shallow conization should not be done. It is commonly believed that glandular disease sometimes has multifocal locations and this condition can recur even when the conization margins are negative. Soutter et al. reported that 16.7 % of patients required further treatment after four years because of recurrent cytological abnormalities [24]. Shallow coin-shaped conization should be performed only for young patients of reproductive age with CIN (1) in whom the SCJ can be completely detected under colposcopy, and (2) who have squamous histology.

How deeply should we cut the cervix? Sadler et al. demonstrated that for excisions of 17 mm or more the risk of pPROM, but not of preterm delivery, was significantly higher. Barretta et al. recently reported that a depth of >1.5 cm increased the occurrence of preterm delivery [25]. On the other hand, Ferenczy et al. reported that a depth of <1.5 cm did not affect the occurrence of preterm delivery, though it was a result from LEEP [26]. In general, the SCJs of young women tend to exist in shallower areas of the cervical canal.

As we demonstrated, our coin-shaped shallow resection with an average height of 11.2 mm reduced cone depth by 3–4 mm compared with the standard cone-shaped conization. If the patients satisfy the above criteria [(1) and (2)], we believe a shallow cutting depth of around 11–12 mm is sufficient for the treatment of CIN. Although there is still no consensus as to the length of a safe free margin, our average free margin of 37 mm and cutting depth seemed sufficient for the treatment because they did not affect recurrence or increase residual disease.

This study evaluated both the obstetrical outcomes and gynecologic prognosis after conservative shallow coin-shaped conization in comparison with those after the original cone-shaped conization with the use of a CO2 laser. We believe our findings are informative because all the procedures of each conization and the follow-up of pregnancies were performed with the same method in a single institute. The limitations of our study are that it is a non-randomized reptrospective one with only a small number of patients included. A larger number of prospective randomized control studies and cooperation between gynecologic oncologists and perinatal obstetricians will be necessary to confirm the effects of this conservative operative procedure.