1. ¹Department of Pathology, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma
2. ²The Native American Prevention Research Center, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma

Correspondence: Rosemary E. Zuna, M.D., Department of Pathology, BMSB 451, University of Oklahoma Health Sciences Center, Oklahoma City, OK 73190; e-mail: rosemary_zuna@ouhsc.edu; fax: 405-271-6573.

Accepted 24 October 2000.

Abstract

HPV DNA testing of the residual sample volume of liquid-based Pap tests has been recommended as a way to determine the appropriate follow-up for women who have equivocal results in routine clinical screening. A major aspect of quality assurance in the cytopathology laboratory consists of correlation of smear interpretation with biopsy or conization results as mandated by CLIA '88. However, the use of histology as the gold standard suffers from similar problems of subjectivity and sampling as the Pap smear. In this study we explore the potential use of HPV DNA testing of the residual volume from the ThinPrep® Pap Test^TM (Cytyc Corporation, Boxborough, Massachusetts) as a substitute gold standard in quality assurance monitoring of a cervical cytology screening program. The residual samples from 397 ThinPrep® Pap cases were retrospectively analyzed for high-risk HPV DNA using the Hybrid Capture II^TM technique. Sensitivity (71.8%), specificity (86.5%), predictive value of positive (77.1%) and negative (82.9%) ThinPrep® Pap interpretations were calculated on the basis of HPV DNA results for 266 cases classed as either squamous intraepithelial lesion (SIL) or negative. Overall, there was agreement between the two tests in 80.8% of cases (Cohen's kappa =.59). The percentage of HPV DNA-positive cases interpreted as atypical squamous cells of uncertain significance (ASCUS) was 43.7%, and the percentage of negative cases was 17.1%. We believe that this approach is an objective adjunct to the traditional quality assurance protocol, with the added benefit that it includes cases interpreted as negative, as well as abnormal cases that do not come to biopsy.