Prognostic impact of second cancer on the survival of early breast cancer patients.

Although numerous reports have been published on the incidence of second cancers after adjuvant therapy for early breast cancer, there have been few studies on the effect of the development of second cancer on overall survival (OS) of the patients. 1857 female patients younger than 80 years of age with operable breast cancer with UICC Stages I, II, and III who entered 3 trials of adjuvant therapy were studied for the detection of recurrence and second cancer. The median follow-up period for surviving patients was 12 years (range, 5-25 years). 384 of recurrence and 119 of second cancers occurred. 465 deaths were recorded, the causes of which were designated to be due to recurrence of breast cancer in 326 patients, second cancers in 57, and due to other causes in 82. Many background factors that were significantly related to recurrence did not influence the incidence of and death by second cancers after mastectomy: age and menopausal status alone were related. The difference in the Kaplan-Meier curves between the event (recurrence and second cancers)-free and relapse-free survivals indicates the incidence of second cancers is a significant event in post-operation course of early breast cancer patients (P=0.0001). The survival curve of patients who were free from recurrence- and second cancer-death is shown to be significantly lower than that of those free from breast cancer-specific death (P=0.0355), suggesting that death by second cancers is significantly influential to the overall survival. According to the Cox regression model using a recurrence or second cancers as time-dependent variables, the diagnosis of a recurrence or second cancer is shown to be highly significant to OS of the patients (hazard ratio: 49.3, and 6.3, respectively). Second cancers are shown to be not statistically significantly influential to the breast cancer specific survival (P=0.0637), nor a recurrence to second cancer specific survival (P=0.2285). In spite of heterogeneous distribution of sites of second primary cancers and their different natural histories, the incidence of second cancer has a significant effect on the postoperative survival of early breast cancer patients. In the survival analysis of early breast cancer patients, we have to estimate the contribution of second cancers to properly evaluate the effect of treatment.