Introduction

Cervical intraepithelial neoplasia (CIN) is a very common disease in young women, typically affecting women between 25 and 35 years of age [1]. Although the management of CIN has evolved to a much more conservative pathway, high-grade lesions are usually treated by conization [2].

A variety of excisional techniques exist: large loop excision of the transformation zone (LLETZ) and cold knife conization (CKC) are two of the most common techniques performed. The advantages and disadvantages of each excisional procedure have been previously published [3]. LLETZ seems to be as effective as CKC concerning complete resection rate. Given that LLETZ was reported to be associated with fewer intra- and post-operative complications, it has become the golden standard of excisional procedures in most parts of the world [3].

The most relevant benefit of LLETZ—compared with other excisional procedures in general and to CKC in particular—seems to be the relatively low rate of pregnancy-related complications [4–6]. CKC is associated with an increased risk of preterm birth (PTB) ranging from risk ratio (RR) 2.6 (1.8–3.7) to 2.9 (1.4–16.7), whereas the risk of PTB after LLETZ reveals conflicting results with RR ranging from 1.7 (1.2–2.4) to 1.2 (0.5–2.9) [4–6]. Regarding cone depth, there is a significant increase of PTB after conization with a cone depth >10 mm [RR 2.6 (1.3–5.3)] [5]. Thus, the risk for preterm birth after conization seems to be modulated by the type of procedure performed and/or the depth of the cervical cone removed [4–6]. However, the crucial factors linking conization and subsequent risk for preterm birth have not been fully elucidated [4–6].

Despite the fact that the extent of resected cervical tissue has become of increasing interest, reports addressing this topic are limited [7–9]. Phadnis et al. investigated the association between height and volume of the cervical tissue removed by laser cone biopsy and LLETZ. They observed a significantly higher median cone volume removal by laser conization than by LLETZ [7].

Although LLETZ has become the standard treatment for women with CIN, CKC is still performed in many patients, even in young women, who still want to conceive. The objective of our study was to compare cone volumes removed by LLETZ and CKC. Furthermore we aimed to identify independent risk factors, which might influence the size of the resected cone volume.

Materials and methods

Patients

Women who underwent LLETZ or CKC between January 2004 and December 2009 at the Vienna General Hospital (tertiary care center), Rudolfstiftung Hospital or Lainz Hospital (secondary care centers), were consecutively included in the present study. Conization was performed on patients with high-grade CIN and micro-invasive cervical cancer, diagnosed pre-operatively by colposcopically guided biopsy (CGB). Paraffin sections of the specimen were cross-sectioned and measured using micrometers. All analyses were performed by two pathologists specialized in gynecological pathology. The resection margins were only considered affected when any grade of CIN or invasive carcinoma was directly present on the resection margin. The result of endocervical curettage (ECC) was given as dysplastic cells absent (negative) or present (positive). Histology reports were examined to determine cone biopsy dimensions. The longitudinal diameter (a), transverse diameter (b), and depth (c) were measured and recorded after formalin fixation.

Given that a, b, and c were frequently disproportionate, we proposed that the shape of the cone was hemi-ellipsoid. Therefore, we used the following equation to calculate come volumes: 1/2 × 4/3 × π × a/2 × b/2 × c, as previously reported by Phadnis et al. [7].

Patients eligible for this trial were included after inclusion criteria were verified by chart review. Patients with available data regarding date, type, and histological result of conization were included. Patients were excluded if they underwent more than one conization, cervical cancer was preoperatively diagnosed by CGB, information about resected cone volume was incomplete or more than one cone was removed. Moreover, patients with adenocarcinoma in situ were excluded from the present analysis.

It is of particular interest, how the two groups (LLETZ vs. CKC) were generated. In the tertiary care center, LLETZ is the standard procedure and is routinely performed in all cases. Therefore, only 6/305 (2.0 %) patients were treated with CKC. In the secondary care centers, LLETZ was introduced and routinely performed at a later point in time—the middle of 2007. Therefore, patients were routinely treated with CKC until middle of 2007 and after the introduction of LLETZ as the standard treatment with LLETZ. This provides the unique setting to compare these two groups with a very little selection bias caused by clinical decision. Approval for this study was obtained by the Institutional Review Boards (IRB) of the Medical University of Vienna (IRB-approval number 991/2009).

Statistical analysis

After testing for normality using Kolmogorov–Smirnov test, values were given as mean [standard deviation (SD)] or median [interquartile range (IQR)] where appropriate. Parameters in the LLETZ and CKC group were compared by Student’s t test, Mann–Whitney-U test or Chi square tests. Values were given as p values for Chi square tests also as odds ratio (95 % confidence interval). A multivariable regression model using 2.5 cm3 as a cut-off cone volume was calculated to assess the association between conization technique and volume by correcting for possible confounding parameters, i.e. type of institution (secondary vs. tertiary care center), age (≤35 vs. >35 years), ECC (negative vs. positive), cone depth (≤1.7 vs. >1.7 cm), presence of cervical cancer (CIN vs. cervical cancer), and clear resection margins (yes vs. no). We used the software PASW Statistics 18.0 (Release 18.0.3, IBM Corporation 2010, Somers, NY 10589, USA) for statistical analysis. Two-sided p values <0.05 were considered statistically significant.

Results

Patients’ characteristics broken down by type of conization are provided in Table 1. The resected cone volume and specimen depth were significantly lower in women with LLETZ. Women with CKC more frequently experienced additional ECC (95.5 vs. 82.0 %). The removal of less cone volume and cone depth was not associated with lower complete resection rates (p = 0.9 and p = 0.9, respectively).

Table 1 Patients’ characteristics broken down by cold knife conization (CKC) and large loop excision of the transformation zone (LLETZ)
Full size table

In the tertiary care center, LLETZ was performed significantly more often than CKC [299/305 (98.0 %)], whereas in secondary care centers the opposite was observed [113/499 (22.6 %)] (p < 0.0001). As mentioned previously, the large proportion of CKC has to be attributed to the fact that in the two secondary care centers LLETZ has only been introduced in the middle of 2007. Women treated with LLETZ were significantly younger [34.1 (8.9) vs. 38.2 (10.7) years] and had a lower prevalence of high-grade CIN (70.2 vs. 82.4 %) or microinvasive cervical cancer (3.9 vs. 6.1 %) compared with patients treated with CKC.

Type of conization, cone depth, patient’s age, type of hospital, and malignant histology were associated with larger cone volume in a univariate analysis (Table 2). In the multivariable regression model, only type of conization, cone depth, and type of hospital revealed to be independent risk factors for larger cone volume (Table 2).

Table 2 Univariate analysis and multivariable regression model to assess independent risk factors for larger cone volume
Full size table

Overall, a strong correlation between cone depth and cone volume was observed (ρ = 0.5, p < 0.0001). In a subgroup analysis, a strong correlation between cone depth and cone volume for LLETZ cases (ρ = 0.9, p < 0.0001), but only a weak correlation for CKC was observed (ρ = 0.3, p < 0.0001) (Fig. 1).

Fig. 1
figure 1

A scatter plot showing the relationship between cone volume and cone depth broken down by type of surgery

Full size image

Discussion

In the present study, total volume of cervical tissue removed by LLETZ was significantly lower compared with CKC. CKC remained to be an independent risk factor for large cone volume [adjusted(a)OR 1.7 (1.1–2.8)] after correcting for possible confounders, such as cone depth, patient’s age, different type of hospital, presence of microinvasive cervical cancer, and clear margins in a multivariable model.

The finding that larger cone volume is removed by CKC than by LLETZ is in line with studies comparing different conization methods with LLETZ [7–9]. Phadnis et al. [7] observed that laser conization (median volume 1.84 cm3, 95 % CI 1.98–2.54 cm3) accounts for a larger volume of cervical tissue removed compared with LLETZ (median volume 0.78 cm3, 95 % CI 0.92–1.02 cm3). Conization volumes observed in this large study were considerably smaller than in a study comprising 63 patients in 1999, which compared CKC (mean volume 4.38 cm3) with loop conization (mean volume 3.96 cm3) [9]. Cone volumes in our investigation were slightly larger than those reported by Phadnis et al. [7].

Larger cone volume was not associated with a higher complete resection rate. This finding is highly interesting, as one might reasonably speculate that the extent of surgery—by resecting a larger cone—might be associated with a higher rate of complete resection. A recently published study has investigated the recurrence rates of high-grade lesions after conization [10]. In women <35 years of age, recurrence rates after conization were similar in the group with a cone depth <10 mm compared with ≥10 mm. Taken together, these findings emphasize the importance of close pre-therapeutic assessment and individual treatment planning. The planning of a conization where the smallest necessary portion of cervix is removed might be an important issue especially in young women, who still intend to conceive.

The value of cone volume regarding the risk of subsequent preterm birth is unclear. Up to date, there are two studies that investigated the association between cone volume and preterm birth with conflicting results [11, 12]. The first study observed an association between cone depth and PTB, but not between cone volume and PTB [11]. Of note, only 67 cases of CKC compared with 572 cases of LLETZ were included in this study. Due to the few cases of CKC the mean cone volume observed in this study was relatively small. The second study observed an association between large cone volume and increased risk of PTB [cone volume >6 cm3 was associated with a PTB risk of RR 3.00 (1.45–5.92)] [12]. This study is limited by the small sample size of patients with PTB and large cone volume removed (n = 10). If and to what extent cone volume might be of additional value compared with cone depth in the prediction of risk for preterm birth still needs further investigation.

In addition to method of surgery, type of institution (Austrian secondary vs. tertiary care center) and cone depth were observed to be independent risk factors for large cone volume in a multivariable regression model. Performing conization at an Austrian secondary care center was associated with a crude 2.0-fold risk (95 % CI 1.5–2.8) for large cone volume. This might be due to the higher rate of CKC performed at secondary care centers and a higher rate of high-grade CIN and microinvasive cancer cases. Nonetheless, after correcting for confounders in the multivariable regression model an adjusted 1.7-fold increase in risk (95 % CI 1.1–2.7) remained. This might be attributable to an increased awareness for the potential risk of preterm delivery after conization in tertiary care centers and, therefore, the tendency to remove as little cervical tissue as necessary—especially in young women who wish to conceive. Of note, the relatively large percentage of CKCs performed at secondary care centers has to be attributed to the early years of this study. Nowadays LLETZ is the golden standard of excisional procedure in all types of hospital in Austria.

High cone depth was associated with an increased risk for large cone volume [aOR 3.8 (2.6–5.7), p < 0.0001]. We aimed to further investigate the association between cone depth and cone volume and observed a relatively strong correlation for conisation in general (ρ = 0.5, p < 0.0001). Interestingly, this correlation was particularly strong for LLETZ cases (ρ = 0.9, p < 0.0001), whereas it was surprisingly weak for CKC cases (ρ = 0.3, p < 0.0001). This supports the findings that the shapes of cone biopsies vary, depending on histology and location of the lesion, patient’s age, and type of conization [7].

Of note, it has to be stated that calculation of cone volume has several limitations, which have been previously discussed [7]. Moreover, this study has particular limitations due to its retrospective study design. Obviously, one might argue that young patients with low-grade lesions would be treated by LLETZ, whereas older patients with large high grade lesions would be treated by CKC. This is only partly true, as the setting of the present study was rather unique with a tertiary care center where nearly all patients were treated by LLETZ (only 6/305 patients had CKC) and two secondary care centers, where LLETZ has only been introduced in the middle of 2007. Therefore, the decision whether LLETZ or CKC had been performed was not only partially based on clinical criteria but also on the date of surgery. Another limitation is the fact that we did not assess data regarding subsequent pregnancy outcome, as we did not intend to investigate pregnancy related postoperative complications.

Nevertheless, this is the first study to compare the cone volume after LLETZ and CKC in a relatively large set of patients. CKC removes significantly more cervical tissue without the benefit of higher complete resection rates. Interestingly, cone depth is strongly correlated with cone volume only in cone specimens removed by LLETZ, but not in those excised with CKC.