Keywords

1 Introduction: Cancer Chemotherapy

Despite the fact that the modern era of cancer chemotherapy began following the World War II, the origins of cancer treatment are recorded in ancient documents.The medicinal utilization of different chemical substances can be dated back to the ancient Indian system of traditional medicine called Ayurveda (the ‘science of life’) that recommended various herbs and metals for patients with different diseases. The Ebers papyrus, the Edwin Smith papyrus and the Ramayana describe malignant diseases and their treatments. Dioscorides, in the first century A.D., compiled a listing of medicinal herbs and botanicals, including topical applications for treatment of tumors and carcinomata. Later, in the 9–10th century, Persian physician and alchemist, Muhammad ibn Zakariya al-Razi, known as Rhazes or Rasis, introduced sulfate, mercuric and arsenic salts, gold, copper, chalk, clay, coral, pearl and other natural agents for medical purposes. In the 11th century an Arabic physician, Ibn Sina, known in the West as Avicenna, used the arsenical therapy systemically, although it was found to be dangerous and received a little attention. Arsenical preparations known as Unguentum Aegypticum were used topically until the 16th century. Many consider the use of potassium arsenite to treat chronic myelogenous leukemia in 1865 by Lissauer as the first instance of effective chemotherapy for malignant disease (Timkin 1942; Papac 2001; DeVita and Chu 2008; Shurin et al. 2012).

The origins of effective chemotherapy for cancer date to World War I when mustard gas (sulfur mustard) was used. The blood and bone marrow findings in cases of mustard gas poisoning were described by Krumbhaar, a captain in the US Medical Corps, noted the development of profound leukopenia in those individuals who survived for several days (Krumbhaar 1919). This probably was the very first suggestion that mustard gas might have the effect on the immune cells and therapeutic possibilities for treating lymphomas (Hirsch 2006). The advent of World War II stimulated further research on chemical warfare. A series of analogues of sulfur mustards were produced as potential offensive agents. It was recognized that β-chlorethyl amines could exert cytotoxic actions on a variety of tissues, particularly related to the degree of their proliferative activity. Since the data were classified during the wartime, these findings were reported following World War II by Louis S. Goodman and Alfred Gilman from the Pharmacology Section, Medical Division of the Chemical Warfare Service of the US Army (Goodman et al. 1946).

Interestingly, major textbooks of cancer medicine have attributed the initial use of nitrogen mustard to findings of exposure to mustard gas following an explosion that occurred in the harbor of Bari, Italy, where a very few survivors developed a profound lymphoid and myeloid suppression (Hirsch 2006). The Bari attack occurred in December 1943, but according to several references, the initial clinical trial occurred at Yale in May 1942 (Papac 2001; Goodman et al. 1946; Gilman 1963). 66 patients with Hodgkin’s disease, lymphosarcoma, leukemia or other types of cancer were treated with either a nitrogen mustard called tris(β-chloroethyl)amine hydrochloride (HN3) or another nitrogen mustard, methyl-bis(β-chloroethyl)amine hydrochloride (HN2) in the early 1940s. The most important observation was that patients with Hodgkin’s disease often showed a prominent response to the nitrogen mustards (Joensuu 2008). Although this effect lasted only a few weeks, this was the first step to realization that cancer could be treated by pharmacological agents (Goodman et al. 1946; Goodman et al. 1984).

Following the introduction of nitrogen mustard into clinical practice, other types of alkylating agents were developed many of which are in the clinical practice today, most notably chlorambucil, melphalan, busufan and cyclophosphamide (Papac 2001). Interestingly, the clinical spectrum of their effectiveness is minimally changed from the initial report, although the toxicities differ. The use of an antifolate compound was reported by Sidney Farber et al. (1948), initiating the development of antimetabolite therapy. The authors used folate analogues aminopterin and amethopterin (now methotrexate) that are antagonists of folic acid, and reported in 1948 that when administered to children with acute lymphoblastic leukemia (ALL), these agents induced remission of the disease. This was followed by the use of purine and pyrimidine analogs. Subsequently, anti-tumor antibiotics, platinum compounds, imidizole compounds, vinca rosea alkaloids, taxols, camptothecin analogs and biologic agents have become a part of the therapeutic schedule for neoplastic diseases (Papac 2001; DeVita and Chu 2008).

2 Cancer Chemotherapy and Immunosuppression

Immunosuppression in cancer patients is a common clinical condition that is induced by the cancer itself and high dose chemotherapeutic agents. Many commonly used chemotherapeutic regimens decrease white blood cell counts in addition to their intended effects (Wijayahadi et al. 2007; Rasmussen and Arvin 1982). Many chemotherapeutic agents are myelosuppressive by inhibiting the bone marrow production of blood cells and resulting in a diminished absolute blood count. The result of myelosuppression is evident during the chemotherapy-induced neutropenia (Kim and Demetri 1996). Immune toxicity, in particular neutropenia, can lead to a considerable morbidity and mortality by predisposing patients with cancer to life-threatening bacterial infections and sepsis. Bacterial infection is one of the most common consequences of neutropenia in cancer patients and is directly related to the degree and duration of neutropenia: the more intensive the chemotherapy regimen, the greater the neutropenia and the higher the risk of infection with accompanying fever and potential sepsis (Mackall 2000; Steele 2002).

In addition to myelosuppression, chemotherapy-induced immunosuppression (Goldman 2000) is commonly detected in treated patients and reflects a reduced ability to mount an immune response due to either decreased lymphocyte numbers or the absence of antigen-specific lymphocyte populations (Steele 2002). For instance, treatment of cancer patients with 2-chlorodeoxyadenosine resulted in severe depletion of both CD4+ and CD8+ T cells (Dmoszynska et al. 1999). While CD8+ T cells recover reasonably fast, CD4+ T cells may recover to only 35 % of pre-therapy levels at three months post-chemotherapy (Mackall et al. 1997). However, rapid up-regulation of the number of CD8+ T cell following chemotherapy does not reflect recovery of their functionality, since low expression of CD28 molecules may be observed during the first year after the cessation of chemotherapy (Mackall 2000). B lymphocytes are also sensitive to dose-intensive chemotherapy, which results in the suppression of immunoglobulin levels: almost 90 % of treated patients were reported to have low serum antibody concentrations following chemotherapy for at least one year post-chemotherapy (Mackall 2000; Mustafa et al. 1998). NK cell numbers are also decreased following chemotherapy, but rebound in numbers and cytotoxic activity relatively quickly as compared to other lymphocyte populations (Alanko et al. 1995). Based on the established role for NK cells in anti-infectious and anti-tumor responses and in regulating an induction of adaptive immunity, the chemotherapy-induced depletion of NK cells is an important component of the general immunosuppression in cancer therapy. Similarly, depletion of NKT and γδT cells can also be involved in diminished immunity induced by high dose chemotherapy (Zocchi and Poggi 2004).

Chemotherapeutic agents affect immune cells by different mechanisms. For instance, paclitaxel, docetaxel and doxorubicin have been shown to achieve their immunosuppressive effects through inhibition of the function or proliferation of immune cells (Ferrari et al. 2003; Nakashima et al. 2005; Ferraro et al. 2000). Human NK cells showed a dose-dependent decrease in cytotoxicity against tumor cell lines upon an incubation with clinically relevant concentrations of paclitaxel, as did peripheral blood mononuclear cells (PBMC), MHC-unrestricted T cells and cytotoxic T cells (CTL) (Chuang et al. 1993; Chuang et al. 1994). In contrast to microtubule stabilizing drugs, microtubule destabilizing drugs, such as colchicines and vinblastine, in addition to inducing apoptosis in dividing cells, functionally repress CTL by inhibiting microtubule-organizing centre translocation (Kupfer and Dennert 1984) with the subsequent killing of target cells (Wolberg et al. 1984). Colchicine also reduces the stability of the immunological synapse (Bunnell et al. 2001) and the continuous recycling of T cell receptors at the immunological synapse (Das et al. 2004). Anthracyclines doxorubicin and daunorubicin could induce apoptosis in both mitogen-activated and non-activated human peripheral blood lymphocytes and induce a marked depletion of B and T cells after the injection in mice (Ferraro et al. 2000). Thus, conventional high-dose chemotherapy is often associated with a profound inhibition of immune cell generation and activity, leading to an acute or prolonged immunosuppression.

3 Cancer Chemotherapy and Immunostimulation

Several recent studies demonstrated that certain chemotherapeutic regimens were able to increase the efficacy of anti-cancer immunotherapy (Nowak et al. 2006; Lake and Robinson 2005), allowing to speculate that the effect of chemotherapy on the immune response depends on the dose of chemotherapeutic agents, their mechanisms of action on immune cells, the type of malignancy and the level of tumor-induced immunosuppression or tolerance (Shurin et al. 2012). For instance, cyclophosphamide could enhance vaccine-induced tumor-specific antibody responses in patients with breast cancer, while in higher doses it suppressed immunity (Emens et al. 2009). Analysis of immune alterations in breast cancer patients undergoing chemotherapy indicated that patients responded to the treatment with paclitaxel or docetaxel dysplayed an increase in serum levels of IFN-γ, IL-2, IL-6 and GM-CSF as well as an enhancement of PBMC, NK and LAK cell activity (Tsavaris et al. 2002). Experimental data revealed that the doxorubicin treatment could enhance tumor antigen-specific proliferation of CD8 T cells in tumor-draining lymph nodes and promote tumor infiltration by activated IFN-γ-producing CD8+ T cells (Mattarollo et al. 2011). Furthermore, macrophages isolated from doxorubicin or cyclophosphamide treated mice displayed increased cytotoxicity for tumor cells (Stoychkov et al. 1979), confirming data on enhanced activity and tumoricidal potential of macrophage and NK cells treated with doxorubicin and cyclophosphamide in vitro (Mihich and Ehrke 2000; Ewens et al. 2006; Ujhazy et al. 2003).

Tumor cell death induced by antineoplastic chemotherapy might be accompanied by the release of “danger” signals assisting engulfment of tumor antigens (Nowak et al. 2003). For instance, chemotherapy with paclitaxel up-regulated CD8+ T cell function, which was attributed to drug-induced apoptosis of malignant cells and accessibility of tumor antigens (Emens and Jaffee 2005; Coleman et al. 2005). Increased immunogenicity of tumor cells due to their apoptosis was also reported for doxorubicin, 5-fluorouracil and gemcitabine (Ehrke et al. 2000; Casares et al. 2005). Anthracyclins, for instance, have been shown to induce the rapid translocation of calreticulin to the surface of dying cells which promoted phagocytosis of cancer cells by DC (Obeid et al. 2007). HMGB 1 alarmin protein secreted by dying tumor cells interacts with TLR4 on DC to cause immunogenic cross-presentation (Apetoh et al. 2007). Also, many chemotherapeutic agents induce the production of uric acid caused by tumor cell death. Uric acid, at least supposedly in crystal form, may activate DC (Melief 2008). Thus, the augmented efficacy of successful combinations of conventional and moderate-dose cytotoxic chemotherapy and DC vaccines was in part due to the accessibility of antigens, expression of calreticulin and release of alarmin and uric acid from dying tumor cells; i.e., this success relies on the cytotoxic effect of chemotherapy on tumor cells (Melief 2008; Liu et al. 2007; Yu et al. 2003).

Interestingly, certain chemotherapeutic agents such as cisplatin, doxorubicin, mitomycin C, fluorouracil and camptothecin, which exert direct cytotoxic effects, also exhibit an indirect cytotoxic activity by “preparing” tumor cells to their distraction by NK cells or cytotoxic T lymphocytes using Fas or TRAIL-dependent pathways (Lacour et al. 2001; Micheau et al. 1997). For example, human colon carcinoma cell lines treated with 5-fluorouracil acquired CD95 and ICAM1 expressions and became more sensitive to lysis by CTL (Tanaka et al. 2002).

Chemotherapy-induced lymphodepletion can also be beneficial because it may eliminate immunosuppressive regulatory T cells (Treg) and poorly functional anti-tumor T cells (Nowak et al. 2006). In the late 60s, a few reports demonstrated that cyclophosphamide might improve the anti-tumor activity of allogeneic tumor-specific splenocytes and prolong the animal survival in leukemia models (Glynn et al. 1969; Fefer 1969). The authors speculated that this effect was due to the immunosuppression induced by cyclophosphamide in the allogeneic system. At the same time, Mihich et al. (1969) demonstrated that other chemotherapeutic agents might up-regulate the formation of anti-tumor lymphocytes and synergized in the anti-tumor activity with serum or spleen cells obtained from mice pretreated with irradiated tumor cells. These findings were confirmed when it was reported that cyclophosphamide in combination with tumor-specific lymphocytes prolonged animal survival in different models of murine leukemia (Vadlamudi et al. 1971; Fass and Fefer 1972; Berenson et al. 1975). Data showing that tumor-bearing animals have T lymphocytes that can inhibit the immune response to tumor antigens (Fujimoto et al. 1976; Takei et al. 1977) allowed hypothesizing that cyclophosphamide might eliminate tumor-induced T suppressor cells and thus potentiate the effect of immunotherapy (North 1982; Mastrangelo et al. 1986), which has been eventually proven in experimental setting (Liu et al. 2007; Lutsiak et al. 2005; Ghiringhelli et al. 2004) and clinical trials (Berd et al. 1986; Berd and Mastrangelo 1988). New data revealed that low-dose cyclophosphamide rendered Treg cells susceptible to apoptosis by down-regulating Bcl-xL and CTLA-4 in these cells and decreased production of IL-2 by effector CD4 T cells (Sharabi et al. 2010).

In addition to elimination and suppression of Treg cells in the tumor environment, chemotherapeutic agents may also affect other subpopulations of immune regulators. Gemcitabine, an antimetabolite agent, has been reported to deplete myeloid-derived suppressor cells (MDSC) (Suzuki et al. 2005; Ko et al. 2007). Docetaxel can polarize MDSC toward the macrophage M1-like phenotype, resulting in the expression of CCR7 (M1 marker) in 40 % of MDSC in vivo and in vitro (Kodumudi et al. 2010). Therefore, alleviation of Treg- and MDSC-mediated immunosuppression and elicitation of immunogenic tumor cell death are involved in known immunopotentiating activity of several chemotherapeutic regimens (Chan and Yang 2000; Javeed et al. 2009). In this context, the administration of selected chemotherapy agents before immunotherapy can have multiple effects, eliminating both immune suppressors and regulators, as well as facilitating tumor antigen cross-presentation (Liao et al. 2007; Emens 2006; Sinkovics and Horvath 2006). Additional mechanisms through which conventional chemotherapeutics may affect the immune system have been recently reviewed (Galluzzi et al. 2012).

4 Chemoimmunomodulation: The Effect of Extra low Noncytotoxic Noncytostatic Doses of Chemotherapeutic Agents on the Immune System

Recently a new approach to immune cell modulation by chemotherapeutic agents has been introduced and termed ‘chemomodulation’ or ‘chemoimmunomodulation’, indicating that this approach aims primarily at the alteration of the tumor immunoenvironment rather than at direct elimination of malignant cells. It was based on new findings demonstrated that different chemotherapeutic agents displayed an unpredicted ability to regulate signal transduction pathways in immune cells without inhibiting cell cycle or inducing cell death when used in ultra low concentrations. For instance, it has been shown that several chemotherapeutic drugs from different groups in low nanomolar concentrations were able to regulate activity of classical and non-classical members of the small Rho GTPase family of proteins, including Rac, RhoA, Cdc42 and RhoE in DC (Shurin et al. 2008). These results suggested for the first time that different classes of chemotherapeutic drugs in ultra low concentrations may directly activate signaling pathways in immune cells and alter their function or differentiation without inducing their apoptosis. This possibility has been tested both in vitro and in vivo by several groups.

For example, it has been reported that several chemotherapeutic drugs in low nontoxic concentrations affect DC maturation in vitro. The authors treated murine bone marrow-derived DC with different agents and revealed that paclitaxel, methotrexate and doxorubicin in ultra low noncytotoxic noncytostatic concentrations significantly up-regulated expression of phenotypic markers of DC maturation (such as CD80, CD86, CD40 or MHC class II), while vinblastin, vincristin, 5-azacytidine and bleomycin did not significantly alter the expression of these molecules on DC (Shurin et al. 2009a). Furthermore, the authors found that doxorubicin significantly increased both spontaneous and induced migration of DC toward CCL3 (MIP-1α) and CCL19 (MIP-3β), and that paclitaxel, methotrexate and doxorubicin increased antigen presentation by DC. Interestingly, all these chemotherapeutics appeared to be the strongest inducers of IL-12 expression and marked inhibitors of IL-10 production by DC (Shurin et al. 2009a, b).

Analysis of chemomodulation phenomenon in human DC prepared from CD14+ monocytes obtained from PBMC of healthy volunteers revealed that several agents (including paclitaxel, doxorubicin and methotrexate) in noncytotoxic concentrations up-regulated phenotypic maturation of DC and stimulate APC function of human DC (Kaneno et al. 2009). It has been also shown that paclitaxel, doxorubicin and methotrexate protected APC function of human DC from inhibition induced by primary HNSCC cells in an allogeneic MLR assay. Furthermore, the authors demonstrated that human DC loaded with tumor lysates prepared from paclitaxel-pretreated colon carcinoma cells, induced CTL with significantly higher cytotoxic activity against tumors than DC loaded with lysates from control (nontreated) tumor cells (Kaneno et al. 2011). Interestingly, CTL induced by DC loaded with nontreated tumor cells displayed significantly higher cytotoxicity against paclitaxel-pretreated tumor cells than against nontreated tumor cells. Altogether, these results suggest that the presence of chemotherapeutic agents in nontoxic concentrations in the tumor microenvironment may augment the development of DC-mediated antitumor immune responses in vivo (Shurin et al. 2012).

Our new data revealed another interesting phenomenon associated with chemoimmunomodulation in the tumor microenvironment. We and others have reported that the development of lung cancer in mice is associated with a rapid accumulation of regulatory DC (regDC) in the lung and lymphoid tissues (Liu et al. 2009; Shurin et al. 2011). Using this model, and several in vitro and in vivo approaches, we demonstrated that paclitaxel in ultra-low noncytotoxic doses abrogated the polarization of cDC into immunosuppressive protumorigenic regDC through the small Rho GTPase signaling and increased the anti-tumor potential of DC vaccine in lung cancer-bearing animals (Zhong et al., submitted). These findings offer novel insights into the therapeutic efficacy and immunomodulatory capacity of chemotherapeutic agents used in noncytotoxic doses.

Besides DC, noncytotoxic paclitaxel application was also studied with regards to its effects on the major myeloid cell subset with immunosuppressive functions, myeloid-derived suppressor cells. This extremely heterogeneous population of immature myeloid cells representing precursors of granulocytes, macrophages, and DC (Gabrilovich et al. 2012; Condamine and Gabrilovich 2011; Peranzoni et al. 2010) has been described to inhibit anti-tumor T cell reactions via different mechanisms including an increased production of nitric oxide (NO) and reactive oxygen species, as well as an enhanced arginase (ARG)-1 activity (Gabrilovich et al. 2012; Condamine and Gabrilovich 2011; Ostrand-Rosenberg 2010; Marigo et al. 2008; Raber et al. 2012). The administration of paclitaxel in noncytotoxic doses into C57BL/6 mice resulted in the decrease in the frequencies of Gr1+CD11b+ immature myeloid cells (IMC) in the spleen and lymph nodes (Sevko et al. 2012). This subset is considered as an MDSC counterpart in healthy mice (Gabrilovich et al. 2012; Gabrilovich and Nagaraj 2009). The observed changes might be due to an augmented differentiation of IMC in the bone marrow into mature granulocytes, macrophages and DC (Gabrilovich and Nagaraj 2009; Naiditch et al. 2011), resulting in reduced migration of IMC to the secondary lymphoid tissue. Interestingly, the levels of NO production and ARG-1 expression in Gr1+CD11b+ IMC in treated mice were not significantly different from those in untreated animals indicating that, in the absence of a growing tumor, IMC are not able to exert the immunosuppressive activity (Youn and Gabrilovich 2010).

To investigate the effect of noncytotoxic paclitaxel in tumor-bearing mice, we applied ret transgenic mouse model that, in contrast to transplantation models, closely resemble human skin malignant melanoma with regard to etiology, genetic alterations, histopathology and clinical development (Kato et al. 1998; Umansky et al. 2008; Lengagne et al. 2011). In this model, mouse melanocytes express the human ret transgene controlled by the mouse metallothionein I promoter-enhancer (Kato et al. 1998). A constant activation of Ret kinase, a member of the receptor tyrosine kinase family (Eng 1999) resulted in the stimulation of other downstream kinases leading to the spontaneous development of skin melanoma with metastases in lymph nodes, lungs, liver, brain, and the bone marrow (Kato et al. 1998; Umansky et al. 2008; Lengagne et al. 2011). Such metastatic profile shows a high similarity to metastasis observed in malignant melanoma patients (Houghton and Polsky 2002). It has been demonstrated by us and others that the melanoma progression in these transgenic mice was characterized by a strong accumulation of highly immunosuppressive MDSC and immature, non-functional DC in skin tumors, metastatic lymph nodes, spleen and bone marrow (Lengagne et al. 2011; Meyer et al. 2011; Zhao et al. 2009).

Upon the paclitaxel administration in tumor-bearing ret transgenic mice, we detected a significant reduction of MDSC frequencies in melanoma lesions (Sevko et al. 2013, in press). Furthermore, we revealed a reduction in frequencies of MDSC producing NO. Importantly, tumor-infiltrating MDSC also displayed a significantly lower ability to suppress the proliferation of stimulated normal T cells indicating their decreased immunosuppressive function after paclitaxel application.

Analysis of signaling pathways in MDSC affected by the paclitaxel-based chemoimmunomodulation revealed a strong down-regulation of the phosphorylation of p38 MAPK in these cells located in skin tumors, the bone marrow and spleen as compared to the untreated group (Sevko et al. 2013, in press). Signaling through p38 MAPK in myeloid cells was previously reported to induce a reduction of DC immunostimulatory capacities in ret transgenic melanoma-bearing mice (Zhao et al. 2009) and to stimulate the production of chronic inflammatory mediators TNF-α, IL-1β, IL-10, TGF-β and IL-6 (Perdiguero et al. 2011; Kim et al. 2008; Ho et al. 2004; Baldassare et al. 1999; Ajibade et al. 2012). Indeed, we have found that a down-regulation of the p38 MAPK signaling in MDSC was strongly associated with reduction of TNF-α production by these cells and decreased levels of the abovementioned mediators of chronic inflammation (Sevko et al. 2013, in press). An autocrine regulation of p38 MAPK activation could also be at work since TNF-α and IL-1β were found to activate p38 MAPK signaling (Zhou et al. 2006; Gao and Bing 2011), stimulating thereby the IL-6 production, an activation of inducible NO synthase (Chae et al. 2001; Chen et al. 1999) and the MDSC accumulation (Lokuta and Huttenlocher 2005). In agreement with these publications, our findings demonstrated that paclitaxel-mediated abrogation of p38 MAPK activation in MDSC resulted in the decrease in their numbers in tumors. Since p38 MAPK is also activated by secreted S100A8/A9 in autocrine manner (Hermani et al. 2006; Gebhardt et al. 2006; Sunahori et al. 2006), we evaluated the intracellular expression of S100A9 in MDSC. A significant reduction in the number of tumor-infiltrating MDSCs expressing S100A9 was found after administration of paclitaxel (Sevko et al. 2013, in press) additionally highlighting the importance of the stimulation of p38 MAPK-S100A8/A9 pathways in MDSC-mediated immunosuppression.

Reduced amounts of tumor-infiltrating MDSC could be also due to the inhibition of MDSC generation, induction of MDSC apoptosis or enhanced MDSC differentiation. We found that the generation of MDSC from the bone marrow precursors in vitro was not altered paclitaxel treatment (Michels et al. 2012). Under these conditions, the rate of apoptosis in MDSC cultures increased. However, our results revealed that paclitaxel in ultra-low concentrations can promote the MDSC differentiation into functional conventional DC (Michels et al. 2012). In addition, we and others provided evidence that the stimulation of MDSC differentiation into DC by paclitaxel in vitro is not mediated by TLR4 signaling (Kodumudi et al. 2010; Michels et al. 2012).

Taking into account a critical role of MDSC and chronic inflammation in tumor progression, we also assessed the anti-tumor potential of non-toxic application of paclitaxel and revealed a significant delay in tumor development indicated by a prolonged survival of treated animals (Sevko et al., 2013, in press). Investigating the mechanism of this effect, we found an increased numbers of tumor infiltrating CD8+ T lymphocytes. Importantly, the expression of TCR ζ-chain in CD8+ cells, which was usually strikingly reduced in tumor-bearing hosts (Meyer et al. 2011; Baniyash 2006), was found to be increased upon the treatment. Importantly, in the metastatic lymph nodes, frequencies of CD8+ T cells specific for melanoma associated antigen tyrosinase related protein (TRP)-2 were also significantly higher, although the total numbers of CD8+ T cells were not changed. Along this line, we have also reported that in healthy C57BL/6 mice, administration of ultra-low dose paclitaxel strongly increased the frequencies of TRP-2 specific spleen T cells upon the vaccination with the respective peptide (Sevko et al. 2012). These results are in agreement with data reported by us and others that the reduction of MDSC numbers and/or immunosuppressive functions resulted in the restoration of effector functions of tumor-specific CD8+ T cells (Meyer et al. 2011; Apetoh et al. 2011). Furthermore, a selective depletion of CD8+ T cells led to the complete abrogation of the beneficial anti-tumor activity of paclitaxel, underlying a critical role of CD8+ T cells in the mechanism of its therapeutic efficiency.

Analyzing effects of paclitaxel chemoimmunomodulation on various lymphoid cell subsets in healthy C57BL/6 mice, we found a significant reduction in numbers of Treg in treated animals, which displayed also an enhanced TRP-2 specific T cell response after peptide vaccination (Sevko et al. 2012). Elimination of Treg has been reported to be essential for the development of immune responses upon vaccination with TRP-2 antigen in mice (Grauer et al. 2008). Moreover, immunopotentiating effects of ultra-low-dose paclitaxel were demonstrated to be associated with a substantial elevation in frequencies of CD8+ and CD4+ T cells, NK, and NKT cells as well as enhanced IFN-γ production in NK cells. Interestingly, it has been reported that paclitaxel in low concentrations stimulates the cytotoxicity of human NK cells against breast carcinoma cells in vitro by increasing the concentration of perforin both at mRNA and protein levels (Kubo et al. 2005).

5 Conclusions

It is well-documented that chemotherapeutic agents can significantly modulate numbers and functions of various subsets of immune cells. The direction of this modulation has been found to critically depend on the doses of one and the same agent. Conventional chemotherapy, which is based on high, maximum tolerated doses (MTD), is widely used as an anti-cancer therapy to eliminate quickly proliferating tumor cells. However, such approach is strongly associated with severe side effects including besides general toxicity also myelotoxicity and inhibition of immune effector cells linked with immunosuppression in tumor bearing hosts. It has been recently reported that the application of chemotherapeutics in moderately low doses (20-33 % MTD) can induce immunogenic death of cancer cells involving the cell surface alterations and the release of soluble immunogenic factors from dying tumor cells. Although this approach involves cytotoxic effect on tumor cells, it induces rather stimulation of immune responses than immunosuppression.

Completely different strategy of therapy with chemotherapeutic agents in ultra-low, non-cytotoxic and non-cytostatic doses (1/20–1/40 of MTD) has been recently proposed and termed chemoimmunomodulation. In contrast to “classical” treatments, this novel approach induces no inhibition of proliferation of tumor cells, hematopoietic cells or immune cells in vitro, but supports the development of anti-tumor responses by altering activity of immune cells. In particular, chemomodulation causes a significant reduction of tumor-associated immunosuppression due to 1) the abrogation of conventional DC polarization into immunosuppressive protumorigenic regDC; 2) the decrease in MDSC numbers and functions in the tumor lesions; 3) the stimulation of anti-tumor activities of conventional DC; and 4) the modulation of intratumoral cytokine network. This suggests that administration of chemotherapeutics in ultra low, non-cytotoxic doses can be considered as a novel neoadjuvant approach for decreasing the protumorigenic, immunosuppressive potential of the tumor microenvironment and increasing the efficacy of various anti-cancer treatments including immunotherapy.