Abstract
Objective:
The distribution of cervical intraepithelial neoplasia (CIN) lesions across the cervix was determined.
Methods:
A total of 575 women whose pathological diagnosis after cervical conization was confirmed as CIN were studied; 146 had low-grade CIN and 429 had high-grade CIN. CIN lesion location on the cervix was recorded using 12-h clock face notation.
Results:
In both groups, 12 o'clock was the most common and 2 o'clock the least common lesion location. The most severe lesions were most often located at 8 o'clock and 7 o'clock, in the low- and high-grade groups, respectively. The 2 o'clock site was the least frequent site for the most severe lesion in both groups. Lesions were found more frequently on the posterior lip of the cervix than on the anterior lip, and on the right side of the cervix than on the left side, in both groups.
Conclusions:
The distribution of CIN lesions is not randomly distributed across the cervix. The 12, 8 and 7 o'clock sites, and the posterior lip and right side of the cervix, should be targeted during colposcopy-directed biopsy of patients with CIN lesions as this may improve diagnostic accuracy.
Keywords
Introduction
Cervical intraepithelial neoplasia (CIN) is a precancerous lesion of the cervix that can progress to become cervical cancer, which is one of the most common gynaecological malignancies. 1 CIN of all grades can potentially develop into invasive cervical cancer. 2 Consequently, early diagnosis and treatment of CIN are essential for the prevention and treatment of cervical cancer. 3
Although human papillomavirus is the major aetiological agent associated with the development of CIN, the diagnosis of infection and/or the identification of cervical lesions is initially performed using the Papanicolaou (Pap) test. 4 An abnormal Pap smear prompts a colposcopic examination with biopsies of abnormal-appearing cervical tissue. 5 It is common practice to use a three-step procedure involving cytology, colposcopy and histopathology to diagnose CIN, colposcopy-directed multiple biopsy sampling being the key step. 6 However, the accuracy of this technique is highly dependent on the skill of the colposcopist; colposcopy-based diagnosis of CIN may be associated with a false-positive diagnosis leading to unnecessary hysterectomy, 7 or to a false-negative diagnosis that may overlook CIN lesions developing into invasive cancer. 8 Clinical decision making is a serious challenge because of these limitations of colposcopy-directed multiple biopsy sampling.
The incidence of CIN varies greatly among different regions and countries. 9 It is not known whether the distribution pattern of CIN lesions across the surface of the cervix is associated with the risk of cancer progression. Little research has been conducted on this subject and no clear conclusions have been drawn. If a pattern to the distribution of CIN lesions across the cervix can be identified, this may be useful in improving the accuracy of clinical biopsies. The present study investigated the distribution pattern of CIN lesions over 12 sites on the cervix in Chinese women who had a pathological diagnosis of CIN after cervical conization.
Patients and methods
Patient Population
In this retrospective histopathological study, data stored in the database of the China– Japan Friendship Hospital, Beijing, China, were analysed. The data were obtained from Chinese women with a pathological diagnosis of CIN after cervical conization that was undertaken between May 2005 and May 2010. Biopsy samples taken as part of the cervical conization procedure were used in the histopathological analysis. Inclusion criteria were: (i) CIN 2 or 3 (explanation of CIN grades is given in the next section); (ii) persistent CIN 1; (iii) cytological/histological discrepancy; (iv) inadequate colposcopy; (v) positive endocervical biopsy; (vi) suspicion of invasive cervical cancer. Exclusion criteria were: pregnancy; undiagnosed uterine bleeding; invasive carcinoma. Specimens were evaluated by three or four gynaecological pathologists.
The study was approved by the Ethics Committee of the China–Japan Friendship Hospital (Ref. No. ZRLW-2) and all study subjects provided verbal informed consent to participation.
Diagnosis and Classification of CIN Lesions
All patients who provided cervical biopsy samples for the study underwent cytological screening, colposcopy-directed cervical biopsy and histopathological examination. The surgeons (all authors of this article) used a 12-h clock face to establish the reference points of the cervix during conization (i.e. 1 o'clock, 2 o'clock, etc.), and cervical conization tissues were divided into 12 equal segments, which were examined in serial sections. Histopathological samples taken during cervical conization were classified as low- or high-grade CIN. The low-grade group included samples with histopathological results that demonstrated CIN grade 1 (mild dysplasia or abnormal cell growth) or changes associated with human papillomavirus (i.e. the presence of koilocytes, which occur as a result of infection of the cell by human papillomavirus), while the high-grade group included samples with pathological changes that were consistent with CIN grade 2 (moderate dysplasia confined to the basal two-thirds of the epithelium) or grade 3 (severe dysplasia that spans more than two-thirds of the epithelium, and may involve the full thickness; also known as cervical carcinoma in situ). 10
In the first part of the study, sites where lesions were detected on the cervix of each patient after conization were counted, to determine the most common sites of CIN lesions. In the second part, the site of the most severe lesion in each patient was identified, to determine the most frequent site of severe CIN lesions. In the third part of the study, the cervix was divided into four quadrants: quadrants 1 – 4 comprised the segments from 1 to 3 o'clock, from 4 to 6 o'clock, from 7 to 9 o'clock and from 10 to 12 o'clock, respectively. The pattern of CIN lesion distribution was compared between the four quadrants and between different combinations of two quadrants, which were equivalent to specific anatomical locations on the cervix: quadrants 1 + 4 represented the anterior lip of the cervix and quadrants 2 + 3 the posterior lip.
Statistical Analyses
The data were analysed with the χ2-test using the SPSS® statistical package, version 10.0 (SPSS Inc., Chicago, IL, USA) for Windows®. A P-value < 0.05 was considered statistically significant.
Results
A total of 575 women with confirmed CIN participated in the study. The mean age of the patients was 38.4 years (range 19 – 66 years). Of the 575 patients, 146 were classified as having low-grade CIN lesions and 429 as having high-grade CIN lesions after cervical conization. According to the requirements of the χ2-test statistical analysis (sample size > 40, expected frequency of 5 CIN lesions at each location), the number of patients (n = 575) was considered adequate for analysis.
The locations and frequencies of the CIN lesions, based on histopathological results after conization, are presented in Table 1. In both the low- and high-grade groups, the site at which CIN lesions were most frequent was 12 o'clock, and the site where they were least frequent was 2 o'clock. Lesion frequency was significantly different between the 12 o'clock and 2 o'clock sites in the high-grade group (P < 0.05), but not in the low-grade group.
Frequencies of low- and high-grade cervical intraepithelial neoplasia (CIN) lesions at 12 sites on the cervix in women diagnosed with CIN after conization
Data presented as n (%) of patients.
The cervix, viewed as a circle, was divided into 12 segments denoted 1 – 12 by analogy with the 12 positions of the hours on a clock face (i.e., 1 – 12 o'clock).
Not statistically significant compared with 2 o'clock in the low-grade group (χ2-test).
P < 0.05 compared with 2 o'clock in the high-grade group (χ2-test).
In the 146 patients with low-grade CIN, the most severe lesion occurred with the highest frequency at the 8 o'clock location and with the lowest frequency at the 2 o'clock location (18/146 [12.3%] versus 4/146 [2.7%], respectively; P < 0.01; Fig. 1). In the 429 patients with high-grade CIN, the most severe lesion occurred with the highest frequency at the 7 o'clock location and with the lowest frequency at the 2 o'clock location (49/429 [11.4%] versus 19/429 [4.4%]; P < 0.01; Fig. 1).
Frequency distribution at the 12 sites on the cervix where the most severe lesion occurred in each patient stratified according to low- and high-grade cervical intraepithelial neoplasia (CIN) diagnosed after conization (aP < 0.01 for 8 o'clock versus 2 o'clock in the low-grade group; bP < 0.01 for 7 o'clock versus 2 o'clock in the high-grade group; χ2-test)
When the distribution of CIN lesions among the four cervical quadrants (Table 2) was examined, the frequency of CIN lesions was significantly lower in quadrant 1 than in quadrants 2 – 4 in the in high-grade group (P < 0.001 for all comparisons). In the low-grade group, the quadrant 1 had a significantly lower proportion of CIN lesions than quadrants 2 (P < 0.05) and 3 (P < 0.01).
Frequency distribution of low- and high-grade cervical intraepithelial neoplasia (CIN) lesions at 12 sites on the cervix, grouped into four quadrants, in women diagnosed with CIN after conization
Data presented as n (%) of patients.
The cervix, viewed as a circle, was divided into 12 segments denoted 1 – 12 by analogy with the 12 positions of the hours on a clock face (i.e., 1 – 12 o'clock); these segments were then grouped into the four quadrants shown.
P < 0.05,
P < 0.01 and
P < 0.001 compared with quadrant 1 (χ2-test).
Table 3 presents CIN frequency distribution data for combinations of the four quadrants that are equivalent to the different anatomical areas of the cervix. The prevalence of CIN was significantly higher in the posterior lip of the cervix compared with the anterior lip in both the low- and high-grade groups (P < 0.01 for both comparisons). The prevalence of CIN was also significantly higher on the right side of the cervix compared with the left side in both the low- and high-grade groups (P < 0.05 and P < 0.01, respectively).
Frequency distribution of low- and high-grade cervical intraepithelial neoplasia (CIN) lesions on the four quadrants of the cervix in women diagnosed with CIN after conization
Data presented as n (%) of patients.
P < 0.01 for posterior lip versus anterior lip in both groups (χ2-test).
P < 0.05 for right side versus left side in the low-grade group (χ2-test).
P < 0.01 for right side versus left side in the high-grade group (χ2-test).
Discussion
The goal of colposcopy-directed cervical biopsy is to obtain a representative tissue specimen from the lesion that is suspected to be the most severe. The process of taking a cervical biopsy is inherently subjective because it relies on the skill of the colposcopist, and has been demonstrated to be fairly inaccurate even for the most skilled of practitioners. 11 The diagnostic accuracy of colposcopy-directed biopsy has been reported to be 53.4%, with 23.6% under-diagnosis and 23.0% over-diagnosis. 12 A critical review reported false-negative diagnosis rates of 0 – 8.9%, 13 so there are clear limitations associated with colposcopy-directed biopsy.
In the absence of better diagnostic methods, most investigations of CIN are still based on colposcopy-directed biopsy samples. To date, there is no clear consensus on the importance of the distribution pattern of CIN across the cervix. In 1948, a study found that biopsies from the anterior and posterior cervix were more likely to have CIN grade 3 lesions than biopsies taken from the lateral angles. 14 Heatley, 15 using the method that had been used in the 1948 study to section the cervix sagittally, found similar results in 1995. Colposcopists may not select sites for biopsy at random. For example, they may select sites based on the mechanical ease of performing the biopsy, or it may be easier to perform a biopsy by holding the instrument straight up and down rather than at a lateral angle. A study using colpomicroscopy demonstrated that CIN grade 3 lesions occurred more frequently on the anterior than on the posterior lip of the cervix. 16 Similar findings were reported by Guido et al. 17 in an investigation of the topographic distribution of precancerous intraepithelial lesions on the cervix. It is possible that colpomicroscopy and colposcopy detect more lesions on the anterior cervical hemisphere because the cervix is usually oriented downwards and it may be easier to obtain a perpendicular view of the anterior hemisphere. As a consequence, some lesions at other sites could have been missed in the previous studies mentioned above. A study of 69 cases using colposcopy-directed cervical biopsy demonstrated a random distribution of CIN across the cervix, 18 which supported the findings of an earlier study of CIN distribution. 19 In contrast, CIN lesions of grade 2 or higher were found slightly more frequently on the posterior lip than on the anterior lip of the cervix (426/806 [52.9%] versus 380/806 [47.1%]; P < 0.05). 20
It is clear from the evidence presented above that the results of previous histopathological studies have been inconsistent,13–19 largely because they all used histopathological results from colposcopical biopsies. A study of cervix conization specimens with detailed histological data on the entire specimen would help to determine the true distribution of CIN lesions. Compared with colposcopic biopsies, which are subject to the colposcopist's skill and site-selection preferences, the present study, which was based on histopathological results from 12 sites after conization, should provide a more objective and accurate assessment of CIN lesion distribution.
The prevalence of CIN in the present study was significantly greater on the posterior lip of the cervix compared with the anterior lip of the cervix in both the low- and the high-grade groups, which was in agreement with a previous study. 19 Although the results of the present study were statistically meaningful, the small sample size and the number of comparisons could lead to false positives. Future studies will need to include larger numbers of samples to confirm these findings and to prevent methodological errors influencing the results.
In conclusion, the findings of the present study demonstrated that CIN lesions were not randomly distributed across the cervix: (i) the most common CIN location was at 12 o'clock and the least common was at 2 o'clock; (ii) the most frequent sites for the most severe CIN lesions were at 8 o'clock and 7 o'clock in the low- and high-grade groups, respectively, and the least frequent site was at 2 o'clock in both groups; (iii) the frequency of CIN lesions was lower in quadrant 1 (1 – 3 o'clock) of the cervix than in the other three quadrants; (iv) the frequency of CIN lesions was significantly higher on the posterior lip of the cervix than on the anterior lip, and on the right side of the cervix than on the left side.
The nonrandom distribution pattern found during this study has important implications related to the pathogenesis and clinical management of CIN, and we make the following suggestions: (i) during colposcopic examination, the 7, 8 and 12 o'clock sites of the cervix should be targeted for biopsy in preference to other sites in the event that multiple, diffuse, or questionable changes are found, so as to improve diagnostic accuracy; (ii) when performing cervical conization, clinicians should pay close attention to the resection range of the posterior lip and the right side of the cervix, to reduce residual lesions at the margins so as to improve clinical efficacy.
Footnotes
The authors had no conflicts of interest to declare in relation to this article.