Analysis of lymphocyte subsets between patients and NCs
The principle of ACL detected by the single-platform was that the known total number of fluorescent microbeads were used as the standard internal parameters and fluorescent-labeled antibodies added into the trucount tubes, then applied acquisition and analysis software in the flow cytometry to calculate data.
Using the method, we primitively compared both the PL and AC of CD3+, CD3+CD4+, CD3+CD8+, B, and NK cells in all patients with BC and NCs. There was no difference in percentages between the two groups (P > 0.05, Fig. 1a). But the AC of CD3+, CD3+CD4+, CD3+CD8+, B and NK cells were significantly decreased in patients with BC (P < 0.001, Fig. 1b). The results suggested that it's ACL but not PL decreased in patients with BC. To our knowledge, PL represents the proportion or composition of each subsets, indicating the development and differentiation of lymphocytes, while ACL demonstrates the exact number of peripheral lymphocyte subsets, indicating the proliferation of lymphocyte progenitor. The results showed that the proliferation ability of lymphocytes was impaired distinctly.
In addition, compared with NCs,the PL in patients with BC at stage Ⅰ-Ⅱ and Ⅲ-Ⅳ showed no significant difference(P > 0.05, Fig. 1c, e), but significant differences were observed in the ACL (P < 0.001, Fig. 1d, f). The results suggested that significant decrease of ACL was one of important characteristics of immune impairment or one sign of immunodepletion.
ACL in patients with different clinical stages
There was no significant difference in PL of CD3+, CD3+CD4+, CD3+CD8+, B and NK cells among patients at different clinical stages (Fig. 2f). Compared with the AC of CD3+, CD3+CD4+, CD3+CD8+, B and NK cells in patients at different clinical stages, we discovered that all of them had varying degrees of decline (Fig. 2). The AC of CD3+ in patients at stage IV was lower than patients at stage I and II(P<0.001), but patients at stage III had no significant difference compared with patients at stage I and II (Fig. 2a). With the progression of clinical stages, the decline of AC of CD4+ was particularly significant in BC patients. Compared with patients at stage I, II and III, the AC of CD4+ in patients at stage IV decreased most strikingly (P < 0.001), then followed by stage III (P<0.01), stage II (P<0.01), and stage I (P < 0.001). However, there was no obvious difference of AC between stage I and stage II in patients (Fig. 2b). Compared to patients at stage I, the AC of CD8+ was decreased in patients at stage II (P < 0.05) and stage III-IV (P < 0.01) (Fig. 2c). Compared with patients at different stages, the variation of B cells AC was only observed in patients at stage I and IV (P < 0.01) (Fig. 2d). Meanwhile, the difference in NK cell AC was only found in patients at stage II and IV (P < 0.001) (Fig. 2e). In a word, our data showed that the decrease of CD4+ and CD8+ AC was closely related to clinical stages, which indicated that it declined upon the exacerbations of BC.
ACL between CR + PR and SD + PD group
To further study the relationship of ACL and efficacy, we divided the patients into two groups according to curative effect, one was effective group (CR + PR), and the other was ineffective group(SD + PD). Independent sample t-test revealed that AC of CD3+, CD3+CD4+, CD3+CD8+, B, and NK cells in CR + PR group were significantly higher than that in SD + PD group (P < 0.001, Fig. 3a, b, d, e; P < 0.01, Fig. 3c). In brief, the results indicated that ACL, especially AC of CD3+CD4+ was closely related to the curative effect, the higher the ACL, the better the efficiency.
ACL in different treatments
To further study the relationship of ACL and different treatments, we divided the patients in effective(CR + PR) and ineffective group(SD + PD) into surgery, chemotherapy and endorinotherapy according to curative ways. The common characteristics of effective and ineffective groups were that AC of CD3+, CD3+CD4+, CD3+CD8+, B, and NK cells were all the highest treated by endocrinotherapy, then followed by surgery and chemotherapy (Fig. 4). The differences were that mean numbers of ACL of every treatment in effective group were higher than that in ineffective group. Strikingly, the mean value of AC of CD3+CD4+ showed the most obvious difference with 518 cells/uL in effective group and 393 cells/uL in ineffective group (Fig. 4b), suggesting AC of CD3+CD4+ was closely related to the efficacy, and may be served as potent blood biomarker to evaluate the prognosis.
Effect of ACL on the progression of BC
In order to ascertain whether ACL can influence the progression of BC (freedom from any events as follows: other newly diagnosed tumors; distant organ metastasis; concurrent infection), we further conducted binary logistic regression analysis to determine the influencing factors of disease progression. From the forest plots of subgroups analysis for progression, we could see AC of CD3+CD4+ cells (95% confidence interval 0.015-0.202, P < 0.001), distant metastasis (95% confidence interval 1.628-28.296, P = 0.009), clinical stages (95% confidence interval 1.252-48.192, P = 0.028), pathological category (95% confidence interval 0.015-0.469, P = 0.005), tumor size (95% confidence interval 0.027-0.601, P = 0.009), menopause (95% confidence interval 2.159-86.12, P = 0.005) could affect the progression of disease(Fig.5). Of them, median AC of CD3+CD4+≥451cells/μL, tumor size<2cm, and pathological category (invasive lobular carcinoma) helpfully contributed to favorable prognosis, which showed in the picture were on the left side of the line of OR = 1. On the contrary, distant metastasis, clinical stages, menopause led to unfavorable prognosis. However, age, family history, age of menarche, differentiation, lymphatic metastasis, vessel carcinoma embolus, treatment, ACL of CD3+CD8+, B cells, and NK cells were not the factors of BC progression (P > 0.05).
Relationship between AC of CD3+CD4+ and clinicopathologic parameters of BC
As can be seen from the figure above, a high AC of CD3+CD4+ cells is beneficial to the prognosis. But, the correlation between CD3+CD4+ cells and clinicopathologic parameters of BC patients is unclear. In 237 patients enrolled, 209(88.2%) were invasive lobular carcinomas (ILC), 28(11.8%) were invasive ductal carcinomas (IDC). Patients with stage I-II and III-IV were 89(37.6%), 148(63.4%), respectively. 76.4% of the patients were medium/low differentiation, 66.7% had lymph node metastasis, 49.4% had distant metastasis, 48.9% had family history, 84.0% had a previous history, such as cyclomastopathy, diabetes, hypertension. The median (cut-off value) AC of CD3+CD4+ cells were 451 cells/μL. According to the cut-off value 451 cells/μL, the patients were divided into two groups,one was CD3+CD4+ cells < 451 cells/μL, there were 118 cases, the other was CD3+CD4+cells ≥ 451 cells/μL, there were 119 cases. Further analysis showed that CD3+CD4+ level was not correlated with age, age of menarche, menopause, tumor size, and differentiation degree of BC patients (P > 0.05), but significantly correlated with family history, previous history, pathological category, clinical stages, lymphatic metastasis, distant metastases, vessel carcinoma embolus, and treatments (P < 0.05) (Table 1).
Table 1
Relationship between AC of CD3+CD4+ and clinicopathologic parameters of BC
Characteristics
|
Overall, n(%)
|
AC of CD3+CD4+
|
χ2
|
P-value
|
|
< 451 cells/µL
|
≥ 451 cells/µL
|
|
|
Age
≥ 64
|
113(47.7%)
|
60
|
50
|
3.072
|
0.08
|
< 64
|
124(52.3%)
|
55
|
69
|
|
|
Family history
Yes
No
Previous history
Yes
No
|
116(48.9%)
121(51.1%)
199(84.0%)
38(16.0%)
|
68
50
105
13
|
48
71
94
25
|
7.089
4.393
|
0.008
0.036
|
Age of menarche
< 12
≥ 12
|
121(51.1%)
116(48.9%)
|
56
62
|
65
54
|
1.217
|
0.270
|
Menopause
|
|
|
|
1.031
|
0.310
|
Yes
No
|
208(87.8%)
29(12.2%)
|
101
17
|
107
12
|
|
|
Tumor size
< 2cm
=2-5cm
> 5cm
|
61(25.7%)
127(53.6%)
49(20.7%)
|
27
66
25
|
34
61
24
|
1.016
|
0.602
|
Pathological category
|
|
|
|
10.215
|
0.001
|
IDC
ILC
|
28(11.8%)
209(88.2%)
|
6
112
|
22
97
|
|
|
Differentiation
High
Medium/low
|
56(23.6%)
181(76.4%)
|
25
93
|
31
88
|
0.777
|
0.378
|
Clinical stages
I + II
III + IV
Lymph node metastasis
Yes
No
|
89(37.6%)
148(63.4%)
158(66.7%)
79(33.3%)
|
16
102
100
18
|
73
46
58
61
|
57.692
34.566
|
< 0.001
< 0.001
|
Distant metastasis
|
|
|
|
48.305
|
< 0.001
|
Yes
No
|
117(49.4%)
120(50.6%)
|
85
33
|
32
87
|
|
|
Vessel carcinoma embolus
Yes
No
|
59(21.6%)
178(78.4%)
|
47
71
|
12
107
|
28.040
|
< 0.001
|
Treatments
Surgery
Surgery + chemotherapy
Surgery + endocrinotherapy
|
83(35.0%)
82(34.6%)
72(30.4%)
|
42
55
21
|
41
27
51
|
22.069
|
< 0.001
|
Bold value represents that P value was significant |
ACL affecting progression-free survival
Moreover, kaplan-Meier survival and multivariate Cox regression model were used to investigate whether ACL could affect PFS. Table 2 showed the results of univariate analysis and multivariate analysis of prognostic factors of PFS. Univariate analysis suggested that 14 clinicopathological parameters contributed to be important predictors of PFS, including AC of CD3+CD4+ (cut-off value <451 cells/μL vs ≥451 cells/μL, P < 0.001)(Fig. 6a), AC of CD3+CD8+ ( cut-off value <324 cells/μL vs ≥324 cells/μL, P = 0.001) (Fig. 6b), AC of B cells (cut-off value <155 cells/μL vs ≥155 cells/μL, P < 0.001)(Fig. 6c), AC of NK cells (cut-off value <162 cells/μL vs ≥162 cells/μL, P < 0.001) (Fig. 6d), family history(yes vs no, P < 0.001), previous history(yes vs no, P = 0.001), smoke(yes vs no, P = 0.043), age of menarche(<12 vs ≥12, P = 0.022) , clinical stages (I-II vs III-IV, P < 0.001) (Fig. 6e) , tumor size (<2cm vs ≥2cm, P = 0.011) (Fig. 6f), pathological category (ILC vs IDC, P = 0.022), lymphatic metastasis (yes vs no, P < 0.001), distant metastasis (yes vs no, P < 0.001) (Fig. 6g), vessel carcinoma embolus(yes vs no, P < 0.001) (Fig. 6h). Multivariate analysis revealed that clinical stages (OR = 5.706, 95% confidence interval 1.884-17.277, P = 0.002) and distant metastasis (OR = 1.929, 95% confidence interval 1.122-3.318, P = 0.018) were the unfavorable prognostic factor, whereas AC of CD3+CD4+ (OR = 0.435, 95% confidence interval 0.269-0.703, P = 0.001) and AC of CD3+CD8+ (OR = 0.435, 95% confidence interval 0.414-0.896, P = 0.012) were the favorable prognostic factors. Kaplan–Meier survival curve of PFS and the cut-off value were shown in Fig.6. The PFS of patients with AC of CD3+CD4+ ≥ 451 cells/μL were longer than that of CD3+CD4+ < 451 cells/μL. Meanwhile, The patients with AC of CD3+CD8+ ≥ 324 cells/μL had longer PFS than that of CD3+CD8+ < 324 cells/μL. Conversely, patients with later clinical stages or distant metastasis predicted poor prognosis. Surprisingly, AC of CD3+ and different treatments had no effect on PFS (Fig. 5i, j).
Table 2
Univariate and multivariate analysis of prognostic factors of PFS
Characteristics
|
|
Univariate
|
|
Multivariate
|
Log-rank χ2
|
P-value
|
OR(95%CI)
|
P-value
|
Age
|
< 64
|
0.318
|
0.573
|
---
|
|
|
|
≥ 64
|
|
|
|
|
|
Family history
|
No
Yes
|
15.503
|
< 0.001
|
|
1.115(0.766–1.622)
|
0.57
|
Previous history
|
No
Yes
|
10.976
|
0.001
|
|
1.107(0.572–2.143)
|
0.763
|
Smoke
|
No
Yes
|
4.079
|
0.043
|
|
0.765(0.485–1.208)
|
0.25
|
Drink
|
No
Yes
|
0.046
|
0.830
|
---
|
|
|
Age of menarche
|
< 12
≥ 12
|
5.214
|
0.022
|
|
1.044(0.725–1.504)
|
0.817
|
Menopause
|
No
Yes
|
1.135
|
0.287
|
---
|
|
|
Tumor size
|
< 2cm
≥ 2cm
|
6.694
|
0.011
|
|
1.003(0.631–1.595)
|
0.99
|
Pathological category
|
ILC
IDC
|
9.251
|
0.002
|
|
0.577(0.261–1.276)
|
0.174
|
ER
|
Negative
Positive
|
1.755
|
0.185
|
---
|
|
|
PR
|
Negative
Positive
|
0.921
|
0.337
|
---
|
|
|
Her2
|
Negative
Positive
|
0.001
|
0.979
|
---
|
|
|
Differentiation
|
High
Medium/low
|
1.322
|
0.250
|
---
|
|
|
Clinical stages
|
I + II
III + IV
|
111.916
|
< 0.001
|
|
5.706(1.884–17.277)
|
0.002
|
Lymphatic metastasis
|
No
Yes
|
80.679
|
< 0.001
|
|
1.224(0.450–3.330)
|
0.692
|
Distant metastasis
|
No
Yes
|
115.315
|
< 0.001
|
|
1.929(1.122–3.318)
|
0.018
|
Vessel carcinoma embolus
|
No
Yes
|
62.756
|
< 0.001
|
|
1.425(0.953–2.131)
|
0.085
|
Treatment
|
Surgery
Surgery + chemotherapy
Surgery + endocrinotherapy
|
1.927
|
0.382
|
---
|
|
|
AC of CD3+
|
< 924
≥ 924
|
0.032
|
0.858
|
---
|
|
|
AC of CD3+CD4+
|
< 451
|
93.634
|
< 0.001
|
|
0.435(0.269–0.703)
|
0.001
|
|
≥ 451
|
|
|
|
|
|
AC of CD3+CD4+
|
< 324
≥ 324
|
12.045
|
0.001
|
|
0.609(0.414–0.896)
|
0.012
|
AC of B cells
|
< 155
≥ 155
|
16.644
|
< 0.001
|
|
0.948(0.641–1.401)
|
0.788
|
AC of NK cells
|
< 162
≥ 162
|
23.792
|
< 0.001
|
|
1.011(0.671–1.522)
|
0.96
|
Bold value represents that P value was significant |