We identified 5,758 patients with invasive breast cancer who underwent conserving breast surgery or total mastectomy at Samsung Medical Center in Seoul, Korea, from 1995 to 2012. For inclusion in the study, patients needed to meet the following criteria: no distant metastasis at the time of diagnosis, no neoadjuvant therapy prior to surgery, and a follow-up period longer than 36 months. The mean age of the patients was 47 years (age range, 21 to 86 years), and the median follow-up period was 64 months. This study was approved by the Institutional Review Board (IRB) of Samsung Medical Center (IRB No. 2018-06-098-001). Formal written informed consent was not required due to a waiver by the appropriate IRB.

Clinicopathological information, including multiplicity, age, tumor size, axillary nodal status, and histological grade, was obtained from electronic medical records or surgical pathology reports. According to the eighth edition of the American Joint Committee on Cancer (AJCC) staging, a patient with multiple breast cancer was defined if two or more separate masses were grossly or microscopically identified in a resection specimen no matter whether they were present in the same or different quadrants. In some cases, through assistance of careful gross examination and correlation with imaging findings, we can determine multiple breast cancer. Pathological tumor stage was assessed according to the eighth AJCC TNM classification [16]. If an invasive carcinoma has been transected by vacuum-assisted biopsy or excisional biopsy, then the sizes in each fragment were not added together, and correlation with the size on breast imaging was helpful to determine the best size for classification. If there had been a prior core needle biopsy or incisional biopsy showing a larger area of invasion than in the excisional specimen, the largest dimension of the invasive carcinoma in the prior specimen should be used for T classification. Histological grade was evaluated according to the Scarff-Bloom-Richardson classification modified by Elston and Ellis [16]. The expression status of estrogen receptors (ER), progesterone receptors (PR), and human epidermal growth factor receptor 2 (HER2) were evaluated by immunohistochemistry based on the surgical specimen. For ER and PR, only nuclear (not cytoplasmic) staining was scored. A positive test was defined as positive staining greater than or equal to 1% of tumor cells. A negative test was defined as staining of less than 1% of tumor cells. HER2 was scored as 0, 1+, 2+, or 3+. Only membrane staining intensity and pattern were evaluated using the recommendations of the American Society of Clinical Oncology/College of American Pathologists [17,18]. A positive test was defined as a staining score of 3+. Tumors with a 2+ score were submitted for silver in situ hybridization. The tumor was considered positive for HER2 amplification if the HER2/chromosome 17 probe signal ratio was greater than 2.0 and/or the average HER2 copy number was greater than 6.0 signals per cell. Molecular subtypes of breast cancer were classified into luminal A, luminal B1, luminal B2, HER2, and triple-negative subtypes based on histological grade and the results of ER, PR, and HER2 immunochemistry as follows: luminal A (ER-positive and/or PR-positive, HER2-negative, and low histological grade [grade 1 or 2]); luminal B1 (ER-positive and/or PR-positive, and HER2-positive); luminal B2 (ER-positive and/or PR-positive, HER2-negative, and high histological grade [grade 3]), HER2-positive (ER-negative, PRnegative, and HER2-positive); and triple-negative (ER-negative, PR-negative, and HER2-negative) [19].

The primary outcome was disease-free survival (DFS), defined as the time interval from the date of surgery to the date of first recurrence, including local or distant. Survival curves were estimated using the Kaplan-Meier method, and survival differences were analyzed by log-rank test. The clinicopathological variables were analyzed in univariate and multivariate analyses of DFS with Cox proportional hazards model. Statistical analysis was performed using the R v3.5.1 (R Foundation for Statistical Computing, Vienna, Austria).

Patients were divided into two groups according to multiplicity. We found breast cancers involving a single mass in 4,744 cases (82.4%) and breast cancers involving multiple masses in 1,014 cases (17.6%). Table 1 shows the results of the comparison between patients with a single mass and patients with multifocal or multicentric masses.

Patients with multiple cancers were more likely to be young and have undergone total mastectomy. Statistically significant differences in lymph node positivity (single 38.0% vs multiple 47.3%, p < .001) and lymphatic invasion (single 24.7% vs multiple 32.6%, p < .001) were found between the two groups. In addition, multiplicity was associated with non–high histological grade (p < .001), ER positivity (p < .001), PR positivity (p < .001), and HER2 negativity (p = .003) of tumor. Therefore, breast cancers with multiple masses were more likely to have luminal A molecular subtype and less likely to be triple-negative subtype compared to those with a single mass (p < .001).

DFS was evaluated in patients with single breast mass and multiple masses. Kaplan-Meier survival analysis indicated that patients with multiple masses had significantly poorer DFS than did those with a single mass (5-year rate, 88.2% vs 85.2%; p = .016) (Fig. 1A). When patients were subclassified according to T-category, Kaplan-Meier survival analysis in the T1 category group revealed significantly worse DFS for multiple breast cancer (5-year rate, 91.3% vs 87.4%; p = .033) (Fig. 1B). There was no significant prognostic difference in T2 and T3 category groups (p = .093 and p = .619, respectively) (Fig. 1C, D). Using the anatomic stage group table in the AJCC eighth edition for tumor staging, breast cancer with multiplicity had poor prognosis in stage I (5-year rate, 92.7% vs 90.3%; p = .019) (Fig. 2A). When using the prognostic stage group table in the AJCC eighth edition, multiple breast masses were found to have significantly shorter DFS than single breast masses in stage group IA (5-year rate, 94.9% vs 88.7%; p = .032) (Fig. 2B). However, no significant difference was found between single and multiple tumors in the other stage groups (i.e., anatomic staging group II or III and prognostic staging group IB, II, or III) (Fig. 2C).

Patients were divided into five molecular subtypes (i.e., luminal A, B1, and B2; HER2-positive; and triple-negative). The prognostic significance of multiplicity was only seen in patients with luminal A and HER2-positive groups in terms of DFS (5-year rate, 92.8% vs. 88.6%; p = .013 and 5-year rate, 86.9% vs 77.8%; p = .003, respectively) (Fig. 3A, B). There was no significant difference among the luminal B1 and B2 and triple-negative subtypes (p = .937, p = .453, and p = .411, respectively) (Fig. 3C–E). In addition, when targeting patients with T1–2 N0 M0, hormone-receptor-positive, and HER2-negative cancer, Kaplan-Meier survival analysis revealed a significantly reduced DFS of multiple breast cancer (5-year rate, 95.2% vs 88.6%; p = .031) (Fig. 4).

Univariate analysis using Cox proportional hazard model indicated that high tumor stage (T3) (hazard ratio [HR], 2.44; 95% confidence interval [CI], 1.84 to 3.23; p < .001), positive lymph node metastasis (HR, 2.06; 95% CI, 1.8 to 2.36; p < .001), high anatomic staging group (i.e., stage III) (HR, 3.47; 95% CI, 2.89 to 4.18; p < .001), positive lymphatic emboli (HR, 2.16; 95% CI, 1.88 to 2.49; p < .001), high histological grade (i.e., grade 3) (HR, 1.52; 95% CI, 1.33 to 1.74; p < .001), negative ER status (HR, 1.23; 95% CI, 1.07 to 1.42; p < .001), positive HER2 status (HR, 1.21; 95% CI, 1.05 to 1.40; p = .004), and the presence of multiplicity (HR, 1.24; 95% CI, 1.04 to 1.48; p = .016) are significant variables associated with lower DFS (Table 2).

These significant factors in the univariate model were included in multivariate analysis, which demonstrated that tumor multiplicity correlated independently with worse DFS (adjusted HR, 1.23; 95% CI, 1.05 to 1.47; p = .021). Other independent factors were high tumor stage (T3) (adjusted HR, 1.81; 95% CI, 1.35 to 2.41; p < .001), positive lymph node metastasis (adjusted HR, 1.84; 95% CI, 1.60 to 2.13; p < .001), and high histological grade (i.e., grade 3) (adjusted HR, 1.33; 95% CI, 1.14 to 1.55; p < .001) (Table 3).