[Human papillomavirus typing in routine cervical smears. Results from a series of 3778 patients]

As human papillomavirus (HPV) infection is a significant risk factor for cervical cancer, the aim of this study was to assess the efficacy of HPV detection techniques in predicting the presence or the development of a high-grade squamous intra epithelial lesion. Additional reasons for carrying out this investigation were as follows. It was considered that the clinical value of HPV typing may have been underestimated in certain previous reports, which based their findings on first-generation assays. Moreover, until the present investigation most studies only investigated the cases with abnormal smear findings and did not include long-term follow-up, nor detailed follow-up of women with a normal smear but positive HPV typing, although this patient subgroup is of particular interest. The patient population included 3778 women who underwent routine cytological cervical screening and who gave their informed consent to participate in this study, consisting of a cervical smear, HPV testing via a second-generation Hybrid Capture II assay (which is more sensitive than first-generation tests, and can detect 18 different HPV types, i.e., 13 oncogenic and five non-oncogenic types); and in the case of abnormal smear findings, colposcopic examination and directed biopsies of the sites of suspected lesions. The women with cytomorphologically normal cervical smears but which who HPV-positive were reexamined six months later.

HPV findings were positive in 66 cases where the initial smear indicated the presence of a high-grade lesion with the diagnosis confirmed by biopsy in 65 cases; in 155 cases (76.1%) where the initial smear indicated the presence of a low-grade lesion; in 44 (57.1%) out of 77 smears indicating the presence of ASCUS-type (atypical squamous cells of undetermined significance) lesions; and in 366 cases (10.5%) of apparently normal initial smears. After colposcopy and follow-up, a final diagnosis of high-grade lesion was made for 85 patients. All these patients were found to have high-risk (oncogene-positive) HPV at the first examination, but in only 65 cases (76.5%) was the initial smear indicative of a high-grade lesion. Finally, in the patient groups with low-grade or ASCUS-type smear findings or with an apparently normal smear, a high-grade lesion was only found in those patients with persistent high-risk HPV infection. In all, the sensitivity of the Hybrid Capture II assay in detecting high-grade lesions was 100% versus 85.9% for standard cytology. However, its specificity (86.3%) and positive predictive value were not as high as the latter.

The results of this study indicate that HPV typing via this second-generation assay displays good sensitivity for the detection of at-risk cases, i.e., those involving the presence or development of high-grade cervical lesions. However, a number of issues have to be addressed before HPV typing is adopted in clinical practice, such as the fairly high HPV prevalence in women aged over 60 years, and whether the implications of this are the same as for younger age groups. Also, the present cost of complete typing remains too high for it to be used on a widespread scale. Nevertheless, HPV typing may have a useful role to play in the detection of cervical cancer.