Keywords:
Pathology, Epidemiology, Cancer, Statistics, Screening, Outcomes analysis, Mammography, Oncology, Breast
Authors:
L. Sturesdotter1, H. Sartor2, M. Sandsveden1, K. Johnson2, S. Zackrisson1; 1Malmö/SE, 2Lund/SE
DOI:
10.26044/ecr2019/C-1178
Methods and materials
Data was collected from the Malmö Diet and Cancer Study (MDCS),
a population-based prospective cohort enrolling 28 098 patients from 1991 to 1996,
of whom 17 035 were women6.
At baseline examination anthropometric variables were collected and a comprehensive questionnaire filled in7.
Participants were thereafter continuously followed up8.
In the female population 1242 women were diagnosed with breast cancer from 1991 to 2014.
After exclusion of cancer in situ (n = 105) and bilateral breast tumours (n = 21) a total of 1116 women with invasive breast cancer were eligible for further analyses.
The most dominant mammographic feature of the tumour was retrospectively assessed from the radiology report at time of diagnosis and classified into the following categories; well defined mass,
partly ill-defined mass,
ill-defined/diffuse mass,
spiculated mass,
calcifications of comedo type,
non-specific calcifications,
architectural distortion,
and asymmetric density.
For the purpose of the current study,
the mammographic features were organized into five categories; distinct mass (figure 1),
ill-defined mass,
spiculated mass (figure 2),
calcifications,
and tissue abnormality (the latter category comprised the less frequent features architectural distortion and asymmetrical density).
IHC status for oestrogen receptor (ER),
progesterone receptor (PgR),
and HER2 was available for a majority of the breast cancer cases.
The tumours were classified into molecular breast cancer subtypes in a modified St Gallen 20139 manner using the following categories; luminal A-like,
luminal B-like,
HER2 positive,
and TNBC.
The relation between mammographic features and IHC profile was analysed with logistic regression, adjusted for age at diagnosis,
tumour size,
breast density,
and mode of detection,
generating odds ratios (OR) and 95% confidence intervals (CI).
Chi-square test was performed comparing mammographic features to molecular breast cancer subtypes.
A p-value of <0.05 was considered statistically significant.
All analyses were performed with SPSS version 25.
The study was approved by the Ethical Committee at Lund University (Dnr 652/2005 and Dnr 166/2007).