Purpose: Regulatory T cells (Treg) accumulate in tumor tissues and the peripheral blood of cancer patients and may persist after therapies. This cross-sectional study examines effects of adjuvant chemoradiotherapy (CRT) on Treg numbers and function in head and neck squamous cell carcinoma (HNSCC) patients.

Experimental Design: The frequency and absolute numbers of CD4+, ATP-hydrolyzing CD4+CD39+ and CD8+ T cells, and expression levels of CD39, CD25, TGF-β–associated LAP and GARP on Treg were measured by flow cytometry in 40 healthy donors (NC) and 71 HNSCC patients [29 untreated with active disease (AD); 22 treated with surgery; 20 treated with CRT]. All treated subjects had no evident disease (NED) at the time of phlebotomy. In an additional cohort of 40 subjects with AD (n = 15), NED (n = 10), and NC (n = 15), in vitro sensitivity of CD4+ T-cell subsets to cisplatin and activation-induced cell death (AICD) was tested in Annexin V–binding assays.

Results: CRT decreased the frequency of circulating CD4+ T cells (P < 0.002) but increased that of CD4+CD39+ Treg (P ≤ 0.001) compared with untreated or surgery-only patients. Treg frequency remained elevated for >3 years. CRT increased surface expression of LAP, GARP, and CD39 on Treg. In vitro Treg were resistant to AICD or cisplatin but conventional CD4+ T cells (Tconv) were not. CRT-induced Treg from AD or NC subjects upregulated prosurvival proteins whereas Tconv upregulated proapoptotic Bax.

Conclusions: Highly suppressive, cisplatin-resistant Treg increase in frequency and persist after CRT and could be responsible for suppression of antitumor immune responses and recurrence in HNSCC. Clin Cancer Res; 19(23); 6585–96. ©2013 AACR.

This article is featured in Highlights of This Issue, p. 6333

Translational Relevance

Despite radical surgery and various regimens of chemoradiotherapy (CRT), survival rates remain lower than 50% for head and neck squamous cell carcinoma (HNSCC). The patients initially responsive to therapy develop recurrent disease, second primaries, or metastases. Current interest has been focused on effects of CRT on the host immune system. In this cross-sectional study, absolute numbers, frequency, and functions of CD4+CD39+ regulatory T cells (Treg) were monitored in patients with HNSCC and were compared in cohorts of patients who were either untreated or treated with surgery-only or adjuvant CRT. The data show that in HNSCC patients, CRT increases the frequency of highly suppressive Treg that are resistant to activation-induced cell death and to cisplatin. Mechanisms responsible for this resistance include enhanced expression of prosurvival proteins BcL-2/BcL-xL in Treg. The persistence of Treg in the patients' circulation long after CRT is finished could be responsible for post-CRT immunosuppression, leading to cancer recurrence in HNSCC patients.

Standard of care therapy for patients with head and neck squamous cell carcinoma (HNSCC) includes radical surgery and various regimens of chemoradiotherapy (CRT; ref. 1). However, survival rates remain at less than 50% in HNSCC because of disease recurrence and the development of distant metastases (2). It has been suggested that CRT eliminates subsets of immune cells, thus reducing antitumor effector functions and contributing to disease recurrence (3). For this reason, current interest has been focused on the influence of CRT on the host immune system and on alterations in various hematopoietic cell populations that occur during CRT. We and others have reported that CD4+ T cells are especially sensitive to CRT, and that their absolute number decreases after treatment (3, 4). In contrast, the frequency of regulatory T cells (Treg) was reported to be increased after chemotherapy in some studies (5), and was linked to poor prognosis and shorter patient survival (6). In our earlier studies, we observed an increased frequency and elevated suppressor function of Treg in the peripheral blood of HNSCC patients and at the tumor sites. Furthermore, the Treg frequency was found to be elevated in the blood of patients treated with adjuvant CRT (7). These findings suggesting that in contrast to conventional CD4+ T cells (Tconv), Treg seemed to be resistant to CRT provided a rationale for a more detailed examination of Treg attributes in HNSCC patients treated with adjuvant CRT.

Recently, human Treg have been reported to express surface CD39, an ectonucleotidase, which is a component of the adenosine pathway (8,10) and is considered as a reliable marker for isolation of human Treg (11). We reported that the CD4+CD39+ T-cell subset consists of 2 ATP-hydrolyzing subsets: CD25+FOXP3+ cells, which suppress cytokine expression and proliferation of autologous CD4+ Tconv; and CD25negFOXP3neg cells, which are not suppressive in CFSE-based assays, but are capable of mediating suppression via the adenosine pathway (12). These 2 CD4+ T-cell subsets share the ability to hydrolyze exogenous (e)ATP to ADP and AMP (12, 13). Effects of CRT on these Treg subsets are unknown, and we considered the possibility that their functions and interactions with each other or with CD4+ Tconv are influenced by oncological therapies.

In this cross-sectional study, we prospectively examine the frequency and functions of these 2 Treg subsets in cohorts of HNSCC patients who were either untreated or treated with surgery alone or CRT. The objective was to evaluate short-term and long-term effects of CRT on these Treg subsets relative to CD4+ Tconv or CD8+ Tconv and to consider the mechanisms responsible for persistently elevated frequency of Treg after CRT and its potential impact on disease recurrence.

Patients and normal donors

Peripheral venous blood samples obtained from 20 normal donors (NC) and 71 HNSCC patients were used for phenotyping studies. An additional cohort of 25 patients and 15 NC donated peripheral blood for in vitro experiments with cisplatin. All subjects signed an informed consent approved by the Institutional Review Board of the University of Pittsburgh (IRB #991206). The first patient cohort included 23 females and 48 males with a mean age of 59.3 ± 10.1 years (range: 31–78 years). The 29 patients with active disease (AD) were untreated at the time of blood draws; 22 patients were treated with surgery alone (SRG); and 20 patients had received adjuvant CRT 14 ± 9 months (mean ± SD) before the phlebotomy for this study. All treated patients were no evident disease (NED) at the time of blood draws. Chemotherapy was platinum-based and consisted of cisplatin or carboplatin. Panitunumab and paclitaxel each were added in one case, respectively. All patients received radiotherapy, which ranged from 44 to 70 Gy. Clinicopathologic and demographic data for the patient cohorts are listed in Table 1. The age-matched NC cohort included 9 females and 31 males with a mean age of 51 ± 6 years (range 39–69 years).

Table 1.

Clinicopathologic characteristics of patients studied for the Treg frequency

ControlUntreatedTreated with surgery aloneaTreated with CRT
N (female/male) 40 (9/31) 29 (6/23) 22 (6/16) 20 (2/18) 
Age (±SD)b range (y) 51 ± 6 (39–69) 58 ± 9 (39–74) 60 ± 10 (38–78) 55 ± 12 (31–76) 
Stage (n)c 
 T1  10 
 T2  10 
 T3  
 T4  
Nodal status (n)c 
 N0  13 20 
 N1  
 N2  11 12 
Location (n
 Tongue  10 
 Oral cavity  
 Pharynx  
 Larynx  
 Other  
Days since treatment (range)d   321 ± 389 (8–1680) 408 ± 279 (1–975) 
ControlUntreatedTreated with surgery aloneaTreated with CRT
N (female/male) 40 (9/31) 29 (6/23) 22 (6/16) 20 (2/18) 
Age (±SD)b range (y) 51 ± 6 (39–69) 58 ± 9 (39–74) 60 ± 10 (38–78) 55 ± 12 (31–76) 
Stage (n)c 
 T1  10 
 T2  10 
 T3  
 T4  
Nodal status (n)c 
 N0  13 20 
 N1  
 N2  11 12 
Location (n
 Tongue  10 
 Oral cavity  
 Pharynx  
 Larynx  
 Other  
Days since treatment (range)d   321 ± 389 (8–1680) 408 ± 279 (1–975) 

aTwo patients had received adjuvant radiation without chemotherapy.

bAge at diagnosis.

cThe patients' TNM status was based on clinical/pathological diagnoses. In the CRT cohort, diagnostic biopsies were performed in 12 of 20 patients.

dThe patients who donated blood for the Treg studies were NED at the time of phlebotomy.

The cohort of 25 patients who donated blood for the in vitro sensitivity/resistance studies included 17 males and 8 females with a mean age of 60 years (range 23–82 years); 10 patients had AD (7 of 10 had primary untreated tumors and 3 of 10 had a recurrent disease); all 15 NED patients underwent surgery and were treated with CRT. Their therapy was terminated from 3 weeks to 12 months before phlebotomy for this study.

Collection of peripheral blood mononuclear cells

Blood samples (20 mL) were drawn into heparinized tubes, and plasma was collected after centrifugation at low speed. The cell pellets were resuspended in 40 mL PBS and centrifuged on Ficoll-Hypaque gradients (GE Healthcare Bioscience). Peripheral blood mononuclear cells (PBMC) were recovered, washed in AIM-V medium (Invitrogen), counted in a trypan blue dye, and immediately used for experiments. A fraction of harvested cells was cryopreserved for cytokine expression assays by flow cytometry.

Flow cytometry reagents

The following antihuman monoclonal antibodies (mAb) were used for staining: anti-CD19-ECD, anti-CD4-PC5, anti-CD8-PE, anti-HLA-DR-ECD (all Beckman Coulter); anti-CD73-PE, anti-Bcl-2-FITC, anti-Bcl-2-PE, anti-Bcl-xL-FITC, and anti-Bax-FITC (all BD Pharmingen); anti-CD39-FITC, anti-CD38-FITC, FOXP3-FITC (Clone PCH101), LAP-PE, GARP-APC (all eBioscience), and anti-CD25-PE (Miltenyi). Isotype controls were included in all assays and served as negative controls for surface as well as intracellular staining. All Abs were pretitered using activated as well as nonactivated PBMC to determine the optimal staining dilution for each.

Surface and intracellular staining

Freshly isolated cells to be used for flow cytometry were incubated with mAbs specific for each surface marker in 50 μL PBS for 30 minutes at room temperature in the dark and washed with PBS before acquisition for surface marker detection. For intracellular staining of prosurvival BcL-2 and BcL-xL proteins or proapoptotic Bax, cells were first incubated with relevant mAbs specific for surface markers. After washing, cells were fixed with 4% (v/v) paraformaldehyde in PBS for 20 minutes at room temperature, washed once with PBS, and permeabilized with PBS containing 0.5% bovine serum albumin (BSA) and 0.1% (v/v) saponin. Next, pretitered antibodies specific for BcL-1, BcL-xL, or Bax were added and incubated with the cells for 30 minutes at room temperature. Cells were further washed twice with PBS containing 0.5% BSA and 0.2% saponin, resuspended in FACS flow solution and analyzed by flow cytometry. Expression of intracellular FOXP3 was evaluated using a staining kit available from eBioscience as previously described (14).

Flow cytometry

Flow cytometry was performed using an EPICS XL-MCL or Gallios flow cytometer equipped with Expo32 software (Beckman Coulter). The acquisition and analysis gates were restricted to the live cells based on forward and side scatter properties of the cells. At least 1 × 105 events were acquired for analysis and, where applicable, gates were restricted to the CD4+, CD4+CD39+, or CD8+ T-cell subsets. The cell frequency and mean fluorescence intensity (MFI) were determined. Absolute numbers of CD4+ T cells were determined using the whole blood counts; absolute numbers of Treg were calculated by multiplying the Treg frequency values by the absolute number of CD4+ T cells.

Isolation of Treg

CD4+ T cells were isolated from buffy coats of NC by negative magnetic bead separation according to the manufacturer's protocol (AutoMACS, Miltenyi). CD4+CD39+ T cells were then isolated by magnetic immunobeads as previously described (11). Briefly, CD4+ T cells were incubated with anti-CD39 biotinylated Ab and anti-biotin magnetic beads before isolation by AutoMACS. The purity of isolated cell populations was >85%.

Ex vivo activation of T cells

PBMC obtained from NC or patients with HNSCC were incubated in anti-CD3/antiCD28 Abs-coated wells in flat-bottom 96-well plates in the presence of interleukin (IL)-2 (150 IU/mL) for 48 hours at 37°C in the atmosphere of 5% CO2 in air. In some experiments, PBMC were pretreated with physiologic doses of cisplatin determined as described by van den Bongard and colleagues (15) for 4 hours before further incubation with activating Abs and IL-2.

Detection of Annexin V-binding (ANXV+) to activated T cells

PBMC immediately after their isolation; PBMC after ex vivo activation as described earlier; or PBMC preincubated with cisplatin ± activating Abs and IL-2 were analyzed for ANXV binding to Treg or Tconv by flow cytometry. Samples were stained with FITC-conjugated or PE-conjugated ANXV (Molecular Probes), according to instructions provided by the manufacturer. Flow cytometry was performed with gates restricted to CD3+CD4+CD25high and CD3+CD4+CD25neg T-cell subsets.

In vitro sensitivity of T-cell subsets to cisplatin

CD4+CD39+ and CD4+CD39neg T-cell subsets were isolated from PBMC of NC using magnetic beads as described earlier. These T-cell subsets were then incubated with cisplatin (3–5 μg/mL) in the presence of Staphylococcal Protein B (SEB, 1 μg/mL; Sigma-Aldrich) or activating Abs and IL-2 in RPMI media. On day 2 and 5, cells were harvested and stained for Annexin-FITC, 7AAD (Invitrogen), and CD4-PE (Beckman-Coulter) for 20 minutes on ice as previously described (16).

Statistics

The data are presented as mean values ± SD. For samples with nonparametric distribution, P-values were calculated by Kruskal–Wallis and two-tailed exact Wilcoxon–Mann–Whitney tests. Correlations were calculated by the Spearman test. P-values < 0.05 and R2 values > 0.5 were considered to be significant.

Frequency and absolute numbers of T-cell subsets

The frequency and absolute numbers of CD4+ and CD8+ T cells in the peripheral blood specimens obtained from HNSCC patients who were not yet treated (i.e., with AD before surgery) as well as patients treated with surgery alone (SRG) or with CRT were compared (Fig. 1). The frequency and absolute numbers of both CD4+ and CD8+ T cells were significantly decreased in the blood of patients with AD as well as those treated with surgery alone relative to NC values. In the cohort of CRT patients, percentages and absolute numbers of CD4+ T cells were dramatically decreased (P < 0.002 for both; Fig. 1A). However, the frequency of CD8+ T cells increased following CRT in this patient cohort relative to NC values, although the absolute number of CD8+ T cells significantly decreased (P < 0.001) as shown in Fig. 1B. The data suggest that although CRT adversely affects both T-cell subsets, it has especially detrimental effects on the CD4+ T cells. CD8+ T cells seem to be less sensitive to CRT or they recover faster than CD4+ T cells after CRT.

Figure 1.

The frequency and absolute numbers of CD4+ and CD8+ T lymphocyte subsets in the peripheral blood of HNSCC patients. Box plots show median, 25 and 75 percentiles, and the highest and lowest values. NC, normal control (n = 40); AD, untreated active disease (n = 29); SRG, treated with surgery alone (n = 22); CRT, treated with adjuvant chemoradiotherapy (n = 20). *, P < 0.05; **, P < 0.01.

Figure 1.

The frequency and absolute numbers of CD4+ and CD8+ T lymphocyte subsets in the peripheral blood of HNSCC patients. Box plots show median, 25 and 75 percentiles, and the highest and lowest values. NC, normal control (n = 40); AD, untreated active disease (n = 29); SRG, treated with surgery alone (n = 22); CRT, treated with adjuvant chemoradiotherapy (n = 20). *, P < 0.05; **, P < 0.01.

Close modal

Frequency and absolute numbers of CD4+CD39+ Treg

Because CD4+ T cells contain Treg, we were especially interested in therapy-induced changes in the CD4+ T-cell subset. We have previously shown that in humans, functional CD4+ Treg are CD39+CD25+, express FOXP3 and hydrolyze (e)ATP to 5′AMP (12). Within the CD4+ T-cell compartment, the frequency of these cells was found to be increased (P < 0.01) in the AD and SRG patient cohorts relative to that seen in NC (Fig. 1C). However, the most dramatic increase in the frequency, although not the absolute number (Fig. 1C), of Treg was seen in the CRT patients (P < 0.001). These results suggest that in the shrinking CD4+ T-cell compartment, CD4+CD39+CD25+ Treg are relatively resistant to CRT as compared with CD4+ Tconv. Their frequency increases whereas that of CD4+ T cells declines following CRT. We have also calculated the CD8+/Treg ratio for all patient cohorts, because this ratio has been often used in evaluating changes in the immune profile in situ and in the peripheral circulation of cancer patients treated with various therapies (6, 17). As shown in Fig. 1D, the CD8+/Treg ratio was lower in all patient groups than in NC, but it was most prominently decreased in the CRT patient cohort, which is consistent with the significant increase in the frequency of Treg after CRT and the resistance of these cells to CRT as described earlier. Here, the significantly decreased CD8+/Treg ratio after therapy suggests that Treg are more resistant to CRT than CD8+ T cells. Interestingly, the frequency of CD19+ B cells was also increased after CRT (Supplementary Fig. S1), an indication that B cells might also be resistant to CRT.

A decline in the size of the CD4+ T-cell compartment following oncological therapies was further evaluated, and Fig. 2A shows that it is accompanied by increasing percentages of CD4+CD39+ Treg in representative patients selected from each of the cohorts. Importantly, CD4+CD39+CD25+ Treg seem as a distinct, well-defined population among CD4+ T cells, which can be easily distinguished from CD4+CD39negCD25+ Tconv by flow cytometry. For patients in the SRG and CRT cohorts, the frequency and absolute numbers of total circulating CD4+ T cells negatively correlate with the percentage of CD4+CD39+CD25+ Treg at P < 0.001 and P < 0.002, respectively (Fig. 2B). We have previously shown that based on expression of surface markers and functional characteristics, human Treg can be divided into (i) strongly suppressive CD4+CD39+CD25+ T cells that are FOXP3+ and (ii) CD4+CD39+CD25neg T cells that do not express FOXP3 and mediate little suppression (12). Within the Treg subpopulation, a strong positive relationship exists between these 2 subsets in NC (12) and, as shown in Fig. 2C, this relationship is not altered in untreated patients (AD) or in patients with NED who were treated with SRG or CRT. The 2 Treg subsets track each other and the frequency of both was found to be equally increased in all patient cohorts compared with NC (Fig. 2C). As expected, the frequency of CD4+CD39+CD25+ Treg significantly correlated with that of CD4+CD25high T cells in the peripheral blood of HNSCC patients, indicating that these 2 subsets of Treg closely overlap (Fig. 2D).

Figure 2.

Changes in the CD4+ T-cell compartment after oncological therapies. A, decreased total CD4+ T-cell frequency is accompanied by increased frequency of CD4+CD39+ Treg. Density plots show data from representative HNSCC subjects selected from each study group. Gates are set on the lymphocyte window (left) and on the CD4+ T-cell subset (right). B, negative associations between percentages (left) or absolute numbers (right) of total CD4+ T cells and the percentages of CD4+CD39+CD25+ Treg in the circulation of HNSCC patients treated with SRG or CRT. C, a positive correlation between the frequency of CD4+CD39+CD25+ Treg and CD4+CD39+CD25neg cells in the peripheral circulation. D, a positive correlation between the frequency of CD4+CD25high and CD4+CD39+CD25+ Treg subsets in the peripheral circulation of NC as well as all HNSCC patients.

Figure 2.

Changes in the CD4+ T-cell compartment after oncological therapies. A, decreased total CD4+ T-cell frequency is accompanied by increased frequency of CD4+CD39+ Treg. Density plots show data from representative HNSCC subjects selected from each study group. Gates are set on the lymphocyte window (left) and on the CD4+ T-cell subset (right). B, negative associations between percentages (left) or absolute numbers (right) of total CD4+ T cells and the percentages of CD4+CD39+CD25+ Treg in the circulation of HNSCC patients treated with SRG or CRT. C, a positive correlation between the frequency of CD4+CD39+CD25+ Treg and CD4+CD39+CD25neg cells in the peripheral circulation. D, a positive correlation between the frequency of CD4+CD25high and CD4+CD39+CD25+ Treg subsets in the peripheral circulation of NC as well as all HNSCC patients.

Close modal

Functional characteristics of Treg in patients treated with CRT

Suppressor cell assays based on proliferation inhibition require large cell numbers and, therefore, they are not useful for monitoring of patient specimens, which are often limited in volume. For this reason, we used flow cytometry–based assays to measure expression of functionally relevant proteins, such as the latency-associated peptide (LAP) and the glycoprotein A repetitions predominant (GARP), known to be expressed on the surface of cells that produce and release TGF-β (18). Because Treg utilize TGF-β as a suppressive mechanism (19), and LAP as well as GARP are established surrogate markers for TGF-β (20, 21), we evaluated expression of LAP and GARP on the surface of CD4+CD39+ Treg, as surrogate markers for Treg suppressor activity in all patient cohorts (Fig. 3A and B). We have previously reported that suppressive CD4+CD39+FOXP3+ Treg expressed LAP and GARP, whereas CD4+CD39+FOXP3neg T cells did not (12). The highest percentages of LAP+ or GARP+ Treg were seen in CRT-treated patients. (Fig. 3C). Also, expression of these markers was highly and significantly upregulated on Treg in the CRT patient cohort (Fig. 3A and B). In addition, the MFI of CD39 was upregulated on Treg in all the HNSCC groups relative to that in NC, and the highest MFI for CD39 was observed in Treg of patients in the CRT cohort (Fig. 3D). These data suggest that Treg of the CRT-treated patients have a better ability to hydrolyze exogenous ATP to 5′-AMP than Treg of patients in the other cohorts. We have previously shown that CD39 expression levels on the cell surface positively correlate with enzymatic hydrolysis of exogenous ATP by Treg (12).

Figure 3.

Analysis of functional surface markers on Treg. Expression of TGF-β–associated LAP and GARP on the surface of CD4+CD39+FOXP3+ T cells stimulated with SEB as described in Materials and Methods. A, density plots are gated on CD4+CD39+ T cells of one representative subject from each study group. B, surface expression (MFI) of LAP and GARP in Treg of patients or NC (*, P = 0.01). C, box plots show percentages of LAP+ or GARP+ Treg among total CD4+CD39+ Treg in NC and HNSCC patient cohorts (*P = 0.02). D, box plots show CD39 expression (MFI) in Treg obtained from HNSCC patients (*, P = 0.04).

Figure 3.

Analysis of functional surface markers on Treg. Expression of TGF-β–associated LAP and GARP on the surface of CD4+CD39+FOXP3+ T cells stimulated with SEB as described in Materials and Methods. A, density plots are gated on CD4+CD39+ T cells of one representative subject from each study group. B, surface expression (MFI) of LAP and GARP in Treg of patients or NC (*, P = 0.01). C, box plots show percentages of LAP+ or GARP+ Treg among total CD4+CD39+ Treg in NC and HNSCC patient cohorts (*P = 0.02). D, box plots show CD39 expression (MFI) in Treg obtained from HNSCC patients (*, P = 0.04).

Close modal

Persistent upregulation of the Treg frequency in CRT-treated patients

Blood specimens were obtained from NED patients treated with CRT at various times after the end of therapy (n = 20). The average time period elapsed between the end of treatment and phlebotomy for this study was 14 ± 9 months (mean ± SD). The increased Treg frequency observed in the CRT-treated patients was independent of the time of blood draws after the end of therapy (Fig. 4A and B). This suggests that the increase in the Treg frequency represents a long-lasting effect in these patients. A subset of HNSCC patients (N = 22) treated with surgery alone also showed an increase in the Treg frequency shortly after surgery (1–6 months). However, in patients tested 6 to 15 months after surgery, the Treg frequency was normal, suggesting that in contrast to CRT, surgery was not associated with persistent elevation of Treg (Fig. 4). Importantly, in some of the NED patients who were treated with CRT, the frequency of Treg was elevated even after 3 years after end of therapy. In 3 NED patients treated with CRT, serial blood specimens were obtained at 6, 8, and 12 months after termination of therapy. The Treg frequency remained elevated, ranging between 12 and 16%, in these patients' blood, further suggesting that a stable and persistent increase in Treg characterizes this patient cohort. Figure 4 shows that the percentages of total CD4+ T cells remained within the normal range in patients treated with surgery, whereas CRT-treated patients had low percentages of CD4+ T cells months after therapy.

Figure 4.

Persistence of Treg in the circulation of NED patients after therapy and in vitro resistance of Treg to cisplatin. A, percentages of CD4+CD39+CD25+ Treg in the peripheral circulation of HNSCC patients who are NED after CRT. The elevated Treg frequency in patients was independent of the time period elapsed between the last treatment and phlebotomy for this study. B, in HNSCC patients treated with SRG alone, the Treg frequency was increased shortly after treatment but later seemed to normalize. C, percentages of total CD4+ T cells were persistently decreased in CRT-treated patients. A–B, the solid line represents the mean Treg (3.2%) or CD4+ T-cell (42%) frequency established for NC, and the dotted lines show ± 1 SD of the mean. D, isolated CD4+CD39+ and CD4+CD39neg T cells obtained from 4 different NC were incubated in the presence of physiological concentrations of cisplatin for 48 hours. At the cisplatin concentration of 3 μg/mL, the increase in frequency of ANXV+ cells was greater (P < 0.03) among CD4+CD39neg Tconv than among CD4+CD39+ Treg.

Figure 4.

Persistence of Treg in the circulation of NED patients after therapy and in vitro resistance of Treg to cisplatin. A, percentages of CD4+CD39+CD25+ Treg in the peripheral circulation of HNSCC patients who are NED after CRT. The elevated Treg frequency in patients was independent of the time period elapsed between the last treatment and phlebotomy for this study. B, in HNSCC patients treated with SRG alone, the Treg frequency was increased shortly after treatment but later seemed to normalize. C, percentages of total CD4+ T cells were persistently decreased in CRT-treated patients. A–B, the solid line represents the mean Treg (3.2%) or CD4+ T-cell (42%) frequency established for NC, and the dotted lines show ± 1 SD of the mean. D, isolated CD4+CD39+ and CD4+CD39neg T cells obtained from 4 different NC were incubated in the presence of physiological concentrations of cisplatin for 48 hours. At the cisplatin concentration of 3 μg/mL, the increase in frequency of ANXV+ cells was greater (P < 0.03) among CD4+CD39neg Tconv than among CD4+CD39+ Treg.

Close modal

Clinical significance of persistently high Treg

The cohort of 20 CRT patients was followed for 37 ± 13 months (mean ± SD from the end of treatment) to determine whether anyone experienced disease recurrence. During this period of time, 4 patients recurred and 3 others developed second primary HNSCC (35%). Two of the second primary tumors occurred at 25 months and the third at 45 months after CRT ended, and all 3 patients had elevated percentages of Treg (6–19%) in the peripheral circulation. Within the CRT cohort, there was no significant difference in the Treg frequency (P = 0.18) between patients who recurred or developed new primary tumors (12 ± 4%) and those who did not recur (9 ± 5%; means ± SD). In the similarly followed surgery-only cohort, 4 of 22 patients experienced recurrence (18%), and there were no second primary cancers observed. There was also no significant difference in the Treg frequency (P > 0.2) between patients who had recurrence and those who did not.

In vitro sensitivity of CD4+ T-cell subsets to cisplatin or AICD

Our data suggest that the high frequency of Treg seen in the blood of HNSCC patients treated with CRT could be because of resistance of these cells induced by therapy. To test the hypothesis that the long-term Treg survival after CRT is a characteristic of these cells, sensitivity/resistance to cisplatin of CD4+CD39+ Treg freshly isolated from the PBMC of NC and activated ex vivo (i.e., incubated either with SEB or anti-CD3/CD28 Abs and IL-2) was tested in ANXV-binding assays. CD4+CD39+ Treg and CD4+CD39neg Tconv were incubated with physiologic concentrations of cisplatin as previously described by van den Bongard and colleagues (15). The CD4+CD39+ Treg subset was found to be more resistant (P < 0.03) to cisplatin than CD4+CD39neg Tconv (Fig. 4D).

Using PBMC obtained from NC and HNSCC patients treated or not treated with CRT, we studied resistance/sensitivity of Treg to activation-induced cell death (AICD) following in vitro activation of PBMC with anti-CD3/CD28 Abs and IL-2 (150 IU/mL). After 48-hour incubation, we used flow cytometry to measure ANXV binding to CD4+CD24high versus CD4+CD25neg T-cell subsets and to compare their resistance/sensitivity of these cells to AICD or cisplatin. Figure 5A shows that after ex vivo activation, only CD4+CD25high Treg of patients treated with CRT were resistant to activation-induced cell death, as evidenced by the lack of ANXV binding (left), whereas activated Treg of untreated patients or of NC were highly sensitive to AICD. In contrast, no difference in ANX binding to CD4+CD25neg Tconv was seen among the 3 cohorts: only 20–40% of Tconv bound ANXV (right). Furthermore, as shown in Fig. 5B, preincubation with cisplatin (4 hours) followed by PBMC activation for 48 hours induced resistance of CD4+CD25high Treg obtained from NC or untreated HNSCC patients to AICD (left), and this resistance was also evident in Treg of patients treated with CRT. In contrast, pretreatment with cisplatin induced greater sensitivity to AICD in CD4+CD25neg Tconv of NC as well as HNSCC patients, as nearly all of these cells became ANXV+ as shown in Fig. 5B (right).

Figure 5.

In vitro resistance of CD4+CD25high Treg to cisplatin and upregulated expression of the prosurvival proteins in Treg treated with cisplatin. A, ANXV binding to Treg (CD4+CD25high) (left) or Tconv (CD4+CD25neg) (right) in the in vitro activated PBMC obtained from NC (n = 15), untreated HNSCC patients (pre; n = 10) or HNSCC patients treated with CRT (post; n = 15). PBMC were incubated with anti-CD3/CD28 Abs and 150 IU/mL of IL-2 for 48 hours and studied for the evidence of AICD (ANXV binding) by flow cytometry, gating on Treg and Tconv cells. Only Treg obtained from the CRT-treated patients (post) were resistant to AICD. B, ANXV binding to Treg (CD4+CD25high) (left) or Tconv (CD4+CD25neg) (right) in PBMC, which were pretreated in vitro with cisplatin for 4 hours before incubation of the cells in the presence of anti-CD3/CD28 Abs and IL-2 for 48 hours. Only Treg became resistant to AICD as determined by measuring ANXV binding. C, expression of Bcl-2, Bcl-xL, and Bax are shown as MFI ± SD in CD4+CD25high Treg and CD4+CD25neg Tconv. PBMC were obtained from the blood of NC, untreated, or CRT-treated patients and were preincubated or not with cisplatin and then activated as described earlier in A. The solid lines (PRE to POST panels) show increased expression levels of Bcl-2 and Bcl-xL in POST Treg as compared with those in PRE Treg that were incubated with in vitro cisplatin. *, P < 0.001; **, P < 0.005.

Figure 5.

In vitro resistance of CD4+CD25high Treg to cisplatin and upregulated expression of the prosurvival proteins in Treg treated with cisplatin. A, ANXV binding to Treg (CD4+CD25high) (left) or Tconv (CD4+CD25neg) (right) in the in vitro activated PBMC obtained from NC (n = 15), untreated HNSCC patients (pre; n = 10) or HNSCC patients treated with CRT (post; n = 15). PBMC were incubated with anti-CD3/CD28 Abs and 150 IU/mL of IL-2 for 48 hours and studied for the evidence of AICD (ANXV binding) by flow cytometry, gating on Treg and Tconv cells. Only Treg obtained from the CRT-treated patients (post) were resistant to AICD. B, ANXV binding to Treg (CD4+CD25high) (left) or Tconv (CD4+CD25neg) (right) in PBMC, which were pretreated in vitro with cisplatin for 4 hours before incubation of the cells in the presence of anti-CD3/CD28 Abs and IL-2 for 48 hours. Only Treg became resistant to AICD as determined by measuring ANXV binding. C, expression of Bcl-2, Bcl-xL, and Bax are shown as MFI ± SD in CD4+CD25high Treg and CD4+CD25neg Tconv. PBMC were obtained from the blood of NC, untreated, or CRT-treated patients and were preincubated or not with cisplatin and then activated as described earlier in A. The solid lines (PRE to POST panels) show increased expression levels of Bcl-2 and Bcl-xL in POST Treg as compared with those in PRE Treg that were incubated with in vitro cisplatin. *, P < 0.001; **, P < 0.005.

Close modal

Differential effects of cisplatin on Treg versus CD4+ Tconv

To investigate the mechanisms responsible for resistance of Treg in CRT-treated patients to AICD, expression of prosurvival Bcl-2/Bcl-xL and of proapoptotic Bax was compared in Treg versus Tconv preincubated with or without cisplatin before in vitro activation (Fig. 5C). Cisplatin upregulated expression of the prosurvival proteins in Treg, whereas it significantly downregulated prosurvival protein expression in Tconv (Fig. 5C, right) Also, the pretreatment with cisplatin decreased expression of Bax in Treg, whereas it dramatically upregulated expression of this proapoptotic protein in Tconv. These CD4+CD25+ Treg in the blood of HNSCC patients treated with CRT (POST) upregulated expression levels of survival proteins relative to those in PRE Treg (P < 0.001) and were highly resistant to in vitro cisplatin, perhaps because of cumulative direct or indirect in vivo effects of CRT. In contrast, CD4+CD25+ Treg in the blood of NC incubated in vitro with cisplatin for 4 hours (Fig. 4D) showed a relatively small, albeit statistically significant, increase in resistance versus that observed in CD4+CD25neg Tconv.

Effects of oncological therapies on lymphocyte subsets have been of considerable interest, because of the possibility that therapy-induced changes in immune-cell homeostasis might interfere with antitumor activity. In this respect, in vivo studies in mice indicate that radiation and chemoradiation exert strong effects on the host immune system, including Treg (22). The limited data that are available for patients with cancer suggest that CRT reduces the number of circulating lymphocytes (23). In HNSCC patients, CD4+ T helper cells were previously reported to be particularly sensitive to CRT, as they were found to be significantly depleted relative to other T-cell subsets (3, 24). Recently, attention has focused on Treg, a subset of CD4+ T cells responsible for mediating suppression of antitumor immune responses (25). The presence and accumulations of Treg at tumor sites and in the peripheral circulation of cancer patients, including HNSCC patients, have been widely reported (6, 7, 26). These high Treg frequencies have been negatively or positively correlated with disease outcome in various malignancies (6, 27, 28). Studies also indicate that the Treg frequency tends to increase after oncologic therapies (23), introducing a possibility that Treg are more resistant to CRT than CD4+ Tconv. This prospective cross-sectional study in a large cohort of HNSCC patients was performed to examine this possibility and to provide insights into potential mechanisms responsible for Treg resistance to oncological therapies.

Confirming our previous observations, we show that in comparison to NC, untreated HNSCC patients or those treated with surgery or surgery plus CRT have reduced percentages and absolute numbers of CD4+ T cells but increased percentages of CD4+CD39+ Treg. As the CD4+ T cell compartment shrinks in disease, the proportion but not absolute number of circulating CD4+CD39+ Treg increases. This contraction was especially magnified in the CRT patient cohort, implying that Treg, unlike CD4+ Tconv, are resistant to CRT. Treg resistance to radiation and drugs has been reported (29, 30). However, the exact mechanisms responsible for Treg resistance to CRT remain unclear. This resistance might reflect the slow proliferation rate of Treg in the periphery or existence of protective mechanisms, for example, multidrug transporter pumps, and the enhanced capability for DNA repair (31). A short exposure of Treg to cisplatin before activation seems to protect them from AICD and is mediated by elevated expression levels of survival proteins, Bcl-2 and Bcl-xL, and lower levels of proapoptotic Bax than those observed in Tconv (Fig. 5). Still another possibility is that toll-like receptors (TLR) present on Treg (32) and signaling in response to ligands released by radiation-induced tissue damage activate the PI3K/Akt survival pathway, protecting Treg from effects of CRT. Such TLR-mediated chemoresistance has been previously described for tumor cells (33–35). Others report that platinum-based chemotherapies decrease immune suppression acting not on Treg but on dendritic cells (DC) or tumor cells, reducing expression of the T-cell inhibitory molecule PD-L2 on these cells and upregulating recognition of tumor targets by T cells (36).

Here, we advance the hypothesis that CRT induces tissue changes that have indirect effects on Treg, favoring their survival and suppressor functions. For example, increased TGF-β levels in irradiated tissues could promote Treg differentiation (12, 22). The persistently enlarged post-CRT Treg compartment includes cells with increased immunosuppressive capabilities (e.g., upregulated LAP, GARP, and CD39). These CRT-induced Treg produce more 5′-AMP and ADO, acquire and maintain the ‘killer phenotype’ associated with expression of Fas, FasL, and GrB/perforin, as previously reported by us (37), and upregulate prosurvival proteins. We show here that activated CD4+CD39+ Treg are more resistant to cisplatin than CD4+ Tconv and that CD4+CD25high Treg obtained from HNSCC patients with NED after CRT become resistant to AICD, whereas CD4+CD25neg Tconv remain sensitive (Fig. 5). Together, our data support the hypothesis that Treg induced in the CRT environment acquire new characteristics that make them into long-surviving and potentially more dangerous immune-regulatory cells.

In preclinical models, radiotherapy has been reported to either interfere with antitumor immune responses by the elimination of immune effector cells (38) or promote these responses by enhancing the processing and presentation of tumor antigens by APC (39, 40). These opposing radiotherapy-induced effects are related to alterations in tissues and cells, which seem to be dependent on the dose and type of radiotherapy (38–40). For example, the ATP concentration is increased in damaged irradiated tissues (22), providing therapy-induced Treg with an opportunity to produce 5′-AMP and ADO. These factors acting via the adenosine receptors potentially influence survival and activity of Treg. Because isolated CD4+25+ T cells contain significantly higher levels of mRNA coding for A1R, A2AR, and A3R than CD4+CD25neg T cells (our unpublished data), we surmise that autocrine and paracrine signaling via these receptors has profound effects on functions and survival of CRT-induced Treg. Our experiments with CD4+CD39+ or CD4+CD25high Treg subsets suggest that environmentally generated signals, as exemplified by TGF-β, ATP, or ADO, regulate functions of CRT-induced Treg and might contribute to their resistance and survival after CRT. Thus, resistance of these Treg to cisplatin or AICD might not be their intrinsic characteristic but rather reflects an environmentally regulated requirement for their activity to regulate immune homeostasis disturbed by therapy that alters the immune cell numbers and balance. To what extent this induced resistance to drugs such as cisplatin is related to immune escape mechanisms, which are responsible for the promotion of tumor growth in cancer patients, remains to be determined. Our data suggest that human Treg are able to utilize various survival mechanisms to protect themselves from exogenous stress signals. Treg survival seems to be critically important for the host, emphasizing the significant role these T cells play in health and disease.

The observed elevations in the Treg frequency and functions in patients with active (pretherapy) and inactive (posttherapy) disease were independent of the time elapsed between the treatment and blood draws. In fact, the Treg frequency remained elevated for weeks and months in many, although not all, HNSCC patients who were NED after CRT. The suppressive potential of Treg is known to be related to the Treg/Tconv ratio. The functional consequence of an increase in the proportion of Treg after CRT is that this ratio changes in favor of Treg. These Treg acquire stronger suppressor function after CRT and express more TGF-β–associated LAP and GARP as well as ATP-hydrolyzing CD39 on their surface. Our data are provocative, as they suggest that persistent elevations in the CD4+CD39+ Treg frequency induced in the blood of cancer patients by potentially curative therapies, such as surgery and adjuvant CRT, result in suppression of antitumor immune effector cells and thus promote disease recurrence or the development of secondary cancers. The fact that in our study, 3 of the CRT-treated patients with elevated Treg percentages developed second primary cancers within 2 to 4 years of therapy suggests that the imbalance in immune regulation might promote disease progression.

It has also been suggested that CRT induces a state of chronic inflammation in tissues, which has been linked to cancer development (41). The presence of iTreg in this setting could be beneficial to the host by normalizing T-cell responses and maintaining immune responses in balance. The role of Treg that accumulate after CRT in the promotion or inhibition of tumor progression is currently unclear. Prospective serial monitoring will be necessary to determine the biologic and clinical consequences of increased Treg frequency in HNSCC patients treated with CRT.

No potential conflicts of interest were disclosed.

Conception and design: P.J. Schuler, L. Strauss, J.T. Johnson, T.L. Whiteside

Development of methodology: M. Harasymczuk, L. Strauss

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Harasymczuk, B. Schilling, Z. Saze, J.T. Johnson

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): P.J. Schuler, B. Schilling, L. Strauss, S. Lang, T.L. Whiteside

Writing, review, and/or revision of the manuscript: P.J. Schuler, B. Schilling, L. Strauss, J.T. Johnson, T.L. Whiteside

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): P.J. Schuler, T.L. Whiteside

Study supervision: P.J. Schuler, T.L. Whiteside

This study was supported in part by NIH grant PO1 CA109688 to T.L. Whiteside and by the Pittsburgh-Essen-Partnership Program to P.J. Schuler. The services of the Immunologic Monitoring and Cellular Products Laboratory (IMCL) and the Flow Cytometry Laboratory were supported in part by the NCI CCSG 5P30 CA047904.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Brockstein
BE
,
Vokes
EE
. 
Head and neck cancer in 2010: maximizing survival and minimizing toxicity
.
Nat Rev Clin Oncol
2011
;
8
:
72
4
.
2.
Oddone
N
,
Morgan
GJ
,
Palme
CE
,
Perera
L
,
Shannon
J
,
Wong
E
, et al
Metastatic cutaneous squamous cell carcinoma of the head and neck: the immunosuppression, treatment, extranodal spread, and margin status (ITEM) prognostic score to predict outcome and the need to improve survival
.
Cancer
2009
;
115
:
1883
91
.
3.
Kuss
I
,
Hathaway
B
,
Ferris
RL
,
Gooding
W
,
Whiteside
TL
. 
Decreased absolute counts of T lymphocyte subsets and their relation to disease in squamous cell carcinoma of the head and neck
.
ClinCancer Res
2004
;
10
:
3755
62
.
4.
Sewell
HF
,
Halbert
CF
,
Robins
RA
,
Galvin
A
,
Chan
S
,
Blamey
RW
. 
Chemotherapy-induced differential changes in lymphocyte subsets and natural-killer-cell function in patients with advanced breast cancer
.
Int J Cancer
1993
;
55
:
735
8
.
5.
Ersvaer
E
,
Liseth
K
,
Skavland
J
,
Gjertsen
BT
,
Bruserud
O
. 
Intensive chemotherapy for acute myeloid leukemia differentially affects circulating TC1, TH1, TH17 and TREG cells
.
BMC Immunol
2010
;
11
:
38
.
6.
Sato
E
,
Olson
SH
,
Ahn
J
,
Bundy
B
,
Nishikawa
H
,
Qian
F
, et al
Intraepithelial CD8+ tumor-infiltrating lymphocytes and a high CD8+/regulatory T cell ratio are associated with favorable prognosis in ovarian cancer
.
Proc Natl Acad Sci USA
2005
;
102
:
18538
43
.
7.
Mandapathil
M
,
Szczepanski
MJ
,
Szajnik
M
,
Ren
J
,
Lenzner
DE
,
Jackson
EK
, et al
Increased ectonucleotidase expression and activity in regulatory T cells of patients with head and neck cancer
.
Clin Cancer Res
2009
;
15
:
6348
57
.
8.
Deaglio
S
,
Dwyer
KM
,
Gao
W
,
Friedman
D
,
Usheva
A
,
Erat
A
, et al
Adenosine generation catalyzed by CD39 and CD73 expressed on regulatory T cells mediates immune suppression
.
J Exp Med
2007
;
204
:
1257
65
.
9.
Borsellino
G
,
Kleinewietfeld
M
,
Di Mitri
D
,
Sternjak
A
,
Diamantini
A
,
Giometto
R
, et al
Expression of ectonucleotidase CD39 by Foxp3+ Treg cells: hydrolysis of extracellular ATP and immune suppression
.
Blood
2007
;
110
:
1225
32
.
10.
Mandapathil
M
,
Hilldorfer
B
,
Szczepanski
MJ
,
Czystowska
M
,
Szajnik
M
,
Ren
J
, et al
Generation and accumulation of immunosuppressive adenosine by human CD4+CD25highFOXP3 +regulatory T cells
.
J Biol Chem
2010
;
285
:
7176
86
.
11.
Schuler
PJ
,
Harasymczuk
M
,
Schilling
B
,
Lang
S
,
Whiteside
TL
. 
Separation of human CD4+CD39 +T cells by magnetic beads reveals two phenotypically and functionally different subsets
.
J Immunol Methods
2011
;
369
:
59
68
.
12.
Schuler
PJ
,
Schilling
B
,
Harasymczuk
M
,
Hoffmann
T
,
Johnson
J
,
Lang
S
, et al
Phenotypic and functional characteristics of ATP-hydrolysing CD4(+) CD39(+) FOXP3(+) and CD4(+) CD39(+) FOXP3(neg) T-cell subsets in patients with cancer
.
Eur J Immunol
2012
;
42
:
1876
85
.
13.
Moncrieffe
H
,
Nistala
K
,
Kamhieh
Y
,
Evans
J
,
Eddaoudi
A
,
Eaton
S
, et al
High expression of the ectonucleotidase CD39 on T cells from the inflamed site identifies two distinct populations, one regulatory and one memory T cell population
.
J Immunol
2010
;
185
:
134
43
.
14.
Law
JP
,
Hirschkorn
DF
,
Owen
RE
,
Biswas
HH
,
Norris
PJ
,
Lanteri
MC
. 
The importance of Foxp3 antibody and fixation/permeabilization buffer combinations in identifying CD4+CD25+Foxp3+ regulatory T cells
.
Cytometry A
2009
;
75
:
1040
50
.
15.
van den Bongard
HJ
,
Mathot
RA
,
Beijnen
JH
,
Schellens
JH
. 
Pharmacokinetically guided administration of chemotherapeutic agents
.
Clin Pharmacokinet
2000
;
39
:
345
67
.
16.
Strauss
L
,
Whiteside
TL
,
Knights
A
,
Bergmann
C
,
Knuth
A
,
Zippelius
A
. 
Selective survival of naturally occurring human CD4+CD25+Foxp3+ regulatory T cells cultured with rapamycin
.
J Immunol
2007
;
178
:
320
9
.
17.
Chikamatsu
K
,
Sakakura
K
,
Whiteside
TL
,
Furuya
N
. 
Relationships between regulatory T cells and CD8 +effector populations in patients with squamous cell carcinoma of the head and neck
.
Head Neck
2007
;
29
:
120
7
.
18.
Stockis
J
,
Colau
D
,
Coulie
PG
,
Lucas
S
. 
Membrane protein GARP is a receptor for latent TGF-beta on the surface of activated human Treg
.
Eur J Immunol
2009
;
39
:
3315
22
.
19.
Zheng
SG
,
Wang
JH
,
Gray
JD
,
Soucier
H
,
Horowitz
DA
. 
Natural and induced CD4+CD25+ cells educate CD4+CD25 cells to develop suppressive activity: the role of IL-2, TGF-β, and IL-10
.
J Immunol
2004
;
172
:
5213
21
.
20.
Tran
DQ
,
Andersson
J
,
Hardwick
D
,
Brebis
L
,
Illei
GG
,
Shevach
EM
. 
Selective expression of latency-associated peptide (LAP) and IL-1 receptor type I/II (CD121a/CD121b) on activated human FOXP3+ regulatory T cells allows for their purification from expansion cultures
.
Blood
2009
;
113
:
5125
33
.
21.
Sun
J
,
Tang
DN
,
Fu
T
,
Sharma
P
. 
Identification of human regulatory T cells in the setting of T-cell activation and anti-CTLA-4 immunotherapy on the basis of expression of latency-associated peptide
.
Cancer Discov
2012
;
2
:
122
30
.
22.
Schaue
D
,
Xie
MW
,
Ratikan
JA
,
McBride
WH
. 
Regulatory T cells in radiotherapeutic responses
.
Front Oncol
2012
;
2
:
90
.
23.
Rech
AJ
,
Mick
R
,
Kaplan
DE
,
Chang
KM
,
Domchek
SM
,
Vonderheide
RH
. 
Homeostasis of peripheral FoxP3(+) CD4 (+) regulatory T cells in patients with early and late stage breast cancer
.
Cancer Immunol Immunother
2010
;
59
:
599
607
.
24.
Hathaway
B
,
Johnson
JT
,
Piccirillo
JF
,
Snyderman
CH
,
Wagner
RL
,
Labriola
S
, et al
Chemoradiation for metastatic SCCA: role of comorbidity
.
Laryngoscope
2001
;
111
:
1893
5
.
25.
Golovina
TN
,
Vonderheide
RH
. 
Regulatory T cells: overcoming suppression of T-cell immunity
.
Cancer J
2010
;
16
:
342
7
.
26.
Strauss
L
,
Bergmann
C
,
Gooding
W
,
Johnson
JT
,
Whiteside
TL
. 
The frequency and suppressor function of CD4+CD25highFoxp3+ T cells in the circulation of patients with squamous cell carcinoma of the head and neck
.
Clin Cancer Res
2007
;
13
:
6301
11
.
27.
Petersen
RP
,
Campa
MJ
,
Sperlazza
J
,
Conlon
D
,
Joshi
MB
,
Harpole
DH
 Jr
, et al
Tumor infiltrating Foxp3+ regulatory T-cells are associated with recurrence in pathologic stage I NSCLC patients
.
Cancer
2006
;
107
:
2866
72
.
28.
Suzuki
H
,
Onishi
H
,
Morisaki
T
,
Tanaka
M
,
Katano
M
. 
Intratumoral FOXP3(+)VEGFR2(+) regulatory T cells are predictive markers for recurrence and survival in patients with colorectal cancer
.
Clin Immunol
2013
;
146
:
26
33
.
29.
Szczepanski
MJ
,
Szajnik
M
,
Czystowska
M
,
Mandapathil
M
,
Strauss
L
,
Welsh
A
, et al
Increased frequency and suppression by regulatory T cells in patients with acute myelogenous leukemia
.
Clin Cancer Res
2009
;
15
:
3325
32
.
30.
Kachikwu
EL
,
Iwamoto
KS
,
Liao
YP
,
DeMarco
JJ
,
Agazaryan
N
,
Economou
JS
, et al
Radiation enhances regulatory T cell representation
.
Int J Radiat Oncol Biol Phys
2011
;
81
:
1128
35
.
31.
Schuler
PJ
,
Trellakis
S
,
Greve
J
,
Bas
M
,
Bergmann
C
,
Bolke
E
, et al
In vitro chemosensitivity of head and neck cancer cell lines
.
Eur J Med Res
2010
;
15
:
337
44
.
32.
Wild
CA
,
Brandau
S
,
Lindemann
M
,
Lotfi
R
,
Hoffmann
TK
,
Lang
S
, et al
Toll-like receptors in regulatory T cells of patients with head and neck cancer
.
Arch Otolaryngol Head Neck Surg
2010
;
136
:
1253
9
.
33.
Szajnik
M
,
Szczepanski
MJ
,
Czystowska
M
,
Elishaev
E
,
Mandapathil
M
,
Nowak-Markwitz
E
, et al
TLR4 signaling induced by lipopolysaccharide or paclitaxel regulates tumor survival and chemoresistance in ovarian cancer
.
Oncogene
2009
;
28
:
4353
63
.
34.
Soderlund
K
,
Perez-Tenorio
G
,
Stal
O
. 
Activation of the phosphatidylinositol 3-kinase/Akt pathway prevents radiation-induced apoptosis in breast cancer cells
.
Int J Oncol
2005
;
26
:
25
32
.
35.
Sun
Q
,
Zheng
Y
,
Liu
Q
,
Cao
X
. 
Rapamycin reverses TLR4 signaling-triggered tumor apoptosis resistance by disrupting Akt-mediated Bcl-xL upregulation
.
Int Immunopharmacol
2008
;
8
:
1854
8
.
36.
Lesterhuis
WJ
,
Punt
CJ
,
Hato
SV
,
Eleveld-Trancikova
D
,
Jansen
BJ
,
Nierkens
S
, et al
Platinum-based drugs disrupt STAT6-mediated suppression of immune responses against cancer in humans and mice
.
J Clin Invest
2011
;
121
:
3100
8
.
37.
Czystowska
M
,
Strauss
L
,
Bergmann
C
,
Szajnik
M
,
Rabinowich
H
,
Whiteside
TL
. 
Reciprocal granzyme/perforin-mediated death of human regulatory and responder T cells is regulated by interleukin-2 (IL-2)
.
J Mol Med (Berl)
2010
;
88
:
577
88
.
38.
Kaur
P
,
Asea
A
. 
Radiation-induced effects and the immune system in cancer
.
Front Oncol
2012
;
2
:
191
.
39.
Ma
Y
,
Kepp
O
,
Ghiringhelli
F
,
Apetoh
L
,
Locher
C
,
Tesniere
A
, et al
Chemotherapy and radiotherapy:cryptic anti-cancer vaccines
.
Semin Immunol
2010
;
22
:
113
24
.
40.
Kwilas
AR
,
Donahue
RN
,
Bernstein
MB
,
Hodge
JW
. 
In the field: exploiting the untapped potential of immunogenic modulation by radiation in combination with immunotherapy for the treatment of cancer
.
Front Oncol
2012
;
2
:
104
.
41.
Lin
WW
,
Karin
M
. 
A cytokine-mediated link between innate immunity, inflammation, and cancer
.
J Clin Invest
2007
;
117
:
1175
83
.