Reducing or eliminating use of the category of atypical squamous cells of undetermined significance decreases the diagnostic accuracy of the Papanicolaou smear

Abstract

BACKGROUND. The diagnosis of "atypical squamous cells of undetermined significance" (ASCUS) is controversial, not only for the clinical utility of its subcategories (favor reactive vs. favor dysplasia), but for its very existence as an expression of uncertainty. In the current study, the authors investigated the impact of reducing and eliminating this category on the sensitivity and predictive values for detecting squamous intraepithelial lesions (SILs). METHODS. One hundred cervical Papanicolaou (Pap) smears originally diagnosed as ASCUS, all of which had histologic follow-up, were reviewed blindly and reclassified as either negative (within normal limits/benign cellular changes), low-grade SIL (LSIL), or high-grade SIL (HSIL) by 1 reviewer who eliminated ASCUS as a diagnostic possibility entirely. A second reviewer reinterpreted the cases but attempted to use the ASCUS diagnosis (favor reactive or favor dysplasia) sparingly. All histologic diagnoses were reviewed, and an adjudicated final diagnosis was established. Reviewed smear interpretations were correlated with the histologic diagnosis (original, reviewed, and adjudicated). Statistical analysis was performed using the Fisher exact test. RESULTS. Thirty-eight women had histologically confirmed SIL (21 LSIL cases and 17 HSIL cases [including 1 case of endocervical adenocarcinoma]); 31 of these 38 cases originally were classified as ASCUS, not otherwise specified, 1 case was classified as ASCUS favor reactive, and 6 cases were classified as ASCUS favor dysplasia. The reviewer who used the ASCUS diagnosis sparingly reclassified the smears as negative (62 cases); ASCUS, favor reactive (3 cases); ASCUS, favor dysplasia (13 cases); LSIL (19 cases); and HSIL (3 cases). The reviewer who eliminated the ASCUS category reclassified the smears as negative (59 cases), LSIL (29 cases), and HSIL (12 cases). The rate of SIL/HSIL in those cases interpreted as abnormal was 38%/17% originally, 42%/24% with a reduced ASCUS interpretation, and 37%/17% when the ASCUS category was eliminated. In those ASCUS smears that were reclassified as negative, the SIL/HSIL rate was 35%/13% with the reduced ASCUS interpretation and 39%/17% when the ASCUS category was eliminated. The sensitivity for detecting a SIL/HSIL was reduced from 100%/100% for the original ASCUS interpretation to 42%/53% for the reduced ASCUS interpretation to 39%/ 41% with the elimination of the ASCUS interpretation. CONCLUSIONS. Although in the current study utilization of the ASCUS diagnosis was found to result in a 62% negative or reactive outcome on biopsy, a significant number of patients with SIL were detected (38% in the current series, 17% with HSIL). Despite the improved correlation with negative biopsies, reducing or eliminating the ASCUS diagnosis appears to decrease the sensitivity of the Pap smear significantly and appears to be no better than chance at predicting a diagnosis of SIL on biopsy, including HSIL. © 2002 American Cancer Society.