Original Article

Potential Predictive Factors for Breast Cancer Subtypes from a North Cyprus Cohort Analysis

10.5152/cjms.2020.2291

  • Ayse Ulgen
  • Özlem Gürkut
  • Wentian Li

Received Date: 12.06.2020 Accepted Date: 28.11.2020 Cyprus J Med Sci 2020;5(4):339-349

BACKGROUND/AIMS

We present the first epidemiological survey from North Cyprus to determine the predictive factors for breast cancer subtypes.

MATERIAL and METHODS

More than 300 patients with breast cancer, with 90% of them having the cancer subtype information, were examined at the State Hospital in Nicosia between 2006 and 2015 for their demographic, reproductive, genetic, and epidemiological factors. The breast cancer subtypes and the estrogen receptor (ER) +/- progesterone receptor (PR) +/- and human epidermal growth factor 2 (HER2) +/- status were determined. Single and multiple variable regularized regressions, with predictive factors as independent variables and breast cancer subtypes as dependent variables, were conducted.

RESULTS

Our cohort differed significantly from larger cohorts (e.g., the Breast Cancer Family Registry) in terms of age, menopause status, age at menarche, parity, education, oral contraceptive use, and breastfeeding, but the distribution of breast cancer subtypes was not significantly different. The subtype distribution in our cohort was also not different from that of another Turkish cohort. We found that the ER+ subtype was positively related to age/postmenopause, ER+/PR+ subtype was associated positively with age but negatively with cancer stage, and HER2+ subtype that negatively correlated with ER+ and ER+/PR+ was associated positively with cancer stage but negatively with age/postmenopause.

CONCLUSION

Assuming ER+ and ER+/PR+ to have better prognostic, HER+ to have worse prognostic, then older age and postmenopause seem to be beneficial, smoking and family history of cancer seem to be detrimental. Further steps include exploring potential biomarkers and using cure models to determine long-term breast cancer survivors.

Keywords: Breast cancer subtypes, predictive factors, estrogen, progesterone, human epidermal receptors, regularized regression, LASSO, ridge, elastic nets