Endocervical glandular atypia: does a preneoplastic lesion of adenocarcinoma in situ exist?

Some believe endocervical glandular atypia (EGA), purportedly composed of cells that are less atypical than cells of adenocarcinoma in situ (AIS), is a preneoplastic precursor of AIS. We examined 246 neoplastic and nonneoplastic cervical cone biopsy and hysterectomy specimens from 221 patients for lesions composed of glandular cells with less atypia than AIS to define and characterize their association with other glandular processes. To avoid the circular argument of high-grade EGA (dysplasia) vs AIS, we set the minimum degree of AIS cells as the degree of atypia of the cells constituting a moderately differentiated invasive adenocarcinoma. Only 4 endocervical glandular lesions with mild atypia were found, 3 in patients with AIS or invasive endocervical-primary adenocarcinoma and 1 in a patient with invasive endometrial-primary, adenocarcinoma with endocervical extension. There were no lesions with high-grade atypia, nor was there a morphologic spectrum of cells with less atypia than AIS. Of the specimens, 14% had benign endocervical cell changes. The percentage of specimens in each group with benign endocervical cell changes was approximately equal. Although our study is small and retrospective, it suggests that no morphologic evidence exists to support the existence of a spectrum of endocervical glandular changes that culminates in AIS.