The association between body mass index and immunohistochemical subtypes in breast cancer.

BACKGROUND: Body mass index (BMI) is defined as a poor prognostic factor in patients with breast cancer (BC). However, there are controversial results regarding the various effects of BMI on BC, hence the exact pathophysiology of the relation between obesity and BC is still under debate, and remains unclear. This paper aims to investigate the association between BMI at presentation and BC subtypes defined according to the immunohistochemical classification in both premenopausal and postmenopausal patients with BC. PATIENTS AND METHODS: This study is a retrospective and explorative analysis of the 3767 female BC patients from a single center. All patients' BMI at the time of initial diagnosis and tumor demographics were recorded. BMI was stratified into 3 groups as normal-weighted (BMI <25 kg/m2), over-weighted (BMI = 25-29.9 kg/m2), and obese (BMI ≥30 kg/m2). Immunohistochemical classification of the tumors was categorized into 4 groups as follows; luminal-like, HER2/luminal-like, HER2-like, and triple-negative according to the ER/PR and HER2 status. Distribution of Immunohistochemical subtypes, tumor characteristics, and overall survival (OS) analysis were evaluated according to the BMI groups in both premenopausal and postmenopausal patients.
RESULTS: Median BMI of premenopausal and postmenopausal patients was 25.5 (kg/m2) and 28.8 (kg/m2), respectively (P < 0.001). In parallel with the increasing age, patients were more obese at diagnosis in both premenopausal (P < 0.001) and postmenopausal period (P < 0.001). Triple-negative subtype was significantly more frequent in premenopausal patients with BMI ≥30 kg/m2 compared to BMI <30 kg/m2 (P = 0.007). Additionally, premenopausal patients with BMI ≥30 kg/m2 had less common luminal-like subtype (P = 0.033) and more frequently presented with higher tumor stage (P = 0.012) and tumor grade (P = 0.004) compared to patients with BMI <25 kg/m2. On the other hand, premenopausal patients with BMI <25 kg/m2 had significantly more ER-positive tumors (P < 0.001) and lower stages of disease (P = 0.01) compared to their counterparts with BMI ≥25 kg/m2. Premenopausal obese patients with triple-negative (P = 0.001) and luminal-like subtype (P = 0.002) had significantly shorter OS duration compared to overweight counterparts. HER2/luminal-like subtype was found to be significantly greater in postmenopausal overweight patients (P = 0.005). However, BMI had no any other significant effect on survival and immunohistochemical subtypes in postmenopausal patients. Multivariate analysis revealed that triple-negative subtype, grade III tumor, BMI ≥30 kg/m2, T3-4 (P < 0.001), nodal involvement, metastatic disease, and lymphovascular involvement were significantly associated with poorer OS.
CONCLUSION: Our data indicated that BMI was an independent factor in patients with BC, with an association indicating a decreased incidence for luminal-like subtype and increased incidence for triple-negative subtype among premenopausal patients. However, this significance was not found in postmenopausal patients. Accordingly, a plausible etiological heterogeneity in BC might play a role among immunohistochemical subtypes in every life stage of women.