You are viewing a javascript disabled version of the site. Please enable Javascript for this site to function properly.
Go to headerGo to navigationGo to searchGo to contentsGo to footer
In content section. Select this link to jump to navigation

Infiltration of γ⁢δ T cells, IL-17+ T cells and FoxP3+ T cells in human breast cancer

Abstract

BACKGROUND:

Tumor-infiltrating lymphocytes (TILs) have a strong prognostic value in various forms of cancers. These data often refer to use of the pan-T cell marker CD3, or the cytotoxic T lymphocyte marker CD8α. However, T cells are a heterogeneous group of cells with a wide array of effector mechanisms ranging from immunosuppression to cytotoxicity.

OBJECTIVE:

In this study we have investigated the prognostic effects of some unconventional T cell subtypes in breast cancer; γδ T cells, IL-17+ T cells and FoxP3+ T cells (Tregs) in relation to the conventional CD3 and CD8α T cell markers.

METHODS:

This was done using immunohistochemistry on a human breast cancer tissue microarray consisting of 498 consecutive cases of primary breast cancer.

RESULTS:

Infiltration of γδ T cells and T cell infiltration in general (CD3), correlated with a good prognosis, while Treg infiltration with a worse. Infiltration of γδ T cells was associated with a significantly improved clinical outcome in all breast cancer subtypes except triple negative tumors. Only infiltration of either CD3+ or CD8α+ cells was independently associated with better prognosis for all breast cancer patients.

CONCLUSIONS:

This study sheds further light on the prognostic impact of various T cell subtypes in breast cancer.

1.Introduction

Breast cancer is a heterogeneous disease consisting of different subtypes with varying prognosis [1]. It is however not only the breast cancer subtype that determines the prognostic outcome, but also the tumor microenvironment cell composition [2]. The cells of the immune system are an important part of the tumor microenvironment, where presence of tumor infiltrating lymphocytes (TILs) usually is associated with a better prognosis, while infiltration of myeloid cells is associated with a worse prognosis [3]. T lymphocytes is an important TIL population [3]. In breast cancer, infiltration of T cells has been linked to different outcomes in different breast cancer subtypes. In HER2+ and triple negative breast cancers (TNBCs; ER-PR-HER2-), infiltration of T cells has been associated with an improved prognosis [4], a finding that was even more evident in patients receiving neoadjuvant chemotherapy [5, 6, 7]. It has been postulated that this effect may be due to tumor cell death and expression of neoantigens that initiate tumoricidal immune responses [8, 9].

Table 1

Correlations between CD3, T cell subtypes and breast cancer molecular subtypes

Breast cancer molecular subtypesCD3γδ T cellsTregIL-17+ T cellsCD8
Triple-negativeCorrelation coefficient0.181**0.107*0.216**-0.125**0.151**
P-value< 0.0010.03< 0.0010.0120.001
N 407412401404493
Luminal ACorrelation coefficient-0.131**-0.116*-0.234**0.021-0.05
P-value0.0080.018< 0.0010.6780.267
N 406411400403492
Luminal BCorrelation coefficient0.0000.034.052-0.006-0.033
P-value0.9960.4930.3030.9040.462
N 406411400403492
HER2Correlation coefficient0.144**0.137**0.231**-0.0230.033
P-value (2-tailed)0.0040.005< 0.0010.6500.466
N 406411400403492
ER statusCorrelation coefficient-0.234**-0.182**-0.300**0.096-0.119*
P-value< 0.001< 0.001< 0.0010.0540.008
N 408413402405498

Spearman’s rho. 2-tailed P-value.

T cells can grossly be divided into cytotoxic T lymphocytes (CTLs; CD8+), T helper cells (Th [Th1, Th2, Th9, TFH, Th17 and Tregs]; CD4+), γδ T cells and NKT cells [10]. Regulatory T cells (Tregs) are defined as CD4+ CD25+FoxP3+ T cells, and Th17 cells as CD4+RORγt+ T cells with high production of the pro-inflammatory cytokine; IL-17A [10, 11]. Tregs are associated with immune suppression [12] and infiltration of Tregs in ER+ breast cancers has been shown to correlate with a worse prognosis [4, 13], while studies regarding the prognostic effects of Th17 infiltration in breast cancer are limited [14, 15, 16]. When it comes to CTLs, infiltration of CD8+ lymphocytes does not seem to have a prognostic impact in ER+ breast cancer, but has been shown to correlate with an improved breast cancer specific survival (BCSS) in ER-negative tumors [17]. γδ T cells are unconventional T cells that express invariant, canonical TCRγ and TCRδ chains. They are either CD4-CD8- or express CD8αα+ homodimers, and recognize antigen in an MHC/HLA independent manner [18]. The fact that γδ T cells often express CD8 puts previous studies concerning the prognostic value of CD8+ CTLs in breast cancer in a different light. γδ T cells are T cells with dual functions and can thus be both tumor promoting and suppressing [19]. In breast cancer, γδ T cell infiltration was reported to be associated with the HER2 subtype and poor prognosis in a small patient cohort [20]. However, contrasting data have been shown in recent publications where elevated expression levels of genes associated with γδ T cells had a positive impact on patient survival [21, 22].

Figure 1.

IHC staining of T cell subpopulations in breast cancers and association to survival outcome. A) IHC stainings in breast cancer TMA showing CD3; brown staining, γδ TCR; red membranous staining, FoxP3; brown staining and IL-17; brown staining. B) BCSS and RFS according to the infiltration of pan-T cell marker CD3, γδT cells, Tregs and IL-17+ T cells. Log-rank P value < 0.05 was considered significant.

IHC staining of T cell subpopulations in breast cancers and association to survival outcome. A) IHC stainings in breast cancer TMA showing CD3; brown staining, γ⁢δ TCR; red membranous staining, FoxP3; brown staining and IL-17; brown staining. B) BCSS and RFS according to the infiltration of pan-T cell marker CD3, γ⁢δ⁢T cells, Tregs and IL-17+ T cells. Log-rank P value < 0.05 was considered significant.

There are many reports concerning the prognostic and predictive impact of infiltrating T cells on breast cancer survival, but often only CD3, CD8 or FoxP3-positive T cells have been evaluated [23]. Furthermore, the T cell subpopulations γδ T cells, Th17 cells and Tregs all have been reported to have dual and opposing effects in different tumor types, therefore making them important to study for each cancer type [24, 25]. Also, the presence of IL-17γ+δ T cells has lately been proposed thus complicating the Th17 nomenclature [26]. In this study, we therefore decided to evaluate the prognostic impact of infiltrating γδ T cells, IL-17+ T cells and FoxP3+ T cells (Tregs), as compared to the conventional TIL markers CD8α+ and CD3+ T cells, in tumors from a retrospective, consecutive cohort of 498 patients with primary breast cancer [27]. Our findings highlight the prognostic effect of each T cell subpopulation in breast cancer and will be important for the future understanding and use of novel drugs like immune checkpoint inhibitors in this disease.

2.Materials and methods

2.1Breast cancer patients

The study cohort has been previously described [27, 28, 29, 30] and included 498 patients that were diagnosed with invasive breast cancer between 1 January 1988 and 31 December 1992 at the Department of Pathology, Skåne University Hospital, Malmö. Patient characteristics are provided in Supplementary Table 1. Ethical approval for this study was obtained from the Ethics Committee at Lund University (Dnr 613/02). Informed consent was not required and patients were offered the option to opt out.

Table 2

Crosstaba for CD8α+, FOXP3+, IL-17+ and γδ TCR expression in breast cancer

γδ T cells
01P-value
Numbers (all breast cancers)
CD8α+ T cells
 031  70
 152257  0.003**,a
FOXP3+ T cells
 062142
 120169< 0.001***,a
IL-17+ T cells
 047173
 1331430.520a
Numbers (HER2+)
CD8α+ T cells
 02  1
 1020< 0.001***,a

aPearson’s χ2-test.

2.2TMA, immunohistochemistry and staining assessment

Tissue microarrays (TMA) were constructed as previously described [27, 28, 29, 30]. Analysis of ER, PR and HER2 status of the tumors in the TMA, was performed according to current Swedish guidelines. For antibodies and staining procedures see Supplementary Table 2. Anti-TCRγ specificity was evaluated using sorted peripheral blood γδ T cells as positive control (Supplementary Fig. 1). CD3 and TCRγ were manually annotated using a semiquantitative scoring system and denoted as 0 = none, 1 = low, 2 = moderate and 3 = high in each core. CD8 had been scored previously [31]. The total number of IL-17 and FoxP3 positive cells with lymphocytic morphology was annotated in each core using automated image analysis (Halo image analysis software, Indica Labs, Corrales, NM, USA). The total number of positive cells was then manually categorized as 0 = none, 1 = low, 2 = moderate and 3 = high. The core with the highest number of positive cells within each case was used in the subsequent statistical analyses.

2.3Statistical analysis

Spearman’s Rho test was applied using non-dicho- tomized CD3, TCRγ, FoxP3 and IL-17 scoring for associations between T cell populations and breast cancer subtypes. For all other analyses dichotomized variables were constructed; CD3 was dichotomized into low (0, 1) or high (2, 3), TCRγ into absence (0) or presence (1, 2, 3), FoxP3 into low (0, 1) or high (2, 3) and IL-17 into low (0, 1) or high (2, 3). Pearson χ2-test was used for crosstabs. Kaplan-Meier analysis was used to evaluate the impact of different T cell populations on breast cancer specific survival (BCSS) and recurrence free survival (RFS). Log rank test was applied to analyze any significant differences in Kaplan-Meier survival plots. Cox regression proportional hazards analysis was used to obtain hazard ratios (HR) for BCSS and RFS according to CD8α, CD3, TCRγ, FoxP3 and IL-17 density in both uni- and multivariable analysis, adjusted for dichotomized clinicopathological parameters (age, lymph node metastasis, tumor size, Nottingham histological grade [NHG], ER status, HER2 status, triple negative, luminal A and luminal B), and Ki67. All P values were two-tailed. P value 0.05 was considered significant. All calculations and statistical analyses were performed with IBM SPSS Statistics version 23.0 (SPSS Inc).

3.Results

3.1Infiltration of alternative T cell subpopulations and associations to molecular subtypes of primary breast cancer

Representative staining patterns of the analyzed T cell-subpopulations and correlations between different T cell populations and breast cancer subtypes are shown in Fig. 1a and Supplementary Fig. 1.

As shown in Table 1, infiltration of CD8α positive cells correlated positively with TNBC and inversely with ER-positive breast cancers. Infiltration of both CD3 and γδ T cells was associated with TNBC and HER2+ breast cancers, but inversely associated with both the luminal A subtype as well as with ER-positive breast cancers. Treg infiltration was associated with the TNBC and HER2+ breast cancers, but inversely associated with both the luminal A subtype and ER-positive breast cancers. IL-17+ T cell infiltration was inversely associated with the TNBC subtype. It is known that γδ T cells can express CD8αα homodimers [18], but also IL-17A and the transcription factor FoxP3 [19]. As shown in Table 2, there was a significant correlation between CD8α and γδ TCR (p= 0.003), as well as for FoxP3 and γδ TCR (p 0.001), but not with IL-17 expression and γδ TCR (Table 2).

Table 3

Univariable Cox regression analysis for BCSS and RFS

BCSSRFS
HR (CI 95%)P-value N HR (CI 95%)P-value N
Univariable Cox regression analysis with no stratification
CD8
 01.001221.00120
 10.78 (0.49–1.24)0.293680.82 (0.56–1.19)0.29362
CD3
 01.00  991.00  96
 10.56 (0.35–0.89)0.015*3060.50 (0.35–0.72)< 0.001***304
γδ T cells
 01.00  851.00  84
 10.53 (0.33–0.86)0.01**3250.56 (0.38–0.81)0.002**321
Treg
 01.002101.00206
 11.67 (1.08–2.60)0.022*1891.30 (0.92–1.83)0.134188
IL-17+ T cells
 01.002261.00224
 10.90 (0.58–1.40)0.6431761.02 (0.72–1.44)0.935173
Univariable Cox regression analysis stratified for treated patients
CD8
 01.00  381.00  37
 10.47 (0.25–0.88)0.019*1200.58 (0.34–0.99)0.044*120
CD3
 01.00  391.00  39
 10.72 (0.37–1.39)0.3241060.59 (0.35–0.99)0.046*105
γδ T cells
 01.00  341.00  34
 10.53 (0.28–1.01)0.0541150.59 (0.35–1.02)0.058114
Treg
 01.00741.0073
 11.22 (0.66–2.27)0.528681.20 (0.73–1.98)0.46968
IL-17+ T cells
 01.00891.0088
 10.98 (0.52–1.85)0.952560.96 (0.57–1.62)0.88656
Univariable Cox regression analysis stratified for untreated patients
CD8
 01.00  561.00  55
 10.84 (0.35–2.01)0.691610.71 (0.39–1.31)0.28160
CD3
 01.00  381.00  37
 10.35 (0.14–0.87)0.023*1300.36 (0.20–0.68)0.002**130
γδ T cells
 01.00  321.00  31
 10.50 (0.20–1.30)0.1571370.46 (0.24–0.89)0.021*137
Treg
 01.00841.0083
 12.59 (1.00–6.67)0.050*811.23 (0.67–2.24)0.50281
IL-17+ T cells
 01.00861.0086
 11.02 (0.43–2.37)0.988801.23 (0.67–2.26)0.50179

Abbreviations: BCSS. breast cancer specific survival; RFS. recurrence free survival; HR. hazard ration.

Table 4

Multivariable Cox regression analysis for BCSS and RFS with no stratification

BCSSRFS
HR (CI 95%)P-value N HR (CI 95%)P-value N
Age (yrs)
< 501.00  521.00  52
500.44 (0.22–0.87)0.019*2630.65 (0.38–1.13)0.124261
Lymph node status
 Negative1.001821.00181
 Positive6.53 (3.43–12.41)< 0.001***1332.58 (1.69–3.95)< 0.001***132
Ki67 grade
 0–101.001201.00120
 11– 252.27 (1.08–4.76)0.031*1951.46 (0.89–2.39)0.131193
Size (mm)
< 201.001861.00186
202.13 (1.22–3.73)0.008**1291.55 (1.02–2.36)0.039*127
NHG
 I and II1.002081.00208
 III2.59 (1.34–5.02)0.005**1072.46 (1.51–4.00)< 0.001***105
HER2
 Negative1.002991.00297
 Positive0.79 (0.21–2.96)0.722  161.26 (0.44–3.66)0.666  16
Triple negative
 Negative1.002911.00289
 Positive6.06 (1.32–27.70)0.02*  2413.36 (2.78–64.31)0.001***  24
ER status
 Negative1.00  451.00  44
 Positive3.20 (0.70–14.73)0.1362706.40 (1.30–31.48)0.022*269
Luminal A
 Negative1.001521.00151
 Positive0.34 (0.18–0.64)0.001***1630.91 (0.57–1.46)0.69162
Luminal B
 Negative1.002861.00284
 Positive0.63 (0.26–1.55)0.315  291.09 (0.54–2.20)0.814  29
CD8
 01.00  811.00  80
 10.44 (0.24–0.82)0.009**2340.57 (0.36–0.92)0.022*233
CD3
 01.00  741.00  72
 10.49 (0.25–0.98)0.043*2410.42 (0.25–0.71)0.001***241
γδ T cells
 01.00  601.00  59
 10.68 (0.35–1.32)0.2522550.92 (0.55–1.54)0.748254
Treg
 01.001561.00154
 11.62 (0.88–2.98)0.1211591.38 (0.88–2.16)0.159159
IL-17+ T cells
 01.001721.00171
 11.31 (0.74–2.32)0.3521431.41 (0.92–2.16)0.115142

Abbreviations: BCSS. breast cancer specific survival; RFS. recurrence free survival; HR. hazard ration; NHG. Nottingham histologic grade.

3.2Prognostic significance of alternative T cell subpopulations in the entire cohort

We next investigated the prognostic impact of individual T cell subsets (CD3, CD8α+, γδ T cells, Tregs and IL-17+ T cells) on BCSS and RFS in the entire cohort. Kaplan-Meier analysis revealed that tumor infiltration of any kind of T cells (CD3) correlated with a significantly improved BCSS and RFS (Fig. 1B). There was no significant association between CD8α+ lymphocyte infiltration and BCSS or RFS. γδ T cell infiltration was associated with a significantly prolonged BCSS and RFS. In contrast, Tregs were associated with poor BCSS, while IL-17+ T cells did not have any prognostic impact (Fig. 1B).

Univariable Cox regression analysis showed that presence of T cells (CD3), had an overall significant positive impact on BCSS (HR = 0.56; 95% CI = 0.35–0.89; p= 0.015) and RFS (HR = 0.50; 95% CI = 0.35–0.72; p 0.001). This was also true for γδ T cells (BCSS [HR = 0.53; 95% CI = 0.33–0.86; p= 0.01] and RFS [HR = 0.56; 95% CI = 0.38–0.81; p= 0.002]), but not for CD8α. In contrast, tumor infiltration of Tregs was associated with a decreased BCSS (HR = 1.67; 95% CI = 1.08–2.60; p= 0.022). IL-17+ T cell infiltration showed no significant impact on BCSS or RFS (Table 3).

When adjusted for clinicopathological parameters in multivariable Cox regression analysis, CD8α+ T cells were independently associated with an improved BCSS (HR = 0.44 95% CI = 0.24–0.82; p= 0.009) and RFS (HR = 0.57; 95% CI = 0.36–0.92; p= 0.022), while CD3+ lymphocytes were independently associated with an improved BCSS (HR = 0.49 95% CI = 0.25–0.98; P= 0.043) and RFS (HR = 0.42; 95% CI = 0.25–0.71; p= 0.001; Table 4).

Figure 2.

Kaplan-Meier estimates of breast cancer specific survival according to different infiltrating T cell subpopulations in breast cancer. Impact of pan-T cell CD3, γδ T cells, Tregs and IL-17+ T cells on BCSS in different breast cancer subtypes. Log-rank P value < 0.05 was considered significant.

Kaplan-Meier estimates of breast cancer specific survival according to different infiltrating T cell subpopulations in breast cancer. Impact of pan-T cell CD3, γ⁢δ T cells, Tregs and IL-17+ T cells on BCSS in different breast cancer subtypes. Log-rank P value < 0.05 was considered significant.

Figure 3.

Kaplan-Meier estimates of recurrence free survival according to different infiltrating T cell subpopulations in breast cancer. Impact of pan-T cell CD3, γδ T cells, Tregs and IL-17+ T cells on RFS in different breast cancer subtypes. Log-rank Pvalue < 0.05 was considered significant.

Kaplan-Meier estimates of recurrence free survival according to different infiltrating T cell subpopulations in breast cancer. Impact of pan-T cell CD3, γ⁢δ T cells, Tregs and IL-17+ T cells on RFS in different breast cancer subtypes. Log-rank Pvalue < 0.05 was considered significant.

Figure 4.

Kaplan-Meier estimates on survival according to different infiltrating T-cell populations in patients receiving and not receiving adjuvant endocrine therapy. Impact of pan-T cell CD3, γδ T cells, Tregs and IL-17+ T cells on both BCSS and RFS in breast cancer patients receiving and not receiving adjuvant endocrine therapy. The study cohort was conceived before clinical use of ER-testing, hence both groups includes both ER-positive and ER-negative patients. Log-rank P value < 0.05 was considered significant.

Kaplan-Meier estimates on survival according to different infiltrating T-cell populations in patients receiving and not receiving adjuvant endocrine therapy. Impact of pan-T cell CD3, γ⁢δ T cells, Tregs and IL-17+ T cells on both BCSS and RFS in breast cancer patients receiving and not receiving adjuvant endocrine therapy. The study cohort was conceived before clinical use of ER-testing, hence both groups includes both ER-positive and ER-negative patients. Log-rank P value < 0.05 was considered significant.

3.3Prognostic value of alternative T cell subpopulations according to breast cancer subtype

Kaplan-Meier analyses were also performed in strata according to different subtypes of breast cancer. This revealed that tumor infiltration of T cells overall (CD3) was associated with an improved prognosis specifically in Luminal A and ER-positive breast cancers (BCSS and RFS; Figs 2 and 3). CD8α+ and γδ T cell lymphocyte infiltration was associated with a prolonged BCSS only in the HER2+ breast cancers (Fig. 2), and interestingly, 20 out of 23 HER2+ cases scored positive for both CD8α+ and γδ T cell infiltration (p 0.001; Table 2). Importantly, γδ T cell infiltration was associated with an improved RFS in Luminal A, Luminal B, HER2 and ER-positive breast cancers (Fig. 3). Tumor infiltration of Tregs was associated with a poor BCSS only in ER-positive breast cancers (Fig. 2). In contrast, infiltration of IL-17+ T cells was associated with a poor RFS in TNBCs (Fig. 3).

3.4Prognostic impact of T cell infiltration in relation to endocrine therapy

Next, we evaluated whether tumor infiltration of different T cell populations had any impact on prognosis in relation to endocrine therapy. The study cohort was conceived before clinical use of ER-testing, hence both groups include both ER-positive and ER-negative patients [27]. Kaplan-Meier analysis in strata according to treatment revealed that tumor infiltration of CD8α+ T cells only had a positive impact on BCSS and RFS in the treated group (Fig. 2B). CD3, however, was significantly associated with a prolonged RFS in the treated group, and with a prolonged BCSS and RFS in the untreated group (Fig. 4). Infiltration of γδ T cells was significantly associated with an improved BCSS in patients receiving adjuvant treatment, and with a significantly improved RFS, but not BCSS, in the untreated group (Fig. 4). Infiltration of Tregs in tumors of untreated patients was significantly associated with a poor outcome on BCSS, while IL-17+ T cells showed no impact on BCSS or RFS in either group (Fig. 4).

Univariable Cox regression analysis of the endocrine therapy treated group revealed that CD8α+ T cell infiltration was associated with an improved BCSS (HR = 0.47; 95% CI = 0.25–0.88; p= 0.019) and RFS (HR = 0.58; 95% CI = 0.34–0.99; p= 0.044). In the endocrine therapy untreated group, CD3 was associated with improved BCSS (HR = 0.35; 95% CI = 0.14–0.87; p= 0.023) and RFS (HR = 0.36; 95% CI = 0.20–0.68; p= 0.002), while in the endocrine therapy treated patient group, CD3 was associated only with a prolonged RFS (HR = 0.59; 95% CI = 0.35–0.99; p= 0.046; Table 3). Presence of tumor infiltrating γδ T cells was significantly associated with an improved RFS in the untreated group only (HR = 0.46; 95% CI = 0.24–0.89; p= 0.021) while Tregs were significantly associated with an impaired BCSS (HR = 2.59; 95% CI = 1.00–6.67; p= 0.050; Table 3).

Table 5

Multivariable Cox regression analysis for BCSS and RFS stratified for treated patients

BCSSRFS
HR (CI 95%)P-value N HR (CI 95%)P-value N
Age (yrs)
< 501.00    61.00    6
500.33 (0.07–1.64)0.1761050.19 (0.06–0.58)0.004**105
Lymph node status
 Negative1.00261.0026
 Positive27.39 (4.93–152.07)< 0.001***857.22 (2.39–21.82)< 0.001***85
Ki67 grade
 0–101.00401.0040
 11– 252.36 (0.71–7.85)0.162711.29 (0.58–2.89)0.53671
Size (mm)
< 201.00471.0047
202.64 (1.07–6.51)0.035*642.32 (1.18–4.55)0.014*64
NHG
 I and II1.00621.0062
 III1.76 (0.56–5.52)0.333491.82 (0.82–4.03)0.14249
HER2
 Negative1.001061.00106
 Positive0.99 (0.13–7.91)0.99550.69 (0.09–5.07)0.7155
Triple negative
 Negative1.001021.00102
 Positive12.21 (0.82–182.39)0.07915.78 (1.11–224.73)0.042*9
ER status
 Negative1.00151.0015
 Positive3.78 (0.21–67.21)0.365966.39 (0.37–108.91)0.296
Luminal A
 Negative1.00641.0064
 Positive0.24 (0.09–0.66)0.006**470.66 (0.32–1.37)0.26547
Luminal B
 Negative1.001011.00101
 Positive0.34 (0.07–1.70)0.189101.09 (0.37–3.22)0.88210
CD8
 01.00271.0027
 10.21 (0.09–0.52)0.001***840.41 (0.20–0.84)0.014*84
CD3
 01.00301.0030
 10.73 (0.26–2.06)0.55810.66 (0.32–1.36)0.26281
γδ T cells
 01.00211.0021
 10.32 (0.12–0.82)0.018*900.68 (0.31–1.50)0.34290
Treg
 01.00551.0055
 11.09 (0.39–3.06)0.865561.71 (0.83–3.55)0.14856
IL-17+ T cells
 01.00641.0064
 11.57 (0.66–3.71)0.309471.53 (0.79–2.96)0.2147

Abbreviations: BCSS. breast cancer specific survival; RFS. recurrence free survival; HR. hazard ration; NHG. Nottingham histologic grade.

Table 6

Multivariable Cox regression analysis for BCSS and RFS stratified for untreated patients

BCSSRFS
HR (CI 95%)P-value N HR (CI 95%)P-value N
Age (yrs)
< 501.00371.0037
500.47 (0.12–1.89)0.2881051.15 (0.47–2.81)0.767104
Lymph node status
 Negative1.001231.00123
 Positive16.11 (3.95–65.65)< 0.001***192.95 (1.24–7.02)0.014*18
Ki67 grade
 0–1021.00621.0062
 11– 256.20 (2.33–295.19)0.008**801.92 (0.82–4.47)0.13179
Size (mm)
< 201.001161.00116
2013.97 (3.99–65.23)0.001***262.52 (0.97–6.51)0.05725
NHG
 I and II1.001081.00108
 III5.22 (1.20–22.81)0.028*343.82 (1.66–8.81)0.002**33
HER2
 Negative1.001331.00132
 Positive0.44 (0.04–4.98)0.51192.55 (0.60–10.81)0.2039
Triple negative
 Negative1.001311.00130
 Positive10.20 (0.76–137.56)0.081116.33 (1.58–169.16)0.019*11
ER status
 Negative1.00231.0022
 Positive7.87 (0.66- 93.92)0.1031198.02 (0.79–81.24)0.078119
Luminal A
 Negative1.00601.0059
 Positive0.11 (0.02–0.57)0.009**821.78 (0.66–4.82)0.25682
Luminal B
 Negative1.001311.00130
 Positive0.28 (0.03–2.51)0.255113.14 (0.73–13.50)0.12411
CD8
 01.00401.0039
 10.36 (0.06–2.03)0.2441020.67 (0.31–1.44)0.303102
CD3
 01.00321.0031
 10.03 (0.00–0.24)0.001***1100.21 (0.07–0.64)0.006**110
γδ T cells
 01.00281.0027
 12.64 (0.33–20.96)0.3581141.30 (0.43–3.92)0.644114
Treg
 01.00681.0067
 117.31 (2.45–122.16)0.004**741.50 (0.68–3.31)0.31774
IL-17+ T cells
 01.00741.0074
 12.12 (0.54–8.28)0.279681.65 (0.77–3.52)0.19567

Abbreviations: BCSS. breast cancer specific survival; RFS. recurrence free survival; HR. hazard ration; NHG. Nottingham histologic grade.

In endocrine therapy treated patients, CD8α+T cells remained an independent prognostic factor for a prolonged BCSS and RFS in multivariable Cox regression analysis (HR = 0.21; 95% CI = 0.09–0.52; p= 0.001 and HR = 0.41; 95% CI = 0.20–0.84; p= 0.014, respectively; Table 5). Infiltration of γδ T cells was independently associated with an improved BCSS (HR = 0.32; 95% CI = 0.12–0.82; p= 0.018; Table 5) only in endocrine therapy treated patients. In endocrine therapy untreated patients, however, CD3 was an independent factor of a prolonged BCSS (HR = 0.03; 95% CI = 0.00–0.24; p= 0.001) and RFS (HR = 0.21; 95% CI = 0.07–0.64; p= 0.006), whereas presence tumor infiltrating Tregs was revealed to be an independent factor of an impaired BCSS (HR = 17.31; 95% CI = 2.45–122.16; p= 0.004; Table 6).

4.Discussion

In this study, we investigated the prognostic value of infiltrating γδ T cells, Tregs and IL-17+ T cells, as compared to the common TIL markers CD3 and CD8α, in invasive breast cancer. The analyses were also performed in relation to adjuvant endocrine therapy.

We show that infiltration of T cells (CD3) had a positive effect on prognostic outcome in breast cancer, which supports previous studies evaluating TILs as a prognostic parameter. However, since CD3 is a pan-T cell marker, further investigation was needed to identify specific T cell subtypes that may play a role in breast cancer progression and as potential responders to immune-therapies.

The only T cell subpopulation with results similar to that of CD3, turned out to be γδ T cells that were associated with an improved survival in general. When stratified according to breast cancer subtype, infiltration of γδ T cells was the only T cell population that correlated with a positive outcome in all subtypes (luminal A, luminal B, HER2-positive and ER-positive breast cancers), except TNBCs. It is likely that the TNBC environment not only recruits more γδ T cells, but also has a negative effect on γδ T cell activity. This is a controversy that would merit further investigation at the functional level in the future.

In multivariable Cox regression analyses however, both CD3 and CD8α+ T cell infiltration remained independent factors of a prolonged BCSS as well as RFS. Interestingly, in contrast to general T cell infiltration (CD3), both infiltrating CD8α+ and γδ T cells were revealed to be independent prognostic factors in patients receiving endocrine treatment. The fact that several immune cells can express CD8α, including γδ T cells, makes it difficult to evaluate whether it is conventional CTLs that has the major effect on prognosis in our investigation [32, 33, 34, 35]. Some studies have proposed that γδ T cells can acquire both pro- or anti-tumor properties depending on the tumor microenvironment [25]. Furthermore, the regulation of γδ T cells by inhibitory co-receptors is slightly different as compared to conventional αβ T cells [36, 37]. Indeed, the majority of tumors that scored positive for γδ T cell infiltration in our cohort also scored positive for CD8α+ T cells and these may therefore be γδ T cells, rather than conventional TCR αβ+ CTLs. This was especially clear in the HER2 group, however, due to the small size of this patient group no conclusions should be drawn. In line with our findings, it has already been shown that γδ T cell gene expression signatures were associated with a favorable prognosis across multiple malignancies, also in breast cancer [21, 22].

In the herein studied cohort, infiltration of Tregs was associated with a worse prognosis, but not in multivariable Cox regression analyses. When stratified according to breast cancer subtypes it was revealed that infiltration of Tregs was associated with worse prognosis primarily in ER-positive breast cancers, indicating that Tregs could serve as a potential therapeutic target in ER-positive breast cancers specifically. Also, a significant association between presence of FoxP3+ and γδ T cells was found. However, the fact that the prognostic impact of infiltrating γδ T cells and Tregs was contradictory suggests that these are not regulatory FoxP3γ+δ T cells [38].

We show that infiltration of IL-17+ T cells in general has no prognostic impact. However, when stratified according to breast cancer subtype it was revealed that infiltration of IL-17+ T cells was associated with a worse prognosis in TNBC, specifically. TNBCs often produce inflammatory mediators and it is likely that Th17 cells are either attracted to or even play a role in regulating the TNBC microenvironment [39]. Studies on the clinical relevance of Th17 infiltration in breast tumors has been performed previously, but on small patient cohorts and with opposing results [14, 15, 16, 40]. Nonetheless, these studies point out the complex nature of Th17 cell behavior in a tumor context. Tumor infiltrating Th17 cells have been associated with both tumor promoting and inhibiting properties [39]. Indeed, reports investigating the plasticity of Th17 in breast cancer have shown that Th17 even might be immunosuppressive [16, 41]. Lately, the presence of IL-17+γδ T cells that promotes myeloid cell accumulation and polarization has been reported as an important mechanism in breast cancer progression and metastasis in mice [26] and supporting evidence was recently shown also for human colorectal cancers [42]. The strength of our study lies in the size of the breast cancer cohort, but also that IL-17+ T cell and γδ T cell infiltration was analyzed in parallel. We found no correlation between presence of IL-17+ cells and γδ T cells in this study.

5.Conclusions

In conclusion, our results demonstrate that in breast cancer, infiltration of T cells (CD3) and γδ T cells are associated with good clinical outcome, while Tregs are associated with worse clinical outcome. Infiltration of γδ T cells was associated with a significantly improved prognosis in all breast cancer subtypes except TNBC. However, when adjusted for clinicopathological parameters, only CD8α T cells and CD3 infiltration proved to be independently associated with clinical outcome. Furthermore, except for CD3, infiltration of γδ T cells was the only marker that correlated with an improved clinical outcome in both endocrine treated and untreated patients. When adjusted for clinicopathological parameters, both CD8α T cells and γδ T cell infiltration was independently associated with an improved prognosis for patients receiving adjuvant endocrine therapy, while CD3 and Treg infiltration was associated with clinical outcome in untreated patients. Finally our findings indicate that in breast tumors, the infiltrating CD8α+ T cells may indeed be γδ T cells, which merit further investigation concerning regulation and checkpoint inhibitors. This study renders new light on previously unexplored aspects of tumor infiltrating lymphocytes, and highlights potential prognostic aspects of γδ T cells, IL-17+ T cells and Tregs in breast cancer, with particular reference to molecular subtypes and endocrine adjuvant treatment.

Acknowledgments

The authors wish to thank Ms Elise Nilsson and Mrs Kristina Ekström-Holka for professional help with IHC. This work was generously supported by grants from the Swedish Cancer Society, Gunnar Nilsson Cancer Foundation, MAS Cancer Foundation, Åke Wibergs Foundation and Percy Falks Foundation.

Conflict of interest

The authors declare no conflict of interest.

Supplementary data

The supplementary files are available to download from http://dx.doi.org/10.323/CBM-170026.

References

[1] 

T. Sorlie, C.M. Perou, R. Tibshirani, T. Aas, S. Geisler, H. Johnsen, T. Hastie, M.B. Eisen, M. van de Rijn, S.S. Jeffrey, T. Thorsen, H. Quist, J.C. Matese, P.O. Brown, D. Botstein, P.E. Lonning and A.L. Borresen-Dale, Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications, Proc Natl Acad Sci U S A 98 (2001), 10869–10874.

[2] 

C. Medrek, F. Ponten, K. Jirstrom and K. Leandersson, The presence of tumor associated macrophages in tumor stroma as a prognostic marker for breast cancer patients, BMC Cancer 12 (2012), 306.

[3] 

A. Mantovani, P. Allavena, A. Sica and F. Balkwill, Cancer-related inflammation, Nature 454 (2008), 436–444.

[4] 

S.E. Stanton and M.L. Disis, Clinical significance of tumor-infiltrating lymphocytes in breast cancer, J Immunother Cancer 4 (2016), 59.

[5] 

C. Denkert, G. von Minckwitz, J.C. Brase, B.V. Sinn, S. Gade, R. Kronenwett, B.M. Pfitzner, C. Salat, S. Loi, W.D. Schmitt, C. Schem, K. Fisch, S. Darb-Esfahani, K. Mehta, C. Sotiriou, S. Wienert, P. Klare, F. Andre, F. Klauschen, J.U. Blohmer, K. Krappmann, M. Schmidt, H. Tesch, S. Kummel, P. Sinn, C. Jackisch, M. Dietel, T. Reimer, M. Untch and S. Loibl, Tumor-infiltrating lymphocytes and response to neoadjuvant chemotherapy with or without carboplatin in human epidermal growth factor receptor 2-positive and triple-negative primary breast cancers, J Clin Oncol 33 (2015), 983–991.

[6] 

S. Adams, R.J. Gray, S. Demaria, L. Goldstein, E.A. Perez, L.N. Shulman, S. Martino, M. Wang, V.E. Jones, T.J. Saphner, A.C. Wolff, W.C. Wood, N.E. Davidson, G.W. Sledge, J.A. Sparano and S.S. Badve, Prognostic value of tumor-infiltrating lymphocytes in triple-negative breast cancers from two phase III randomized adjuvant breast cancer trials: ECOG 2197 and ECOG 1199, J Clin Oncol 32 (2014), 2959–2966.

[7] 

M. Ono, H. Tsuda, C. Shimizu, S. Yamamoto, T. Shibata, H. Yamamoto, T. Hirata, K. Yonemori, M. Ando, K. Tamura, N. Katsumata, T. Kinoshita, Y. Takiguchi, H. Tanzawa and Y. Fujiwara, Tumor-infiltrating lymphocytes are correlated with response to neoadjuvant chemotherapy in triple-negative breast cancer, Breast Cancer Res Treat 132 (2012), 793–805.

[8] 

L. Apetoh, F. Ghiringhelli, A. Tesniere, M. Obeid, C. Ortiz, A. Criollo, G. Mignot, M.C. Maiuri, E. Ullrich, P. Saulnier, H. Yang, S. Amigorena, B. Ryffel, F.J. Barrat, P. Saftig, F. Levi, R. Lidereau, C. Nogues, J.P. Mira, A. Chompret, V. Joulin, F. Clavel-Chapelon, J. Bourhis, F. Andre, S. Delaloge, T. Tursz, G. Kroemer and L. Zitvogel, Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy, Nat Med 13 (2007), 1050–1059.

[9] 

R.A. Lake and B.W. Robinson, Immunotherapy and chemotherapy – a practical partnership, Nat Rev Cancer 5 (2005), 397–405.

[10] 

B. Lakshmi Narendra, K. Eshvendar Reddy, S. Shantikumar and S. Ramakrishna, Immune system: A double-edged sword in cancer, Inflamm Res 62 (2013), 823–834.

[11] 

D. Alizadeh, E. Katsanis and N. Larmonier, The multifaceted role of Th17 lymphocytes and their associated cytokines in cancer, Clin Dev Immunol 2013 (2013), 957878.

[12] 

S. Sakaguchi, T. Yamaguchi, T. Nomura and M. Ono, Regulatory T cells and immune tolerance, Cell 133 (2008), 775–787.

[13] 

S. Liu, W.D. Foulkes, S. Leung, D. Gao, S. Lau, Z. Kos and T.O. Nielsen, Prognostic significance of FOXP3+ tumor-infiltrating lymphocytes in breast cancer depends on estrogen receptor and human epidermal growth factor receptor-2 expression status and concurrent cytotoxic T-cell infiltration, Breast Cancer Res 16 (2014), 432.

[14] 

L. Yang, Y. Qi, J. Hu, L. Tang, S. Zhao and B. Shan, Expression of Th17 cells in breast cancer tissue and its association with clinical parameters, Cell Biochem Biophys 62 (2012), 153–159.

[15] 

W.C. Chen, Y.H. Lai, H.Y. Chen, H.R. Guo, I.J. Su and H.H. Chen, Interleukin-17-producing cell infiltration in the breast cancer tumour microenvironment is a poor prognostic factor, Histopathology 63 (2013), 225–233.

[16] 

M. Thibaudin, M. Chaix, R. Boidot, F. Vegran, V. Derangere, E. Limagne, H. Berger, S. Ladoire, L. Apetoh and F. Ghiringhelli, Human ectonucleotidase-expressing CD25high Th17 cells accumulate in breast cancer tumors and exert immunosuppressive functions, Oncoimmunology 5 (2016), e1055444.

[17] 

S.M. Mahmoud, E.C. Paish, D.G. Powe, R.D. Macmillan, M.J. Grainge, A.H. Lee, I.O. Ellis and A.R. Green, Tumor-infiltrating CD8+ lymphocytes predict clinical outcome in breast cancer, J Clin Oncol 29 (2011), 1949–1955.

[18] 

S.R. Carding and P.J. Egan, Gammadelta T cells: Functional plasticity and heterogeneity, Nat Rev Immunol 2 (2002), 336–345.

[19] 

S. Paul and G. Lal, Regulatory and effector functions of gamma-delta (gammadelta) T cells and their therapeutic potential in adoptive cellular therapy for cancer, Int J Cancer 139 (2016), 976–985.

[20] 

C. Ma, Q. Zhang, J. Ye, F. Wang, Y. Zhang, E. Wevers, T. Schwartz, P. Hunborg, M.A. Varvares, D.F. Hoft, E.C. Hsueh and G. Peng, Tumor-infiltrating gammadelta T lymphocytes predict clinical outcome in human breast cancer, J Immunol 189 (2012), 5029–5036.

[21] 

A.J. Gentles, A.M. Newman, C.L. Liu, S.V. Bratman, W. Feng, D. Kim, V.S. Nair, Y. Xu, A. Khuong, C.D. Hoang, M. Diehn, R.B. West, S.K. Plevritis and A.A. Alizadeh, The prognostic landscape of genes and infiltrating immune cells across human cancers, Nat Med 21 (2015), 938–945.

[22] 

R.D. Bense, C. Sotiriou, M.J. Piccart-Gebhart, J.B. Haanen, M.A. van Vugt, E.G. de Vries, C.P. Schroder and R.S. Fehrmann, Relevance of tumor-infiltrating immune cell composition and functionality for disease outcome in breast cancer, J Natl Cancer Inst 109 (2017).

[23] 

K. Wang, J. Xu, T. Zhang and D. Xue, Tumor-infiltrating lymphocytes in breast cancer predict the response to chemotherapy and survival outcome: A meta-analysis, Oncotarget (2016).

[24] 

W.H. Fridman, F. Pages, C. Sautes-Fridman and J. Galon, The immune contexture in human tumours: Impact on clinical outcome, Nat Rev Cancer 12 (2012), 298–306.

[25] 

B. Silva-Santos, K. Serre and H. Norell, gammadelta T cells in cancer, Nat Rev Immunol 15 (2015), 683–691.

[26] 

S.B. Coffelt, K. Kersten, C.W. Doornebal, J. Weiden, K. Vrijland, C.S. Hau, N.J. Verstegen, M. Ciampricotti, L.J. Hawinkels, J. Jonkers and K.E. de Visser, IL-17-producing gammadelta T cells and neutrophils conspire to promote breast cancer metastasis, Nature 522 (2015), 345–348.

[27] 

S. Borgquist, C. Holm, M. Stendahl, L. Anagnostaki, G. Landberg and K. Jirstrom, Oestrogen receptors alpha and beta show different associations to clinicopathological parameters and their co-expression might predict a better response to endocrine treatment in breast cancer, J Clin Pathol 61 (2008), 197–203.

[28] 

F. Lanigan, G. Gremel, R. Hughes, D.J. Brennan, F. Martin, K. Jirstrom and W.M. Gallagher, Homeobox transcription factor muscle segment homeobox 2 (Msx2) correlates with good prognosis in breast cancer patients and induces apoptosis in vitro, Breast Cancer Res 12 (2010), R59.

[29] 

O.L. PC, S.A. Penny, R.T. Dolan, C.M. Kelly, S.F. Madden, E. Rexhepaj, D.J. Brennan, A.H. McCann, F. Ponten, M. Uhlen, R. Zagozdzon, M.J. Duffy, M.R. Kell, K. Jirstrom and W.M. Gallagher, Systematic antibody generation and validation via tissue microarray technology leading to identification of a novel protein prognostic panel in breast cancer, BMC Cancer 13 (2013), 175.

[30] 

S. Svensson, K. Jirstrom, L. Ryden, G. Roos, S. Emdin, M.C. Ostrowski and G. Landberg, ERK phosphorylation is linked to VEGFR2 expression and Ets-2 phosphorylation in breast cancer and is associated with tamoxifen treatment resistance and small tumours with good prognosis, Oncogene 24 (2005), 4370–4379.

[31] 

D.G. DeNardo, D.J. Brennan, E. Rexhepaj, B. Ruffell, S.L. Shiao, S.F. Madden, W.M. Gallagher, N. Wadhwani, S.D. Keil, S.A. Junaid, H.S. Rugo, E.S. Hwang, K. Jirstrom, B.L. West and L.M. Coussens, Leukocyte complexity predicts breast cancer survival and functionally regulates response to chemotherapy, Cancer Discov 1 (2011), 54–67.

[32] 

G. Churlaud, F. Pitoiset, F. Jebbawi, R. Lorenzon, B. Bellier, M. Rosenzwajg and D. Klatzmann, Human and mouse CD8(+)CD25(+)FOXP3(+) regulatory T cells at steady state and during interleukin-2 therapy, Front Immunol 6 (2015), 171.

[33] 

T.F. Gajewski, H. Schreiber and Y.X. Fu, Innate and adaptive immune cells in the tumor microenvironment, Nat Immunol 14 (2013), 1014–1022.

[34] 

M. Ferrarini, S. Heltai, G. Chiesa and M.G. Sabbadini, V delta 1+ gamma/delta T lymphocytes infiltrating human lung cancer express the CD8 alpha/alpha homodimer, Scand J Immunol 40 (1994), 363–367.

[35] 

E.G. Addison, J. North, I. Bakhsh, C. Marden, S. Haq, S. Al-Sarraj, R. Malayeri, R.G. Wickremasinghe, J.K. Davies and M.W. Lowdell, Ligation of CD8alpha on human natural killer cells prevents activation-induced apoptosis and enhances cytolytic activity, Immunology 116 (2005), 354–361.

[36] 

M. Iwasaki, Y. Tanaka, H. Kobayashi, K. Murata-Hirai, H. Miyabe, T. Sugie, M. Toi and N. Minato, Expression and function of PD-1 in human gammadelta T cells that recognize phosphoantigens, Eur J Immunol 41 (2011), 345–355.

[37] 

J. Gertner-Dardenne, C. Fauriat, F. Orlanducci, M.L. Thibult, S. Pastor, J. Fitzgibbon, R. Bouabdallah, L. Xerri and D. Olive, The co-receptor BTLA negatively regulates human Vgamma9Vdelta2 T-cell proliferation: A potential way of immune escape for lymphoma cells, Blood 122 (2013), 922–931.

[38] 

N. Kang, L. Tang, X. Li, D. Wu, W. Li, X. Chen, L. Cui, D. Ba and W. He, Identification and characterization of Foxp3(+) gammadelta T cells in mouse and human, Immunol Lett 125 (2009), 105–113.

[39] 

C.M. Wilke, I. Kryczek, S. Wei, E. Zhao, K. Wu, G. Wang and W. Zou, Th17 cells in cancer: Help or hindrance? Carcinogenesis 32 (2011), 643–649.

[40] 

V. Kaewkangsadan, C. Verma, J.M. Eremin, G. Cowley, M. Ilyas and O. Eremin, Crucial contributions by T lymphocytes (effector, regulatory, and checkpoint inhibitor) and cytokines (TH1, TH2, and TH17) to a pathological complete response induced by neoadjuvant chemotherapy in women with breast cancer, J Immunol Res 2016 (2016), 4757405.

[41] 

J. Ye, X. Su, E.C. Hsueh, Y. Zhang, J.M. Koenig, D.F. Hoft and G. Peng, Human tumor-infiltrating Th17 cells have the capacity to differentiate into IFN-gamma+ and FOXP3+ T cells with potent suppressive function, Eur J Immunol 41 (2011), 936–951.

[42] 

P. Wu, D. Wu, C. Ni, J. Ye, W. Chen, G. Hu, Z. Wang, C. Wang, Z. Zhang, W. Xia, Z. Chen, K. Wang, T. Zhang, J. Xu, Y. Han, T. Zhang, X. Wu, J. Wang, W. Gong, S. Zheng, F. Qiu, J. Yan and J. Huang, gammadeltaT17 cells promote the accumulation and expansion of myeloid-derived suppressor cells in human colorectal cancer, Immunity 40 (2014), 785–800.