The sensitivity of cervical smears for adenocarcinoma in situ (AIS) is not known, nor is it known whether false-negative smears are due to sampling or to screening or interpretive errors. In 16 of 34 patients with AIS, 38 negative smears were reported 2 weeks to 7 years before biopsy. Thirty-one of these negative smears were rescreened, and 17 (55%) were retrospectively diagnosed as abnormal. The purpose of this study was to obtain an estimate of Pap smear sensitivity for AIS and to uncover difficulties in interpretation of smears that may have led to false-negative diagnoses.
MATERIALS AND METHODS
The surgical p a t h o l o g y files of Brigham andWomen's Hospital, Boston, were searched for case reports of AIS or AIS combined with squamous cervical intraepithelial neoplasia (CIN) with a complete cone biopsy, hysterectomy, or both for the years 1986 to 1995. All available smears diagnosed as negative before an initial biopsy confirming AIS were confirmed sampling errors. The sensitivity of cervical smears for AIS was 55% to 72%. Improved sampling of the endocervical canal offers cytologists the opportunity to diagnose AIS. This study demonstrates that this opportunity may not be fully exploited. Small "endometrioid" AIS cells and AIS cells resembling reactive endocervical cells may be mistaken for benign cells, thus decreasing sensitivity.