The current study examines 1) the sensitivity of detection and 2) sampling and screening/diagnostic error in the cytologic diagnosis of adenocarcinoma in situ (AIS) of the cervix. The data were taken from public and private sector screening laboratories reporting 25,000 and 80,000 smears, respectively, each year.
The study group was comprised of women with a biopsy diagnosis of AIS or AIS combined with a high-grade squamous intraepithelial lesion (HSIL) who were accessioned by the Western Australian Cervical Cytology Registry (WACCR) between 1993-1998. Cervical smears reported by the Western Australia Centre for Pathology and Medical Research (PathCentre) or Western Diagnostic Pathology (WDP) in the 36 months before the index biopsy was obtained were retrieved. A true measure of the sensitivity of detection could not be determined because to the authors' knowledge the exact prevalence of disease is unknown at present. For the current study, sensitivity was defined as the percentage of smears reported as demonstrating a possible or definite high-grade epithelial abnormality (HGEA), either glandular or squamous. Sampling error was defined as the percentage of smears found to have no HGEA on review. Screening/diagnostic error was defined as the percentage of smears in which HGEA was not diagnosed initially but review demonstrated possible or definite HGEA. Sensitivity also was calculated for a randomly selected control group of biopsy proven cases of Grade 3 cervical intraepithelial neoplasia (CIN 3) accessioned at the WACCR in 1999.
For biopsy findings of AIS alone, the diagnostic "sensitivity" of a single smear was 47.6% for the PathCentre and 54.3% for WDP. Nearly all the abnormalities were reported as glandular. The sampling and screening/diagnostic errors were 47.6% and 4.8%, respectively, for the PathCentre and 33.3% and 12.3%, respectively, for WDP. The results from the PathCentre were better for AIS plus HSIL than for AIS alone, but the results from WDP were similar for both groups. For the CIN 3 control cases, the "sensitivity" of a single smear was 42.5%.
To the authors' knowledge epidemiologic studies published to date have not demonstrated a benefit from screening for precursors of cervical adenocarcinoma. However, in the study laboratories as in many others, reasonable expertise in diagnosing AIS has been acquired only within the last 10-15 years, which may be too short a period in which to demonstrate a significant effect. The results of the current study provide some encouraging baseline data regarding the sensitivity of the Papanicolaou smear in detecting AIS. Further improvements in sampling and cytodiagnosis may be possible.