Trends in the distribution of breast cancer over time in the southeast of Scotland and review of the literature

Introduction: Breast cancer is the most common form of malignancy in Scottish women, and its incidence appears to be increasing with time. It is therefore important to identify factors associated with risk and outcome. Whilst breast cancer occurs equally in the right and left breasts, tumours most commonly affect the upper outer quadrant (UOQ) of the breast. However, there is only limited information as to whether the incidence has changed over time. Materials and patients: We investigated two cohorts of women diagnosed with breast cancer in the south-east of Scotland between either 1957-1959 or 1997-1999 (i.e., 40 years apart). The earlier cohorts represent 1158 of 1207 women referred to radiation oncologists in the region and the latter group comprised 1477 of about 1600 women referred to the Edinburgh Breast Unit. Results: Whilst the mean age, menopausal status, and laterality of the patients were similar in both groups, the tumour size and tumour location within the breast were significantly different in the two groups. Thus, there was significant reduction in T stage with year of diagnosis (p < 0.0001), the incidence of T1, T2, and T3/4 being 15.6%, 51.9%, and 25.6% in the earlier cohort compared with 49.3%, 36.8%, and 13.7% in the later cohort. The overall distribution within the breast was significantly different by chi-squared analysis (p < 0.0001). In terms of individual quadrants 469 of 1158 (40.5%) tumours were located in the UOQ, whereas in the more recent cohort it was 788 of 1477 (53.4%), this increase in proportion being statistically significant (p < 0.0001). Occurrence in the lower outer quadrant also significantly increased (p < 0.028) but was significantly reduced in the upper inner quadrant and centrally (both p < 0.0001). Conclusion: Analysing data on location for each T stage separately showed that the increased incidence in the UOQ with time was apparent for each subgroup. The increased incidence in UOQ tumours over time is therefore not a simple reflection of decreased size between the two time groups. The underlying reason(s) for this change in distribution with time requires further study.