Comparison of cervical cytology and the polymerase chain reaction for HPV 16 to identify women with cervical disease in a general practice population.

A comparison of the ability of cervical cytology and the polymerase chain reaction (PCR) for human papilloma virus type 16 (HPV 16) to identify women with cervical disease has been performed in a general practice population of 249 women, none of whom were believed to have current cervical disease prior to examination. Within this population, 29 women were found by colposcopy and subsequent histopathology to have evidence of cervical disease [5 with cervical intraepithelial neoplasia (CIN) 3; 8 with CIN 2; and 16 with CIN 1]. The prevalence of HPV 16 in this population was 18.9% (CIN 3, 80%; CIN 2, 50%, CIN 1, 12.5%, normal, 16.8%). Women with severe disease (CIN 2 and CIN 3) had a significantly higher incidence of HPV 16 DNA than those with mild cervical disease (CIN 1) or no cervical abnormality (P = 0.001). There was no significant difference in the ability of either PCR for HPV 16 or cytology to identify women with cervical disease. The combination of screening by cytology and the presence of HPV 16 DNA resulted in the identification of a higher proportion of the women with disease, but this observation did not reach statistical significance. Although the failure to detect disease by the two screening methods was similar, HPV 16 DNA positivity was associated with a higher false-positive rate for disease detection than cytology (P less than 0.03). The PCR assay for detecting HPV 16 in this investigation was shown to have a false-positive rate of 2.4% and a false-negative rate of 10.4%. The prospect of screening women for cervical disease using PCR for HPV 16 is discussed.