Introduction: One of the main issues in breast cancer treatment is metastasis, which reflects on the patient's health worsening. Studying the factors affecting this problem could prevent and ultimately treat breast cancer diseases. In this respect, this study intends to investigate the factors associated with recurrence and metastasis of breast cancer patients with marginal bivariate logistic model.

Methods: In this retrospective longitudinal study, demographic and disease status information for all patients with breast cancer referred to Iranian Center for Breast Cancer from their files were collected. Incomplete records were excluded. The marginal bivariate logistic model for both recurrence and metastasis events was performed in SPSS/21.

Results: In total, data of 360 patients with average age of 45.2 ± 10.2 was observed. The recurrence and metastasis in patients were 5% and 16% respectively. The odds ratio of recurrences and metastasis was 2.92. ER and lymph node involvement were influenced on recurrence and lymph node involvement was influenced on metastasis. In general, these factors had more effect on recurrence than on metastasis.

Conclusions: In this study, ER and lymph node involvement were risk factors on metastasis and recurrence. By identifying risk factors of the metastasis and recurrence, indicating the progression of the disease and the lack of response to treatment, necessary treatment can earlier start for high risk patients. Use of this model is suggested for other diseases.