The Appropriate of Cone Depth in Loop Electrical Excision Procedure (LEEP) for Negative Pathological Margin from High Grade Precancerous Lesion of Cervix, Retrospective Study.

OBJECTIVE: To determine the appropriate cone depth for treating high grade precancerous lesions to achieve negative pathological margins of cones from LEEPs. Other factors associated with positive pathological margin were also investigated.
METHODS: A Retrospective study recruited 170 patients who received indications for LEEP during January 2015 to July 2020 were enrolled. The participants were operated by a single cut of LEEP and not had previously conization before. All patient data were collected into two groups, including negative and positive cone margin groups. Then, we used the cone depth by calculating from cone tissue after formalin fixation to eliminate shrinkage effect. The appropriate cut-off points for cone depth were calculated by ROC and analyzed factors that influence positive cone margin.
RESULTS: The depth of cone (mm ±SD) of negative margin group was 8.70 (±3.36) and 6.13 (±2.28) mm in positive margin group. The appropriate cut-off points for cone depth were calculated by ROC presented at resection depth of 7.21 mm, which displayed proper cone depth with a sensitivity of 63.53% and specificity of 71.76%. Elderly age (adjusted OR 1.061, 95%CI 1.008-1.117, p=0.002), number of quadrants of lesion involvement (adjusted OR 1.182, 95%CI 1.312-2.513, p=<0.001) and glandular involvement (adjusted OR 3.648, 95%CI 1.605-8.292, p=0.002) were the significant risk factors for positive margin.
CONCLUSION: The appropriate cone depth for treating high grade precancerous lesions was at least 7.21 mm to achieve a negative cone margin from LEEP. The significant factors associated with positive cone margin include elderly age, more quadrants of lesion involvement and glandular involvement.

Introduction

Cervical cancer has the third highest incidence of cancer among Thai women. There are approximately 5,500 new cases per year, an estimated 11.79 cases of cervical cancer cases per 100,000 Thai population and 93.5% are found in women older than 35 to 60 years (Sangrajrang et al., 2020). However developed countries, has a much smaller incidence. The reduced incidence is due to the provision of education to the population, raised awareness by comprehensive cervical cancer screening program, effective treatment and monitoring from the onset of abnormalities with low precancerous lesions and high precancerous lesions until it progressed to cancer (Sherris, 2001). Usually, lesions preceding cervical cancer have a gradual disease progression, and cure may be possible if there are abnormalities in cytology screening (Šarenac and Mikov, 2019). Therefore, no new cervical cancer patients should occur ideally.

When the cervical cytological screening test has abnormal results, the physicians can perform additional tests to confirm diagnosis and then has perform various treatment methods. Cervical conization was the one method used to treat high grade precancerous lesion of cervix (CIN2,3) which can be performed with different technique, for example; loop electrosurgical excision procedure (LEEP), cold knife conization, and carbon dioxide laser (CO2 laser) (Khan and Smith-McCune, 2014). In the gynecological department of HRH Princess Maha Chakri Sirindhorn Medical Center, diagnostic and therapeutic operations have performed mainly using a loop electrosurgical excision procedure (LEEP). Consequently, the excised tissue is sent to pathologist for examination of the margin and severity of lesions.

From the current study, the complication of cervical excision procedure can occur such as blood loss, infection at the excision site of cervix, cervical stenosis and long-term effects on fertility (Kietpeerakool et al., 2006). The consequent pregnancy may be increased risk for miscarriage and preterm birth (Kietpeerakool et al., 2006; Ciavattini et al., 2015; Kyrgiou et al., 2016; Liverani et al., 2016). The wide of cone did not associate with a bad consequent pregnancy outcome; however, this dimension depended on size and location of lesion on cervix of individual patients. Therefore, whether the lesion is large or laterally located, the wide of cone should be wider for negatively ectocervical margin. Due to the consequence pregnancy complications, many studies have been concerned to determine the cutting depth of cone excision. For example, a cervical excision greater than 15 millimeters(mm) in depth would significantly increase preterm birth of consequence pregnancy (Castanon et al., 2014) and will affect various pregnancy complications such as spontaneous preterm birth, premature rupture of membranes, chorioamnionitis, low birth weight, admission to neonatal intensive care, and perinatal mortality (Kyrgiou et al., 2016).

Nowadays, the newest guideline to treat high grade precancerous cervical lesion had raised more awareness of the excisional procedure but still the excisional procedure has some higher benefits (Martin-Hirsch et al., 2010; Perkins et al., 2020). Therefore, the lesser cone depth for adequate treatment result should have been expected. Despite this, we still did not know the appropriate cone depth for negative endocervical and/or deep margin. According to Beyer et al., (2017) reported the cone depth by LEEP in the 10-19.9 mm. range had a 73% chance of negative margin with AUC = 0.657, 79% for sensitivity and 64% for specificity. However, the previous studies did not report a clearly optimal cut-off value of the cone depth to negative cone margin. Therefore, the purpose of our study was to find the appropriate cone depth by LEEP for negative endocervical and/or deep margin and find the risk factors which might have an effect on positive endocervical and/or deep margin.

Materials and Methods

This was a retrospective study which recruited the patients who had high grade precancerous lesion of cervix (CIN 2, 3), who visited the gynecology department for diagnostic and therapeutic operations performed using a loop electrosurgical excision procedure (LEEP) at the HRH Princess Maha Chakri Sirindhorn Medical Center (MSMC) during January 2015 to July 2020. This study was approved from the institution Ethics Committee of Faculty of Medicine, Srinakharinwirot University (registry number: SWUEC-051/2563E) for research involving human subjects.

The inclusion criteria consisted of female patients who i) had abnormal cervical cytology with colposcopic directed biopsy with pathological results of CIN 2 or CIN 3 and treated by using LEEP. ii) had HSIL of cervical cytology underwent colposcopic exam and LEEP at the same time (see and treat technique). iii) who underwent LEEP for diagnosis in case of inconsistent cytological and pathological result. Exclusion criteria were defined as: i) had previously performed LEEP or other cervical conization method was done. ii) There was more than one cut of the cervix with an electric loop in the same treatment cycle iii) the pathological result showed cervical cancer and iv) the pathological results were not high grade precancerous lesions such CIN 1, cervicitis, or HPV infection. We used the maximum of cone depth by the part of gross examination in the pathological report of LEEP specimen that had a shrinkage effect from formalin fixation so we reflect the cone depth of fresh tissue by calculation. Based on Boonstra (1983) reported the shrinkage effect of fresh tissue by formalin fixation was 8%. Therefore, we add 8% of depth which reported on the gross examinations after formalin fixation to reflect the true depth of fresh cone specimen. Surgical margins were considered positive when the high grade dysplasia was closer than 1 millimeter to the endocervical or deep margin, however for >1 millimeter distance to the lesion was considered negative (Beyer et al., 2017; Sopracordevole et al., 2019).

Sample size calculation was performed using a formula for estimating accuracy index of diagnostic studies, involving ROC analysis. The study of 109 patients published by Beyer (2017) showed the cutting depth in the 10-19.9 mm. range had a 73% chance negative margin of cervical dysplasia with AUC = 0.657, 79% of sensitivity and 64% of specificity. This study needed more preciseness. Therefore, the sample size was calculated based on an expected 0.8 of AUC. An extra 10% was added to the calculated sample size for any missing. Thus, the total sample size of at least 170 participants was needed for inclusion.

Statistical analysis was performed using STATA version 13 (StataCorp, College Station, TX, USA). The efficacy of the cone depth including sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for negative cone margin at different cut-off points was examined. The receiver operating characteristic analysis (ROC) was used to find the most appropriate cut-off point of cone depth for negative cone margin. Lastly, a logistic regression model was used to analyze factors that can influence the positive cone margin. The level of significance was defined as α=0.05%.

Results

The 294 patients were eligible for inclusion criteria and enrolled in this study during the period between January 2015 to July 2020. We excluded 124 patients because they i) had previously performed LEEP or other cervical conization methods (20 patients). ii) There was more than one cut of the cervix with an electric loop in the same treatment cycle (24 patients) iii) the pathological result showed cervical cancer (21patients) and iv) the pathological results were not high grade precancerous lesions (59 patients). So, the total number of patients was 170 who enrolled in this study. The demographic characteristics of the both study populations are shown in table 1. The mean (±SD) age of the patients was 38.24 (±7.68) years in the negative cone margin group and 43.52 (±11.24) years in positive cone margin group. The majority was multiparous and had premenopausal status in both groups. The interval time (day ±SD) after diagnosing HSIL of cervical cytologic result or colposcopic exam until LEEP were 44.47 (±25.31) days in the negative cone margin group and 39.61 (±20.22) days in positive cone margin group. In the part of colposcope finding, most of the participants showed satisfactory findings and the lesion involved at least three quadrants. For the histopathologic result, the depth of cone (mm ±SD) of the negative cone margin group was 8.70 (±3.36) and 6.13 (±2.28) mm in the positive cone margin group. Most of them had no glandular involvement and the gynecologic oncologists dominated out of all the surgeons.

Table 1

Demographic Characteristics of Patients

Characteristic	Negative cone margin	Positive cone margin
Age (mean ± SD) (year)	38.24 (±7.68)	43.52 (±11.24)
Parity status
Nulliparous	20	21
Multiparous	65	64
Menopausal status
Menopause	8	17
Pre-menopause	77	68
Interval after diagnosis until LEEP (mean ± SD) (day)	44.47 (±25.31)	39.61 (±20.22)
Satisfactory colposcopic exam
Satisfactory	82	74
Unsatisfactory	3	11
Quadrant of lesion involvement
1	23	11
2	43	33
3	2	3
4	17	38
Depth of cone (mean ± SD) (mm.)	8.70 (±3.36)	6.13 (±2.28)
Glandular involvement
Yes	48	72
No	37	13
Surgeon
Gynecologic oncologist	70	68
Fellow	10	11
Resident	5	6

The level of cone depth for negative margin had presented the performance of each level as shown in table 2. From the ROC (Figure 1), a cone depth at a level of 7.21 mm displayed the most appropriate for negative cone margin with a sensitivity, specificity, positive predictive value, negative predictive value and accuracy was 63.53%, 71.76%, 69.2%, 66.3%, and 67.65% respectively.

Table 2

The Performance of Level of Cone Depth for Negative Margin

Cone depth (mm.)	Sensitivity	Specificity	PPV*	NPV**	Accuracy
3.09	100.00%	0.00%	-	-	50.00%
4.12	97.65%	9.41%	51.90%	80.00%	53.53%
5.15	92.94%	12.94%	51.60%	64.70%	52.94%
6.18	80.00%	57.65%	65.40%	74.20%	68.82%
7.21 †	63.53%	71.76%	69.20%	66.30%	67.65%
8.24	49.41%	87.06%	79.20%	63.20%	68.24%
9.27	45.88%	89.41%	81.30%	67.40%	67.65%
10.3	44.71%	89.41%	80.90%	61.80%	67.06%
11.33	20.00%	95.29%	81.00%	54.40%	57.65%
12.36	20.00%	96.47%	85.00%	54.70%	58.24%
13.39	14.12%	97.65%	85.70%	53.20%	55.88%
14.42	9.41%	97.65%	80.00%	51.90%	53.53%
15.45	7.06%	100.00%	100.00%	51.80%	53.53%
>15.45	0.00%	100.00%	-	-	50.00%

†, the most appropriate cut-off point; *PPV, positive predictive value; * NPV, Negative predictive value

Figure 1

The Level of Cone Depth for Negative Margin

Concerning the secondary analysis, the association between risk factors and positive cone margin was evaluated. The result showed an age of patient, number of quadrants of lesion involvement, satisfactory colposocpic exam and glandular involvement were significant risk factors for positive cone margin by LEEP as shown in table 3. Finally, a multivariate logistic regression model was used. Overall, it is confirmed that elderly age of patient had 6.1% increased risk of positive cone margin (adjusted OR 1.061, 95%CI 1.008-1.117, p=0.002), same as number of quadrants of lesion involvement had increase 18.2% (adjusted OR 1.182, 95%CI 1.312-2.513, p=<0.001) and glandular involvement had 3.648 times higher risk for having positive cone margin (adjusted OR 3.648, 95%CI 1.605-8.292, p=0.002) as shown in Table 4.

Table 3

The Association of Risk Factors and Positive Cone Margin by Bivariate Analysis

Factor	Crude OR	95% CI	P value
Age	1.062	1.025-1.101	0.001*
Parity status	1.291	0.982-1.696	0.067
Menopausal status	2.406	0.977-5.927	0.056
Satisfactory colposcopic exam	4.063	1.091-15.130	0.037*
Quadrant of lesion involvement	1.687	1.271-2.239	0.001*
Glandular involvement	4.269	2.058-8.858	0.001*
Surgeon
Gynecologic oncologist	1	-	-
Fellow	1.132	0.452-2.839	0.791
Resident	1.235	0.360-4.238	0.737

* significant factor

Table 4

The Association of Risk Factors and Not Free Cone Margin by Multivariate Analysis

Factor	Adjusted OR	95% CI	P value
Age	1.061	1.008-1.117	0.002
Quadrant of lesion involvement	1.182	1.312-2.513	< 0.001
Glandular involvement	3.648	1.605-8.292	0.002

*, Factors that include in multivariate analysis were age, parity status, menopausal status, satisfactory colposcopic exam, quadrant of lesion involvement, glandular involvement

Discussion

High grade precancerous lesion of cervix was currently considered the precursor of invasive cervical cancer (Sopracordevole et al, 2019). The incidence was mostly in late reproductive to premenopausal period, the disease progressed in a gradual period and easier to screening than other organs (Šarenac and Mikov, 2019). When the doctors found abnormal cytological screening results such as high grade intraepithelial neoplasia (CIN2,3) can be performed conization for treatment, especially LEEP technique (Šarenac and Mikov, 2019; Apgar et al., 2013). Several previous studies reported more deeper of cone depth than our study for negative pathological margin. Prendiville et al., (1995) found optimal LEEP cone depth of 7 to 15 mm, Bayer et al., (2017) reported negative cone margin was achieved in 100% with a cone depth of ≥20 mm and Kyrgiou (2016) reported resection depth between 10-19.9 mm led to 73.0% negative cone margin. In addition, Kliemann et al., (2012) reported 10 - 14 mm of cone depth considering the percentage chance of free margins in the order of 84–97%, the likelihood of complete resection of CIN 2,3 was approximately 100% with a cone depth of 20.0 mm, 95% with a cone depth of 13.0 mm and 90% with a cone depth of 11.0 mm. In the olden days, Anderson et al., (1980) reported destruction of tissue up to 3.80 mm in depth for eradication of 99.7% of CIN 3 lesions which had an interesting point to consider lesser cone depth. In this decade, many studies attempted to raise awareness about the depth of conization for purpose of recede complications after treatment such as bleeding, cervical stenosis, abortion and preterm birth in subsequent pregnancy (Nam, 2010; Khan and Smith-McCune, 2014; Oz et al, 2016). Therefore, lesser cone depth for adequate treatment should have been expected.

This study showed the performance of cone depth for negative pathological margin with the appropriate cut-off point of cone depth for negative margin was 7.21 mm (AUC 0.737 with 0.663-0.812 of 95%CI) for lesser the cone depth from LEEP while the outcome of treatment for negative pathological margin was still acceptable. The reason to support our results was from Shafi et al., (2007) study, which reported CIN can involve the gland crypt of endocervix range from 1–2 mm to a maximum of 5.22 mm in depth. Therefore, the cone depth of 7.21 mm was adequate for negative pathological margin.

The study which reported the lesser cone depth for adequate treatment of high grade precancerous lesion as similar to our study had many studies in recently. Sopracordevole et al., (2019) reported in his study that from 1996 to 2015 showed a significant decrease in the length of cone excision over the whole study period, with a 30% reduction in the length of cone specimens. Pina et al., (2013) suggested that the shallowest LEEP that will enable complete removal of the lesion should be performed. It may be safer for women to defer pregnancy for several months after conization. Therefore, the trend of lesser cone depth was ensuring adequate treatment and should be considered.

The associated factors for positive pathological margin were elderly age (adjusted OR 1.061, 95% CI 1.008-1.117, p=0.002), more quadrants of lesion involvement (adjusted OR 1.182, 95% CI 1.312-2.513, p<0.001) and the most important factor from our study was glandular involvement status (adjusted OR 3.648, 95% CI 1.605-8.292, p=0.002). This result was consistent with many previous studies (Zun et al., 2007; Kir et al., 2012; Guducu et al, 2013; Erkiling et al, 2018; Newtorn et al., 2020). Kir et al., (2021) reported in previous study, the cone depth on the patient who had a glandular involvement might deeper than without glandular involvement in purpose to negative cone margin. In the subgroup analysis of our study, the patient who had glandular involvement had the same appropriate cut-off point of cone depth for negative margin (7.21 mm) with sensitivity, specificity, positive predictive value and negative predictive value was 66.67%, 72.22%, 61.5% and 76.5% respectively (data shown in appendix). Therefore, even if positive glandular involvement, the cone depth might not need to be deeper than negative glandular involvement with the same level of efficacy. For the same reason, we can use the see and treat technique which omits colposcopic directed biopsy and the need to be evaluate glandular involvement on biopsy specimens before LEEP for the least of cone depth.

The strength of this study was a first study to reveal the appropriate definite cut off point of cone depth from LEEP to achieve negative pathological margins with acceptable sensitivity and specificity and appropriate sample size to a summarize the result. That supporting the treatment of patients nowadays raised awareness of the complications, especially those dealing with obstetric complications, that will occur associated with many previous studies. However, the study had limitation. Our retrospective study included patients who primarily visited gynecology department of the HRH Princess Maha Chakri Sirindhorn Medical Center (MSMC) for diagnostic and therapeutic operations performed using a loop electrosurgical excision procedure (LEEP). The majority were Thai and lived in central of Thailand e.g. Bangkok, Pathum-Thani, Nakhon -Nayok and Prachinburi province. Therefore, there was no diversity among the populations.

In conclusion, the appropriate cone depth from LEEP for treating high grade precancerous lesions was at least 7.21 mm to achieve negative pathological margins. In addition, the significant risk factors associated with positive pathological margin were elderly age, more quadrants of lesion involvement and glandular involvement. On the other hand, parity status, menopausal status, satisfactory colposcopic result and surgeon’s experience was not a risk factor. However, the gynecologist should carefully consider the individual risk factors of each patient and concern for the minimal cone depth to optimize the treatment outcome with the least complication.

Author Contribution Statement

Conceptualization: R.N., S.N.; Data curation: R.N., S.N.; Formal analysis: R.N.; Investigation: R.N., S.N.; Methodology: R.N., S.N.; Project administration: R.N.; Resources: S.N.; Supervision: R.N.; Validation: R.N.; Visualization: R.N.; Writing - original draft: S.N.; Writing - review & editing: R.N.