To assess the impact of large loop excision of the transformation zone (LLETZ) for cervical intraepithelial neoplasia (CIN) on cervical morphology as assessed by three-dimensional ultrasound.
To assess the impact of large loop excision of the transformation zone (LLETZ) for cervical intraepithelial neoplasia (CIN) on cervical morphology as assessed by three-dimensional ultrasound.
Prospective observational study.
University Hospital in Dublin.
Women with CIN who underwent an LLETZ procedure.
All 106 participants had a three-dimensional transvaginal ultrasound scan (3DTVS) performed immediately before and 6 months after LLETZ. The excised LLETZ specimen dimensions were also recorded. Blind analysis of the images was performed. The volume of the uterus and cervix was documented.
The relationship between the extirpated LLETZ dimensions and subsequent cervical and uterine biometry, as assessed by 3DTVS.
LLETZ induced a statistically significant reduction in both the length (mean, –0.46 cm; P < 0.001) and the volume (–6.12 cm3; P < 0.001) of the uterus, and in the volume of the cervix (–1.60 cm3; P < 0.001). The volume of the excised specimen had a significant impact on the reduction of the length of the uterus (β, –0.038; P = 0.012), the volume of the uterus (β, –0.791; P = 0.036) and the volume of the cervix (β, –0.121; P = 0.046). The circumference of the excised specimen appeared to have a significant impact on the length (β, –0.016; 95% CI, –0.028 to –0.003; P = 0.013) and volume (β, –0.413; 95% CI, –0.719 to –0.107; P = 0.009) of the uterus 6 months after LLETZ.
The volume of tissue removed at LLETZ is related to the subsequent cervical volume, as well as the uterine length and volume, 6 months after the procedure.
Initially described in the late 1980s, large loop excision of the transformation zone (LLETZ) is a conservative excisional method for the treatment of cervical intraepithelial neoplasia (CIN).[1, 2] Although success rates of treatment for CIN are very similar between the different destructive and excisional methods, LLETZ has gained widespread popularity amongst colposcopists because of its inherent advantages over the other available methods. It is a simple inexpensive technique, usually performed as an outpatient procedure under local analgesia, and often performed at the time of the assessment colposcopic examination. It facilitates comprehensive histological examination of the excised specimen with precise assessment of both the degree of abnormality and the excision margins. Unsuspected glandular and microinvasive disease may be unveiled.
In broad analysis, excisional procedures for the treatment of CIN have been shown to expose patients to an increased risk of subsequent obstetrical morbidity, especially premature delivery (PD).[4-6] In a recent meta-analysis, Arbyn et al. reported that cold-knife conisation, and probably both laser conisation and radical diathermy, were associated with an increased risk of subsequent perinatal mortality and other serious pregnancy outcomes. Although LLETZ, and ablative treatment with cryotherapy or laser, were not associated with a significantly increased risk of serious adverse pregnancy outcomes, there is growing evidence that the volume of tissue removed at LLETZ relates to the subsequent risk of PD.[6, 7] Because CIN mostly occurs in women of childbearing age, knowledge about the precise obstetrical risk associated with LLETZ and the mechanisms involved is important. The dimensions of the excised specimen have been found to be important predictors for subsequent premature delivery. This risk was shown to be significant in women whose cone depth exceeded 10 mm. Moreover, Noehr and colleagues found that increasing the cone length of the excised LLETZ specimen, as defined in the recent nomenclature classification of the International Federation of Cervical Pathology and Colposcopy (IFCPC), led to a 6% increase in the risk of preterm delivery per extra millimetre excised. Also, the risk of preterm delivery has been shown to increase dramatically in women who have had two or more LLETZ procedures.[9, 10] There are relatively few publications reporting the impact of the volume of the excised specimen. A recent retrospective study reported that both the volume and thickness of the excised LLETZ specimen is associated with a three-fold increased risk of preterm labour.
The aim of this study was to assess the impact of LLETZ on the dimensions of the uterus and the cervix. We also aimed to determine factors that have a significant impact on the reduction of the dimensions of the uterus and the cervix.
A prospective observational cohort study was conducted. From June 2007 to June 2010, women who were referred to the colposcopy clinic of the Adelaide and Meath and National Children's Hospital (AMNCH) for whom an LLETZ was required were invited to participate. The study protocol received approval from the AMNCH ethics committee. Written informed consent was obtained for all participants. After recruitment, patients had a three-dimensional transvaginal pelvic ultrasound scan (3DTVS) performed immediately before LLETZ treatment. According to our routine standard practice, after local anaesthetic infiltration all LLETZ procedures were performed under direct binocular colposcopic vision. Patients were invited for a second 3DTVS 6 months later for the post-LLETZ evaluation of the dimensions and volume of the uterus and cervix.
Immediately before LLETZ was performed, all included patients underwent standard colposcopic examination by one of three British Society for Colposcopy and Cervical Pathology (BSCCP) certified colposcopists. Specific attention was paid to reporting the transformation zone (TZ) type according to the IFCPC classification.[11-13] The length, thickness and circumference dimensions of the excised TZs were also documented by the pathologists (Figure 1).
We performed pre- and post-LLETZ 3DTVS on a Voluson 730 Expert (GE Medical Systems, Kretz, United States) after patients had emptied their bladder. The vaginal probe was introduced and positioned such that a satisfactory sagittal view of the whole uterus, with complete vision of the endometrial line and cervical canal, was obtained. Then, the measurement of uterine volume was performed holding the probe in the uterine sagittal plane. Pre- and post-LLETZ images of the uterus were anonymously stored in the hard drive memory of the machine. Blind analyses of all stored images and assessment of the length and volume of the uterus, and of the cervix, were secondarily performed by two different examiners. Analyses of the stored dimensions were performed after all patients had attended their follow-up scan appointments. For each scan recording, one measurement of uterine length, two of uterine volume and two of cervical volume were recorded by each examiner. The uterine length corresponded to the distance from the serosal surface of the fundus to the external os of the cervix. The uterine and cervical volume measurements were determined using the manual mode of the Voluson 730 volume calculation vocal software, starting from the same sagittal view of the uterus used for the measurement of uterine length. The volume of the uterus was determined after the contour of the uterus was traced on six images of the uterus obtained after six subsequent 30° rotations. Precisely the same technique was employed to determine the volume of the cervix. The upper limit of the cervix was arbitrarily defined as a plane perpendicular to the cervical canal positioned at the inferior limit of the endometrial line. Finally, all measurements were calculated as the mean of all the values obtained from each of the two examiners.
The exact dimensions of the LLETZ specimens were systematically calculated in the histopathology laboratory after the specimen had been opened and stored in formalin. The length, thickness and circumference dimensions of the excised TZ were documented by the pathologists (Figure 1). The thickness of the specimen was defined as the distance from the stromal margins to the surface of the epithelium, and the length as the distance from the external margin to the internal margin. The volume of the LLETZ was calculated by multiplying the three dimensions, i.e. length × thickness × perimeter, as described previously.
Statistical analysis was performed using spss® 17.0 for windows (SPSS Inc., Chicago, IL, USA). Means of the pre- and post-LLETZ uterine measurement values were compared using paired Student's t–tests. Factors with an impact on the reduction in dimensions and volume of the uterus and the cervix were identified using simple linear regression models. Inter- and intra-operator agreement between ultrasound scan measurements was evaluated using the intra-class correlation coefficient (ICC). A two-sided P value of less than 0.05 was considered to be statistically significant.
A total of 154 women were initially recruited to participate in the study. None of them was pregnant at the time of inclusion. Fourteen patients were secondarily excluded: five were pregnant at the time of their post-LLETZ scan and nine declined their second scan. In 34 cases, pre- and/or post-LLETZ stored images were not of suitable quality for proper three-dimensional assessment, and the corresponding patients were therefore subsequently excluded from the study. A total of 106 women with acceptable pre- and post-LLETZ 3DTVS stored images were then available for inclusion. The characteristics of these patients are summarised in Table 1. The mean delay between the pre- and post-LLETZ 3DTVS was 7.8 months (± 2.4 months SD). No patients had a type–3 TZ, and all had a satisfactory initial colposcopic examination. Eight (8.5%) patients had a history of previous LLETZ, but none of the patients had a cervical biopsy before the LLETZ was performed. Histopathological analysis of the LLETZ specimen diagnosed CIN2+ in 100 (94.3%) cases. Although the circumference and the length of the specimen showed heterogeneous distribution, the thickness of the specimen had quite a standardised distribution, and was > 10 mm in only four cases (3.8%). Finally, the mean volume of the excised specimen was 3.74 cm3 (± 3.7 cm3 SD). All of the LLETZ procedures were performed in the colposcopy clinic.
|Age (years) mean (± SD)||31.5 (± 7.0)|
|Previous history of LLETZ||8 (7.5)|
|Delay between the two ultrasound scans (months) mean (± SD)||7.8 (± 2.4)|
|Type of transition zone|
|Type 1||75 (70.8)|
|Type 2||31 (29.2)|
|Size of abnormal transition zone|
|Surgical specimen dimensions (mm) mean (± SD)|
|Thickness||6.2 (± 2.1)|
|Circumference||36.0 (± 8.8)|
|Length||14.7 (± 5.3)|
|Surgical specimen volume (cm3) mean (± SD)||3.74 (± 3.7)|
|Positive endomargins||14 (13.2)|
The evaluation of intra- and inter-examiner agreement between 3DTVS measurements of the uterus and the cervix is summarised in Table 2. The intra-examiner agreement for the evaluation of the volume of the uterus and the cervix was very high, with an ICC value systematically higher than 0.9. Inter-examiner agreement for the evaluation of the pre- and post-LLETZ length and volume of the uterus was also high. Pre- and post-LLETZ inter-examiner variation showed an ICC of 0.813 (95% CI 0.710–0.878) and 0.657 (95% CI 0.463–0.778) for the length of the uterus, and of 0.856 (95% CI 0.197–0.952) and 0.778 (95% CI –0.018 to 0.929) for the volume of the uterus, respectively. The pre- and post-LLETZ inter-examiner ICC values were lowest for the evaluation of the cervical volume: 0.380 (95% CI 0.152–0.557) and 0.253 (95% CI –0.027 to 0.481), respectively.
|Intra-class correlation coefficient (95% CI)|
|Operator A||Operator B|
|Length of the uterus||–||–||0.813 (0.710–0.878)|
|Volume of the uterus||0.984 (0.977–0.989)||0.984 (0.974–0.989)||0.856 (0.197–0.952)|
|Volume of the cervix||0.974 (0.955–0.984)||0.941 (0.877–0.968)||0.380 (0.152–0.557)|
|Length of the uterus||–||–||0.657 (0.463–0.778)|
|Volume of the uterus||0.988 (0.982–0.992)||0.976 (0.963–0.984)||0.778 (−0.018 to 0.929)|
|Volume of the cervix||0.938 (0.911–0.958)||0.943 (0.915–0.961)||0.253 (−0.027 to 0.481)|
Comparison between pre- and post-LLETZ 3DTVS showed a significant reduction in all dimensions measured, following LLETZ (Table 3). The mean reductions in the length and volume of the uterus, and in the volume of the cervix, following LLETZ were –0.46 cm, –6.12 cm3 and –1.60 cm3, respectively (P < 0.001 for all three values).
|Blind ultrasound scan measurements||P|
|Uterine Length (cm)||6.56 (±0.80)||6.10 (±0.80)||–0.46 (±0.58)||< 0.001|
|Volume of the uterus (cm3)||68.07 (±26.82)||61.95 (±24.35)||–6.12 (±14.29)||< 0.001|
|Volume of the cervix (cm3)||8.50 (±3.27)||6.89 (±2.52)||–1.60 (±2.29)||< 0.001|
Neither the delay between the pre- and post-LLETZ 3D TVS scans nor the type, nor size, of the abnormal TZ were found to have a significant impact on the reduction of the length of the uterus, volume of the uterus or volume of the cervix following LLETZ (Table 4). The volume of the surgical specimen was associated with a significant reduction of all measurements. Thus, the volume of the excised specimen had a significant impact on the reduction in the length of the uterus (β –0.038; 95% CI –0.068 to –0.009; P = 0.012), reduction in uterine volume (β, –0.791; 95% CI, –1.527 to –0.055; P = 0.036) and reduction in cervical volume (β, –0.121; 95% CI, –0.239 to –0.002; P = 0.046). Of the excised LLETZ specimen dimensions, only circumference was found to have a significant impact on the reduction in the length (β, –0.016; 95% CI, –0.028 to –0.003; P = 0.013) and volume (β, –0.413; 95% CI, –0.719 to –0.107; P = 0.009) of the uterus. The circumference of the specimen was not found to have a significant impact on the reduction of cervical volume (β, –0.043; 95% CI, –0.093 to 0.007; P = 0.090). Neither the thickness nor the length of the specimen showed a significant impact on the subsequent uterine or cervical dimensions.
|Dimensions of the uterus|
|Length of the uterus||Volume of the uterus||Volume of the cervix|
|βa (95% CI)||P||βa (95% CI)||P||βa (95% CI)||P|
|Delay b||0.005 (–0.043 to 0.052)||0.847||1.034 (–0.118 to 2.186)||0.078||0.042 (–0.145 to 0.230)||0.657|
|Type-2 transition zone c||–0.112 (–0.358 to 0.134)||0.370||–0.948 (–7.026 to 5.130)||0.758||–0.320 (–1.294 to 0.655)||0.517|
|Large size of abnormal transition zone||–0.050 (–0.289 to 0.189)||0.678||–0.585 (–6.465 to 5.296)||0.844||0.0422 (–0.902 to 0.986)||0.6930|
|Surgical specimen dimensions d|
|Thickness||–0.035 (–0.088 to 0.019)||0.200||–0.978 (–2.280 to 0.324)||0.139||–0.198 (–0.406 to 0.009)||0.061|
|Circumference||–0.016 (–0.028 to –0.003)||0.013||–0.413 (–0.719 to –0.107)||0.009||–0.043 (–0.093 to 0.007)||0.090|
|Length||–0.009 (–0.030 to 0.013)||0.421||–0.015 (–0.541 to 0.512)||0.955||–0.036 (–0.121 to 0.048)||0.395|
|Surgical specimen volume e||–0.038 (–0.068 to –0.009)||0.012||–0.791 (–1.527 to –0.055)||0.036||–0.121 (–0.239 to –0.002)||0.046|
The present study demonstrates that LLETZ performed for the treatment of CIN has a significant impact on the subsequently measured cervical and uterine volume. A reduction in uterine and cervical dimensions correlates with the volume of tissue removed at LLETZ. We found six previous studies that have evaluated the impact of LLETZ on cervical regeneration.[14-19] Four of them used a standard transvaginal two-dimensional ultrasound scan to estimate the influence on cervical length, two of which reported a significant reduction of cervical length following LLETZ.[14, 17-19] Only two previous studies have reported the effect of LLETZ on cervical volume. Founta et al. estimated the impact of LLETZ on cervical volume using magnetic resonance imaging (MRI). Papoutsis et al. recently used 3DTVS to estimate cervical regeneration. In keeping with our findings, these two studies demonstrated that cervical regeneration is affected relative to the proportion of excised tissue.
In our study we report the effect of LLETZ dimensions on both cervical and uterine volume. Also, the current study reports dimensions and volume evaluations where the measurement assessments were performed blindly and from anonymised stored 3D TVS after all patients had completed their 6 months post-LLETZ follow-up. 3DTVS volume calculation using a rotation technique has been shown to be reliable and valid; however, we think that blind analysis is likely to reduce observer bias. Moreover, we tried to reduce observer bias further by using two examiners. We hope that these methodological controls add value to this study and reduce subjective observer bias.
The mean reduction in uterine and cervical volume was 6.12 and 1.60 cm3, where the mean volume of the excised specimen was of 3.74 cm3. Perhaps surprisingly the impact of LLETZ appeared to be greater for uterine volume than for cervical volume (Table 3). This result should be interpreted with caution, and might reflect limitations of the 3DTVS technique of volume measurement of the uterus and cervix, particularly cervical volume. Indeed inter-operator agreement was lower for cervical than uterine volume measurements (Table 3). Also, an appreciation of the rotation method employed for the evaluation of volumes using 3DTVS will inform the interested reader in interpreting our results. Furthermore, whereas uterine volume measurement is relatively straightforward, cervical measurements are open to more interpretive subjective error. In order to reduce such subjectivity, we arbitrarily chose to define the upper limit of the cervix as a plane perpendicular to the cervical canal at the exact inferior limit of the endometrial line. This might explain why the mean estimated pre-LLETZ volume of the cervix was lower in the present study than in the two previous studies of cervical volume change after LLETZ.[15, 16] Also, the rotation measurement technique used in this study may not have reliably determined the upper limit of the cervix, as it is impossible to genuinely define the upper limit of the cervix immediately after the stored image of the uterus went on its first rotation.
Finally, unlike the fluid displacement technique performed in a volumetric tube, the assessment of the volume of tissue removed at LLETZ by the standard volume formula, and by assessing the three dimensions of the specimen, might not be the most representative assessment of the true volume, and somehow constitutes a limitation to the present study. Moreover, measurements were calculated in the histopathology department after the specimen had been stored in formalin. One should therefore consider the possibility of under-assessment, as formalin might be associated with shrinkage of the tissue.
Our results illustrate the subjectivity of cervical volume assessment using 3DTVS, and our inter-observer agreement was lowest for the crucial cervical volume measurement (Table 2). Nevertheless, accepting the limitations of the 3DTVS technique used, our results suggest the healing process following LLETZ has a significant impact on both the cervix and the uterus. In a previous two-dimensional transvaginal ultrasound scan study evaluating cervical regeneration after LLETZ, Paraskevaidis et al. showed that the healing process of the cervical crater was almost complete by 6 months after treatment. In their study, the mean crater size at 12 months was significantly smaller than the crater size immediately after LLETZ, and the cervical defect appeared similar whether the LLETZ was reported as small or large. The authors hypothesised that regeneration of the cervical tissue will compensate, at least in part, for the larger tissue loss. However, this study did not evaluate the dimensions of the whole uterus.
If our measurements were confined to cervical volume, we might have come to a similar conclusion, but our data on uterine volume loss reveal greater differences in post-LLETZ volumetric recovery. Surprisingly, we did not show any significant impact of the length, nor the thickness, of excision on subsequent uterine and cervical volume, nor on uterine length. Only the circumference was found to induce significant reduction in the volume and length of the uterus. This result is likely to be explained by the fact that, with a mean thickness of the excised cone of 6.2 mm, deep excisions were rare in our cohort. Indeed, the thickness of the LLETZ specimen exceeded 10 mm in only four (3.8%) patients.
Finally, there is a growing appreciation that the length of LLETZ excision is an important predictor of risk of subsequent premature delivery, but relatively little has been reported about the impact of the volume of tissue excised.[4, 7, 9] This study and that by Khalid et al. suggest that the volume of the excised specimen is an important predictor for the relative risk of subsequent premature delivery following LLETZ. Further work on this relationship is required.
The use of LLETZ for the treatment of CIN results in a significant reduction in cervical volume, as well as in the length and volume of the uterus. A reduction of uterine and cervical dimensions correlates with the volume of tissue removed at LLETZ. These results bring additional arguments, suggesting that every effort should be made to minimise the volume of the surgical specimen when performing a LLETZ, especially on young women with childbearing potential. Long-term studies are needed to assess how these findings correlate with the obstetric outcome.
None of the authors have any conflicts of interest to disclose regarding this publication.
XC and WP were the project leaders, and designed the study in collaboration with JAM, KF and MM. JAM, XC, KF, MM, SK and MJ collected the data. JM and XC analysed the data. XC, JM and WP wrote the article. All authors were responsible for critically reviewing the article. The corresponding and last author had final responsibility for the decision to submit for publication.
This study was approved by the research ethics committee of the AMNCH on 18 April 2007.
XC was supported by a grant from the ‘Fondation de France’.
We are particularly grateful to all of the women who participated in the study.