Cytohistologic Discrepancy of High-Grade Squamous Intraepithelial Lesions in Papanicolaou Smears

The Papanicolaou (Pap) smear is the most widely used screening method for cervical cancer. The 2001 Bethesda system terminology is used for cytologic classification (Solomon et al., 2002). Management of abnormal cervical cytology depends on the degree of abnormalities of cervical cytology, previous history of abnormal Pap test and age of patients. High grade squamous intraepithelial lesion (HSIL) Pap smear carries a high risk for significant cervical pathology. Women with HSIL Pap smear have been reported to have a high-grade cervical intraepithelial neoplasia (CIN) at 53-66% and 84-97% from colposcopic directed biopsy and loop electrosurgical excision procedure (LEEP), respectively (Massad et al., 2001; Dunn et al., 2003; Kantathavorn et al., 2006; Alvarez and Wright, 2007; Sadan et al., 2007). Approximately 1-4% of women with HSIL Pap smear had invasive cervical cancer (Massad et al., 2001; Wright et al., 2007). However, previous studies showed that 8-18 % of women with HSIL Pap smear had low-grade CIN (CIN 1) from colposcopic directed biopsy (Numnum et al., 2005; Cho and Kim, 2009; Li et al., 2009), so-called cytohistologic discrepancy. Management for patients with cytohistologic discrepancy who had


Introduction
The Papanicolaou (Pap) smear is the most widely used screening method for cervical cancer.The 2001 Bethesda system terminology is used for cytologic classification (Solomon et al., 2002).Management of abnormal cervical cytology depends on the degree of abnormalities of cervical cytology, previous history of abnormal Pap test and age of patients.
Observation with colposcopy and cytology has a disadvantage in patients with CIN 2-3 who are missed by colposcopic examination with regard to delaying the treatment.To the best of our knowledge, data on factors associated with cytohistologic discrepancy in HSIL Pap smear are limited.Therefore, we conducted this study to evaluate the factors associated with cytohistologic discrepancy in HSIL Pap smear and to determine the rate of cytohistologic discrepancy.

Materials and Methods
Medical records of women with HSIL Pap smear who were treated at Thammasat University Hospital, the tertiary hospital and referral center during the years 2005-2011 were reviewed.Inclusion criteria were women with HSIL Pap smear and had pathology reports of both colposcopic directed biopsy and LEEP.Age, parity, menopausal status, contraceptive use and pathology reports of colposcopic directed biopsy and LEEP were collected.Exclusion criteria were women who had prior hysterectomy and no histological data.Available cytologic and all histologic slides were reviewed by a single cytopathologist (W.H.).Cytohistologic discrepancy was defined as having HSIL Pap smear but less than CIN 2 histology from colposcopic directed biopsy and/or LEEP.
The study was approved by the Ethics Committee of Thammasat University.
Data were analyzed using SPSS version 15 program.Descriptive statistics were used for demographic data.Chi-square or Fisher's exact test, where appropriate, was used to compare the difference between groups.P <0.05 was considered statistically significant.

Results
There were 223 women who had HSIL Pap smear results recruited.Mean (SD) age was 38.0 (9.4) years.Thirty-six (16.1%) of them were nulliparous.One-hundred and ninety-three (86.5%) women were premenopausal and 46 (20.6%) women used oral contraceptive pills for birth control at the time of HSIL Pap smear discovery (Table 1).
One hundred and eleven (49.8%) women had CIN 2 or greater on both colposcopic directed biopsy and LEEP specimen.Ten (4.5%) women had CIN 2 or greater on colposcopic directed biopsy but no high-grade CIN on LEEP specimen, while 48 (21.5%) women had no highgrade CIN on colposcopic directed biopsy but CIN 2 or greater on LEEP specimen.Fifty-four (24.2%) women had no high-grade CIN on both colposcopic directed biopsy and LEEP.Therefore, the exact number of cytohistologic discrepancy (less than CIN 2 on both colposcopic directed biopsy and LEEP specimens) was 24.2% (Table 2).
Multivariate analysis using logistic regression model was performed and confirmed that nulliparity, postmenopausal status and having no oral contraceptive pills use were associated with cytohistologic discrepancy (Table 3).

Discussion
Cytohistologic discrepancy between Pap test and colposcopic directed biopsy was greater (45.7%) than that   between Pap test and LEEP (29.5%).This result indicated that high-grade CIN could be missed by colposcopic examination.Due to the fact that accurate colposcopic diagnosis depends on many factors such as colposcopic findings i.e. size of lesion; satisfactory or unsatisfactory colposcopy and an experience of colposcopists.
Our study demonstrated the exact percentage of cytohistologic discrepancy as high as 24.2.Previous studies demonstrated a lower frequency.(Numnum et al., 2005;Cho and Kim, 2009;Li et al., 2009).Li et al (Li et al., 2009) and Numnum et al (Numnum et al., 2005) reported the prevalences of cytohistologic discrepancy of 7.8% and respectively.Greater cytohistologic discrepancy found in our study could be a matter of cytologic interpretation error of Pap test, resulting in an overdiagnosis of HSIL.As almost half of the patients were referred from other hospitals, cytologic slides for review were not available.
Cytologic error was reported to be the major cause of cytohistologic discrepancy (Tzeng et al., 1999;Moss et al., 2010).Cytologic errors included cytologic sampling error, poor specimen preservation, suboptimal staining quality and cytologic interpretation error.
By using univariate and multivariate analyses, our study showed that nulliparity postmenopausal status and having no oral contraceptive pills use were associated with cytohistologic discrepancy.Our study agrees with the previous studies which reported multiparity and current use of contraceptive pills were significantly associated with high-grade CIN (Parazzini et al., 1992;De Vet et al., 1993;Clements et al., 2011;Gargano et al., 2012).
The option for managing women with HSIL Pap smear but low-grade CIN on colposcopic directed biopsy is either immediate diagnostic excisional procedure or observation with colposcopy and cytology at 6-month interval (Wright et al., 2007).Diagnostic excisional procedure carries a risk of perioperative hemorrhage, infection and adverse pregnancy outcomes including preterm delivery, premature rupture of membrane and low birthweight (Crane, 2003;Kietpeerakool et al., 2006;Kyrgiou et al., 2006;Sjoborg et al., 2007;Simoens et al., 2012).Therefore, if clinical factors associated with low risk for high-grade CIN are present, observation with colposcopy and cytology may be the appropriate option.We would suggest to observe and perform colposcopy and cytology at 6-month interval in nulliparous, postmenopausal women and women who do not use oral contraceptive pills.
Strength of our study is that final histological diagnosis was confirmed on LEEP specimens.However, there were some limitations.As it was a retrospective study, data on other factors which may be associated with cytohistologic discrepancy such as sexual behavior, previous history of Pap testing, were lacking.In addition, half of patients were referred from other hospitals, therefore, review of cervical cytology was not possible.
In conclusion, cytohistologic discrepancy rate in our study was relatively high at 24.2%.Nulliparity, postmenopausal status and having no oral contraceptive pills use were associated with cytohistologic discrepancy.Therefore, we suggest to observe and perform colposcopy and cytology at 6-month interval in HSIL Pap test patients who have these factors.