skip to main content
Language:
Search Limited to: Search Limited to: Resource type Show Results with: Show Results with: Search type Index

Breast cancer incidence and survival trends by molecular subtypes in Scotland

DOI: 10.7488/era/1458

Digital Resources/Online E-Resources

Citations Cited by
  • Title:
    Breast cancer incidence and survival trends by molecular subtypes in Scotland
  • Author: Mesa-Eguiagaray, Ines
  • Subjects: breast cancer ; epidemiology ; incidence ; oestrogen ; subtypes ; survival
  • Description: Background: Breast cancer (BC) is the most common cancer among women and leading contributor to cancer mortality, hence constitutes a major public health issue worldwide. In Scotland, over 4,000 women are diagnosed with BC every year and around a 1,000 die from this disease. Monitoring incidence, mortality and survival trends is key for surveillance of disease progression. BC is heterogeneous, with multiple subtypes defined by molecular markers, such as the oestrogen receptor (ER), that have different aetiology, targeted treatments and prognosis, yet standard reporting of incidence and mortality rates is not usually done using tumour marker data. The Scottish Cancer Registry was the first registry in the UK to collect molecular marker data and therefore, constitutes an excellent opportunity to explore incidence and survival trends over time by molecular subtypes. This PhD aims to describe temporal trends in BC incidence and survival by molecular subtypes in Scotland to inform public health prevention programmes, diagnostic and therapeutic services. Methods: A systematic review was conducted to determine the extent of available data on BC incidence trends by ER in population-based studies of women of European ancestry. In addition, the Scottish Cancer registry data on over 72,000 women diagnosed with incident primary BC from 1997 to 2016 (the focus of most analyses for this dissertation) was used to describe trends in incidence and survival in Scotland. Age-standardised incidence rates (ASiR) and age-specific incidence were estimated by BC subtype after imputation of molecular marker data. Joinpoint regression and age-period-cohort (APC) models were used to assess whether significant differences were observed in incidence trends by ER, the human epidermal growth factor receptor 2 (HER2) and the immunohistochemistry (IHC) defined molecular subtypes. Kaplan-Meier (KM) estimates and traditional and extended Cox proportional hazards models were computed to assess breast cancer specific survival (BCSS) by BC subtypes. Sensitivity analysis was carried out to compare results for the Cox models from complete case analysis (CCA) and multiple imputation analysis (MIA). The effect of individual, tumour characteristics and treatments on BCSS for each subtype was also investigated. Trends in 5-year survival by age, grade and stage characteristics for the different subtypes (ER+ and ER-) were investigated to identify the characteristics of women showing greatest and lowest improvements over time. Other causes of death were also explored and cumulative incidence functions (CIF) were investigated. Results: The systematic review showed that ER+ BC incidence increased and ER- BC incidence decreased in the last four decades (EAPCs ranging from 0.8% to 3% for ER+ tumours and -2.1% to -3.4% for ER- tumours) and that the rise in overall incidence trends is mainly driven by increases of ER+ tumours in women of screening age. In Scotland, BC incidence rates showed the same divergent pattern between ER+ and ER- tumours observed in other countries. ER+ tumour incidence increased by 0.4% per year from 1997 to 2011 and increases were mainly among routinely screened women aged 50 to 69 years. In contrast, ER- tumour incidence decreased among all ages by -2.5% per year over the study period. Apart from the period effects observed, APC models showed that older cohorts of women born in 1912-1940 had lower incidence rate ratios (IRR) for ER+ tumours, and younger cohort of women born in 1960-1986 had lower IRR for ER- tumours, compared to women from the 1941-1959 birth cohorts. Results for the IHC defined subtypes showed that luminal A tumours, that account for more than half of all tumours, had similar patterns to those observed for ER+ tumours, with increases until 2011. In contrast, luminal B tumours declined over time, particularly in women over 50 years of age. There was no clear trend for HER2-enriched or triple negative breast cancers (TNBC) overall but TNBC tumours seemed to increase in younger women aged 20 to 49 years. BCSS also differed between subtypes with ER+ tumours having better survival than ER- tumours, luminal A tumours having the best survival of all IHC defined subtypes and TNBC having the worst survival. Age, grade, stage, screening and surgery were the most important prognostic factors irrespective of tumour subtype, with women who had older age, higher grade, stages III-IV, tumours not screen detected and who did not have surgery having worse survival. Deprivation was also associated with lower BCSS, with women living in the most deprived areas of Scotland having increased BC-specific mortality when compared to women in the least deprived areas and this relationship was observed for all subtypes with slightly higher HR for HER2-enriched subtypes (but wider CI). Five-year BCSS trends showed improvements in the last two decades, especially for women aged 50 to 69 years. The greatest gains in survival were seeing in women with advanced tumours (high grade or stage III-IV tumours) and ER-tumours seemed to have greatest improvements than ER+ tumours, although their survival remained lower than for ER+ tumours. The improvements observed for women with high grade and stage III-IV tumours were observed in both screen and not screen detected tumours but the rise was sharper amongst women with screen detected tumours. Women younger than 50 years showed similar improvements than those observed in women aged 50 to 69 years. Older women aged 70 years or more showed no consistent survival improvements over time and over 50% of women in that age had a primary cause of death other than BC with cardiovascular diseases (CVD) being a major contributor (22% of all deaths). Conclusions: This project is the first in the UK to describe incidence and survival trends by molecular subtypes of BC using population-based data. Divergent incidence trends found in Scotland are similar to those observed in other countries and confirm different aetiology of BC molecular subtypes. Increases in the incidence of hormone sensitive tumours are likely to be driven by the implementation of mammographic screening programmes, population aging and changes in risk factors (RFs) that have differential effects on the subtypes, such as, reproductive factors and obesity. Survival improvements in Scotland are likely due to multiple contributors with two major factors such as screening and the improvement and development of new treatments likely playing a role. This PhD has allowed us to further understand disease progression of the different subtypes in Scotland and has identified groups of women (those with advanced tumour characteristics, living in the most deprived regions of Scotland or women aged 70 years or older) with lower survival and/or lower improvements in survival trends that could benefit from further prevention and treatment programmes. This PhD also highlights the importance of monitoring future incidence and survival by molecular subtypes to inform clinical planning and cancer control programmes.
  • Publisher: The University of Edinburgh
  • Creation Date: 2021
  • Language: English
  • Identifier: DOI: 10.7488/era/1458
  • Source: University of Edinburgh dspace

Searching Remote Databases, Please Wait