Since Cooper et al. described three types of nonisotopic in situ hybridization (NISH) labelling pattern in uterine cervical lesions harbouring human papillomavirus (HPV), 1 several studies have validated the use of this technique to detect episomal and integrated HPV in cervical carcinomas. 2,3 As Dr Cooper points out in his letter, it was proposed that in low-grade squamous intraepithelial lesions (LSIL/CIN I) HPV predominates in an episomal physical state (diffuse signal), while in high-grade SIL (HSIL/CIN II-III) and squamous cell carcinoma, integrated HPV predominates (punctate signal). 1,4 Our work 5 confirms this hypothesis, since the number of biopsy specimens studied allowed statistical analysis of the results and demonstrated correlations between NISH labelling pattern and both lesion and HPV types.
In our opinion, the histological diagnoses in cervical/ vulvar squamous intraepithelial lesions should be complemented with additional information, in order to choose the more effective treatment in each case. This information should include both cellular (degree of ploidy, 6 analysis of clonality, 7 proliferation index 8 ) and viral parameters (type of infection, productive or nonproductive; 9 physical state of the genome, episomal or integrated 5 ). These data should be obtained by way of techniques easily adaptable to routine use, such as flow cytometry, image analysis, immunohistochemistry, and NISH.