Skip to main content

Advertisement

Table 1 Patient characteristics and incidence of lymph node metastasis

From: Prediction of axillary lymph node metastasis in primary breast cancer patients using a decision tree-based model

Variables Tokyo dataset Kyoto dataset Seoul dataset P-value§
  No % No % No %
No. of patients 148 (100) 143 (100) 174 (100)  
Age        <0.001
 Median 55 60 50  
 Range (31–85) (26–88) (25–74)  
Body mass index 0.019
 Median 22.9 22.3 23.2  
 Range (16.6–43.2) (14.8–31.4) (17.8–37)  
 Unknown 3 (2) 0 (0) 1 (0.6)  
Clinical T classification 0.2621
 T1 102 (68.9) 100 (69.9) 108 (62.1)  
 T2 46 (31.1) 43 (30.1) 66 (37.9)  
Clinical N classification 0.002
 N0 137 (92.6) 135 (94.4) 174 (100)  
 N1 11 (7.4) 8 (5.6) 0 (0)  
Skin dimpling <0.001
 Yes 22 (14.9) 14 (9.8) 2 (1.1)  
 No 109 (73.6) 129 (90.2) 172 (98.9)  
 Unknown 17 (11.5) 0 (0) 0 (0)  
Nipple discharge 0.238
 Yes 6 (4.1) 2 (1.4) 3 (1.7)  
 No 138 (93.2) 141 (98.6) 170 (97.7)  
 Unknown 4 (2.7) 0 (0) 1 (0.6)  
Mammography  
Presence of masses 0.284
 Yes 90 (60.8) 88 (61.5) 102 (58.6)  
 Focal asymmetry 22 (14.9) 20 (14) 39 (22.4)  
 No 35 (23.6) 26 (18.2) 33 (19)  
 Unknown 1 (0.7) 9 (6.3) 0 (0)  
Presence of calcifications 0.037
 Yes 67 (45.3) 44 (30.8) 59 (33.9)  
 No 81 (54.7) 94 (65.7) 115 (66.1)  
 Unknown 0 (0) 5 (3.5) 0 (0)  
Shape of calcifications 0.010
 Fine branching or casting 4 (6) 1 (2.3) 3 (5.1)  
 Pleomorphic 9 (13.4) 11 (25) 21 (35.6)  
 Amorphous or indistinct 43 (64.2) 27 (61.4) 35 (59.3)  
 Round or benign 11 (16.4) 4 (9.1) 0 (0)  
 Unknown 0 (0) 1 (2.3) 0 (0)  
Distribution of calcifications 0.024
 Linear or segmented 26 (38.8) 14 (31.8) 22 (37.3)  
 Grouped or clustered 30 (44.8) 29 (65.9) 36 (61)  
 Regional or diffuse 9 (13.4) 1 (2.3) 1 (1.7)  
 Unknown 2 (3) 0 (0) 0 (0)  
Ultrasonography        
 Presence of masses 0.264
 Yes 142 (95.9) 133 (93) 161 (92.5)  
 No 5 (3.4) 10 (7) 13 (7.5)  
 Unknown 1 (0.7) 0 (0) 0 (0)  
 Multifocality 0.114
 Yes 27 (19) 14 (10.5) 21 (13)  
 No 115 (81) 119 (89.5) 140 (87)  
 Maximum tumor size (mm) 0.004
 Median 16 16.1 19  
 Range (4–37) (5–35) (4–37)  
 Depth/width ratio 0.001
 Median 0.72 0.67 0.64  
 Range (0.31–1.36) (0.22–1.43) (0.33–1.27)  
 Unknown 0 (0) 9 (6.8) 0 (0)  
 Echogenic halo <0.001
 Yes 32 (22.5) 62 (46.6) 38 (23.6)  
 No 109 (76.8) 71 (53.4) 123 (76.4)  
 Unknown 1 (0.7) 0 (0) 0 (0)  
 Interruption of the anterior border of the mammary gland 0.807
 Yes 99 (69.7) 91 (68.4) 106 (65.8)  
 No 43 (30.3) 42 (31.6) 54 (33.5)  
 Unknown 0 (0) 0 (0) 1 (0.6)  
 Detection of LNs 0.130
 Detectable 49 (33.1) 37 (25.9) 56 (32.2)  
 Not detectable 82 (55.4) 105 (73.4) 117 (67.2)  
 Unknown 17 (11.5) 1 (0.7) 1 (0.6)  
 Maximum size (mm) of LNs 0.010
 Median 11 10 10  
 Range (5–22) (3–32) (4–17)  
 Unknown 0 (0) 4 (10.8) 1 (1.8)  
 Hilum of LNs 0.021
 Detectable 43 (87.8) 27 (73) 36 (64.3)  
 Not detectable 6 (12.2) 9 (24.3) 20 (35.7)  
 Unknown 0 (0) 1 (2.7) 0 (0)  
Histological type 0.584
 Invasive ductal carcinoma 135 (91.2) 129 (90.2) 160 (92)  
 Invasive lobular carcinoma 5 (3.4) 3 (2.1) 7 (4)  
 Other specific types 8 (5.4) 11 (7.7) 7 (4)  
Estrogen receptor 0.023
 Positive 119 (80.4) 114 (79.7) 121 (69.5)  
 Negative 27 (18.2) 29 (20.3) 53 (30.5)  
 Unknown 2 (1.4) 0 (0) 0 (0)  
Progesterone receptor 0.427
 Positive 83 (56.1) 89 (62.2) 96 (55.2)  
 Negative 63 (42.6) 54 (37.8) 78 (44.8)  
 Unknown 2 (1.4) 0 (0) 0 (0)  
HER2 0.019
 Positive 18 (12.2) 11 (7.7) 29 (16.7)  
 Negative 121 (81.8) 131 (91.6) 125 (71.8)  
 Unknown 9 (6.1) 1 (0.7) 20 (11.5)  
Histological/nuclear grade <0.001
 1 64 (43.2) 43 (30.1) 4 (2.3)  
 2 47 (31.8) 63 (44.1) 82 (47.1)  
 3 27 (18.2) 36 (25.2) 88 (50.6)  
 Unknown 10 (6.8) 1 (0.7) 0 (0)  
LN metastasis 0.292
 Yes 44 (29.7) 44 (30.8) 41 (23.6)  
 No 104 (70.3) 99 (69.2) 133 (76.4)  
  1. Note:
  2. Abbreviations: LN, lymph node.
  3. Estrogen receptor or progesterone receptor positive was defined as ≥10% positively stained cells on immunohistochemical (IHC) testing.
  4. HER2 positive was defined as IHC 3+ or positive on fluorescence in situ hybridization testing.
  5. §The χ 2 test or Kruskal–Wallis test was used depending on the distribution of patients in each variable and dataset.