Development and Validation of Nomograms for Predicting the Prognosis of Triple-Negative Breast Cancer Patients Based on 379 Chinese Patients

Figures & data

Table 1 Clinicopathological Characteristics of the Training and Validation Groups

Characteristic	Training Group (N=255) No. (%)	Validation Group (N=124) No. (%)	p-value
Age (years)			0.986
≤60	212 (83.1)	103 (83.1)
>60	43 (16.9)	21 (16.9)
BMI (kg/m^2)			0.641
<18.5	11 (4.3)	7 (5.7)
18.5–24	179 (70.2)	81 (65.3)
>24	65 (25.5)	36 (29.0)
Menstrual status			0.291
Premenopausal	134 (52.5)	58 (46.8)
Postmenopausal	121 (47.5)	66 (53.2)
Neutrophils (×10^9/L)			0.312
≤3.6	121 (47.5)	52 (41.9)
>3.6	134 (52.5)	72 (58.1)
Lymphocytes (×10^9/L)			0.133
<1.8	123 (48.2)	70 (56.4)
≥1.8	132 (51.8)	54 (43.6)
Platelets (×10^9/L)			0.394
≤210	125 (49.0)	55 (44.4)
>210	130 (51.0)	69 (55.6)
Total protein (g/L)			0.639
≤74	123 (48.2)	63 (50.8)
>74	132 (51.8)	61 (49.2)
Albumin (g/L)			0.958
<45	120 (47.1)	58 (46.8)
≥45	135 (52.9)	66 (53.2)
AGR			0.302
≤1.3	68 (26.7)	27 (21.8)
>1.3	187 (73.3)	97 (78.2)
PLR			0.194
<145	189 (74.1)	84 (67.8)
≥145	66 (25.9)	40 (32.2)
NLR			0.097
≤3.1	229 (89.8)	104 (83.9)
>3.1	26 (10.2)	20 (16.1)
SII			0.121
≤500	163 (63.9)	69 (55.6)
>500	92 (36.1)	55 (44.4)
Surgery type			0.205
Mastectomy	204 (80.0)	106 (85.5)
BCS	51 (20.0)	18 (14.5)
Histopathological type			0.755
IDC	219 (85.9)	105 (84.7)
Others	36 (14.1)	19 (15.3)
Tumor size (cm)			0.180
≤2	122 (47.8)	70 (56.5)
2–5	126 (49.4)	49 (39.5)
>5	7 (2.8)	5 (4.0)
Positive lymph nodes			0.648
0	160 (62.7)	81 (65.3)
1–3	62 (24.3)	25 (20.2)
≥4	33 (13.0)	18 (14.5)
Ki-67^a (%)			0.635
≤30	76 (29.8)	40 (32.3)
>30	166 (65.1)	78 (62.9)

Notes: ^aKi-67 had missing data.

Abbreviations: AGR, albumin-to-globulin ratio; PLR, platelet-to-lymphocyte ratio; NLR, neutrophil-to-lymphocyte ratio; SII, systemic immune inflammation index, defined as (platelets×neutrophils)/lymphocytes; BCS, breast-conserving surgery; IDC, invasive ductal carcinoma.

Table 2 Univariate and Multivariate Analyses of DFS in the Training Group

Characteristic	Univariate Analysis of DFS		Multivariate Analysis of DFS
HR (95% CI)	p-value	HR (95% CI)	p-value
Age (years)
≤60	1		1
>60	2.031 (1.099–3.755)	0.024	1.660 (0.868–3.174)	0.125
AGR
≤1.3	1		1
>1.3	0.414 (0.238–0.722)	0.002	0.463 (0.264–0.812)	0.007
NLR^a
≤3.1	1		1
>3.1	1.870 (0.911–3.837)	0.088	2.117 (1.026–4.367)	0.042
Tumor size (cm)
≤2	1		1
2–5	1.422(0.797–2.537)	0.234	1.426 (0.791–2.570)	0.237
>5	6.937(2.592–18.567)	0.000	2.644 (0.877–7.975)	0.084
Positive lymph nodes
0	1		1
1–3	1.687(0.868–3.246)	0.124	1.763 (0.909–3.420)	0.093
≥4	3.647(1.885–7.055)	0.000	3.341 (1.709–6.532)	0.000
BMI (kg/m^2)	–	0.893
Menstrual status	1.561 (0.900–2.707)	0.113
Neutrophils (×10^9/L)	1.200 (0.694–2.075)	0.514
Lymphocytes (×10^9/L)	0.608 (0.349–1.057)	0.078
Platelets (×10^9/L)	1.150 (0.666–1.984)	0.617
Total protein (g/L)	1.174 (0.679–2.029)	0.567
Albumin (g/L)	0.604 (0.348–1.047)	0.072
PLR	1.709 (0.972–3.006)	0.063
SII	1.416 (0.819–2.448)	0.213
Surgery type	1.003(0.515–1.951)	0.993
Histopathological type	0.904(0.408–2.005)	0.804
Ki-67 (%)	0.889 (0.495–1.597)	0.694

Notes: ^aConsidering the clinical importance of NLR, it was also included in the multivariate analysis.

Abbreviations: DFS, disease-free survival; HR, hazard ratio; CI, confidence interval; AGR, albumin-to-globulin ratio; NLR, neutrophil-to-lymphocyte ratio; BMI, body mass index; PLR, platelet-to-lymphocyte; SII, systemic immune inflammation index, defined as (platelets×neutrophils)/lymphocytes.

Table 3 Univariate and Multivariate Analyses of OS in the Training Group

Characteristics	Univariate Analysis of OS		Multivariate Analysis of OS
HR (95% CI)	p-value	HR (95% CI)	p-value
Age (years)
≤60	1		1
>60	2.500 (1.264–4.945)	0.008	2.044 (0.987–4.230)	0.054
AGR
≤1.3	1		1
>1.3	0.349 (0.184–0.662)	0.001	0.471 (0.239–0.930)	0.030
NLR
≤3.1	1		1
>3.1	2.352 (1.068–5.059)	0.033	3.072 (1.389–6.796)	0.006
Tumor size (cm)
≤2	1		1
2–5	1.658 (0.829–3.314)	0.153	1.670 (0.824–3.386)	0.155
>5	9.943 (3.529–28.016)	0.000	3.223 (1.008–10.305)	0.048
Positive lymph nodes
0	1		1
1–3	1.771 (0.803–3.904)	0.157	2.066 (0.928–4.602)	0.076
≥4	5.194 (2.496–10.810)	0.000	4.731 (2.227–10.048)	0.000
BMI (kg/m^2)	–	0.430
Menstrual status	1.651 (0.872–3.126)	0.123
Neutrophils (×10^9/L)	1.723 (0.895–3.316)	0.103
Lymphocytes (×10^9/L)	0.561 (0.294–1.069)	0.079
Platelets (×10^9/L)	1.296 (0.688–2.443)	0.422
Total protein (g/L)	1.081 (0.576–2.031)	0.808
Albumin (g/L)	0.571 (0.301–1.081)	0.085
PLR	1.872 (0.982–3.569)	0.057
SII	1.947 (1.039–3.649)	0.038
Surgery type	0.662 (0.277–1.581)	0.353
Histopathological type	0.872 (0.341–2.230)	0.775
Ki-67 (%)	1.004 (0.506–1.992)	0.991

Abbreviations: OS, overall survival; HR, hazard ratio; CI, confidence interval; AGR, albumin-to-globulin ratio; NLR, neutrophil-to-lymphocyte ratio; BMI, body mass index; PLR, platelet-to-lymphocyte ratio; SII, systemic immune inflammation index defined as (platelets×neutrophils)/lymphocytes.

Figure 1 Kaplan–Meier survival curves in subgroups with different AGR or NLR of the training group. (A, B) Disease-free survival. (C, D) Overall survival.

Abbreviations: AGR, albumin-to-globulin ratio; NLR, neutrophil-to-lymphocyte ratio.

Figure 2 Nomogram for the prediction of DFS in triple-negative breast cancer.

Abbreviations: DFS, disease-free survival; AGR, albumin-to-globulin ratio; NLR, neutrophil-to-lymphocyte ratio.

Figure 3 Nomogram for the prediction of OS in triple-negative breast cancer.

Abbreviations: OS, overall survival; AGR, albumin-to-globulin ratio; NLR, neutrophil-to-lymphocyte ratio.

Figure 4 Calibration curves for predicting the OS and DFS for patients with triple-negative breast cancer in the training group at (A, C) 3 years and (B, D) 5 years. Nomogram-predicted probability is plotted on the x-axis and the actual survival is plotted on the y-axis.

Abbreviations: DFS, disease-free survival; OS, overall survival.

Figure 5 Calibration curves for predicting the OS and DFS for patients with triple-negative breast cancer in the validation group at (A, C) 3 years and (B, D) 5 years. Nomogram-predicted probability is plotted on the x-axis and the actual survival is plotted on the y-axis.

Abbreviations: DFS, disease-free survival; OS, overall survival.

Figure 6 Time-dependent AUC of OS and DFS (A, B). Difference of AUC between the prediction models and the TNM staging system (C, D).

Abbreviations: AUC, area under curve; OS, overall survival; DFS, disease-free survival; TNM, Tumor-Node-Metastasis.

Figure 7 Five-year ROC curves.

Abbreviations: ROC, receiver operating characteristic; AUC, area under curve; DFS, disease-free survival; OS, overall survival; TNM, Tumor-Node-Metastasis.