Abstract
We compared the efficacy of human papillomavirus (HPV) DNA detection between a PCR-based genechip (Easychip HPV Blot [hereafter referred to as HPV Blot]; King Car, Taiwan) method and Hybrid Capture II (HCII; Digene, Gaithersburg, MD) in women with previous normal (n = 146) or abnormal (> or =atypical squamous cells of undetermined significance [ASCUS] [n = 208]) cytology. A total of 354 cervical swab samples were collected for HPV DNA assay by both HCII and SPF1/GP6+ PCR followed by HPV Blot tests. Colposcopy-directed biopsy was performed if clinically indicated. Of the 354 samples, HPV-positive rates by these two methods (HCII and HPV Blot) were 12.6% and 18.2% in 143 normal samples, 36.2% and 45.7% in 105 ASCUS samples, 57.4% and 57.4% in 94 low-grade squamous intraepithelial lesion samples, and 83.3% and 75.0% in 12 high-grade squamous intraepithelial lesion samples, respectively. The concordance of HPV Blot and HCII was 80.8% (286/354), and the agreement between the methods (kappa value, 0.68) was substantial. Discrepancies were further investigated by at least one of the following three methods: direct sequencing, type-specific PCR, and HPV Blot genotyping of cervical biopsy tissue. In the 15 HCII-positive samples, HPV Blot detected only non-HCII HPV genotypes; results of further verification methods were consistent with the latter test in the 15 samples. Of the 20 samples with HCII-negative and HPV Blot-positive results, 18 were found to contain the 13 HCII high-risk genotypes by verification methods. In only 16.7% (3/18) of the HCII-positive but HPV Blot-negative samples, further studies detected the 13 HCII genotypes. We conclude that HPV Blot seemed comparable to HCII for detection of HPV DNA in cervical swab samples.