Risk factors for cervical cancer development

Abstract

Cervical cancer remains one of the leading causes of cancer mortality globally, predominantly in less developed countries. It is widely accepted that certain oncogenic types of HPV (Human papillomavirus) are necessary causes of cervical cancer development and a number of co-factors (including smoking) have also been implicated. The introduction of Papanicolaou testing has achieved a large reduction in cervical cancer incidence over the last 4 decades however its use is not without problems, particularly with regards to its relatively low specificity. Recently HPV vaccines targeting HPV-16 and HPV-18 have been developed, which hold the promise of reducing cervical cancer incidence further. In the four studies included in this thesis, we investigate several risk factors related to cervical cancer formation, test the efficiency of Pap smear screening in reducing the incidence of the two main cervical cancer types (cervical squamous cell carcinoma and adenocarcinoma), investigate possible maternal transmission of cervical cancer in situ, and estimate the effects that vaccines against HPV-16/18 might have had on the Swedish population over the last 30 years had they been implemented.

Using a population-based case-control study, we examined the individual risks for cervical cancer in situ (CIS) associated with HPV-16 presence, HPV-16 load, and smoking. After analysis of first cervical smears for 375 case and 363 control women, we found higher increased risks associated with HPV-16 presence in smokers (adjusted OR=14.4; 95% CI 5.6-36.8) than non-smokers (adjusted OR=5.6; 95% CI 2.7-11.2), compared with HPV-16 negative women. Risk for CIS was even higher in smokers with high HPV-16 load (adjusted OR=27.0; 95% CI 6.5-114.2) compared to smokers who were HPV-16 negative. In contrast, non-smokers with high HPV-16 load had a lower risk (adjusted OR=5.9; 95% CI 2.4-14.6). Neither HPV-16 presence nor load were found to significantly interact with smoking status, although significant interaction was found between current smoking status (at first smear) and duration of smoking (p=0.03).

Analysis of Swedish nation-wide data covering 1968-2002 provided evidence for the beneficial nature of Pap smear screening in reducing cervical squamous cell carcinoma (SCC), but not cervical adenocarcinoma (AC). We also found that higher reported CIS incidence rates in certain counties did not lead to future reductions in SCC incidence. This was also the case for adenocarcinoma in situ (AIS) reported, and future AC incidence. On the whole, our results suggest there is over-treatment of CIS. Also it appears likely that inefficiencies exist in the use of Pap smear screening for AC prevention.

In a cohort of mother-daughter pairs, we found a 24% excess risk for CIS in daughters whose mothers exhibited CIS within 10 years after their daughter s conception compared to daughters whose mothers had CIS more than 10 years after their conception. This may conceivably be the result of maternal transmission of HPV during pregnancy or childhood.

Using a population-based cohort of Swedish women, we estimated that a 61.2% reduction in SCC incidence could have been achieved through the removal of HPV-16/18 from the population during 1969-2002. In comparison, removal of all measured high risk HPV types may have reduced SCC incidence by 83.9%. Removal or reduction of HPV-16 and -18 from the Swedish population will give considerable benefit in reducing cervical cancer. However, the risks associated with other non- HPV-16/18 oncogenic types are worthy of consideration in future monitoring and prevention strategies.