Subdividing atypical glandular cells of undetermined significance according to the Australian modified Bethesda, system: analysis of outcomes

The Australian modification of the Bethesda system (TBS) for reporting endocervical cellular abnormalities differs from TBS in having a "high grade" category for prediction of adenocarcinoma in situ (AIS), an "inconclusive" category requiring that AIS be excluded (AISEX), and a "low grade" category (Enatyp) for when only minor changes are seen in endocervical cells. From subsequent follow-up, the authors evaluated the proportion of histologically confirmed high grade lesions (squamous, glandular, or endometrial) in each cytologic reporting category, for cases in which no concurrent squamous abnormality was also predicted. From information was available for 95.9% of 172 AIS predictions, 81.5% of 119 AISEX predictions, and 89.0% of 237 reports of Enatyp. The histologic positive predictive values for high grade lesions were 95.7% for AIS, 71.1% for AISEX, and 7.1% for Enatyp. In the first 2 categories, a minority of lesions proved to be purely squamous (3.8% for AIS and 33.3% for AISEX), whereas in the Enatyp category 87.0% were squamous only. Compared with TBS, the Australian terminology for reporting of endocervical abnormalities better reflects the risk of underlying high grade abnormalities and may result in more appropriate management recommendations. For Enatyp reports, which constitute over 85% of reported endocervical abnormalities, repeat cytology should be considered appropriate management. However, cone biopsy should be the investigative method of choice for women with reports of AIS and AISEX.