Prevention, incidence, and survival of cervical cancer in Sweden

Sammanfattning: Cervical cancer is preventable by screening. In 1967, Sweden introduced a cervical screening program. Screening for high-risk human papillomavirus (hr-HPV), the causative factor of cervical cancer, is more sensitive than cytology and can be performed as a self-sample. Non-attendance to screening is a large risk factor for cervical cancer. Self-samples may improve screening attendance. The aims of this thesis were: To analyze if the cervical screening program in Sweden can be improved by using vaginal self-samples and an HPV mRNA assay, with the main focus on screening non-attendees. To obtain knowledge if cervical cancer incidence and survival has changed since the implementation of the screening program. Study I: Incidence and net survival according to morphology, age, and stage at diagnosis among women diagnosed with invasive cervical cancer between 1960 and 2014 were calculated. The age-standardized incidence of squamous cell carcinoma (SCC) decreased until the year 2000, while the incidence of adenocarcinoma (ADC) increased continuously. Age-standardized 5-year net survival increased. SCC and ADC did not statistically differ in net survival after 2012. Among women >75 years, long-term net survival has decreased since 1960. Study II: The sensitivity and specificity of vaginal and urine self-samples compared to cervical samples analyzed by Aptima HPV mRNA assay were evaluated in a referral population. The sensitivity for detection of high-grade squamous intraepithelial lesions /adenocarcinoma in situ/cancer was 85.5% for the vaginal self-sample, 44.8% for the urinary sample, 100.0% for the cervical sample and 81.7% for cytology. Study III-V: Screening non-attendees or women in the upper age screening limit were sent a vaginal self-sampling kit by mail. In study III, 1,000 women, aged 69-70 years, received a kit. In study IV and V, 6,023 and 19,766 women, aged 30-70 years, received a kit. Returned samples were analyzed for HPV mRNA by Aptima assay. HPV-positive women were invited to follow-up. The response rate of the self-sample was 43.3%, 13.2% and 18.5% for study III, IV and V respectively. The HPV prevalence was 6.2% in study III, and no cases of high-grade dysplasia/cancer were diagnosed. The HPV prevalence was 9.9% and 11.3% in study IV and V respectively. In study IV, the prevalence of cervical cancer was almost seven times higher compared to organized screening, but in study V the prevalence of cancer was not increased. Conclusion: This thesis demonstrated that the incidence of SCC, but not ADC, has decreased since 1960. SCC and ADC did not statistically differ in net survival after 2012. The decreased long-term net survival among women >75 years of age suggests the need for prolonged HPV screening up to 75 years of age. Self-sampling is a promising method since it was accepted among women 69-70 years old, and it increased the attendance to cervical screening by almost one fifth among non-attendees. A vaginal HPV self-sample analyzed by Aptima mRNA assay showed a similar sensitivity as routine cytology and may be used to reach screening non-attendees. Among screening non-attendees and women in the upper age screening limit, around one in 10-20 tested positive for HPV mRNA, with risk of development of cervical dysplasia, although the prevalence of cervical cancer varied between the studies.