High-grade squamous intraepithelial lesions (HSILs) encompassing cervical intraepithelial neoplasia (CIN) 2/3 have considerable potential for evolving into invasive cervical carcinoma and are closely linked to human papillomavirus (HPV) infection (1, 2). A retrospective study revealed that approximately 31% of CIN3 developed into invasive cancer within 30 years without treatment, albeit this ratio declined to 0.7% with regular treatment (3). Conization is the recommended treatment for HSIL (4). This therapy can be performed using cold-knife conization (CKC), laser conization, loop electro-excision procedure (LEEP), or large-loop excision of the transformation zone (5). LEEP and CKC are the most common surgical procedures conducted in China. LEEP has been demonstrated to be equivalent to CKC in preventing persistent CIN (6). When compared with LEEP, CKC is more complex; however, the rate of residual/recurrent lesions can be controlled at a lower level and the later rate of a hysterectomy can be reduced owing to deep surgical incision (7). Moreover, CKC can provide original tissue specimens that have not been electrothermally burned at the excised margin, thereby aiding the pathological diagnosis (8). As such, CKC is an important method for treating cervical lesions, especially HSIL. Nevertheless, the risk factors associated with residual/recurrent lesions post-CKC warrant elucidation (9). Furthermore, patients who have a pathological upgrade or encounter other complications after CKC may opt for a hysterectomy. Hence, the related risk factors must be investigated, guiding further clinical treatment based on these factors after CKC and hysterectomy (10, 11).

In this study, we aimed to determine the risk factors for RLC and RLH, as well as to develop programmed treatment recommendations based on them.

From January 2015 to October 2022, 1236 patients were diagnosed with CIN2/3 and underwent treatment via CKC at the First Affiliated Hospital of Henan University of Science and Technology. Of these, 260 patients were excluded because of incomplete TCT or HPV DNA follow-up data. Ultimately, 976 patients (mean age: 35.3 ± 9.7 years), with an average follow-up duration of 24.7 months (range: 6–60 months, median 22 months), were included in the study.

During the follow-up period, of the 976 patients, 152 underwent hysterectomy, of whom 53 were pathologically diagnosed with CIN2/3 after the surgery. The remaining 824 patients were followed up after the surgery, and 45 were diagnosed with CIN2/3 during the follow-up period. In this study, the mean interval between CKC surgery and the identification of residual/recurrent lesions was 9.7 months (range 6–48 months, median 12 months). Of the 45 patients diagnosed with CIN2/3 postoperatively, 25 were identified within 1 year, whereas 20 were diagnosed between 1 and 2 years. A total of 152 patients underwent a hysterectomy within 1–6 months following CKC.

In this study, RLC was significantly associated with factors such as persistent HPV infection, HPV infection types (preoperative and postoperative), TCT test (6-month postoperative), postoperative margins, and ECC results (positive ECC ≥CIN2/3) (P < 0.05) ( Table 1 ). However, age, menopausal status, pregnancy, parity, transformation zone, the height of excision, glandular involvement, and TCT test (preoperative) were not significantly associated with RLC (P > 0.05) ( Table 1 ).

During the 6-month follow-up period, 824 patients who had undergone treatment underwent TCT testing. TCT test ≥ASC-US was made in 137 patients, which was confirmed via biopsy and found to be CIN2/3 in 36 patients. The likelihood of residual/recurrent lesion was significantly higher in patients with TCT test ≥ASC-US during the 6-month follow-up (P < 0.001) ( Table 2 ). In this investigation, 824 patients underwent HPV DNA testing at the 6-month follow-up after CKC, and HPV infection was detected in 210 of the 824 cases. Of these, 35 were pathologically confirmed to have CIN2/3. The likelihood of being diagnosed with CIN2/3 was significantly higher in patients infected with HPV 16/18 than in those who were HPV-negative at the 6-month follow-up (P < 0.001) ( Table 2 ).

In patients who underwent a hysterectomy, RLH was associated with transformation zone, the height of excision, glandular involvement, HPV infection types (preoperative), TCT test (preoperative), postoperative margins, and ECC results (P < 0.05). However, age, menopausal status, pregnancy, and parity were not significantly associated with RLH (P > 0.05) ( Table 3 ).

Based on the results of univariate analysis, a multivariate logistic regression analysis with adjusted variables was performed to determine the independent high-risk factors for RLC and RLH. Of these variables, persistent HPV infection (odds ratio [OR] 9.237, 95% confidence interval [CI] 1.858–41.519), HPV 16/18 infection (preoperative, OR 7.614, 95% CI 1.345–29.472), positive margins (OR 9.751, 95% CI 1.681–42.886), TCT test ≥ ASC-US (6-month postoperative, OR 14.527, 95% CI 2.373–52.721), and HPV 16/18 infection (6-month postoperative, OR 8.886, 95% CI 1.514–35.459) emerged as independent risk factors for RLC (P < 0.05) ( Table 4 ). HPV 16/18 infection (preoperative, OR 2.613, 95% CI 1.226–5.569), transformation zone (type 3, OR 2.315, 95% CI 1.112–4.820), positive margins (OR 2.959, 95% CI 1.276–6.866), and positive ECC (OR 2.400, 95% CI 1.044–5.518) were independent risk factors for RLH (P < 0.05) ( Table 5 ).

Past studies have mostly focused on the risk factors for residual/recurrent lesions after cervical cone biopsy (including LEEP and CKC) (17). In contrast, in this study, only those patients who underwent CKC were included so as to reduce the interference of confounding factors. Furthermore, CKC has a lower postoperative infection rate, a more comprehensive cone cut, a lower incidence of residual/recurrent lesions, more accurate pathological examination results, and more precise pathological information, especially for evaluating squamous intraepithelial lesion grading and stromal invasion (18, 19). In CKC, the entire uterus is not removed, which renders the possibility of residual/recurrent lesions after the procedure (20). Long-term meticulous follow-up and postoperative treatment are thus crucial in such cases (21). Moreover, the risk factors for RLC and RLH should be analyzed to provide reasonable suggestions for clinical work-up.

In this study, a follow-up analysis was conducted on 976 patients who received CKC treatment for CIN2/3, with an average follow-up time of 24.7 months. This duration exceeded the follow-up time of most relevant studies, which permitted the detection of residual/recurrent lesions in a comprehensive manner (22). Throughout the follow-up period, TCT testing and HPV infection testing were implemented as a part of the monitoring protocol. TCT test is recognized as the most precise method for diagnosing high-grade intraepithelial neoplasia, whereas HPV infection is acknowledged as a key contributor to precancerous alterations under gynecological conditions (23). Furthermore, we selected a combination of residual and recurrent lesions for analysis, as persistent lesion progression typically follows a continuum that is challenging to strictly delineate at a single time point. Regardless of residual or recurrent status, patients require comparable follow-up strategies (e.g., HPV testing, TCT, and colposcopic biopsy) as well as appropriate therapeutic interventions (e.g., hysterectomy).

Several factors can affect RLC (9). For instance, in our study, RLC was significantly associated with factors such as persistent HPV infection, HPV infection types (preoperative and postoperative), TCT test (6-month postoperative), postoperative margins, and ECC results (P < 0.05). Nonetheless, age, menopausal status, pregnancy, parity, transformation zone, the height of excision, glandular involvement, and TCT test (preoperative) were not significantly associated with RLC (P > 0.05). Previous studies suggest persistent HPV infection as an important risk factor for RLC, which was confirmed in this study through a long-term follow-up. During the follow-up period, especially in patients diagnosed with HPV16/18 infection, the risk of developing RLC was higher. Consistent with other studies, women with positive margins exhibited a relatively higher risk of recurrence (24, 25). In addition, TCT test ≥ASC-US (6-month postoperative) has been reported to be a possible risk factor that necessitates further exploration for residual/recurrent cases (26). To the best of our knowledge, this is the first study to investigate the correlation between preoperative ECC results and residual/recurrent lesions during follow-up. A positive ECC result indicates deeper penetration of the lesion into the cervical canal, which substantially increases the risk of recurrence and is statistically significant. To assess the independent risk factors for RLC, a multivariate logistic regression analysis was conducted. The findings indicated that persistent HPV infection, HPV 16/18 infection (preoperative), positive margins, TCT test ≥ASC-US (6- month postoperative), and HPV 16/18 infection (6-month postoperative) were independent risk factors for RLC.

In patients undergoing a hysterectomy after CKC, RLH was associated with the transformation zone, the height of excision, glandular involvement, HPV infection types (preoperative), TCT test (preoperative), postoperative margins, and ECC results (P < 0.05). HPV 16/18 infection (preoperative), transformation zone (type 3), positive margins, and positive ECC as independent risk factors.

TCT is necessary for preoperative examination and postoperative follow-up and is related to the initial diagnosis and the selection of surgical approach for patients. The findings of this study indicated that TCT test ≥ASC-US (preoperative) is a risk factor for RLH in patients who have undergone a hysterectomy. This result signifies that the higher the malignancy grade diagnosed preoperatively, the greater is the risk of RLH (27, 28). Qing reported that a cervical excision length of 10–15 mm is reasonable for patients with TZ1 and TZ2, whereas 17–25 mm is optimal for TZ3 excision with more negative internal margins (5). The type of TZ2 and TZ3 indicates whether the cervical transformation zone is partly visible or not visible. Therefore, our results suggest that the transformation zone and the height of the excision are the risk factors that prompt further examinations for residual lesions. Kang observed that women infected with HPV16 and HPV18 are more likely to have residual lesions after LEEP (29). Our results also revealed that positive HPV16/18 status is an independent predictive factor for RLH. In line with numerous other studies, our research asserted that a positive margin is an independent risk factor for RLH. Furthermore, it is worth noting that among patients with positive margins who underwent a hysterectomy, the proportion of pathologically confirmed nonresidual lesions were 58.25% (60/103). Hence, we contend that a total hysterectomy is not an optimal choice merely because the margins are positive. Our observations imply that positive ECC are independent risk factors for RLH, which are related to the fact that they suggest the presence of abnormal cells or pathological tissues within the cervical canal. These pathologies may include CIN, especially high-grade CIN, or even cervical cancer. Similarly, glandular involvement signifies the infiltration of the lesional tissue into the subepithelial glandular structures, contributing to the incomplete excision of the lesion and, consequently, increased risk of residual/recurrent lesions. Our findings revealed a significant association between glandular involvement and RLH, indicating that pathological evidence of glandular involvement after hysterectomy is an important factor influencing residual lesions. However, logistic regression analysis demonstrated that glandular involvement was not an independent risk factor for RLH, suggesting that its impact on RLH loses independence when comprehensively analyzed alongside other factors, potentially being masked or confounded by them.

According to the guidelines of the American Society for Colposcopy and Cervical Pathology and the National Comprehensive Cancer Network in the United States, patients with positive margins following initial cone biopsy may opt for close follow-up, undergo further diagnostic procedures for cervical lesions, or proceed to total hysterectomy (30). In our research, of the 824 patients who were followed up, 116 presented with positive resection margins. Of these 116 patients, 18 (15.52%) exhibited residual/recurrent lesions. Of the 152 patients who underwent a hysterectomy, 103 showed positive margins before the operation, and 43 (41.75%) of these 103 patients (41.75%) had residual lesions. Positive margins act as independent risk factors for RLC and RLH. However, the proportion of patients who ultimately developed CIN2/3 in both groups of patients with positive margins was <50%. Consequently, the treatment plan and surgical approach for patients could not be determined merely based on this single factor.

Our study specifically focused on patients who underwent CKC treatment, thereby minimizing confounding interference from different surgical approaches in risk factor assessment. Additionally, this investigation comprehensively evaluated RLC and RLH patients, enhancing the precision and inclusiveness of clinical decision-making. However, our study is limited by its retrospective design and the single-center nature, as well as a lack of data related to preoperative and follow-up HPV viral load and the extended follow-up after CKC treatment. Furthermore, considering that this is a retrospective study, postoperative immunohistochemical outcomes, such as the expression of Ki-67 and P16, were unavailable, which may also function as risk factors for recurrent or residual lesions (31, 32).

In conclusion, our findings suggest that patients should be managed in a stratified manner, with those having high-risk factors requiring more frequent follow-ups with shorter intervals to mitigate the risk of recurrence: an initial follow-up at 3 months postoperatively (combining TCT and HPV testing), followed by assessments every 3–6 months (for at least 2 years). After 2 years, if results remain persistently negative, the interval may be extended to annually. Whether patients require additional surgery should also be dealt with differently. Those with high-risk factors, no fertility desires, and advanced age can opt for a hysterectomy. This integrated management approach can contribute to establishing a comprehensive treatment plan and alleviating patients’ anxiety.