Cervical cancer has high morbidity and mortality across the world, accounting for an estimated 604,000 new cancer cases and 342,000 deaths worldwide in 2018 (1, 2). For cervical precancerous lesion treatment, excision procedures comprise three methods, i.e., loop electrosurgical excision procedure [LEEP; also called large loop excision of the transformation zone (LLETZ)], laser, and cold knife (3, 4).

According to the 2011 International Federation of Cervical Pathology and Colposcopy (IFCPC) terminology, the cervical transformation zone (TZ) is categorized into three types, with the squamous–columnar junction (SCJ) also separated into three types, and it is proposed that the cervical conization length of each transformation zone is different (5). In the 2011 IFCPC colposcopy terminology, TZ3 patients’ conization requires longer and larger excision of cervical tissue than type 1 or type 2 with a significant amount of endocervical epithelium. However, the precise conization length of different TZs was not defined and thus needs further investigation. The 2017 American Society for Colposcopy and Cervical Pathology (ASCCP) terminology categorized SCJ into two types (namely, fully visible SCJ and incompletely visible SCJ) and two types of treatment principles for cervical lesions (6, 7). The British National Health Service (NHS) cervical screening program gives more detailed information for three types of TZs with a wide range of conization lengths, i.e., 7–10, 10–15, and 15–25 mm for TZ1, TZ2, and TZ3, respectively (8).

However, currently, the LEEP conization length has not been standardized (9–11). According to a previous retrospective study carried out in our center in 2019, comparing the different margin status of conization specimens, the persistence rate of endocervical and stromal margin status has no significant statistical difference for internal margins, while internal margin and other margins (ectocervical margins status and negative margin) differ significantly (12). Therefore, to clarify, we group the endocervical and stromal margins as internal margins while the ectocervical margin as external margins, which is also demonstrated in Figure 1 .

In this study, we investigated cervical conization lengths and margin status in LEEP patients to explore a reasonable excision length for different types of TZs and to help achieve better outcomes of treatment simultaneously.

From July 2016 to September 2019, we prospectively enrolled patients with or suspicious of cervical precancer, including 1) cytology high-grade squamous intraepithelial lesion (HSIL)/ASC-H with positive human papillomavirus (HPV) results; 2) colposcopy-directed punch biopsy histological HSIL, regardless of the results of cytology and colposcopy; 3) cytology atypical glandular cells (AGCs) with positive HPV results; 4) colposcopy impression of HSIL with biopsy histopathology low-grade squamous intraepithelial lesion (LSIL). Exclusion criteria: 1) biopsy HSIL with stromal invasion, with invasion depth not determined; 2) cytology HSIL with biopsy confirmed invasive cancer. Cases were included in strict accordance with the inclusion criteria and exclusion criteria.

A total of 618 cases were included from five medical centers, i.e., Obstetrics and Gynecology Hospital of Fudan University, Renji Hospital Affiliated to Shanghai Jiaotong University School of Medicine, Shanghai First Maternal and Child Health Hospital Affiliated to Tongji University, Cancer Hospital Affiliated to Fudan University, and Shanghai Tongji Hospital. According to the LEEP histopathological report, the negative excision margin group includes cervical chronic inflammation, LSIL, HSIL with a negative margin, and adenocarcinoma in situ (AIS) with a negative resection margin. The margin with LSIL belonged to a negative margin. Margins equal to or higher than HSIL (HSIL, AIS, and malignancy) belonged to the positive margins. Five types of margin status were analyzed, including positive external, positive internal (endocervical or stromal), positive undetermined, and negative margin.

Stata 15.0 statistical software was used to analyze the data. t-Test and chi-square test analysis were used to analyze the differences in the mean value and proportion, respectively. The logistic regression model was used for univariate and multivariate analyses. p < 0.05 was considered statistically significant.

This multi-center prospective study recruited patients who underwent LEEP conization (13–15). After reviewing their histopathological results, the LEEP specimen margin was negative for 90.0% (556/618), and the margin was positive for 10.0% (62/618), 0.81% (5/618) of which were further diagnosed as invasive carcinoma. In this study, most patients with cervical lesions (90.0%) can obtain early diagnosis and precise treatment through a single LEEP conization. Patients with positive margins (9.2%) were triaged to follow-up within 3–6 months after surgery and repeat conization if necessary (16, 17). As a result, most patients can preserve their cervix through cervical conization, without the need for total hysterectomy. Red House Hospital’s single-center retrospective large sample data showed that LEEP specimen histopathology diagnosed 5.97% (759/12,713) as early cervical cancer (18), and the retrospective single-center data from Cancer Hospital affiliated to Fudan University of LEEP specimen histopathology diagnosed 8.60% (112/1,303) as early cervical cancer (19). LEEP specimen histopathology diagnosis rate of cervical cancer (0.81%) was lower than in that of two single centers. The reason might be that this study excluded patients with biopsy results suggesting cervical cancer with an uncertain depth of invasion, and this group of patients was found to have high proportions of invasive cervical cancer diagnosed by conization 61.9%.

Multivariate logistic regression analysis further confirmed that cytology ≥ HSIL and age are high-risk factors for LEEP positive internal margins, which indicates that the severity and duration of the disease might be related to the positive internal margin status. During the 3–6 months’ follow-up after conization, our previously published data and the cancer hospital team’s data showed that age, menopause, positive conization margin, diagnosis of microinvasive carcinoma by conization, abnormal cytology in postoperative follow-up, and positive HPV were high-risk factors for persistent HSIL. In the 3–6 months’ follow-up after conization, the proportion of persistent HSIL in patients with positive margin was 7.6% (46/609) and only 1.9% in patients with negative margin (55/2,964). The risk of persistent HSIL varies in the different positive margins, with positive external margin in 4.7% (13/278), positive undetermined (including stromal) margin in 9.7% (22/227), and positive endocervical margin in 13.2% (23/174). The data show that positive margins do not mean that lesions always persist after surgery, and positive internal margin had significantly higher risks of persistent HSIL when compared with positive external margin. During the 3–6 months’ follow-up after conization, only 7.6% of patients have persistent HSIL after conization on average, which means that up to 92.4% of patients with positive margins for HSIL lesion regressed naturally.

In this study, the LEEP margin of type 1, 2, and 3 transformation zones under different conization lengths was thoroughly investigated. In the condition of excision length of 7–15 mm for TZ1 and 10–15 mm for TZ2, this study showed that both the total positive margin rates and positive rate of external margin of TZ1 were similar to those of TZ2 (10.7% (16/150) vs. 10.6% (19/179), and 2.7% (4/150) vs. 5.1% (9/179), respectively). Hence, clinicians can simply choose a loop with an excision length of 10–15 mm for fully visible SCJ (TZ1 and TZ2) patients.

In 27 patients with TZ3 with positive margins, most (74.1%) were found when the excision length was 15 mm, and only 26.0% were found in the 16–25-mm group. Both the total positive margin rate and positive internal margin rate of the 15-mm group were significantly higher than those of the 16–25-mm group (12.2% (20/164) vs. 5.6% (7/125), 9.7% (16/164) vs. 2.7% (4/150)), showing that the excision length of TZ3 should be greater than 15 mm, and the clinician should choose loops with excision length greater than 15 mm. When excision length was 15, 16, 17, 18, 19, 20, 21, and 25 mm in TZ3, the positive rate of internal margin was 9.8% (16/164), 11.1% (3/27), 0% (0/9), 0% (0/47), 0% (0/1), 2.7% (1/37), 0% (0/1), and 0% (0/3), respectively. Hence, 17–20 mm can achieve more negative internal margins and retain a more normal cervix and thus might be optimal for TZ3 excision.

According to the trend line in Figure 2 , excision of 7–10 mm of the type 1 transformation zone patients is reasonable and safe for HSIL, except for the case of HSIL combined with AIS. If cytology and colposcopy suggest glandular disease, the excision depth needs to be >10 mm. For women who have completed childbirth, the excision length can reach 18–20 mm. The treatment of AIS should follow the latest Society of Gynecologic Oncology Evidence-Based Review and Recommendations of Adenocarcinoma in Situ (20–22).

The limitation of the study is that the sample size of each excision length was not evenly distributed in TZ3, while relatively limited cases were in the 17-mm, 19-mm, and longer than 20-mm groups. In the future, more data from these groups will help to achieve a more accurate conclusion.

In conclusion, the cervical LEEP aims to resect cervical precancerous lesions. However, up to now, the study of conization length is scant. In this study, we explored the reasonable and optimal conization length in patients with different types of TZs. The ideal cervical conization length for patients with TZ1 and TZ2 is 10–15 mm, while 17–25 mm is optimal for TZ3 excision with more negative internal margins.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

QC and YY had full access to all data and take responsibility for the integrity of the data and the accuracy of the analysis. Study concept and design: LS. Acquisition, analysis, and interpretation of data: QC, YY, YX, and YL. Drafting of the manuscript: QC and YY. Critical revision of the manuscript for important intellectual content: all authors. All authors contributed to the article and approved the submitted version.