Abstract

AIMS: To determine the relative diagnostic performance of non-isotopic in situ hybridisation (NISH) and a dot-blot assay for detecting human papillomavirus (HPV) on exfoliated cervical cells; and to correlate the results with cytopathological assessment. METHODS: Cervical smears and cytological samples were obtained from 122 patients during the same clinical examination and the presence of HPV sequences determined by NISH and dot-blot analysis, respectively. RESULTS: Dot-blot analysis gave an autoradiographic signal in 15 of 121 (12.4%) cases, while NISH detected viral genomes in 38 of 114 (33.3%) cases. Even in the presence of koilocytosis, where vegetative replication of the virus occurs, NISH was positive in over twice as many cases as dot-blot analysis (NISH 90%, dot-blot 40%), while in smears within normal cytological limits, where the viral copy number is likely to be considerably lower, the differences were more striking (NISH 31%, dot-blot 5%). CONCLUSIONS: These data show that NISH on cytological smears is more sensitive than a standardised dot-blot hybridisation assay for detecting HPV infection in cytological material and is therefore a more appropriate screening tool.