Epidemiology of ductal carcinoma in situ

Introduction: Almost 7,000 people are diagnosed with ductal carcinoma in situ (DCIS) in the United Kingdom each year, but there remains uncertainty regarding its natural history and optimal management. The aim of this thesis was to evaluate factors contributing to the epidemiology of DCIS and its outcomes. Methods: 1) A cohort study comparing risk factors for DCIS and invasive breast cancer (IBC) using UK Biobank; 2) A cohort study examining the accuracy of preoperative biopsy in DCIS using clinical records from the Netherlands Cancer Institute; 3) A cohort study examining the rate of invasive breast cancer following treatment for screen-detected DCIS in England using the National Health Service Breast Screening Programme (NHSBSP) audit; 4) A methodological study to develop an algorithm to identify invasive breast cancer recurrences, which in the future may used to identify DCIS recurrences, using all relevant routinely collected data stored within Public Health England (PHE). Results: (1) For both DCIS and IBC, postmenopausal BMI was associated with an increased risk of developing disease, and the number of live births was associated with a decreased risk of developing disease. However, the magnitude of the effect differed between DCIS and IBC. The increased risk from postmenopausal BMI ≥35 kg/m2 was larger for DCIS than for IBC (RR 2.35, 95% CI 1.14 – 4.82), and the trend of reduction in risk with each additional live birth was greater for DCIS than for IBC (p for trend = 0.03). (2) Consideration of mammographic lesion size and the absence of necrosis on biopsy may be helpful in selecting low-risk women for non-operative management of DCIS in the future, as may use of the 9G vacuum-assisted method of biopsy. (3) The cumulative risks of IBC at 5, 10 and 15 years after screen-detected DCIS in England were 3.5%, 7.1%, and 9.4% respectively. Women with clear surgical margins of 1 – 2 mm had a higher IBC rate than women with clear margins of 5+ mm (RR 1.85, 95% CI 1.20 – 2.84). Women given breast-conserving surgery (BCS) without radiotherapy had a higher ipsilateral IBC rate than women given BCS with radiotherapy (RR 1.63, 95% CI 1.27 – 2.10). Women given hormone therapy had a lower rate of any IBC compared with oestrogen receptor (ER) positive women not given hormone therapy (RR 0.76, 95% CI 0.63 – 0.93). (4) There was good agreement between the number of recurrences indicated by the developed algorithm using routinely collected data sources and the number of recurrences recorded in the test dataset. This finding supports the potential value of compiling recurrence information on a nationwide basis from routinely collected data, for use in future descriptive and epidemiological studies and in follow-up for randomised trials. Conclusions: Using a variety of methods these studies have all succeeded in adding to knowledge about the epidemiology of DCIS. This knowledge can be used to help the future management of women with DCIS. In addition, each of the studies has planned extensions and will continue to contribute further knowledge periodically into the future.