Abstract
BACKGROUND: The outcomes of patients with postmenopausal breast cancer(PBC) can be improved through the early detection of intraocular metastases(IOMs). In this study, we investigated patients with PBC, and compared those with IOMs with those with non-intraocular metastases(NIOMs) in terms of blood lipid levels, and then differentiated the risk factors associated with IOMs. METHODS: Student's t-test and a chi-square test were used to discriminate between the IOMs and NIOMs groups. After establishing a Poisson regression model to analyze risk factors, we plotted receiver operating characteristic curves(ROC) to assess the quality of risk factors predicting IOMs. RESULTS: The incidence of IOMs in PBC was 1.16%. There was no significant difference in terms of histopathology between the 2 groups. The levels of total cholesterol (TC), apolipoprotein A1(APOA1) and low-density lipoprotein(LDL) in IOMs were significantly lower than in NIOMs groups. Poisson regression suggested that low levels of APOA1 and LDL were risk factors for IOMs (P = 0.002 and P < 0.001, respectively). ROC curve analysis demonstrated that the cut-off values of APOA1 and LDL were 1.025 g/L and 2.415 mmol/L. The highest prediction accuracy for IOMs involved the combination of APOA1 and LDL (AUC = 0.881, P < 0.001). CONCLUSION: Our research demonstrates that low levels of APOA1 and LDL efficiently predict IOMs in PBC as risk factors, and the combination of APOA1 and LDL was more predictive than single factors.