The predictability of invasive cervical neoplasia by the presence of specific human papillomavirus sequences in preinvasive cervical neoplasia

Abstract

<p>Cancer of the cervix is the second most common form of cancer in women worldwide. The natural history of cervical cancer is thought to involve sequential changes from varying grades of precursor intraepithelial lesions called CIN. However, little is known of the risk factors that can predict the oncogenic potential of a particular CIN lesion. Based on their preferential occurrence in cervical cancers and their potential oncogenic properties, the human papillomaviruses (HPV) particularly the genotypes 16, 18, 31, 33, 35 and 42 have been implicated in the etiology of invasive cervical cancer. However, these viruses could occur as either secondary pathogen of cancer or as predictor of those CIN lesions that are likely to progress to invasive disease. The hypothesis that was verified in this study was that HPV 16,33 and 18 are likely to be predictive of CIN lesions that progress to invasive cancer.</p> <p>First of all, in order to characterize the prevalent type of HPV in the target study population of B.C., a cross-sectional study was conducted and the presence of specific HPV types ascertained by the tissue in situ hybridization. The frequency of HPV types 16, or 33, was found to vary with the severity of the CIN grades, in contrast to the frequency of HPV types 6/11 and 18 that segregated independently of the CIN grades.</p> <p>Next, a case-control study was undertaken to verify the main hypothesis of HPV being predictive of CIN lesion progression to invasion. It was reasoned that the particular HPV would occur at higher frequency in CIN biopsies of cervical cancer cases than in CIN biopsies of noncases (controls). A total of 47 cases and 94 controls were enrolled from patients registered by the Cancer Control Agency of the Province of British Columbia. A case was defined as a post-pubertal woman with invasive disease and who had a CIN diagnosis at least two years prior to the invasive disease. For each case an attempt was made to enrol two control matched on grade of CIN and year of diagnosis. On each subject attempt was made to gather demographic informations that are known to be associated with cervical cancer. The HPV probes that were used included HPV 16/33 and HPV 6/11. The relative frequency of occurrence of specific HPV in the preinvasive biopsies of cases and controls were as folow: HPV 16/33 occurred in 10.6% of controls and 12.8% of cases; HPV 18 was found in 3.2% of controls and 8.7% of cases and HPV 6/11 in 2.2% of controls and 8.7% of cases. Conditional Chi-square analysis showed that the difference in the proportions of HPV positivity between cases and controls was compatible with sampling variation. Hence, with a statistical power approximately 60%, it was concluded that particular HPV could not be predictive of CIN lesions progression in the sample of population that was studied.</p> <p>However an excess risk for incurring cervical cancer, by being exposed to particular HPV at the CIN stage, was noted. The relative risk for HPV 16/33 was 2.34, [95% CI 0.70 to 7.66]; for HPV 18 was 2.45, [95% CI 0.22 to 27.80]; for HPV 6/11 2.19, [95% CI 0.39 to 12.42] or for all HPV combined was 1.87, [95% CI 0.55 to 6.28].</p> <p>Interestingly, a comparison of the frequency of HPV occurrences in the case-control study with that in the cross-sectional study revealed a lower rate of HPV positivity in the case-control component. This could possibly be due to a cohort effect.</p>