Effectiveness and equity of cervical cancer prevention : real-life evidence from organised programmes in Sweden

Abstract: Cervical cancer incidence has substantially declined since cervical screening was implemented five decades ago. The long-term hope of eliminating cervical cancer is promising with the development of effective Human papillomavirus (HPV) vaccines. Optimising effectiveness and promoting high and equal uptake of preventive approaches are essential for achieving this goal. This thesis aimed to use register data in Sweden to perform in-depth evaluation of the effectiveness of cervical screening and investigate the factors associated with HPV vaccine uptake. The work focused in detail upon the uncertain preventive effect against cervical adenocarcinoma in relation to the finding of glandular abnormalities in screening, the effectiveness of screening for the conspicuous incidence of cervical cancer at older ages, as well as social disparity of HPV vaccine uptake in different modes of vaccination delivery. Effectiveness of cervical screening in preventing invasive cervical cancer has been reported under a case-control audit framework in Sweden, based on cervical cancer cases diagnosed in 1999-2001. Yet, assessing long-term screening history and controlling for confounding factors were hampered by unavailable data, and statistical power was limited to stratify the evaluation by histopathological type. We therefore performed an updated case-control audit based on cervical cancer cases diagnosed in 2002-2011 (Study III), to examine the risk of invasive cervical cancer in relation to screening history in the past two screening rounds, adjusted for education and stratified by the two main histopathological types. We found that non-routine participation to cervical screening was associated with increased risk of invasive cervical cancer. Having an abnormality in previous two screening rounds was associated with elevated risk, particularly if not being screened in the subsequent screening round after an abnormality. The lower effectiveness of cervical screening in preventing adenocarcinoma compared to squamous cell carcinoma resulted from both lower assurance from normal screening results, and higher risk following abnormalities. These findings reinforce the evidence of the cancer-preventive effect of cervical screening and emphasise the importance of routine participation in screening. We also identify relative weaknesses of the screening programme that would guide future research to address improvement. Atypical glandular cells (AGC) found in cervical screening is a cytological abnormality of the same type of cells giving rise to cervical adenocarcinoma. However, the long-term risk of cervical cancer following AGC has not been comprehensively investigated due to its rarity. We used the Swedish National Cervical Screening Registry (NKCx) to identify all AGC diagnosed in cervical screening from 1980-2011, examined the risk of cervical cancer by histopathological type for up to 15 years, and compared the subsequent histological assessment and risk of cancer to that after high-grade squamous intraepithelial neoplasia (HSIL) (Study I). We found that AGC was associated with a moderate-high proportion of prevalent cancer, and a high long-term risk of incident cervical cancer, especially adenocarcinoma. Only 54% of AGC were followed by histology within six months, and among those being followed with histology, the cancer incidence after AGC was still statistically significantly higher than that after HSIL. Our findings confirm the considerable risk associated with AGC and revealed suboptimal management following this specific abnormality. We highlight one of the deeper causes of unsatisfactory preventive effects of screening against cervical adenocarcinoma, and the necessity of improving management practice for AGC. Cervical cancer is generally associated with middle-aged women, but the first Swedish case-control audit revealed that cervical cancer in women over 60 years of age accounted for more than one-third of annual cervical cancer cases. Data from other Nordic countries have also exhibited similar to higher incidence of cervical cancer among older-aged women compared to middle-aged women, leading to uncertainty on the underlying reason of biological or screening effect and the effectiveness of cervical screening after age 60. Therefore, we used NKCx and the Swedish Cancer Registry to investigate the risk of cervical cancer from age 61-80 years by screening history at ages 51-60 years, and evaluated the effectiveness of cervical screening at ages 61-65 stratified by screening history (Study II). We found that screening at ages 61-65 was associated with substantial risk reduction up to age 80 in women unscreened or having abnormalities in their 50s. Yet in women screened with normal results in their 50s, the subsequent risk of cervical cancer was remarkably lower than that in women unscreened or having abnormalities in the past, and in these women screening after age 60 was not associated with any statistically significant risk reduction. Our results should inform the current debate regarding when and how to discontinue cervical screening in older-aged women. HPV vaccines have been available worldwide since 2006-2007. Their efficacy and effectiveness in preventing cervical precursor lesions have been shown repeatedly. Pursuing high and equal coverage of HPV vaccination is the common goal to reduce inequality of cervical cancer development in the future. Various modes of delivery of HPV vaccination were implemented worldwide and in Sweden. We used Swedish vaccination registers and social-demographic registers to examine girls’ HPV vaccine uptake in relation to parental country of birth, education and family income, by three delivery modes of HPV vaccination (Study IV). We found that free-of-charge school-based delivery achieved the highest uptake of HPV vaccination with the lowest social disparity, suggesting the importance of reducing individual payment and providing easy access to promote equality of cervical cancer prevention. In conclusion, this thesis confirms the overall effectiveness of cervical screening in preventing invasive cervical cancer, addresses specific questions regarding unsatisfactory prevention for cervical adenocarcinoma and cervical cancer in older women, as well as identifies suboptimal aspects in screening for further investigation. It optimises the audit framework as an evaluation tool for screening programme quality assurance, and provides a benchmark for future comparisons with new screening practices. The thesis also verifies the role of delivery mode of HPV vaccination on reaching high and equal uptake of the vaccine, bolstering the hope of ultimate elimination of cervical cancer.