Abstract
Metabolic diseases like diabetes and obesity are major risk factors for breast cancer. Aberrant expression of metabolic effectors such as fibroblast growth factor 19 (FGF19) could be therefore associated with the disease. The expression of FGF19 was examined in 193 archival breast tumor samples by immunohistochemistry and evaluated semi-quantitatively by determining the staining index and correlating it with clinicopathological parameters using Fisher's exact test. The correlation between FGF19 expression and 5-year disease-specific survival rate was determined using the univariate Kaplan-Meier analysis. The prognostic value of FGF19 expression was evaluated using the multivariate Cox regression analysis. Of the 193 tumors analyzed, 40 % were classified with low FGF19 expression, whereas 60 % were categorized as tumors with high FGF19 expression. There was a highly significant correlation between high FGF19 expression and patients' age (p = 0.008) as well as 5-year disease-specific survival (p = 0.001). However, FGF19 expression did not show any significant correlations with other clinicopathological parameters, including hormonal status, tumor grade, tumor size, or lymph node status. Univariate Kaplan-Meier log rank analysis showed that patients with high FGF19 expression exhibited a significantly shorter disease-specific 5-year survival (p = 0.007). This effect was exacerbated by lymph node metastasis (p = 0.001), negative estrogen receptor (ER) status (p = 0.002), or old age (p = 0.013). Multivariate analysis showed that high FGF19 expression could be an independent prognostic marker of disease-specific survival in breast cancer patients (p = 0.030). Quantification of FGF19 expression appears to provide valuable prognostic information in breast cancer, particularly in older patients with lymph node metastasis and negative ER status.