Conventional and liquid-based cervicovaginal cytology: a comparison study with clinical and histologic follow-up

Increased rates of squamous intraepithelial lesion (SIL) diagnosis with liquid-based cervicovaginal cytology (CVC) methods are well documented. This retrospective study compares the ability of the ThinPrep Pap Test (TP) and the conventional Pap smear (CP) to detect biopsy-proven SIL and to exclude nonneoplastic disease. All CVC reports from January 1999 through December 2000 from seven community Family Medicine clinics affiliated with the University of Nebraska were reviewed. For women with at least one CVC diagnosis of epithelial cell abnormality (ECA), follow-up histology, cytology, and clinical data were obtained. Statistical analysis was performed using the chi-square method. SIL was diagnosed in 166 of 3,286 patients by TP (5.1%) and in 169 of 4,872 patients by CP (3.5%) (P < 0.001); 32 of the TP diagnoses (1.0%) and 34 of the CP diagnoses (0.7%) were high-grade SIL (HSIL). Atypical squamous or glandular cells of undetermined significance (ASCUS/AGUS) was the most severe abnormality diagnosed by TP in 218 patients (6.6%) and by CP in 279 patients (5.7%). Follow-up histology data on CVC SIL diagnoses showed evidence of cervical intraepithelial neoplasia in 94 patients screened by TP (2.9%) and in 79 patients screened by CP (1.6%) (P < 0.001); the biopsy diagnoses were CIN 2 or CIN 3 in 34 patients in the TP group (1.0%) and in 28 patients in the CP group (0.6%) (P < 0.025). Follow-up of patients in whom the first ECA CVC diagnosis was ASCUS or AGUS disclosed a positive predictive value for CIN of 22.8% for TP ASCUS/AGUS diagnoses and 11.9% for CP ASCUS/AGUS diagnoses (P < 0.005). In this population, TP was significantly better than CP in detecting biopsy- proven disease and in screening out benign abnormalities.