Atypical Glandular Cells of Undetermined Significance: Diagnostic Accuracy and Interobserver Variability Using Select Cytologic Criteria

Abstract

Histologic follow-up of patients with the Bethesda system cervicalvaginal diagnosis of atypical glandular cells of undetermined significance (AGUS), "favor endocervical origin", or "not otherwise specified" (NOS) shows a high percentage of clinically significant (neoplastic or preneoplastic) lesions. Using the criteria of atypical single cells, irregular nuclear membranes, and decreased cytoplasm, eight observers retrospectively reclassified 88 AGUS, "favor endocervical", or NOS smears using a probabilistic scheme. Follow-up showed 46 clinically significant and 42 benign lesions. The mean accuracy for all observers and the experienced observers was 65% and 72%, respectively. For the experienced observers, the mean specificity of a "favor clinically significant" category was 72%; the mean sensitivity of a "favor benign" category was 90%. For the less experienced observers, subclassification had poor predictive value. We conclude that experienced observers may use specific criteria to correctly subclassify AGUS lesions, and this may aid in patient management. (Key words: Cytology; Accuracy; Cervical-vaginal smear; Dysplasia; Probability; Carcinoma) Am J Clin Pathol 1997,107:299-307.The Bethesda system diagnostic category of atypical glandular cells of undetermined significance (AGUS) is important for two reasons: (1) AGUS has a relatively high incidence that reportedly ranges from 0.46% to 2.5% of all cervical-vaginal smears 1-3 ; and (2) a high percentage (42%-83%) of patients with an AGUS diagnosis have a clinically significant lesion. [1][2][3][4][5][6][7][8] Surprisingly, histologic follow-up has shown that some of the clinically significant lesions are pure squamous dysplasias without glandular neoplasia.Problems with the AGUS diagnosis are that its defining criteria are somewhat nebulous, and it is used as a wastebasket for lesions that cannot be definitely classified as reactive or neoplastic.9-11 The Bethesda system recommends subclassification of AGUS into the categories of "favor endocervical origin", "favor endometrial origin", and "not otherwise specified" (NOS); "favor endocervical origin" lesions are further classified into the categories of "favor neoplastic" and "favor reactive". [9][10][11] This classification scheme is supposed to aid in patient triage, which, given the AGUS incidence and the high rate of clinically significant lesions, is exceedingly important. However, subclassification of AGUS has yet to be proved clinically effective, and although the Bethesda Committee and many others have studied cytologic criteria important in subclassification, these criteria have not been tested rigorously. This has resulted in the AGUS category being confusing for the pathologist and the clinician.To more thoroughly investigate the utility of AGUS subclassification, our laboratory used a rigorous statistical technique known as logistic regression analysis to identify cytologic criteria important in differentiating a clinically significant AGUS lesion from a benign AGUS lesion.3 The presence of the crit...

Interobsenrer Variability of a Papanicolaou Smear Diagnosis of Atypical Glandular Cells of Undetermined Significance

“…The fourth cytopathologist (P.A.T) was a participant in several AGUS projects defining cytologic criteria for subclassification. 33 ' 34 This cytopathologist had made the original diagnosis in 7 cases but had not viewed these cases within 4 years. The cytopathologists were told that the study was designed to measure the interobserver variability of a diagnosis of AGUS, and they were not given any clinical history or follow-up information.…”

“…The diagnosis of atypical glandular cells of undetermined significance (AGUS) has been poorly characterized and has not been evaluated in terms of interobserver diagnostic variability. 33 " 36 An estimated 230,000 to 1.25 million…”

“…34 Follow-up studies show that a clinically significant lesion, ranging from squamous dysplasia to adenocarcinoma, is present in up to 83% of women with a diagnosis of AGUS. 33 " 41 Only 12% to 46% of these lesions are true glandular neoplasias, and up to 68% are high-grade squamous intraepithelial lesions (HSILs). 33^11 Thus, AGUS is a high-risk diagnosis.…”

“…This confirms the supposition that the AGUS category is a misnomer inasmuch as many AGUS lesions are not glandular at all, and a qualification should be added to cytology reports that reflects this finding. 33 Correctly diagnosing those squamous lesions that currently fall under the AGUS umbrella may be exceedingly difficult, and the effort to do so may be misguided. 33 Wilbur has hypothesized that the difficulty in recognizing truly glandular lesions may have arisen secondarily to newer sampling methods, such as the use of the endocervical brush.…”

Clinical evaluation of atypical glandular cells of undetermined significance upon cervical cytologic examination in Israeli Jewish women

“…Some physicians choose a conservative approach to AGUS and treat it much like ASCUS, with repeated cytologic examination alone, whereas others take a more comprehensive approach, performing colposcopy and biopsy of the endocervix. The diagnosis of AGUS is made in routine clinical practice in 0.13 -2.5% of the smears; most rates quoted in the literature are less than 1% (Goff et al, 1992;Nasu et al, 1993;Lee et al, 1995;Ferris et al, 1996;Kennedy et al, 1996;Raab et al, 1997;Zweizig et al, 1997). Thus, our diagnostic rate of 0.38% in this study is in agreement with the current literature.…”