Abstract
Objective
To evaluate the risk factors of recurrent high‐grade cervical intraepithelial neoplasia grade 2 or worse (CIN2+) after loop electrosurgical excision procedure (LEEP).
Methods
This retrospective study included patients with histopathologically confirmed CIN2/3 who underwent LEEP in 2015–2020. Cox regression analysis was used to evaluate the risk factors of recurrence.
Results
Recurrent CIN2+ was found in 268 patients after LEEP (268/4369, recurrence rate, 6.1%). High‐risk (hr‐) HPV infection (hazard ratio [HR] 12.09, 95% confidence interval [CI] 7.78–18.79), margin status (HR 6.48, 95% CI 4.75–8.84), baseline diagnosis (HR 1.45, 95% CI 1.08–1.95), smoking (HR 3.17, 95% CI 2.27–4.43), and immunosuppression (HR 1.96, 95% CI 1.33–2.91) were significant independent risk factors of recurrence. HPV16 (HR 3.61, 95% CI 2.43–5.37), HPV33 (HR 2.62, 95% CI 1.12–6.12), and HPV52 (HR 1.61, 95% CI 1.02–2.55) infection showed a higher risk of recurrence. High‐risk HPV had the highest accuracy (sensitivity 88.5%; negative predictive values 98.7%) in predicting recurrence compared with liquid‐based cytology test and margins.
Conclusion
Given that positive margins present a higher risk, wide excision may be required to avoid residual lesions. More attention should be paid to the correlation between recurrence and hr‐HPV genotypes. After treatment for high‐grade CIN, HPV‐based testing is recommended at 6 months. Timely identification of high‐risk factors enables risk stratification, and enables individual management or individual follow‐up and recall strategies.
Keywords: cervical intraepithelial neoplasia, conization, HPV, loop electrosurgical excision procedure
Synopsis
Post‐treatment CIN2+ was associated with high‐risk HPV infection, margin status, baseline diagnosis, smoking, and immunosuppression.
1. INTRODUCTION
Cervical cancer causes high morbidity and mortality in low‐ and middle‐income countries. 1 China launched an organized, population‐based cervical screening program in 2009, and expanded the national program in 2012 to offer annual screening to 10 million women. 2 The purpose of cervical screening is to identify and treat women with precancerous lesions before progression to invasive cancer. 3 Currently, women with high‐grade cervical intraepithelial neoplasia (CIN) are treated typically by loop electrosurgical excision procedure (LEEP) to prevent the development of cervical cancer. However, previous studies have shown that compared with women with a normal liquid‐based cytology test (LCT) or women in the general population, women with high‐grade CIN (CIN2/3) still have an increased risk of developing recurrent CIN2 or worse (CIN2+). 4 , 5
A few studies evaluate the risk of recurrence after LEEP for women with CIN2/3, but they report discordant results. 6 , 7 , 8 , 9 , 10 There is still uncertainty and debate as to which risk factors or combination of factors most accurately predict recurrence. Most post‐treatment surveillance reports were limited in follow‐up duration, data quality, and study size. In the present study, we aimed to examine the risk factors of recurrent CIN2+ after LEEP in women with CIN2/3. As secondary outcomes, we sought to identify the accuracy of risk factors in predicting recurrence.
2. MATERIALS AND METHODS
This was a retrospective study approved by the Institutional Review Board, which determined that participant informed consent was not required. The study collected clinical information on patients with high‐grade CIN (CIN2/3) who underwent LEEP at the West China Second Hospital of Sichuan University from 2015 to 2020. The inclusion criteria were: (1) baseline cone biopsy was diagnosed as CIN2/3; (2) squamous cell lesions; (3) cervical conization performed with LEEP; (4) patients with at least 6 months of follow‐up data. Exclusion criteria were: (1) patients without follow‐up data; (2) diagnosis of invasive cancer at the time of conization; (3) patients treated by other methods; (4) history of hysterectomy. Demographic details, data from the pathologic diagnosis report, as well as data on treatment for the occurrence of cervical dysplasia were retrospectively reviewed. Comorbidities included autoimmune diseases, hepatitis B and/or C, malignant tumors, diabetes, and immunosuppression. Immunosuppression included HIV, use of immunosuppressive drugs, hematologic disorders, and organ transplantation.
All patients had a follow up scheduled including HPV test and/or co‐testing in the outpatient clinic every 6 months for the first 2 years, and annually thereafter. If HPV‐based testing was positive, colposcopy was performed and any suspicious lesions were biopsied. Histologic specimens were graded as normal (no intraepithelial lesion), CIN1, CIN2, CIN3, adenocarcinoma in situ, adenocarcinoma, or squamous cell carcinoma, and classified according to the highest histologic abnormality found after biopsy or conization. The extracted information of HPV and LCT were the test results at the time of LEEP and 6 months after surgery, respectively. All HPV/LCT examinations, colposcopies and biopsies were performed by professional gynecologic oncologists and colposcopy physicians. The diagnosis of cervical lesions was confirmed by two independent pathologists, and reviewed by senior experts for inconclusive diagnoses. Follow‐up time was calculated from the date of conization to the date of recurrent CIN2+ diagnosis, death, deregistration, or end of study, whichever occurred first.
Pearson χ2 test and Fisher exact test were used for the relation between categorical variables. The relation between categorical variables and continuous variables was assessed by analysis of variance or Kruskal‐Wallis tests. Through Cox regression analysis, we calculated the hazard ratio (HR) and 95% confidence interval (CI) of recurrent disease. P values less than 0.05 were considered statistically significant. Statistical analysis was performed with graphpad prism version 6.0 (GraphPad) and SPSS 20.0 (IBM).
3. RESULTS
We identified 4369 women, including 2056 (47.1%) women with CIN2 and 2313 (52.9%) women with CIN3. The median age was 37 years (range 15–83 years), and the median follow‐up time was 13 months (range 6–72 months). Characteristics of all patients are summarized in Table 1. Before initial treatment, 2708 (96.3%) of 2811 patients who had an available hr‐HPV test tested positive. Of the 2708 women, 2043 (75.4%) were infected with one hr‐HPV type, 509 (18.8%) had a double infection, and 124 (4.6%) had a triple infection. HPV16 was the most prevalent genotype (637/2043, 31.2%), followed by HPV58 (438/2043, 21.4%), HPV52 (426/2043, 20.9%), and HPV33 (137/2043, 6.7%). During follow up, 268 (6.1%) CIN2+ cases were identified: 115 (42.9%) were CIN2, 135 (50.4%) were CIN3, 4 (1.5%) were adenocarcinoma in situ, and 14 (5.2%) were cervical cancer (2 adenocarcinoma; 12 squamous cell carcinoma) (Table 2). In all, 214 (79.8%) of the lesions were diagnosed within the first 2 years after LEEP. The 2‐year recurrence rate of post‐treatment CIN2+ was 5.0%, the 3‐year recurrence rate was 5.7%, and the 5‐year recurrence rate was 6.0%. The median age of patients in the recurrence group was older than that in the non‐recurrence group (40.0 years vs 37.0 years). HPV‐positive patients had a higher recurrence rate than those with abnormal LCT (20.0% vs 8.3%, P < 0.001), but not statistically different from co‐testing (20.0% vs 15.4%, P > 0.05).
TABLE 1.
Baseline characteristics of all 4369 patients a
| No recurrent CIN2+ | Recurrent CIN2+ | Total | P value | |
|---|---|---|---|---|
| Follow‐up time, mo | ||||
| Median (min–max) | 13.0 (6–70) | 10.0 (6–72) | 13.0 (6–72) | |
| IQR | 26–8 | 18–8 | 25–8 | |
| Age, years | ||||
| Median (min–max) | 37.0 (15–80) | 40.0 (19–83) | 37.0 (15–83) | |
| Mean (SD) | 38.54 (9.63) | 40.79 (9.97) | 38.68 (9.67) | |
| Age (categorized at 40 years) | 0.001 b | |||
| <40 | 2354 (95.0) | 124 (5.0) | 2478 | |
| >40 | 1747 (92.4) | 144 (7.6) | 1891 | |
| BMI | 0.114 | |||
| <18.5 | 421 (95.0) | 22 (5.0) | 443 | |
| 18.5–24 | 2667 (93.3) | 191 (6.7) | 2858 | |
| >24 | 1013 (94.9) | 55 (5.1) | 1068 | |
| Menopause | 0.055c | |||
| No | 3235 (94.2) | 198 (5.8) | 3433 | |
| Yes | 866 (92.5) | 70 (7.5) | 936 | |
| Smoking | 0.001 | |||
| No | 3105 (96.4) | 115 (3.6) | 3220 | |
| Yes | 272 (78.4) | 75 (21.6) | 347 | |
| Unknown | 724 (90.3) | 78 (9.7) | 802 | |
| Comorbidities | 0.001 | |||
| No | 3950 (94.9) | 214 (5.1) | 4164 | |
| Autoimmune disease | 37 (84.1) | 7 (15.9) | 44 | |
| Hepatitis B and/or C | 16 (94.1) | 1 (5.9) | 17 | |
| Malignant tumors | 37 (85.7) | 6 (14.3) | 42 | |
| Diabetes | 8 (80.0) | 2 (20.0) | 10 | |
| Immunosuppression c | 54 (58.7) | 38 (41.3) | 92 | |
| Age at first sex | 0.310 | |||
| <18 years | 278 (94.9) | 15 (5.1) | 293 | |
| >18 years | 2791 (94.1) | 175 (5.9) | 2966 | |
| Unknown | 1032 (93.0) | 78 (7.0) | 1110 | |
| Number of sexual partners | 0.238 | |||
| 0–3 | 2335 (93.7) | 156 (6.3) | 2491 | |
| >3 | 239 (96.4) | 9 (3.6) | 248 | |
| Unknown | 1527 (93.7) | 103 (6.3) | 1630 | |
| Number of pregnancies | 0.001 | |||
| ≤3 | 1230 (95.9) | 52 (4.1) | 1282 | |
| >3 | 595 (95.0) | 31 (5.0) | 626 | |
| Unknown | 2276 (92.5) | 185 (7.5) | 2461 | |
| Birth control method | 0.310 | |||
| No | 3722 (93.7) | 250 (6.3) | 3972 | |
| Copper‐IUD | 282 (95.9) | 12 (4.1) | 294 | |
| Mirena | 97 (94.2) | 6 (5.8) | 103 | |
| Preoperative hr‐HPV result | 0.071 | |||
| Negative | 101 (98.1) | 2 (1.9) | 103 | |
| Positive | 2550 (94.2) | 158 (5.8) | 2708 | |
| Unknown | 1450 (93.1) | 108 (6.9) | 1558 | |
| Preoperative LCT result | 0.001 | |||
| Normal | 72 (87.8) | 10 (12.2) | 82 | |
| LSIL | 1061 (96.2) | 42 (3.8) | 1103 | |
| ASC‐US | 1635 (94.8) | 90 (5.2) | 1725 | |
| ASC‐H | 307 (91.4) | 29 (8.6) | 336 | |
| HSIL | 1026 (91.4) | 97 (8.6) | 1123 | |
| First hr‐HPV result after LEEP | 0.001 | |||
| Negative | 1870 (98.7) | 24 (1.3) | 1894 | |
| Positive | 739 (80.0) | 185 (20.0) | 924 | |
| Unknown | 1492 (96.2) | 59 (3.8) | 1551 | |
| First LCT result after LEEP | 0.005 | |||
| Normal | 2812 (94.1) | 175 (5.9) | 2987 | |
| LSIL | 314 (94.0) | 20 (6.0) | 334 | |
| ASC‐US | 388 (90.2) | 42 (9.8) | 430 | |
| ASC‐H | 28 (87.5) | 4 (12.5) | 32 | |
| HSIL | 53 (91.4) | 5 (8.6) | 58 | |
| Unknown | 506 (95.8) | 22 (4.2) | 528 | |
| Baseline diagnosis | 0.001c | |||
| CIN2 | 1977 (96.2) | 79 (3.8) | 2056 | |
| CIN3 | 2124 (91.8) | 189 (8.2) | 2313 | |
| Glandular involvement | 0.058 c | |||
| No | 2072 (94.6) | 119 (5.4) | 2191 | |
| Yes | 2029 (93.1) | 149 (6.8) | 2178 | |
| Margin status | 0.001 | |||
| Negative | 3166 (97.2) | 92 (2.8) | 3258 | |
| Positive | 245 (70.2) | 104 (29.8) | 349 | |
| Unknown | 692 (90.6) | 72 (9.4) | 762 | |
| Total | 4110 (93.9) | 268 (6.1) | 4369 | |
Abbreviations: ASC‐H, atypical squamous cells cannot exclude high grade squamous intraepithelial lesion; ASC‐US, atypical squamous cells of undetermined significance; BMI, body mass index (calculated as weight in kilograms divided by the square of height in meters); CIN, cervical intraepithelial neoplasia; hr‐HPV, high‐risk human papillomavirus; HSIL, high‐grade intraepithelial lesion; IUD, intrauterine device; LEEP, loop electrosurgical excision procedure; LSIL, low‐grade squamous intraepithelial lesion.
Values are presented as median (minimum–maximum) (interquartile range) or as number (percentage). Pearson χ2 analysis was used to compare the women with recurrent disease and non‐recurrent disease after conization.
Fisher exact test P value.
Immunosuppression including HIV, drugs, blood system diseases, and organ transplantation.
TABLE 2.
Recurrent CIN2/3 or cervical cancer a
| Recurrent disease | Baseline diagnosis | ||
|---|---|---|---|
| CIN2 | CIN3 | Total | |
| Normal | 1977 (48.2) | 2124 (51.8) | 4101 |
| CIN2 | 47 (40.9) | 68 (59.1) | 115 |
| CIN3 | 27 (20.0) | 108 (80.0) | 135 |
| AIS | 2 (50.0) | 2 (50.0) | 4 |
| ADCA | 2 (100.0) | 0 (0.0) | 2 |
| SCC | 1 (8.3) | 11 (91.7) | 12 |
| Total | 2056 (47.1) | 2313 (52.9) | 4369 |
Abbreviations: ADCA, adenocarcinoma; AIS, adenocarcinoma in situ; CIN, cervical intraepithelial neoplasia; SCC, squamous cell carcinoma.
Values are presented as number (percentage).
Using multivariate analysis, hr‐HPV infection (HR 12.09, 95% CI 7.78–18.79), margin status (HR = 6.48, 95% CI: 4.75–8.84), baseline diagnosis (HR 1.45, 95% CI 1.08–1.95), immunosuppression (HR 1.96, 95% CI 1.33–2.91), and smoking (HR 3.17, 95% CI 2.27–4.43) were correlated with an increased risk of recurrent CIN2+ (Table 3). Patients with CIN3 had a higher risk of positive margins compared with CIN2 (odds ratio [OR] 2.97, 95% CI 2.31–3.18); 11.2% (260) of CIN3 patients and 4.3% (89) of CIN2 patients had positive margins (Table S1). Compared with women with negative margins (587/3258, 18.0%), women with positive margins (105/349, 30.1%) were at higher risk of hr‐HPV infection after LEEP (OR 2.28, 95% CI 1.72–3.02) (Table S2).
TABLE 3.
Factors predicting recurrence in women having cervical conization
| Univariate | Multivariate | |||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Age, years | ||||
| <40 | Ref. | Ref. | ||
| >40 | 1.35 (1.06–1.71) | 0.015 | 0.96 (0.75–1.22) | 0.721 |
| BMI | ||||
| <18.5 | Ref. | – | ||
| 18.5–24 | 1.37 (0.87–2.16) | 0.173 | – | – |
| >24 | 1.04 (0.63–1.73) | 0.883 | – | – |
| Menopause | ||||
| No | Ref. | – | ||
| Yes | 1.32 (0.99–1.75) | 0.054 | – | – |
| Smoking | ||||
| No | Ref. | Ref. | ||
| Yes | 10.21 (7.61–13.69) | 0.001 | 3.17 (2.27–4.43) | 0.001 |
| Unknown | 4.58 (3.42–6.14) | 0.001 | 2.65 (1.92–3.64) | 0.001 |
| Comorbidities | ||||
| No | Ref. | Ref. | ||
| Autoimmune disease | 4.20 (1.97–8.95) | 0.001 | 1.07 (0.49–2.34) | 0.852 |
| Hepatitis B and/or C | 1.23 (0.17–8.75) | 0.839 | 0.23 (0.30–1.70) | 0.149 |
| Malignant tumors | 3.75 (1.66–8.46) | 0.001 | 0.68 (0.29–1.57) | 0.363 |
| Diabetes | 6.98 (1.73–28.14) | 0.006 | 1.70 (0.42–6.99) | 0.459 |
| Immunosuppression a | 11.32 (8.01–16.01) | 0.001 | 1.96 (1.33–2.91) | 0.001 |
| Age at first sex, years | ||||
| <18 | Ref. | – | ||
| >18 | 1.16 (0.67–1.99) | 0.587 | – | – |
| Unknown | 1.40 (0.79–2.47) | 0.245 | – | – |
| Number of sexual partners | ||||
| 0–3 | Ref. | – | ||
| >3 | 0.56 (0.28–1.12) | 0.101 | – | – |
| Unknown | 1.01 (0.78–1.31) | 0.942 | – | – |
| Number of pregnancies | ||||
| <3 | Ref. | – | ||
| >3 | 1.16 (0.74–1.80) | 0.524 | – | – |
| Unknown | 1.98 (1.45–2.69) | 0.001 | – | – |
| Birth control method | ||||
| No | Ref. | – | ||
| Copper‐IUD | 0.63 (0.35–1.14) | 0.131 | – | – |
| Mirena | 0.92 (0.40–2.12) | 0.847 | – | – |
| Preoperative hr‐HPV result | ||||
| Negative | Ref. | – | ||
| Positive | 3.29 (0.82–13.26) | 0.095 | – | – |
| Unknown | 3.41 (0.84–13.81) | 0.086 | – | – |
| Preoperative LCT result | ||||
| Normal | Ref. | – | ||
| LSIL | 0.50 (0.24–1.32) | 0.231 | – | – |
| ASC‐US | 0.39 (0.19–1.21) | 0.152 | – | – |
| ASC‐H | 0.68 (0.31–1.46) | 0.322 | – | – |
| HSIL | 0.68 (0.34–1.36) | 0.277 | – | – |
| First hr‐HPV result after LEEP | ||||
| Negative | Ref. | Ref. | ||
| Positive | 18.79 (12.28–28.77) | 0.001 | 12.09 (7.78–18.79) | 0.001 |
| Unknown | 3.53 (2.19–5.68) | 0.001 | 3.28 (2.02–5.33) | 0.001 |
| First LCT result after LEEP | ||||
| Normal | Ref. | Ref. | ||
| LSIL | 1.06 (0.67–1.69) | 0.791 | 0.75 (0.46–1.21) | 0.237 |
| ASC‐US | 1.85 (1.32–2.59) | 0.001 | 1.00 (0.70–1.42) | 0.991 |
| ASC‐H | 2.15 (0.79–5.79) | 0.130 | 1.73 (0.63–4.79) | 0.291 |
| HSIL | 1.55 (0.64–3.77) | 0.335 | 1.31 (0.13–1.89) | 0.291 |
| Unknown | 0.63 (0.41–0.98) | 0.043 | 0.92 (0.58–1.46) | 0.728 |
| Baseline diagnosis | ||||
| CIN2 | Ref. | Ref. | ||
| CIN3 | 2.05 (1.57–2.66) | 0.001 | 1.45 (1.08–1.95) | 0.014 |
| Glandular involvement | ||||
| No | Ref. | Ref. | ||
| Yes | 1.48 (1.16–1.88) | 0.002 | 1.11 (0.83–1.48) | 0.500 |
| Margin status | ||||
| Negative | Ref. | Ref. | ||
| Positive | 11.76 (8.88–15.57) | 0.001 | 6.48 (4.75–8.84) | 0.001 |
| Unknown | 2.08 (1.51–2.85) | 0.001 | 1.45 (1.03–2.04) | 0.032 |
Abbreviations: ASC‐H, atypical squamous cells cannot exclude high grade squamous intraepithelial lesion; ASC‐US, atypical squamous cells of undetermined significance; BMI, body mass index (calculated as weight in kilograms divided by the square of height in meters); CI, confidence interval; CIN, cervical intraepithelial neoplasia; HR, hazard ratio; hr‐HPV, high‐risk human papillomavirus; HSIL, high‐grade intraepithelial lesion; IUD, intrauterine device; LEEP, loop electrosurgical excision procedure; LSIL, low‐grade squamous intraepithelial lesion.
Hazard ratios are adjusted for all variables in the analysis.
Six months after treatment, 924 (21.1%) of 2818 available hr‐HPV tests were positive, of which 727 (78.7%) women had a single infection and 197 (21.3%) women had mixed infections. Table 4 shows hr‐HPV genotypes with various clinical outcomes. Compared with a single infection, women with mixed infections had no statistically significant difference in recurrence (P = 0.916). HPV16, HPV58, and HPV52 were the most common genotypes, detected in 161 (22.1%), 123 (17.0%), and 119 (16.4%) patients, respectively. More women who were positive for HPV16 developed CIN2+ than did those infected with other hr‐HPV genotypes (60/161, 37.3% vs 80/566, 14.1%; P = 0.001). Using univariate analysis, HPV16, HPV33, and HPV52 positivity were high‐risk factors for post‐treatment CIN2+ (HR 3.61, 95% CI 2.43–5.37; HR 2.62, 95% CI 1.12–6.12; and HR 1.61, 95% CI 1.02–2.55, respectively) (Table S3).
TABLE 4.
High‐risk HPV genotypes in recurrent disease a
| Total | No recurrent CIN2+ | Recurrent CIN2+ | P value b | |
|---|---|---|---|---|
| Type of infection | ||||
| Single infection | 727 | 587 (80.7) | 140 (19.3) | 0.916 |
| Multiple infections | 197 | 160 (81.2) | 37 (18.8) | |
| Single infection classification | ||||
| HPV16 | 161 | 101 (62.7) | 60 (37.2) | 0.001 |
| HPV18 | 18 | 15 (83.3) | 3 (16.7) | 1.000 |
| HPV31 | 19 | 18 (94.7) | 1 (5.3) | 0.146 |
| HPV33 | 24 | 15 (62.5) | 9 (37.5) | 0.032 |
| HPV35 | 20 | 19 (95.0) | 1 (5.0) | 0.148 |
| HPV39 | 13 | 13 (100.0) | 0 (0.00) | 0.084 |
| HPV45 | 15 | 15 (100.0) | 0 (0.00) | 0.089 |
| HPV51 | 29 | 26 (89.7) | 3 (10.3) | 0.334 |
| HPV52 | 119 | 88 (73.9) | 31 (26.1) | 0.043 |
| HPV53 | 60 | 54 (90.0) | 6(10.0) | 0.061 |
| HPV56 | 56 | 51 (91.1) | 5 (8.9) | 0.051 |
| HPV58 | 123 | 107 (87.0) | 16 (13.0) | 0.060 |
| HPV59 | 18 | 17 (94.4) | 1 (5.6) | 0.222 |
| HPV66 | 21 | 2017 (95.2) | 1 (4.8) | 0.098 |
| HPV68 | 23 | 21 (91.3) | 2 (8.7) | 0.282 |
| HPV73 | 2 | 2 (100.0) | 0 (0.0) | 1.000 |
| HPV82 | 6 | 5 (83.3) | 1 (16.7) | 1.000 |
Abbreviations: CIN, cervical intraepithelial neoplasia; HPV, human papillomavirus.
Bold values indicate that the P value is statistically significant.
Values are presented as number (percentage).
Fisher exact test P value.
Table 5 shows the sensitivity, specificity, positive predictive value, negative predictive value and likelihood ratio for LCT, margins or hr‐HPV as predictors of post‐treatment disease. The sensitivity and specificity of predicting recurrence were 88.5% and 71.7% for hr‐HPV, 53.1% and 92.8% for margins, and 28.9% and 78.2% for LCT, respectively. High‐risk HPV had the highest sensitivity (88.5%) and negative predictive value (98.7%). Margin status combined with hr‐HPV had the highest predictive specificity (97.8%). In general, the addition of LCT or margins to hr‐HPV results did not substantially improve the prediction accuracy of recurrence (P > 0.05).
TABLE 5.
Sensitivity, specificity, PPV, NPV, LR, and NR of recurrent CIN2+ after conization a
| Sensitivity | Specificity | PPV | NPV | PLR | NLR | |
|---|---|---|---|---|---|---|
| Margin status | 104/196 (53.1) | 3166/3411 (92.8) | 30.0 | 97.2 | 7.4 | 0.5 |
| hr‐HPV results after LEEP | 185/209 (88.5) | 1870/2609 (71.7) | 20.0 | 98.7 | 3.1 | 0.2 |
| LCT result after LEEP | 71/246 (28.9) | 2812/3595 (78.2) | 8.3 | 94.1 | 1.3 | 0.9 |
| hr‐HPV and margin status b | ||||||
| HPV+ & Positive | 60/147 (40.8) | 2039/2084 (97.8) | 57.1 | 95.9 | 18.9 | 0.6 |
| HPV+ & Negative | 68/147 (46.3) | 1565/2084 (75.1) | 11.6 | 95.2 | 1.9 | 0.7 |
| HPV− & Positive | 13/147 (8.8) | 1983/2084 (95.2) | 11.4 | 93.7 | 1.8 | 1.0 |
| HPV− & Negative | 6/147 (4.1) | 665/2084 (31.9) | 0.4 | 82.5 | 0.06 | 3.0 |
| hr‐HPV or margin status b | ||||||
| HPV+ or Positive | 141/147 (95.9) | 1419/2087 (68.0) | 17.5 | 99.6 | 3.0 | 0.06 |
| HPV+ or Negative | 134/147 (91.2) | 101/2084 (4.8) | 6.3 | 88.6 | 1.0 | 1.8 |
| HPV− or Positive | 79/147 (53.7) | 519/2084 (24.9) | 4.8 | 88.4 | 0.7 | 1.9 |
| HPV− or Negative | 87/147 (59.2) | 45/2084 (2.2) | 4.1 | 42.8 | 0.6 | 18.9 |
| hr‐HPV & first LCT after LEEP b | ||||||
| HPV+ & LCT+ | 61/201 (30.3) | 2062/2398 (86.0) | 15.4 | 93.6 | 2.2 | 0.8 |
| HPV+ & LCT− | 119/201 (59.2) | 2024/2398 (84.4) | 24.1 | 96.1 | 3.8 | 0.5 |
| HPV− & LCT+ | 4/201 (2.0) | 2010/2398 (83.8) | 1.1 | 91.1 | 0.1 | 1.2 |
| HPV− & LCT– | 17/201 (8.5) | 1098/2398 (45.8) | 12.9 | 85.6 | 0.2 | 2.0 |
| hr‐HPV or first LCT after LEEP b | ||||||
| HPV+ or LCT+ | 184/201 (91.5) | 1300/2398 (54.2) | 14.4 | 98.7 | 2.0 | 0.2 |
| HPV+ or LCT− | 197/201 (98.1) | 388/2398 (16.2) | 8.9 | 98.9 | 1.2 | 0.1 |
| HPV− or LCT+ | 82/201 (40.8) | 374/2398 (15.6) | 3.9 | 75.9 | 0.5 | 3.8 |
| HPV− or LCT− | 140/201 (69.7) | 336/2398 (14.0) | 6.4 | 84.6 | 0.8 | 2.2 |
Abbreviations: hr‐HPV, high‐risk human papillomavirus; NLR, negative likelihood ratio; NPV, negative predictive value; PLR, positive likelihood ratio; PPV, positive predictive value; LCT, liquid‐based cytology; LEEP, loop electrosurgical excision procedure.
Values are presented as number (percentage).
Only considering hr‐HPV positive or negative patients, excludes cases with unknown results.
Incidentally, of the 268 women with recurrent disease, 14 had cervical cancer. Patient characteristics of recurrent CIN2/CIN3/adenocarcinoma in situ and cervical cancer are summarized in Tables S4 and S5. The median age was 46 years (range 33–72 years), and the median follow‐up time was 19.5 months (range 7–61 months). The incidence of cervical cancer was increased in patients older than 40 years (8.3% vs 1.6%, P = 0.014). The incidence of cervical cancer in patients previously diagnosed with CIN2 and CIN3 was 3.8% and 5.8% (P > 0.05), respectively. Of the 13 HPV‐positive women, 10 had a single infection and 3 had multiple infections. HPV16 was the most prevalent genotype (8/13, 61.5%). Moreover, five patients had positive margins and six patients had negative margins. Characteristics of women with and without recurrence are summarized in Table S6. Univariate analysis revealed a significant association between incident recurrent cervical cancer and age (HR 6.86, 95% CI 1.53–30.68), hr‐HPV (HR 41.96, 95% CI 5.48–321.59), and margin status (HR 10.85, 95% CI 3.31–35.59) (Table S7).
4. DISCUSSION
To the best of our knowledge, this is the largest study investigating the impact of different risk factors for recurrence after LEEP in patients with high‐grade CIN. Recurrent CIN2+ was detected in nearly 6.1% of women, which was lower than the overall average of approximately 7% in recent studies. 11 In order to explain the contradictions between the results, several key points need to be further discussed. First, the lower recurrence rate may be due to insufficient follow‐up time, which prevented us from detecting late recurrence. A previous study reported a high recurrence rate (12.0%), with a median follow‐up time of 10 years and a maximum follow‐up time of 16 years. 7 Huang et al. 12 reported that the 5‐year risk of recurrence was 14.8%, and the median follow‐up time was 74.3 months. Second, patients with CIN3 had a higher risk of recurrence compared with CIN2 (P = 0.014). Reich et al. 13 reported a higher rate (22.0%) of treatment failure because they only included patients treated with CIN3. 13 However, Alder et al. 7 found that the severity of pathology at baseline cone biopsy was not a significant predictor of recurrent CIN2+ (P = 0.269). 7 The reason may be that the proportion of CIN2 patients included in that study was relatively low, only 214 (21.6%) of 991 were CIN2 patients.
Consistent with other studies, we found that women with positive margins have a higher risk of recurrence. Among the 4417 women with clear margins who were followed up for 30 years after conization, post‐treatment CIN2+ was found in a very small percentage (0.35%). 14 Of the 390 patients who received CIN3 treatment but with positive margins, 84 (22%) had recurrent CIN3. 13 In addition, the proportion of positive/negative margins in the meta‐analysis was 23% (range 3%–60% or 3%–35%), while the proportion in the Alder et al. study exceeded 30%. 7 , 15 In contrast, the proportion in our study was relatively low (10.7%), so another reason for the low recurrence rate may be the low proportion of positive/negative margins. Women with positive margins had more severe pathology at baseline diagnosis (P < 0.001), and had a higher risk of hr‐HPV infection after LEEP (P < 0.001). These findings suggest that wide excision may be required for treatment to avoid residual lesions, they also indicate the importance of complete excision for patients with CIN2/3.
Few studies have analyzed the impact of comorbidities on the disease progression of women with CIN2/3. Previous studies have mostly focused on analyzing the risk of recurrence in HIV‐infected women. 16 , 17 A study found that HIV‐positive patients with CIN2+ have an increased risk of recurrence (7.1%, 95% CI 4.4–10.7). 8 In our study, women with comorbidities had a higher risk of recurrence, especially those with immunosuppression. People with weakened immune function are more susceptible to HPV infection and have difficulty clearing the virus, so they have a greater chance of progressing to high‐grade CIN or invasive cancer. This result indicates the importance of understanding the medical history and comorbidities of women with high‐grade CIN, and the need for more stringent post‐treatment monitoring for women with weakened immune systems.
The ASCCP Risk‐Based Management Consensus Guidelines recommend HPV‐based testing at 6 months after treatment for patients with histologic high‐grade intraepithelial lesions, and annual HPV testing or co‐testing thereafter until three consecutive negative tests. 18 After the initial intensive surveillance period, continued surveillance at 3‐year intervals is recommended for at least 25 years after treatment of high‐grade histology. 18 In this study, being hr‐HPV positive was an important predictor of post‐treatment CIN2+. Univariate analysis showed that HPV16, HPV33, and HPV52 positivity increased the risk of recurrence. HPV testing and co‐testing are more sensitive in detecting CIN2+ than LCT alone in both the post‐colposcopy and post‐treatment settings. 19 , 20 , 21 HPV testing and co‐testing showed little difference in detecting recurrent CIN2+, so either test could be used for surveillance. 21 , 22 Similarly, in our study, the recurrence rate of hr‐HPV infection was higher than abnormal LCT, but not statistically different from co‐testing. Considering cost‐effectiveness, HPV‐based testing is recommended for patients with high‐grade CIN at 6 months after LEEP. A retrospective study that analyzed data from two clinical trials found that quadrivalent HPV vaccination in women who had previously undergone surgery for HPV‐related disease significantly reduced the incidence of subsequent HPV‐related disease, including high‐grade disease. 23
The conclusion of a recent meta‐analysis was that hr‐HPV results were more accurate than margin status in predicting recurrence, with higher sensitivity (91% vs 56%) and equivalent specificity (84%). 11 A retrospective study showed that postoperative hr‐HPV results had high sensitivity (88.8% vs 33.3%) and specificity (80% vs 68.8%), and that the combination of HPV results and margin status or cytology did not significantly improve the accuracy of prediction. 9 In addition, a study that combined the status of cervical endocervical margins with hr‐HPV testing found that the positive predictive value for predicting recurrent disease was 94%, and the negative predictive value was 96%, proving the clinical applicability of the test. 10 We found that the sensitivity of margin status for predicting recurrence was only 53.1%. The poor ability of margins to predict recurrence argues against differentiating follow‐up testing by margin status alone. Patients with positive margins can undergo re‐excision or HPV‐based follow up, but premature recalls are not recommended. In contrast, the ability of HPV testing to predict recurrent CIN2+ was 88.5%. The addition of margin status or LCT to hr‐HPV results did not substantially improve the prediction accuracy. Based on these findings, HPV‐based follow up is recommended, and if the test is positive, colposcopy is recommended.
This study is distinguished from most others by its large size, the availability of covariate information in the same hospital, providing all included women with reliable information about recurrent CIN2+. West China Second Hospital of Sichuan University is a medical center for the diagnosis and treatment of cervical precancerous lesions and cervical cancer. Professional diagnosis and treatment techniques and a number of senior pathologists enable us to determine the true recurrence rate. However, our study has several limitations. One limitation is the short follow‐up time, and our study did not detect late recurrence. Another limitation is the lack of detailed information about the status of the margins, and approximately 17.4% of patients had unknown margin status. We only analyzed the difference between the positive and negative margins, but did not further subdivide them into endocervical margins or ectocervical margins. These differences probably had an impact on detection of recurrence among women with endocervical/ectocervical margins.
In conclusion, nearly 6.1% of women developed recurrent CIN2+. High‐risk HPV, margins, baseline diagnosis, immunosuppression, and smoking were risk factors for recurrence. HPV16, HPV33, and HPV52 infection were associated with increased risk of recurrence. During follow up, more attention should be paid to the correlation between cervical lesions and hr‐HPV genotypes. Given the higher risk of recurrence with positive margins, wider excisions could be indicated to avoid residual lesions. High‐risk HPV has the highest accuracy in predicting recurrence. After treatment, HPV‐based testing at 6 months is recommended, regardless of margins. Timely identification of these high‐risk factors enables risk stratification and enables individualized management and follow‐up strategies.
AUTHOR CONTRIBUTIONS
TD and MX conceived and designed the work. TD, LL, RD, YC, and BY contributed to the acquisition, analysis, and interpretation of data. TD wrote the original draft, and TD, LL, RD, and MX contributed to the reviewing and editing. All authors approved the final manuscript.
FUNDING INFORMATION
The Key Research and Development Program of the Science and Technology Department of Sichuan Province (no. 2019YFS0407) provided funding for the present study.
CONFLICT OF INTEREST
The authors declare no conflicts of interest.
Supporting information
Table S1
Table S2
Table S3
Table S4
Table S5
Table S6
Table S7
Ding T, Li L, Duan R, Chen Y, Yang B, Xi M. Risk factors analysis of recurrent disease after treatment with a loop electrosurgical excision procedure for high‐grade cervical intraepithelial neoplasia. Int J Gynecol Obstet. 2023;160:538‐547. doi: 10.1002/ijgo.14340
Ting Ding and Lin Li contributed equally to this work.
DATA AVAILABILITY STATEMENT
All data are available in this paper.
REFERENCES
- 1. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394‐424. Erratum in: CA Cancer J Clin. 2020 Jul;70(4):313. doi: 10.3322/caac.21492 [DOI] [PubMed] [Google Scholar]
- 2. Wen C. China's plans to curb cervical cancer. Lancet Oncol. 2005;6(3):139‐141. doi: 10.1016/S1470-2045(05)01761-4 [DOI] [PubMed] [Google Scholar]
- 3. Andrae B, Kemetli L, Sparén P, et al. Screening‐preventable cervical cancer risks: evidence from a nationwide audit in Sweden. J Natl Cancer Inst. 2008;100(9):622‐629. doi: 10.1093/jnci/djn099 [DOI] [PubMed] [Google Scholar]
- 4. Soutter WP, Sasieni P, Panoskaltsis T. Long‐term risk of invasive cervical cancer after treatment of squamous cervical intraepithelial neoplasia. Int J Cancer. 2006;118(8):2048‐2055. doi: 10.1002/ijc.21604 [DOI] [PubMed] [Google Scholar]
- 5. Strander B, Andersson‐Ellström A, Milsom I, Sparén P. Long term risk of invasive cancer after treatment for cervical intraepithelial neoplasia grade 3: population based cohort study. BMJ. 2007;335(7629):1077. doi: 10.1136/bmj.39363.471806.BE [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Sand FL, Frederiksen K, Munk C, Jensen SM, Kjaer SK. Long‐term risk of cervical cancer following conization of cervical intraepithelial neoplasia grade 3‐A Danish nationwide cohort study. Int J Cancer. 2018;142(9):1759‐1766. doi: 10.1002/ijc.31202 [DOI] [PubMed] [Google Scholar]
- 7. Alder S, Megyessi D, Sundström K, et al. Incomplete excision of cervical intraepithelial neoplasia as a predictor of the risk of recurrent disease‐a 16‐year follow‐up study. Am J Obstet Gynecol. 2020;222(2):172.e1‐172.e12. doi: 10.1016/j.ajog.2019.08.042 [DOI] [PubMed] [Google Scholar]
- 8. Huchko MJ, Leslie H, Maloba M, Bukusi EA, Cohen CR. Factors associated with recurrence of cervical intraepithelial neoplasia 2+ after treatment among HIV‐infected women in Western Kenya. J Acquir Immune Defic Syndr. 2014;66(2):188‐192. doi: 10.1097/QAI.0000000000000130 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Fernández‐Montolí ME, Tous S, Medina G, Castellarnau M, García‐Tejedor A, de Sanjosé S. Long‐term predictors of recurrent cervical intraepithelial neoplasia 2‐3 after treatment with a large loop excision of the transformation zone: a retrospective study. BJOG. 2020;127(3):377‐387. doi: 10.1111/1471-0528.15996 [DOI] [PubMed] [Google Scholar]
- 10. Kong TW, Son JH, Chang SJ, Paek J, Lee Y, Ryu HS. Value of endocervical margin and high‐risk human papillomavirus status after conization for high‐grade cervical intraepithelial neoplasia, adenocarcinoma in situ, and microinvasive carcinoma of the uterine cervix. Gynecol Oncol. 2014;135(3):468‐473. doi: 10.1016/j.ygyno.2014.09.022 [DOI] [PubMed] [Google Scholar]
- 11. Arbyn M, Redman CWE, Verdoodt F, et al. Incomplete excision of cervical precancer as a predictor of treatment failure: a systematic review and meta‐analysis. Lancet Oncol. 2017;18:1665‐1679. doi: 10.1016/S1470-2045(17)30700-3 [DOI] [PubMed] [Google Scholar]
- 12. Huang HJ, Tung HJ, Yang LY, et al. Role of human papillomavirus status after conization for high‐grade cervical intraepithelial neoplasia. Int J Cancer. 2021;148(3):665‐672. doi: 10.1002/ijc.33251 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Reich O, Lahousen M, Pickel H, Tamussino K, Winter R. Cervical intraepithelial neoplasia III: long‐term follow‐up after cold‐knife conization with involved margins. Obstet Gynecol. 2002;99(2):193‐196. doi: 10.1016/s0029-7844(01)01683-0 [DOI] [PubMed] [Google Scholar]
- 14. Reich O, Pickel H, Lahousen M, Tamussino K, Winter R. Cervical intraepithelial neoplasia III: long‐term outcome after cold‐knife conization with clear margins. Obstet Gynecol. 2001;97(3):428‐430. doi: 10.1016/s0029-7844(00)01174-1 [DOI] [PubMed] [Google Scholar]
- 15. Ghaem‐Maghami S, Sagi S, Majeed G, Soutter WP. Incomplete excision of cervical intraepithelial neoplasia and risk of treatment failure: a meta‐analysis. Lancet Oncol. 2007;8(11):985‐993. doi: 10.1016/S1470-2045(07)70283-8 [DOI] [PubMed] [Google Scholar]
- 16. Lodi CT, Michelin MA, Lima MI, et al. Factors associated with recurrence of cervical intraepithelial neoplasia after conization in HIV‐infected and noninfected women. Arch Gynecol Obstet. 2011;284(1):191‐197. doi: 10.1007/s00404-010-1611-1 [DOI] [PubMed] [Google Scholar]
- 17. Forhan SE, Godfrey CC, Watts DH, Langley CL. A systematic review of the effects of visual inspection with acetic acid, cryotherapy, and loop electrosurgical excision procedures for cervical dysplasia in HIV‐infected women in low‐ and middle‐income countries. J Acquir Immune Defic Syndr. 2015;68(Suppl 3):S350‐S356. doi: 10.1097/QAI.0000000000000488 [DOI] [PubMed] [Google Scholar]
- 18. Perkins RB, Guido RS, Castle PE, et al. 2019 ASCCP Risk‐Based Management Consensus Guidelines Committee. 2019 ASCCP risk‐based management consensus guidelines for abnormal cervical cancer screening tests and cancer precursors. J Low Genit Tract Dis. 2020;24(2):102‐131. Erratum in: J Low Genit Tract Dis. 2020 Oct;24(4):427. doi: 10.1097/LGT.0000000000000525 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Kocken M, Uijterwaal MH, de Vries AL, et al. High‐risk human papillomavirus testing versus cytology in predicting post‐treatment disease in women treated for high‐grade cervical disease: a systematic review and meta‐analysis. Gynecol Oncol. 2012;125(2):500‐507. doi: 10.1016/j.ygyno.2012.01.015 [DOI] [PubMed] [Google Scholar]
- 20. Arbyn M, Ronco G, Anttila A, et al. Evidence regarding human papillomavirus testing in secondary prevention of cervical cancer. Vaccine. 2012;30(Suppl. 5):F88‐F99. Erratum in: Vaccine. 2013 Dec 16;31(52):6266. doi: 10.1016/j.vaccine.2012.06.095 [DOI] [PubMed] [Google Scholar]
- 21. Clarke MA, Unger ER, Zuna R, et al. A systematic review of tests for postcolposcopy and posttreatment surveillance. J Low Genit Tract Dis. 2020;24(2):148‐156. doi: 10.1097/LGT.0000000000000526 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Cuschieri K, Bhatia R, Cruickshank M, Hillemanns P, Arbyn M. HPV testing in the context of post‐treatment follow up (test of cure). J Clin Virol. 2016;76(Suppl 1):S56‐S61. doi: 10.1016/j.jcv.2015.10.008 [DOI] [PubMed] [Google Scholar]
- 23. Joura EA, Garland SM, Paavonen J, et al. FUTURE I and II Study Group. Effect of the human papillomavirus (HPV) quadrivalent vaccine in a subgroup of women with cervical and vulvar disease: retrospective pooled analysis of trial data. BMJ. 2012;344:e1401. doi: 10.1136/bmj.e1401 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1
Table S2
Table S3
Table S4
Table S5
Table S6
Table S7
Data Availability Statement
All data are available in this paper.