Multicenter Masked Evaluation of AutoCyte PREP Thin Layers with Matched Conventional Smears

Abstract
To compare the performance of liquid-based cell preparations from the AutoCyte PREP system (AutoCyte, Inc., Elon College, North Carolina) with the conventional cervical smear in masked, split-sample, multisite trials. The AutoCyte PREP system utilizes the CytoRich method, which combines liquid preservation, selective reduction of blood and inflammation, thin-layer cell dispersion and discrete staining. In an eight-site multicenter trial, 8,983 cases were evaluated. Parallel AutoCyte PREP slides and matching conventional cervical cytologic smears were screened in a masked fashion, with all abnormals reviewed in a masked, on-site pathologist review. The conventional smear was always prepared first, with the AutoCyte PREP using the residual cells on the collection device. The Bethesda System was used for reporting diagnosis and specimen adequacy. Of the cases, 7,805 (86.9%) had the same interpretation. In 8,750 cases (97.4%), there was agreement within one diagnostic category. The AutoCyte PREP demonstrated a statistically significant, 31% overall improvement in the detection of squamous intraepithelial lesion and invasive cancer when evaluating cases with more than one diagnostic class difference. Biopsy correlation of available data supported the improvement observed in the cytology-based comparison. Recovery of infectious organisms and endocervical component was comparable between preparations, especially in light of the split-sample study design. There was a 39% reduction in unsatisfactory slides and 44% fewer satisfactory but limited by reports. The AutoCyte PREP and CytoRich method produce excellent cellular presentations with standardized quality, superior sensitivity and improved adequacy as compared to the conventional method.