Abnormal cervical cytology in pregnancy: a 17-year experience.

Abstract

To evaluate the safety and accuracy of colposcopy and colposcopically directed biopsy in pregnant women with abnormal cervical cytology. A retrospective analysis of 612 gravidas with abnormal cervical cytology was conducted. Colposcopy and directed biopsy were performed using standard techniques. Two patients underwent diagnostic conization during the second trimester. One hundred twelve patients had procedures that provided a final specimen. Endocervical curettage was omitted. The transformation zone was fully visualized in all patients by the 20th week of gestation. Directed cervical biopsy was performed on the following patients: 1) with colposcopic evidence of invasion or cervical intraepithelial neoplasia (CIN) III, 2) with discordancy between colposcopy and cytology, 3) electing termination of pregnancy, and 4) whose anticipated reliability was even remotely questioned. A colposcopically directed biopsy was performed in 449 patients (73%). Ninety-one patients (15%) did not have biopsies because of normal colposcopic findings, and the remaining 72 patients (12%) had either CIN I or II. Thirty-nine of these patients (6%) were lost to follow-up. Colposcopically directed biopsy and colposcopic impression had a 95% concordancy within one degree of severity; however, 14% of CIN I colposcopic impressions and 54% of normal colposcopic findings turned out to be CIN III and CIN I or II, respectively. Ninety-five percent of the biopsy diagnoses correlated with the final pathology to within one degree of severity. The data confirm previous findings that colposcopically directed biopsy is a safe and reliable method of evaluating pregnant patients with abnormal cervical cytology.