Anal cytology in an HIV-positive population: a retrospective analysis

OBJECTIVE: To retrospectively evaluate cellularity and correlate the presence of columnar cells with specimen interpretation in conventionally prepared anal cytologic smears from an HIV-positive population. STUDY DESIGN: Two cytopathologists and 1 senior cytotechnologist, blinded to the original diagnosis, screened 114 samples from 110 patients collected between 1997 and 2002. One hundred nine males and 1 female were included, age ranging from 23 to 52 years. Discrepancies were reviewed for consensus. The interpretations, cellularity, and presence or absence of columnar cells were noted. The relationships between diagnosis and presence of columnar cells, visible anal lesions, concurrent HIV viral load and CD4+ T-cell counts were assessed. RESULTS: The cytologic findings were as follows: 7, unsatisfactory (6%); 29, negative (25%); 25, atypical squamous cells of undetermined significance (22%); and 53, dysplasia (47%) (42 anal intraepithelial lesion 1 [37%] and 11 anal intraepithelial lesion 2/3 [10%]). Nearly 50% of the smears, 51, showed the presence of columnar cells (45%); 37 of those specimens had some degree of dysplasia (74%). Of the 63 specimens with no columnar cells, 16 (25%) showed dysplasia. Columnar cells were absent from all unsatisfactory specimens. CONCLUSION: A highly significant association between the presence of columnar cells and anal intraepithelial lesion (p<0.001) and a significant association between the presence of columnar cells and atypical cytology when a visible lesion was absent (p=0.0019) were found. No significant relationship was found between the presence/degree of dysplasia and CD4+ T-cell counts or HIV viral load. Lack of clinical follow-up precluded evaluation of the false negative rates in this data set.