One important mechanism tumors use to escape immune detection is by engaging the immune system's natural mechanisms to avoid self-recognition. Regulatory immune cells are a diverse group found in adaptive and innate immune cell subsets that prevent autoimmunity by suppressing self-recognizing T cells. Tumors hijack this natural mechanism to escape immune detection by secreting particular cytokines into its microenvironment to promote differentiation of many types of regulatory cells [7]. Tumor-induced immune suppression is the consequence of increased proportion of regulatory cells and coinciding reduction in the activity of effector T cells targeted towards the tumor [8]. The two main types of regulatory cells now known to be associated with this process are the CD4⁺CD25^hiFoxP3⁺ T cells (Tregs) and myeloid-derived suppressor cells (MDSCs) [9,10].

TGF-β, produced in abundance by many types of tumor cells, promotes differentiation of naïve CD4⁺ T cells into Tregs [11]. Increased Treg frequency is correlated with poor outcome and in several animal models were Tregs were selectively depleted, tumor regression was enhanced [8,12-14]. Tregs can inhibit antigen presenting cells (APCs) by inducing upregulation of inhibitory B7-H4 molecules or directly killing them through perforin and granzyme release. They engage CD80/86 on APCs with cytotoxic T lymphocyte antigen 4 (CTLA-4), leading to T cell anergy and death. Finally, they secrete immunosuppressive cytokines IL-10 and TGF-β to preserve and spread immunosuppression within the tumor microenvironment [15].

MDSC are a heterogeneous population of precursor myeloid cells that have the ability to cause immune suppression. In healthy individuals, the MDSC population is low as myeloid progenitors differentiate normally into mature myeloid cells, but under some pathological conditions maturation is arrested at various stages and the cells take on a suppressive capacity [16,17]. Tumor-derived factors, such as pro-inflammatory cytokines IL-6 and IL-1β, promote the formation of MDSCs resulting in their accumulation in the blood, lymphoid organs and tumor [18,19]. In cancer patients, the ratio of mature DCs to immature myeloid cells in the blood is inversely proportional to the stage of disease [20,21]. In humans, MDSC are identified by expression of CD33, CD11b and IL-4Rα. In mice, MDSCs universally express CD11b and GR1, for which there is no human homolog [9]. MDSCs can be divided into two groups based on nuclear morphology, the granulocytic MDSC are polymorphonuclear whereas the monocytic are mononuclear. These two subsets may have different functions in cancer [22]. MDSCs express various other surface markers including ICAM-1, CD80 and CD15, and exhibit great variability between individuals depending on the type of tumor.

MDSCs represent a significant hurdle to therapy because of their diverse immune suppression effects, both direct and indirect. They are able to directly inhibit CD8⁺ and CD4⁺ T cells in a cell-contact dependent manner through arginine and cysteine depletion, both amino acids are essential to T cell activation [23,24]. They can also inhibit T cell function though reactive oxygen species production [25]. Monocytic MDSC elevate iNOS, which may play a role in antigen-specific T cell suppression by increasing nitrosylation of MHC [25,26]. MDSC may also inhibit through antigen-independent mechanisms, it was recently shown that they reduce expression of L-selectin on naïve T cells, preventing their circulation through lymph nodes and tumors, thereby reducing the number of active T cells [27]. MDSC also indirectly cause suppression by inducing Tregs [28]. Interestingly, Treg induction may occur through CD40 expressed on MDSCs, and it was shown before this mechanism was discovered that blocking this interaction leads to reversal of CD4⁺ T cell anergy [29,30].

Suppressive subsets of many immune cell types have been found within the tumor microenvironment, including CD8⁺ T cells, NK cells and macrophages [31-34]. This diversity alludes to the intensity of suppression maintained within tumors, and the obstacles in raising an effective immune response for tumor elimination.

Besides inducing immune suppression, tumors have evolved other mechanisms to avoid immune detection. Firstly, tumors down-regulate expression of MHC class I and other proteins involved in antigen presentation [35-37]. Tumors can also decrease, or shed, expression of proteins that are recognized by the immune system, this concept is called immunoediting since it describes how the immune system directly impacts tumor malignancy [38,39]. Thirdly, tumors can by-pass death mechanisms by elevating expression levels of survival factors, such as anti-apoptotic proteins (survivin, BCL-X_L), metastatic proteins (VEGF, MMPs) and proliferation factors (EGFR, c-Myc). The transcription factor STAT3 is upregulated in a number of tumors and controls expression of some of these genes [40].

Tumors contain a heterogeneous population of cancer cells that are at various states of development, allowing it to evolve quickly in response to new stresses. Tumor cells adapt to immune recognition by down regulating expression of antigens, and can also adapt to chemotherapy by increasing expression of adenosine-triphosphate binding cassette (ABC) pumps to actively secrete intracellular drugs [41]. Ironically, a successful chemotherapy regiment can also increase the chance of reoccurrence since there is potential for a few highly resistant cells to survive treatment and seed a secondary malignancy. These cells are referred to as cancer stem cells, and have been identified as a phenotypically distinct subset in some human cancers, such as AML [42].

It was first recognized in the 1980s that low doses of cyclophosphamide (CPA) specifically inhibit a population of suppressor CD4⁺ T cells and enhance immune responses against antigens [45]. It was not until 2005 that Lutsiak et al. showed that CPA treatment specifically affects the CD4⁺CD25⁺ T cells (Tregs) [44]. They found that mice given a low dose of CPA had a reduced Treg population with attenuated suppressor function. The Tregs were shown to undergo apoptosis, but effector CD4⁺CD25^- and CD8⁺ T cell populations were not compromised. The effect was transitory, maximal Treg reduction was observed 4 days after treatment but returned to normal levels by day 10. This landmark study prompted investigation into the combined use of low dose CPA with peptide vaccines. Several reports of CPA combination therapy with various cancer vaccines have demonstrated the feasibility of this treatment in murine models [46-48]. Some have demonstrated that besides reducing Treg cells, CPA therapy can also enhance CD8⁺ T cell activation and memory development through induction of type 1 interferons [49,50].

In humans, low dose CPA treatment also selectively reduces the Treg population, but reports of its augmentation of cancer vaccines have been conflicting [51-53]. In fact, investigation into the effects of CPA and other chemotherapy treatments on the immune system has emphasized the inadequacy of murine models for cancer. Human cancers are heterogeneous in nature and are characterized by a high degree of immunosuppression. In contrast, the majority of murine tumor models rely on use of implanted cell lines that are clonotypic and after years of culture in vitro, have lost some of their initial immunosuppressive capabilities [54]. There are some models of spontaneously arising tumors, but the advantage to using implanted cell lines is their predictability and control. Therefore, while testing cancer immunotherapies in mice does provide some indication of their efficacy, but translation into humans is difficult.

A recent report by Tongu et al. looked at the combination of low dose CPA plus the anthracyline doxorubicin (DR) to therapeutically treat murine CT-26 colon carcinomas [55]. The combination of CPA (i.p.) + DR (i.t.) synergistically reduced tumor growth without vaccine therapy. The effect was shown to be T-cell dependent, since no effect was seen in nude mice, and tumor specific, it could not protect from a second challenge with a different tumor. The authors speculated that CPA treatment removed Treg suppression, enhanced CD8⁺ T cell function and that in combination with DR, which is known to induce immunogenic cell death, the tumors became immunogenic. CPA and DR were combined with a GM-CSF-secreting breast tumor cell vaccine in a small clinical study [56]. Both agents were delivered intravenously and various dose combinations were tested. In twenty-two patients who received CPA + DR and vaccination, serum levels of GM-CSF remained elevated and levels of HER2 antibodies were augmented. Clinical responses were not evaluated, but these results are promising and demonstrate how two chemotherapies with slightly different mechanisms can be combined for enhanced tumor rejection. One important caveat was the effect of CPA treatment was found to be highly dependent on dose, above 200 mg/m² it was immunosuppressive. This highlights the importance of dose selection when considering the immunomodulatory effects of chemotherapy.

Recently, metronomic dosing of CPA has emerged as a promising application of this drug for immune modulation. Continuous low dose CPA treatment was initially investigated for its anti-angiogenic effect since the rapidly dividing vascular intratumor endothelium are most susceptible to treatment [57,58]. It was then demonstrated that a continual low dose schedule of CPA (50–100 mg/day, p.o.) can also specifically reduce Tregs, as well as restore effector T cell and NK cell function [51]. An attractive feature of this approach is the convenience and low toxicity, which increases patient compliance.

Besides reducing Tregs, CPA treatment can also deplete B cells, augment activation and function of DCs, and skew the development of CD4⁺ T cells towards Th1 and Th17 during recovery after CPA induced lymphodepletion [59]. Interestingly, when Liu et al. evaluated the effects on the tumor infiltrating cell population in mice bearing tumors and treated with low dose CPA, they found a concurrent increase in the levels of myeloid derived suppressor cells (MDSCs) with the decreased levels of Tregs [60]. This work could suggest that the desirable effects of CPA treatment on the Treg population may be offset if they actually increase the level of an alternative suppressor cell, MDSCs. However, as MDSCs are loosely defined as a heterogeneous population of progenitor myeloid cells, this could merely be a reflection of enhanced lympho-proliferation following depletion.

A recent study has provided a hypothesis as to the preferential effects of low dose CPA treatment. Zhao et al. found that cells, such as Tregs, which have low levels of intracellular ATP have reduced capacity to detoxify internalized CPA [61]. Defining the mechanism through which CPA can selectively effect a particular population of cells will help in designing best chemotherapy-immunotherapy dosing schedules.

A study by Liu et al. evaluated the tumor infiltrating cell populations in mice bearing large or small tumors after low-dose CPA treatment [60]. They did confirm that CPA reduced the CD4⁺CD25⁺ population of Tregs, and also found an increased level of GR1⁺CD11b⁺ MDSCs, suggesting that in advanced tumors CPA treatment may enhance other suppressive cells. Gemcitabine (GEM) is a nucleoside analog that reportedly suppresses MDSCs specifically and has been used to reduce tumor growth in several murine models [62,63]. Like low dose CPA, GEM treatment is also transient [64]. In murine models, GEM combination with vaccine therapy significantly reduces regulatory T cells and enhances CD8⁺ T cell activation [65,66]. Knowing that MDSC can promote Treg differentiation, GEM could potentially reduce multiple suppressor cell types with a tumor both directly and indirectly.

5-fluorouracil (5-FU), another nucleoside analog, has also been reported to specifically suppress MDSCs. A comprehensive study by Vincent et al. evaluated several types of chemotherapies (GEM, CPA, DR, 5-FU, paclitaxel, oxaliplatin) on MDSC in EL4 thymoma tumor bearing mice [64]. They found that 5-FU specially induced apoptosis of GR⁺CD11b⁺ MDSC, both granulcytic and monocytic subsets were equally affected. 5-FU was more potent than GEM, and in combination with CPA significantly repressed tumor growth in a T-cell dependent manner.

5-FU and GEM have also been reported to increase immunological visibility of tumors by increasing expression of TAA on their surface. 5-FU or GEM were able to synergistically enhance antibody dependent cell-mediated cytotoxic (ADCC) mediated killing of colon cancer cell lines by cetuximab (a monoclonal antibody targeting epidermal growth receptor, EGFR) by increasing expression of EGFR on tumors [67]. Similar findings have been reported in other cancer models [68,69].

Paclitaxel (PX) therapy is common in most standard of care regimens used today because it is efficacious in many different types of cancer [70]. PX arrests cells in mitosis by preventing microtubule formation ultimately resulting in apoptosis. Recently, PX has also been shown to have stimulatory effects on the immune system, especially at lower doses than typically used for chemotherapy [71]. Conversely, standard dose PX treatment is broadly immunosuppressive and inhibits a number of cell types involved in tumor rejection: macrophages, effector T cells and NK cells [70]. The disparity between low and high dose effects has been noted with other chemotherapeutic drugs as well [72]. Interestingly, PX has been shown to be a ligand for TLR4 on murine DCs, which may be indicative of a direct effect on the immune system [73]. PX has also been shown to enhance activation of human DCs, but independently of TLR4 binding, and this effect is partially responsible for its immune-enhancing effect [74]. Investigations by the Gabrilovich group have discovered that PX treatment of cancer cells causes up-regulation of cation-independent mannose-6-phosphate receptor on the surface of tumor cells, which increases the efficiency of Granzyme B mediated cytotoxic killing (reviewed in [75]).

Low dose PX treatment has been combined with a number of vaccine types in murine models to effectively reduce tumor growth [76-78]. Used as metronomic therapy (continuous), low dose PX is a potent inhibitor of angiogenesis and specifically down-regulates expression of VEGF-receptor 2 on endothelial cells in a murine 4T1 breast cancer model [79]. In the clinic, low dose PX has not been tested in combination with cancer vaccines, but the anti-angiogenic effects of metronomic therapy have been confirmed [80-82].

The platinum based drugs, cisplatin and its less toxic analog carboplatin, are often co-administered with PX in standard chemotherapy treatments. Many clinical studies have consistently shown synergism between cisplatin or carboplatin and PX treatment [83,84]. The mechanisms contributing to the synergistic effect are unknown, but addition of a third drug (e.g., GEM or epirubicin) provides no additional benefit and may in fact interfere with primary treatment [85,86]. The mechanism underlying this combinatorial effect may have to do with the unique pathways used by platinum based drugs for import and export at the cellular level, due to the presence of the heavy metal atom [41]. It is less likely that tumors can simultaneously adapt to resisting two completely different drugs.

Carboplatin on its own has little reported evidence of an immunomodulatory effect, but recently an interesting study evaluated the effect of paclitaxel/carboplatin treatment on tumors and the immune system [87]. Preliminary studies in vitro showed the induction of apoptosis in SKOV3 ovarian cell lines by PX/carboplatin treatment. Treated cells were also more likely to be phagocytosed by dendritic cells which acquired activated phenotype (increased MHC II, CD80/86) and were subsequently able to prime CD8⁺ T cells in vitro, indicating the treatment induced immunologic death of the tumors. In the same study, blood samples were collected from 13 patients with ovarian cancer receiving primary therapy with PX/carboplatin before treatment then at regular intervals afterwards. Monitoring the levels of CD4⁺ T cell, CD8⁺ T cell and NK subsets revealed that prior to treatment patients were immunocompromised as evidenced by increased Tregs and decreased Th1, Tc1 and NK cells. A single course of PX/carboplatin treatment reversed the immunosuppression, peaking around 2 weeks after treatment before returning to pre-treatment levels. Therefore, it was suggested that 2 weeks following chemotherapy treatment would be the optimal time for secondary immunotherapy treatment, however this was not studied. This systematic study of the temporal effects on the immune system show how sensitive the timing of combination therapies can be, and how they could be planned for optimal efficacy.

PX/cisplatin treatment has been tested in combination with immunotherapy in a mouse study [88]. Lewis-lung carcinoma tumor bearing mice were treated with a standard course of PX/cisplatin followed by adoptive cell therapy with cytokine-induced killer cells (CIKs). The chemotherapy preconditioning resulted in enhanced tumor rejection which was accompanied by reduced intratumoral Tregs and increased homing of the CIKs to the tumor and spleen. Therefore, even at standard doses this chemotherapy regiment has the potential to enhance immunotherapy.

All types of cancer vaccines stand to benefit from chemotherapy combinations, and many have already been tested in clinical studies. Due to the complexity of these combinations (scheduling and dosing of both components, as well as cancer indication and stage), rarely are two studies the same which makes comparisons difficult. Table 2 summarizes the results of some relevant studies published recently. Gemcitabine, cyclophosphamide and dacarbazine (or temozolomide, which is metabolized to dacarbazine in vivo [89]) in particular have been used. Most trials do not include control arms and instead rely on historical controls. Outcomes have been varied, from no effect whatsoever [52,90] to indication of increase PFS or OS (compared to historical controls) [89,91,92]. Same have noted changes to immune response profile in terms of increased diversity in epitope recognition by T cells (i.e., epitope spreading) [93] or increased cellular and humoral responses [92,94]. Importantly, no studies have reported increased safety risks due to vaccine combinations with chemotherapy.

Somewhat counterintuitive are results from recent clinical studies showing that chemotherapy after vaccination may be a better treatment schedule than chemotherapy pre-treatment or concurrent treatment. Results of a clinical study published by Antonia et al. indicated that patients with extensive stage small cell lung cancer were actually more responsive to second-line chemotherapy treatment after vaccination with dendritic cells transduced with wild-type p53 via adenoviral vector [95]. More recently, the TG4010 viral vector encoding MUC1 and interleukin-2 was tested in a Phase II study in NSCLC patients [96]. The two arm study compared chemotherapy (cisplatin + vinorelbine) administered concurrently with vaccination or administered after vaccination. The results of the study indicated a positive outcome for both treatment arms, but number of evaluable patients was too low to conclude a preference for either schedule. For some types of cancer vaccines, this dosing schedule may be optimal because it primes the immune system before insult with chemotherapy. However, it may not be optimal for all treatment types or indications. Leffers et al. reported no benefits to secondary chemotherapy in ovarian cancer patients that had previously received a p53-synthetic long peptide (SLP)® vaccine, despite observing a significant benefit to NSCLC patients [97].

The most developed mAb of this type target the T cell surface protein CTLA-4. CTLA-4 is a negative regulator of effector T cell activity and is induced upon activation. CTLA-4 out-competes the co-stimulation molecule CD28 for binding B7 molecules on antigen presenting cells and instead delivers an inhibitory signal [111]. Therefore, CTLA-4 is used as a braking mechanism to control T cell responses. It is also used by Tregs for immune suppression; Tregs constitutively express CTLA-4 and induce suppression to DCs when binding through B7 [112]. The DCs in turn induce apoptosis and anergy in T cells [15]. Two fully human antibodies have been developed that target CTLA-4: tremelimumab (by Pfizer) and ipilimumab (by Bristol-Myers Squibb). Potentially, these antibodies could work on two fronts, first by blocking effector T cell CTLA-4 and thereby extending their survival, and second by blocking Treg CTLA-4 to prevent this mechanism of suppression. However, studies have demonstrated that in humans anti-CTLA-4 treatment targets effector T cells only [113,114]. Ipilimumab was recently approved by the FDA for second line treatment of advanced melanoma, but both have been tested in a number of clinical trials targeting various indications, such as melanoma, and have provided positive benefit [115]. Despite being able to induce tumor regression in 10% of patients, Pfizer halted the development of tremelimumab based on a dismal increase of overall survival of only 1 year in a recent phase III trial [116].

The results of a phase III clinical trial of ipilimumab, which supported FDA approval for this mAb, were presented at the American Society of Clinical Oncology (ASCO) meeting in 2010 [117]. The 1:1:3 randomized study containing 750 patients compared ipilimumab treatment alone to vaccination with GVAX (peptide vaccine targeting the melanoma TAA gp100) and to combination treatment with both ipilimumab and GVAX. Patients who received ipilimumab alone or in combination with GVAX were not significantly different and experienced a 10% increase in 2-year survival rates and increased overall survival compared to patients who received GVAX alone. Although these results were used to approve ipilimumab treatment in advanced melanoma patients, they are somewhat controversial because GVAX alone was used as the control arm, and not the common dacarbazine treatment used for advanced melanoma patients [118]. From a vaccine perspective the results are discouraging. Other peptide vaccines targeting gp100 have shown immunogenicity in other small clinical trials, demonstrating that it is possible to break tolerance towards this TAA, yet in this study no effect was attributed to GVAX treatment [119,120]. Pre-clinical research had also indicated that murine anti-CTLA-4 could in fact synergize with peptide cancer vaccines in mice [121-123]. The advanced stage of the patients in the ipilimumab study may have been detrimental to vaccine efficacy, and could show that although ipilimumab does provide some benefit to these patients, it cannot synergize with peptide vaccines in this cohort. Notably, the authors did not report if gp100-specific T cells were raised in any group so it is unclear if the patients immune systems responded at all to vaccination [124]. It is also possible that ipilimumab cannot synergize with cancer vaccines due to the isotype of this antibody. Ipilimumab, like the majority of mAb developed to date, is IgG1 isotype, which induces moderate complement activation and strongly induces phagocytosis by binding to Fc receptors. Although this isotype is ideal for mAb targeting tumor cells for destruction, ipilimumab targeting activated T cells may inadvertently enhance their elimination. In contrast, tremelimumab is IgG2 isotype, which is a poor activator of complement and weak binder of Fc making it an ideal subclass for blocking interactions. It would be interesting to compare both anti-CLTA4 mAb in combination with vaccination to see if tremelimumab can induce a greater synergistic effect than was observed with ipilimumab. Indeed, a better understanding of which antibody isotypes synergize best with vaccines is needed for rational design of future clinical trial protocols involving these two emerging immunotherapies for cancer.

PD-1 (programmed death 1) is a member of the CD28 superfamily, like CTLA-4, and is upregulated on T cells upon activation [112]. PD-1 is a suppressive regulator of T cell activity, ligation with its receptor results in inactivation and apoptosis. The receptors for PD-1, PD-1L and PD-2L, are normally expressed on self-cells to prevent autoimmunity, however PD-1L is upregulated by a number of tumors to quell anti-tumor T cell responses [125-127]. Accordingly, tumor infiltrating CD8⁺ and CD4⁺ T cells have been shown to have increased expression of PD-1 and are anergic [128,129]. Combined treatment of anti-PD-1 treatment and a GM-CSF secreting whole cell vaccine significantly prolonged mice challenged with B16 melanoma or with CT26 colon cancer, whereas monotherapy with either treatment had no effect [130]. The combined treatment was associated with increased antigen-specific CD8⁺ T cell infiltration of the tumor. Another study by Mongsbo et al. also found that monotherapy with anti-PD-1 is not as effective as anti-CTLA-4 monotherapy, but together may have an additive effect in prevention of MB49 murine bladder cancer [131]. The combined blockade of both PD-1 and CTLA-4 was found to synergize with a vaccine in treating of B16-B6 melanoma tumors [132]. The synergistic effect on tumor growth was mirrored with increased tumor infiltration of CD8⁺T cells expressing CTLA-4 and PD-1, presumably without treatment these cells would have been anergized. Dual blockade of PD-1 and CTLA-4 signaling eliminates two T cell suppressive mechanisms, therefore this is a logical combination that should increase longevity of T cells. A human PD-1 antibody (MDX-1106) was recently tested in a clinical trial in patients with several types of advanced cancer [133]. In the small phase I study, 39 patients were treated with antibody monotherapy and levels of PD-1 on circulating PBMCs as well as levels of PD-L1 on tumor cells were monitored. They found that tumor expression of PD-L1 may be indicative of responsiveness to MDX-1106 treatment, but overall clinical responses were low.

An alternate, or perhaps additional, mechanism for the synergistic effect of combined PD-1 and CTLA-4 blockade is by inhibition of MDSC suppression. One group reported that MDSCs isolated from mice bearing I8D ovarian tumors had elevated levels of both PD-1 and CTLA-4 [134]. When blocking antibodies were administered in vitro, the MDSCs had reduced arginase I activity; arginase I is a mechanism through which MDSCs attenuate T cell activation. In vivo treatment of tumor bearing mice reduced tumor burden and increased survival.

Complementary to T cell boosting strategies with anti-CTLA-4 or anti-PD-1 would be Treg inhibition using Treg-specific antibodies. Initially, antibodies towards the relatively non-specific CD25 surface marker found on Tregs were used in an attempt to target this T cell subset. However, anti-CD25 mAb clinical trials (daclizumab by Hoffman-LaRoche) have experienced mixed results; although this antibody does deplete Tregs, it also has an effect on activated effector T cells, which also upregulate CD25 [135]. The result is too devastating on the developing anti-tumor immune response unless timed correctly, which could present technical limitations for heterogenous human patients [135,136]. A new target is GITR (glucocorticoid induced TNF receptor), a co-receptor expressed in constitutively high amounts by Tregs and also increased on activated T effectors. Interestingly, while co-stimulation of CD3 and GITR results in proliferation of both Tregs and effector T cells, the expanded Tregs become functionally unresponsive while the effector T cells gain functional activity [137]. A single administration of the murine anti-GITR antibody DTA-1 eradicates or reduces tumor growth in different mouse models [138-140]. Mice challenged with B16 tumors and treated with DTA-1 developed strong antigen-specific T cell responses, and when combined with a melanoma vaccine, DTA-1 treatment enhanced primary and recall CD8⁺ T cell responses [141,142]. The mechanisms underlying DTA-1 treatment are truly two-fold, they can both enhance effector T cells and reduce Tregs. The mechanism through which they reduce Treg function is not clear, in one study Tregs isolated from tumors of DTA-1 treated mice did not have impaired suppressive function and yet relative numbers of Tregs were reduced compared to CD4 or CD8 T cells, suggesting depletion [138]. However, some studies have found no change in the absolute number of CD4⁺ T cells after DTA-1 treatment, and no death observed in vitro. Instead, it has been proposed that DTA-1 treatment reduces the lineage stability of Tregs through loss of FoxP3 expression [139]. This could mean that Tregs are converted to Th17 cells, these cells are known to be reciprocally regulated and instances of Treg conversion into Th17 have been documented [143]. It would be interesting to see if this was the case with DTA-1. In any case, the combined blockade of CTLA-4 and GITR with mAb was recently shown to synergistically reduce tumor formulation in two different murine tumor models, demonstrating that their respective effects on Tregs and effector T cells, in the end, work together [144].