DCIS of the breast : aspects on treatment and prognosis

Breast cancer is the most common cancer form and a leading cause of death in women worldwide. Ductal breast carcinoma in situ (DCIS) is characterized by a proliferation of malignant cells confined within the mammary ducts and is a potential precursor of invasive breast cancer. The risk estimations of a DCIS to develop into invasive cancer over a 10 year period range from 30-50%. In the past 25 years, concomitant with the implementation of screening mammography, the incidence of DCIS has increased dramatically and presently almost 1 000 women are diagnosed with DCIS each year in Sweden. The increased incidence poses concerns of overtreatment and current research aim at identifying clinical or pathological markers that can reliably distinguish hazardous from harmless DCIS. The overall aim of this thesis was to explore the prognostic significance of clinical and tumourbiological characteristics of DCIS and to assess the benefits and harms of adjuvant treatment. In a population-based cohort of 2 952 women with primary DCIS, we analysed trends in incidence, treatment and outcome over a 20-year period (paper I). Information was obtained from the regional breast cancer register in Uppsala-Örebro healthcare region between 1992 and 2012. A validation of 300 randomly selected women revealed high overall completeness and reliability of most key variables, whereas follow-up data were of moderate quality with only 65% of the recurrences reported to the register. The major finding of the study was a trend towards more intensified treatment over time. The frequency of mastectomy increased from 23.0% to 39.0% and the proportion of patients receiving adjuvant radiotherapy after breast-conserving surgery increased from 30.1% to 67.6%. This did not, however, translate into any noteable improvements in outcome. Relative survival was >97% after 10 years with no significant variation over time. In conclusion, these results may reflect adequate treatment selection, but may also indicate a significant overtreatment. In paper II and III, a nested case-control study was conducted from a cohort of 6 964 women with primary DCIS to identify clinical characteristics in DCIS associated with subsequent breast cancer death. Ninety-six women who later died from breast cancer were compared to 318 controls selected by incidence density sampling. Information was obtained from medial records and histopathology reports. Tumour size over 25 mm or multifocal DCIS (OR 2⋅55; 95%CI 1⋅53 to 4⋅25), a positive or uncertain margin status (OR 3⋅91; 95%CI 1⋅59 to 9⋅61) and detection outside the screening programme (OR 2⋅12; 95%CI 1⋅16 to 3⋅86) increased the risk of death from breast cancer. In the multivariable analysis, tumour size (OR 1⋅95; 95%CI 1⋅06 to 3⋅67) and margin status (OR 2⋅69; 95%CI 1⋅15 to 7⋅11) remained significant. More extensive treatment was not associated with lower risk, which may be due to confounding by indication, or indicate that some DCIS have an inherent potential for metastatic spread. In paper III, to further explore the association of tumour biology and risk of breast cancer death, archival tumour blocks were collected. Freshly cut hematoxylin and eosin (H&E) stained sections of the primary DCIS were histopathologically evaluated for nuclear grade, presence of comedonecrosis and lymphocytic infiltration (LI). Tissue microarrays were constructed for immunohistochemical analysis (IHC) of oestrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2) and Ki67. Using the results of the IHC analyses, tumours were classified into surrogate molecular subtypes. Presence of intense periductal LI was associated with an increased risk of subsequent breast cancer death (OR 2.25; 95%CI 1.02 to 4.99). None of the other biomarkers were individually related to breast cancer death, nor were there any statistically significant differences in risk between the molecular subtypes. In multivariable analysis, stepwise adjusting for age, tumour size and treatment, PR negativity in combination with LI; PR negativity, LI and presence of comedonecrosis and the combination of PR negativity, LI, comedonecrosis and HER2 positivity were all independently associated with increased risk of breast cancer death. The significance of features in the peritumoral stroma need further investigation and may have implications for targeted treatments. In paper IV, we studied the risk of ischemic heart disease (IHD) after treatment for DCIS. Postoperative radiotherapy (RT) in DCIS reduces recurrence rates by half but confers no benefits in terms of survival. It is thus of major importance to consider long-term adverse effects. Left-sided breast irradiation may involve exposure of the heart to ionising radiation with an associated risk of subsequent cardiovascular disease. The cumulative incidence of IHD was analysed in a population-based cohort of 6270 women with DCIS compared 31 257 women without a history of breast cancer. Of the women with DCIS, 38.9% had received adjuvant RT. After a median follow-up of 8 years, there was no increased risk of IHD for women with DCIS versus the comparison cohort. The risk was lower for women with DCIS allocated to RT compared to non-irradiated women and to the comparison cohort, probably due to patient selection. Comparison of RT by laterality did not show any over-risk for irradiation of the left breast. These results are reassuring, but longer follow-up may be warranted considering the continuously increasing use of RT in DCIS management