We analyzed 209 (11.1%) very young patients (≤35 years) and 213 (11.3%) elderly patients (≥60, <70 years). As shown in table , tumor size was of significant difference between the 2 groups (p = 0.018). Less T1 tumors (30.1 vs. 38%) but more T2 and T3 tumors were observed in the very young group. Pathological stage showed there were more stage II and III tumors in the very young group (p = 0.037). The relevance analysis between tumor size and lymph node involvement showed that more young patients with T1 stage had positive lymph nodes (p = 0.033). Molecular subtypes were also significantly different between the 2 groups (p = 0.018). Less Luminal A tumors were observed in the very young group than in the elderly group (43.1 vs. 51.7%), whereas, there were more triple-negative tumors in the very young group (27.8 vs. 16.2%). Tumors with histological grade 3 were more prevalent in very young patients than in elderly patients (45.0 vs. 32.9%; p = 0.038).
Clinicopathological features of patients in the 2 groups
We reviewed the breast imaging data of all patients. 85 patients in the very young group and 110 patients in the elderly group accepted both preoperative breast ultrasonography and mammography (table ). The tumor detection rate with ultrasonography showed no difference between the 2 groups (p = 0.723), whereas the tumor detection rate with mammography was significantly lower in the very young group than in the elderly group (p < 0.001).
Findings of breast imaging in the 2 groups
As shown in table , there were significant differences in surgical methods (p = 0.031), more young patients preferred breast-conserving surgery, and more elderly patients chose mastectomy. More patients received chemotherapy and radiotherapy in the very young group (p < 0.001, p = 0.027, respectively).
Therapeutic strategies in the 2 groups
The median follow-up period was 72 months (range 1–123 months). In the very young patient group, recurrences occurred in 71 cases, 36 cases died from breast cancer, and no patient died from non-cancer-related causes. In the elderly patient group, recurrences occurred in 41 cases, 38 cases died from breast cancer, and 2 patients died from non-cancer-related causes. Younger patients had a worse DFS, with the 6-year DFS being 66 and 80% in the very young and the elderly group, respectively (p = 0.001) (fig. ). After adjustment for tumor size, lymph node status, ER status, PR status, Her2 status, histological grade and chemotherapy in a Cox model, the risk of disease progression for the very young patients was 1.557 times higher than that for the elderly patients (95% confidence interval (CI) = 1.12–2.38) (table ). A higher risk of disease progression was also found for patients with a positive lymph node status, Her2 overexpression, and histological grade 3 (hazard ratio (HR) = 1.71, 95% CI = 1.19–2.45; HR = 1.71, 95% CI = 1.19–2.45; HR = 1.71, 95% CI = 1.19–2.45, respectively) (table ). 6-year overall survival (OS) was 85% in the elderly group as compared to 82% in the younger group (p = 0.840) (fig. ). If non-tumor-related deaths were not included, there also was no significant difference between the 2 groups in terms of tumor-specific death rate (p = 0.859). On univariate and multivariate analysis, young age was not an independent prognostic factor for decreased OS. Positive lymph node status was the only prognostic factor associated with decreased OS in our study (trend p < 0.001).
Kaplan-Meier curves of disease-free survival in the very young group (n = 209) and the elderly group (n = 213).
Cox proportional hazards model for disease-free survival and overall survival
Kaplan-Meier curves of overall survival in the very young group (n = 209) and the elderly group (n = 213).
In the subgroup analysis based on pathological stages, the 6-year DFS for stages I, II and III was 85, 76 and 41% in the very young group, and 93, 84 and 54% in the elderly group, respectively. Only the difference in DFS for stage II achieved statistical significance (p = 0.033), and the DFS of the elderly group was better than that of the very young group. The 6-year OS for stages I, II and III was 94, 88 and 64% in the very young group, and 95, 82 and 77% in the elderly group, respectively, without significant differences among all stages (p = 0.500, 0.125 and 0.402, respectively). In addition, in the subgroup analysis based on molecular subtypes, 6-year DFS for subtype Luminal A in the very young group was poorer than in the elderly group (67 vs. 87%; p < 0.001). There was no significant difference between DFS for other subtypes (p = 0.463, 0.704 and 0.225, respectively), and no significant differences occurred for OS with any subtype.