The role of endocervical curettage (ECC) cervical intraepithelial neoplasia 1 (CIN 1) in colposcopy-directed biopsy is controversial. In the 2006 consensus guidelines for the management of CIN or adenocarcinoma in situ, a diagnostic excisional procedure is recommended if an ECC is positive for CIN 1.1 As per the 2012 updated consensus guidelines for the management of abnormal cervical cancer screening tests and cancer precursors, endocervical sampling reported as CIN 1 should be managed as CIN 1.2
The recommendations were based on 3 retrospective studies. Petersen et al3 compared 37 women positive for mild dysplasia on ECC with 66 women positive for mild ectocervical dysplasia and concluded that patients with mild dysplasia on ECC have comparatively no greater risk of severe dysplasia on subsequent loop electrosurgical excisional procedure (LEEP). Fukuchi et al4 studied 54 ECC-CIN 1 women with ectocervical biopsy CIN or less and 170 ectocervical CIN 1 women with negative ECC findings who underwent LEEP. The 24-month risk of CIN 2 or more (CIN 2+) was lower in the ECC-CIN 1 cohort compared with the other cohort (24.1% vs 44.7%, p = .018). The risk was also lower for CIN 3+, though not significant (7.4% vs 14.1%, p = .23).4 Unlike LEEP conization assessment, Gage et al5 assessed the CIN 2+ risk based on repeat colposcopy according to the widely practiced management for CIN 1 biopsy and CIN 1 ECC. The short-term risk of CIN 2+ was demonstrated the same after CIN 1 ECC and CIN 1 biopsy (5.0% of 359 vs 4.9% of 1,389, respectively, p = .37).5 The risk of CIN 2+ in ECC CIN 1 varied greatly from 5.0% to 24.1% in previous literature, along with the methods of CIN 2+ diagnosis (LEEP or repeat colposcopy). Loop electrosurgical excisional procedure conization was found to detect CIN 2+ more precisely than repeat colposcopy. However, the sample size of the study was small.
On the whole, there is lack of a large sample study investigating the role of endocervical CIN 1 in guiding treatment or surveillance protocols based on LEEP. In the largest women hospital of China, the Obstetrics and Gynecology Hospital of Fudan University (OGHFU), ECC is routinely performed during colposcopy. Ectocervical CIN 1 persistent for more than 2 years and endocervical CIN 1 can trigger LEEP conization. Our objective was to estimate the risk of CIN 2+ in LEEP specimens derived from ECC CIN 1 (ECC results of CIN 1 with colposcopy-directed biopsy results ≤CIN 1) compared with biopsy CIN 1 (colposcopy-directed biopsy results of CIN 1 with negative ECC findings).
MATERIALS AND METHODS
A retrospective computer-based search was performed for this case-control study. Subjects were enrolled from January 2013 to December 2018 at OGHFU. Inclusion criteria were a diagnosis of ECC CIN 1 with LEEP within 6 months and a biopsy of CIN 1 with LEEP within 6 months. Exclusion criteria were a biopsy of CIN 1 women with ECC CIN 2+ (including CIN 2/3, adenocarcinoma in situ [AIS], and cancer) and a biopsy of CIN 2+ women with ECC CIN 1. All the subjects who met the previous criteria were included. Approval was obtained from institutional review board of the OGHFU before data extraction was performed. All the pathologic specimens were processed by adhering to a standardized protocol and were examined by an experienced staff pathologist, followed by verification by another pathologist.
The flow chart of the study design is illustrated in Figure 1. The subjects were divided into 2 groups to investigate risk factors of cervical precancer and cancer (CIN 2+). Case group comprised women with ECC CIN 1. Control group comprised women with biopsy CIN 1. Subject variables of included age, cytology, high-risk human papillomavirus (hrHPV) status, and ECC results. Outcome was the LEEP histopathological result. All data including age, cytology, hrHPV status, ECC, and LEEP results are available on the computer-based database of the OGHFU. Quantitative data were expressed as mean ± SD. Variables were analyzed by univariate and multivariate logistic regression, and p < .05 was defined statistically significant. Statistical analyses were performed using Stata 15.0 (StataCorp, TX).
Role of the Funding Source
Data collection and interpretation were supported by the Science and Technology Commission of Shanghai Municipality (16411950200 and 18411963600).
A total of 1,195 women with ECC CIN 1 and/or biopsy CIN 1 who underwent LEEP were included in the study. Baseline characteristics including age, cytology, and hrHPV status are listed in Table 1. ECC CIN 1 group comprised 400 women, with the following LEEP histopathology results: 110 normal (27.50%), 186 CIN 1 (46.50%), 99 CIN 2/3 (24.75%), 3 AIS (0.75%), and 2 squamous cell carcinoma (SCC, 0.50%). Biopsy CIN 1 group comprised 795 women, with the following LEEP histopathology results: 230 normal (28.93%), 415 CIN 1 (52.20%), 149 CIN 2/3 (18.74%), and 1 SCC (0.13%; see Table 2).
Three patients with AIS were diagnosed in ECC CIN 1 group, whereas none were diagnosed in biopsy CIN 1 group. The risk of AIS was significantly higher in ECC CIN 1 compared with biopsy CIN 1 group (3/400 vs 0/795, p = .015). The cytology reports of the 3 AIS patients revealed 1 atypical glandular cells (AGC) and 2 atypical squamous cells of undetermined significance (ASC-US). In total, 0.25% (3/1,195) cervical CIN 1 was diagnosed as invasive cervical cancer by LEEP in this study. Two invasive squamous cell cancers were diagnosed in ECC CIN 1 group with cytology of ASC-US and low-grade squamous intraepithelial lesions (LSILs), whereas 1 invasive squamous cell cancer was diagnosed in biopsy CIN 1 group with cytology of high-grade squamous intraepithelial lesions (HSIL). The risk of invasive cancer in ECC CIN 1 group was higher than biopsy CIN 1 group, but the difference was not significant (0.5% vs 0.13%, p = .223; see Table 2).
According to LEEP histopathology, 941 women were diagnosed with CIN 1 or normal (<CIN 2+) and 254 with CIN 2+. Variables including age, cytology, hrHPV status, and ECC histopathology are listed in Table 3. On cytological examination, 13.4% of no intraepithelial lesion or malignancy (NILM), 19.0% of ASC-US, 17.5% of LSIL, 39.8% of atypical squamous cells, cannot exclude HSIL (ASC-H), and 40.4% of HSIL or worse (HSIL+) were diagnosed with CIN 2+ on LEEP conization; 19.8% of negative hrHPV and 21.4% of positive hrHPV were diagnosed with CIN 2+ on LEEP conization; 18.9% of biopsy CIN 1 and 26% of ECC CIN 1 were diagnosed with CIN 2+ on LEEP conization. Univariate logistic regression showed that cytology (p < .001) and ECC (p = .005) results were significantly different between less than CIN 2+ and CIN 2+, whereas age (p = .650) and hrHPV (p = .721) were not (see Table 3). Multivariate logistic regression showed that cytology of ASC-H (OR = 4.73, 95% CI = 2.78–8.05, p < .001), HSIL+ (OR = 4.88, 95% CI = 3.00–7.94, p < .001), and ECC CIN 1 (OR = 1.80, 95% CI = 1.33–2.44, p < .001) were risk factors for CIN 2+, whereas age and hrHPV were not (see Table 4).
Three women were diagnosed with invasive cervical squamous cell cancer by LEEP conization. Both punch biopsy and ECC were CIN 1 in 2 women with cytology of ASC-US and LSIL. The remaining one had punch biopsy CIN 1 and negative ECC with cytology of HSIL. All 3 hrHPV were positive. In 961 women with cytology LSIL or less, 21.68% (75/346) of ECC CIN 1, and 13.82% (85/615) of biopsy CIN 1 women were diagnosed with CIN 2+ by LEEP conization (OR = 1.73, 95% CI = 1.20–2.47, p = .002). In 234 women with cytology of HSIL or worse, 53.7% (29/54) of ECC CIN 1 women, and 36.11% (65/180) of biopsy CIN 1 were diagnosed with CIN 2+ by LEEP conization (OR = 2.05, 95% CI: 1.06–3.98, p = .021; see Table 5).
Endocervical curettage was considered an important component of colposcopy to evaluate for CIN and occult invasive carcinoma within the endocervical canal.6,7 Endocervical curettage was more likely to find disease in the endocervix among women with high-grade cytology, positive HPV 16 infection, or high-grade colposcopic impressions. Among women with ASC-US or LSIL cytology, those with an unsatisfactory colposcopic examination had a 13.0% chance of CIN 2+ on ECC (95% CI = 6.1%–25.7%); whereas when colposcopic examination was normal or satisfactory with visible abnormal lesions, ECC detected less than 5% CIN 2 or worse in the endocervix. An ASC-H or HSIL or worse cytology was associated with a CIN 2 or worse yield of 25.8% by ECC (95% CI = 16.6%–37.9%).8 Routine ECC was also shown to be necessary in patients with HPV 16 and normal cytology and normal colposcopy. Cervical intraepithelial neoplasia 2+ results were observed in the ECC of 11% of the patients whose colposcopy was normal regardless of the transformation zone type.9
A positive ECC has traditionally been an indication for cervical conization.1 As per 2012 updated consensus guideline, CIN 1 on ECC should be managed as CIN 1, not as a positive ECC.2 However, there are limited studies concerning the role of ECC CIN 1 diagnosis. The sample size of ECC CIN 1 studies ranged from 37 to 54 cases.3,4,8 The results revealed that ECC CIN 1 was not an increased risk factor of CIN 2+ compared with biopsy CIN 1. Castle PE et al analyzed 203 women with an ECC CIN 1 diagnosis who underwent conservative management consisting of cytological examination and repeated ECC and found that the 24-month risk of CIN 2+ and CIN 3+ was 11.8% and 1.5%, respectively.10
The quality control of ECC is essential. According to Atlas of Colposcopy – Principles and Practice, the curette should be introduced about 2 centimeters inside the endocervical canal. All 4 walls should be scraped systematically. The tissues collected should be put into formaldehyde solution for histology.11 An insufficient ECC that fails to reach squamocolumnar junction may fail to find the worst lesions in the endocervical canal. Reportedly, the false-negative diagnosis rate of ECC in CIN patients could reach 45%.12 Therefore, an ECC CIN 1 diagnosis may indicate that more severe CIN can exist deeply inside the endocervical canal. Consequently, an ECC CIN 1 diagnosis can trigger diagnostic LEEP conization or follow-up as CIN 1 in our hospital. Although a CIN 2+ diagnosis was frequent following LEEP, the exact risk remains unknown. To the best of our knowledge, this is the largest study to investigate the risk of a CIN 2+ diagnosis by LEEP in ECC CIN 1 women.
Our study demonstrated that ECC CIN 1 was a greater risk factor for potential CIN 2+ than biopsy CIN 1 (with negative ECC findings) based on the following evidence. Firstly, there was significantly more diagnosis of CIN 2+ in the ECC CIN 1 group than the biopsy CIN 1 group (26.0% vs 18.9%, p = .005), regardless of cytology reports, hrHPV status and age of patients. Secondly, multivariate logistic regression also showed that an ECC CIN 1 diagnosis (OR = 1.80, 95% CI = 1.33–2.44, p < .001) was a risk factor for CIN 2+. Thirdly, the risk of CIN 2+ in ECC CIN 1 was significantly higher than biopsy CIN 1 in both cytology of LSIL or less subgroup and cytology of HSIL or worse subgroup.
Consistent with previous studies, cytology was an important risk factor for CIN 2+ diagnosis. Multivariate logistic regression showed that cytology findings of HSIL+, ASC-H, and ECC CIN 1 were risk factors for CIN 2+, whereas cytology of ASC-US, LSIL, hrHPV status, and age were not. The high-grade cytology findings were more significant than ECC CIN 1 (OR = 4.73–4.88 vs 1.80). The risk of CIN 2+ in ECC CIN 1 with cytology of LSIL or less was lower than biopsy CIN 1 with cytology of HSIL or worse (21.68% vs 36.11%, p < .001). Therefore, a high-grade cytology finding was a higher risk factor than ECC CIN 1.
Fine et al13 evaluated the relationship between the ECC grade and the excisional specimen histological findings, and they demonstrated a perfect agreement in 36.7% of cases and agreement within 1 grade in 60.4% of cases. In our study, ECC CIN 1 reached perfect agreement with the excisional specimen histological findings in 46.5% and agreement within 1 grade in 99.5% of cases. Endocervical curettage during LEEP was important in predicting persistent/recurrent disease. A retrospective study included 330 women with a mean follow-up of 29.4 months; 188 women underwent ECC at the time of LEEP. On multivariate analysis, a positive ECC was the only variable significantly associated with persistence/recurrence (p = .001). In the Cox regression model, positive ECC (p = .001) and positive margins (p = .009) were independently associated with higher failure rates.14
Our study showed that ECC CIN 1 could help detect AIS and occult squamous cancer in addition to cytology. In this study, 0.25% (3/1,195) women with cervical CIN 1 were diagnosed with invasive cervical cancer by LEEP, which was similar to 0.24% (3/1,257) women in our previous study.15 However, this study was retrospective with age of ECC CIN 1 group women older than biopsy CIN 1 (44.9 ± 9.1 vs 42.6 ± 8.4 years old). In addition, we cannot present the results of those who are CIN 3+ separately from those who are CIN 2+, because we have changed from CIN to squamous intraepithelial lesion terminology since 2013. Hence, a large prospective study comparing the risk of CIN 3+ between the 2 groups would be preferable.
In conclusion, ECC CIN 1 has a greater risk of CIN 2+ diagnosis than biopsy CIN 1, but high-grade cytology has a higher risk than ECC CIN 1.
The authors thank Editage (http://www.editage.com) for English language editing.
1. Wright TC Jr., Massad LS, Dunton CJ, et al. 2006 consensus guidelines for the management of women with cervical intraepithelial neoplasia or adenocarcinoma in situ. J Low Genit Tract Dis
2. Massad LS, Einstein MH, Huh WK, et al. 2012 updated consensus guidelines for the management of abnormal cervical cancer
screening tests and cancer precursors. Obstet Gynecol
3. Petersen S, Belnap C, Larsen WI, et al. Grading of squamous dysplasia in endocervical curettage
specimens: the case for conservative management of mild endocervical dysplasia. J Reprod Med
4. Fukuchi E, Fetterman B, Poitras N, et al. Risk of cervical precancer
and cancer in women with cervical intraepithelial neoplasia grade 1 on endocervical curettage
. J Low Genit Tract Dis
5. Gage JC, Duggan MA, Nation JG, et al. Comparative risk of high-grade histopathology diagnosis after a CIN 1
finding in endocervical curettage
versus cervical biopsy. J Low Genit Tract Dis
6. Pretorius RG, Zhang WH, Belinson JL, et al. Colposcopically directed biopsy, random cervical biopsy, and endocervical curettage
in the diagnosis of cervical intraepithelial neoplasia II or worse. Am J Obstet Gynecol
7. Guo LP, Cong Q, Zhang H, et al. The importance of endocervical curettage
in an old post-loop electrosurgical excision procedure woman with abnormal cervical cytology and a normal punch biopsy histology: a case report and literature review. Reprod Dev Med
8. Liu AH, Walker J, Gage JC, et al. Diagnosis of cervical precancers by endocervical curettage
at colposcopy of women with abnormal cervical cytology. Obstet Gynecol
9. Ureyen I, Toptas T, Uysal A. Is routine ECC necessary in patients with HPV16 and normal cytology? Diagn Cytopathol
10. Castle PE, Sideri M, Jeronimo J, et al. Risk assessment to guide the prevention of cervical cancer
. J Low Genit Tract Dis
11. Basu P, Sankaranarayanan R. Atlas of Colposcopy – Principles and Practice: International Agency for Research on Cancer. Available at: https://screening.iarc.fr/atlascolpo.php
. Accessed April 9, 2020.
12. Lastra RR, Meykler SE, Baloch ZW, et al. Increasing the sensitivity of endocervicalcurettings by performing ThinPrep(R) Pap on transport container fluid: is diagnostic material going down the drain? Cytopathology
13. Fine BA, Feinstein GI, Sabella V. The pre- and postoperative value of endocervical curettage
in the detection of cervical intraepithelial neoplasia and invasive cervical cancer
. Gynecol Oncol
14. Cuello MA, Espinosa ME, Orlandini EJ, et al. The value of endocervical curettage
during loop electrosurgical excision procedures in predicting persistent/recurrent preinvasive cervical disease. Int J Gynaecol Obstet
15. Cong Q, Song Y, Wang Q, et al. A large retrospective study of 12714 cases of LEEP conization focusing on cervical cancer
that colposcopy-directed biopsy
failed to detect. Biomed Res Int