RESUMO
INTRODUCTION: Colposcopic endocervical brushing cytology (CEB) is more sensitive than endocervical curettage (ECC) for detecting squamous intraepithelial lesions. There are no data on performance of CEB for detecting endocervical adenocarcinoma. MATERIALS AND METHODS: A total of 151 patients were identified in a word search for "endocervical adenocarcinoma" in surgical pathology reports from January 2007 to June 2019. To measure sensitivity, reports of CEB or ECC samples within 1 year preceding the first surgical pathology diagnosis of at least endocervical adenocarcinoma in situ (AIS+) were examined. Specificity was measured in a cohort in which at least atypical glandular cells (AGC+) was reported in CEB or ECC. RESULTS: Seven CEB preceding diagnosis of AIS were identified: 6 of 7 were positive or suspicious for AIS+. One of 7 was negative and it was negative on re-review. Three of 6 positive CEB cases used cell blocks with immunohistochemistry. Seventy ECC samples preceding diagnosis of AIS were identified: 40 of 70 were diagnosed as AGC+. The sensitivities of CEB and ECC for detecting AIS+ at a threshold of AGC+ are 86% and 57% (too few patients for statistics), respectively. For specificity, 12 of 18 CEB and 9 of 25 ECC reports with AGC+ were false positive by follow-up surgical pathology. The specificities of CEB and ECC are 99.4% and 99.9%, respectively. CONCLUSION: Sensitivity of CEB for detecting AIS+ (86%) is at least as high as ECC (57%). Specificity of CEB is similar to ECC. Addition of a cell block to CEB may be useful. CEB appears to be an appropriate test for follow-up of atypical glandular cells reported on Papanicolaou tests.