The Biopsy Correlates of Abnormal Cervical Cytology Classified According to the Bethesda System

Kim, Yoon Sook; Bae, Dong Han; Sunwoo, Jae Gun; Kim, Jong Soo

Abstract

OBJECTIVE
The goal of this study was to examine the colposcopic findings underlying referred cytologic abnormalities classified according to the Bethesda system. Further determine the strength of correlations between the referral smear performed at colposcopy, and the histologic diagnosis.
METHODS
The charts of 462 patients who had referred to our hospital due to abnormal cytology at another hospitals from Jan. 1997 to Dec. 2001 at Soonchunhyang University Chunan Hospital were retrospectively reviewed. According to the necessities, repeat Pap, colposcopy-guided biopsy and conization was done. Results were compared both with referral Pap. One way ANOVA and chi-square test were used for statistical analysis.
RESULTS
The results were as follows. 1. Mean age of the patients was 46.3 +/- 3.7 years. 2. Mean interval between referral smear and repeat smear or colposcopy was 3-82 days. 3. The agreement between referral and repeat Pap smears was 48.4%, no AGUS smears were conformed by repeat cytology and abnormal referral smears reverted to negative in 36.7%: 28.6% for the women with ASCUS, 50% with AGUS, 60% with LSIL, 36.4% with HSIL, and 13% with cancer. 4. The likelihood of a negative colposcopy guided biopsy of referral Pap smears was 14.3% for the women with ASCUS, 50% with AGUS, 25% with LSIL, 15.4% with HSIL, and 0% with cancer and more severe lesion than LSIL was found in 71.4% with ASCUS, 100% with AGUS, 68.8% with LSIL, 84.6% with HSIL, 100% with cancer. 5. The likelihood of a negative colposcopy guided biopsy of repeat Pap smears was 9.1% for the women with ASCUS, 50% with LSIL, 9.5% with HSIL, and 0% with cancer and more severe lesion than LSIL was found in 90.9% with ASCUS, 50% with LSIL, 90.5 with HSIL, 100% with cancer. 6. The likelihood of a negative conization of referral Pap smears was 12.5% for the women with ASCUS, 33.3% with AGUS, 0% with LSIL, 16.4% with HSIL, and 0% with cancer and more severe lesion than LSIL was found in 87.5% with ASCUS, 66.6% with AGUS, 100% with LSIL, 83.6% with HSIL, 100% with cancer.
CONCLUSION
The agreement of referral smears with colposcopic biopsy and conization was 54.3% and 51.9%, respectively. Since the cytology classified according to the Bethesda system does not accurately predict histologic diagnosis, cytologic findings should not be used to determine treatment. Further we should use other methods as colposcopy, punch biopsy and conization to determine treatment.