Human papillomavirus and cervical carcinoma in situ : implications for future screening

Abstract

The principal aim of this thesis was to study the temporal relationship between human papillomavirus type 16 (HPV 16) infection and cervical carcinoma in situ, particularly focusing on the effect of viral load. Other aims were to study the role of additional factors, especially smoking and oral contraceptive use, in cervical carcinogenesis. We conducted a case-control study, nested within a population-based cohort of women participating in cytological screening in Uppsala county in Sweden during 1969 to 1995. All incident cases of cervical carcinoma in situ during this period were identified. For each case, one or two controls, matched by age and date of entry to the cohort, were selected randomly from the cohort.

A sensitive PCR assay was used to detect and quantitate HPV 16 load in repeatedly taken smears during Lip to 26 years before diagnosis in cases and their matched controls. Altogether 2553 archival smears from 495 cases and 2164 smears from 649 controls were tested. In addition, detailed information on sexual practice, smoking habits and oral contraceptive use were collected through telephone interviews with 422 cases and 422 controls. We developed a single-tube nested PCR-based system for detection of 19 different genital HPV types in archival smears. Our detection system had a high sensitivity, comparable to the most commonly used PCR-based HPV typing systems. We found an age-dependent increased risk of cervical carcinoma in situ in relation to current smoking, especially confined to women younger than 45 years. An increased risk was further demonstrated with increasing tobacco consumption (duration, intensity, pack-years). Current use of oral contraceptives was positively associated with the risk of cervical carcinoma in situ, with a monotonic increase in risk with increasing duration of use. The number of sexual partners was significantly, positively associated with risk among HPV 16/18-negative but not among HPV 16/18-positive women. We failed to demonstrate any increased risk in relation to age at first intercourse or parity.

The prevalence of HPV 16 positivity among cases was 56% at the time of diagnosis. Having an HPV 16-positive smear at entry to the cohort was associated with a moderately increased risk, whereas having persistent infection with HPV 16 in two subsequent smears was strongly associated with risk of cervical carcinoma in situ. We estimated that among HPV 16-positive women the median incubation period from first detected HPV infection to cervical carcinoma in situ was between 7 and 12 years. Among cases, we found a consistently increased load of HPV 16, detected already 13 years before diagnosis, and often when the smear was still cytologically normal. Women with a high load of HPV 16, were at an over thirty-fold relative risk compared to HPV 16-negative women more than a decade prior to diagnosis. Women with the 20% highest viral load in their first smear, taken on average 7.8 years before diagnosis, had a sixty times higher risk compared to women negative for HPV in their first smear. We further estimated that approximately 114 of women infected with a high viral load before age 25 years, developed cervical carcinoma in situ within 15 years.