Introduction

Combination between immunotherapy and chemotherapy has long been viewed as incompatible as chemotherapy, especially at high doses meant to increase the anti-tumor efficacy, has induced immunosuppression. Possible mechanisms of immune suppression by chemotherapy are induction of lymphopenia, immunosuppressive cytokines, immune tolerance by high doses of antigens released by the dying tumor cells, and inhibition of immune effector cell function [3, 90, 94, 155]. However, in the 1960s, Mihich already demonstrated in murine leukemia model that the curative effects of chemotherapy are due to the induction of immune response directed against the tumor cells [91–93]. Immunoaugmentation has also been shown in later studies following chemotherapy with some drugs at low doses [3, 47, 90, 94, 155]. Treatment of cytotoxic T lymphocytes (CTL) with certain chemotherapeutic drugs enhanced their capacity to lyse Epstein Barr virus (EBV)-transformed lymphocytes, whereas other drugs showed inhibitory activities [86]. Experimental evidence has shown that direct effects of chemotherapy on tumor and host environment, which are discussed in detail below, may counteract its immunosuppressive effects, leading to enhancement of anti-tumor immune response.

We have reviewed here experimental and clinical approaches to combining active specific immunotherapy, or adoptive antibody or cellular immunotherapy with chemotherapy in the treatment of cancer. Most of the previous review articles did not cover combination of adoptive antibody or cellular immunotherapy with chemotherapy in pre-clinical and clinical studies and, in contrast to our article, none (including also articles on combined active specific immunotherapy and chemotherapy) describe experimental details, which are important to better understand differences in the results obtained with similar combination therapies by different investigators [3, 18, 21, 32, 45, 48, 58, 73–75, 77, 83, 90, 95, 96, 101, 117, 123, 132, 137, 143, 144]. The experimental approaches in this review include only studies which are carefully controlled to demonstrate that a combination of both therapies is superior to the use of either therapy alone. Clinical trials with combination therapies are also included in this review although they were not randomized and have not yet reached phase III. This review article does not include studies in which non-specific immune modulators such as cytokines were combined with chemotherapeutic agents. These studies have recently been reviewed by Zitvogel et al. [155].

Pre-clinical and clinical studies of combined mAb IT and CT

MAb therapy, which has long been viewed as unsuccessful, has been greatly rejuvenated by its combination with chemotherapeutics. Naked and radiolabelled mAb in combination with chemotherapeutics, or mAb linked to drugs have been used for the treatment of various malignancies in mice and cancer patients (Tables 1, 2). In mice, the anti-tumor effects of these combination therapies were significantly greater compared to either therapy alone. Of note, in each of the experimental studies (Table 1), significant effects were seen against established tumors. In cancer patients, impressive clinical responses were reported with combination therapies targeting specifically CD33 in leukemias, CD20 in B cell lymphomas, HER-2 in breast carcinomas, and epidermal growth factor receptor (EGF-R) in head and neck carcinomas (Table 2). The possible mechanisms underlying therapeutic effects of this combination therapy are discussed below.

Table 1 Effect of combined mAb IT and CT on tumor growth and/or survival in mice
Full size table
Table 2 Clinical trials of combined mAb IT and CT
Full size table

Pre-clinical and clinical studies of combined active specific IT and CT

The possible mechanisms underlying synergistic effects of active specific IT and CT are quite well understood, but selection of optimal dose of chemotherapy and timing of administration of the two therapies remain a challenge (see below). Various forms of vaccine delivery, such as irradiated tumor cells, tumor cell extract, tumor proteins or antigens expressed in naked plasmids or viral vectors have been used in combination with chemotherapeutics in several tumor models in mice (Table 3). In some of these studies, combination therapy was able to inhibit growth of established tumors [2, 19, 28, 46, 49, 61, 65, 67, 69, 70, 72, 76, 109, 130, 140, 141, 153, 154]. In clinical trials in which combined vaccine/chemotherapy was compared with either therapy or IT alone, promising clinical responses have been reported. Thus, the number of glioblastoma patients demonstrating 2 year disease-free survival was increased after treatment with dendritic cells (DC) loaded with tumor peptides or lysates, followed by chemotherapy with Temozolomide and BCNU as compared to treatment with either therapy alone [148] (Table 4). Clinical response rates of prostate cancer patients were increased following immunization with tumor peptides in combination with chemotherapy (Estramustine phosphate) as compared to IT alone [103] (Table 4). In another trial in prostate cancer patients, median time to tumor progression was increased after combination therapy (recombinant vaccinia virus expressing prostate specific antigen, followed by doxorubicin), compared to IT alone [6] (Table 4).

Table 3 Effect of combined active specific IT and CT on tumor growth and/or survival in experimental animals
Full size table
Table 4 Clinical trials of combined active specific IT and CT
Full size table

Pre-clinical and clinical studies of adoptive lymphocyte or active specific IT in combination with lymphodepletion by CT

The combination of adoptive lymphocyte IT with lymphodepletion by CT in patients with refractory metastatic (stage IV) melanoma has resulted in remarkable clinical response rates of approximately 50% [44] (Table 5), whereas clinical response rates with various CTs or adoptive lymphocyte transfer alone usually ranged between 10 and 34% in historical control patients [128, 129]. Various mechanisms may underly the synergistic effects of lymphodepletion on adoptive lymphocyte IT (see below). Lymphodepletion also has been combined with both active specific and adoptive lymphocyte IT in six metastatic melanoma patients. Thus, each patient received all three therapies [5] (Table 5). Only one of six patients showed a partial response to this combination therapy and it is unclear which form of therapy this response may be attributed to.

Table 5 Pre-clinical and clinical studies of combined adoptive lymphocyte or active specific IT and CT
Full size table

Treatment of well established tumors in mice with a chemotherapeutic drug, followed by adoptive lymphocyte IT resulted in tumor regression [152] (Table 5). Interestingly, synergism between the two therapies was dependent on the tumor microenvironment (see below).

Discussion and conclusions

The major possible direct effects of chemotherapy on tumor and/or host environment, which provide a rationale for combining CT with active and/or adoptive cellular IT, are:

Induction of tumor cell death

In the early studies by Bonmassar, it was shown that various types of immunogenic modification of tumor cells might occur in tumor-bearing hosts after treatment with drugs in vivo [15, 52, 68, 102, 125]. The molecular mechanism of drug-mediated immunogenic changes could be related to somatic mutations [51, 56]. Notably, chemotherapy of tumor-bearing mice and breast cancer patients was followed by induction of immune responses to the tumors [66, 97, 104, 109, 125]. Induction of necrosis and/or apoptosis in tumor cells in vitro has frequently been shown to increase their immunogenicity in vivo [3, 20, 54, 78, 90, 94, 107, 124]. Most likely, necrotic or apoptotic tumor cells induced by chemotherapy were phagocytosed by antigen-presenting cells (APC), presented to immune lymphocytes, followed by the stimulation of an anti-tumor responses in the lymphocytes [3, 55, 79, 90, 94]. Through induction of cell death by chemotherapeutics, a tumor could become its own cellular vaccine by crosspresentation of the apoptotic cells to APC, or induction of pro-inflammatory mediators such as heat shock proteins or interleukin (IL)-6, followed by crosspriming of immune effector cells [80, 145]. Although different chemotherapeutic agents may kill tumor cells through an apparently homogeneous apoptotic pathway, they may differ in the mechanism underlying the induction of immunogenic cell death. Thus, the chemotherapeutic agent anthracyclin induced an immune response to tumor cells only when apoptosis was preceded by translocation of calreticulin to the plasma membrane. Blockade or knock-down of calreticulin suppressed the phagocytosis of anthracyclin-treated tumor cells by dendritic cells and abolished their immunogenicity in mice [3, 90, 106, 107].

In principle, any therapy that delivers higher levels of cross-presented tumor antigens to the draining lymph nodes could synergize with immunotherapy. Thus, anti-tumor immunity induced by apoptotic tumor cells following chemotherapy can be boosted by active specific immunotherapy (see Tables 3, 4).

Elimination of regulatory T (Treg) cells

Cyclophosphamide (Cy) may down regulate the activity of Treg, especially when used in low doses [3, 82, 84, 90, 94, 99], whereas high doses may have direct tumor-cytotoxic effects [97–99]. Cy has been widely used in conjunction with active specific IT to enhance anti-tumor immune responses by down regulation of Treg, and this combination therapy has been pioneered by Berd et al. [12, 13] (Table 3).

Enhancement of tumor cell sensitivity to lysis by CTL

Active specific immunotherapy often induces low avidity CTL which do not effectively lyse tumors. However, when melanoma cells were treated with chemotherapeutic agents in vitro, they became highly sensitive to lysis by low avidity CTL. Cytotoxic drug-mediated sensitization primed both perforin/granzyme and Fas-mediated killing by the CTL [151]. In a related study, treatment of cancer cells with 5-aza-2′-deoxycytidine restored the expression of major histocompatibility complex (MHC) class I molecules and cancer testis antigens on tumor cells, rendering the tumor cells susceptible to CTL attack [133].

In a reverse manner, immunotherapy may directly modulate the tumor’s sensitivity to chemotherapy:

  1. a.

    Monoclonal antibody Rituximab, used for passive immunotherapy of B cell lymphoma and non-Hodgkin’s lymphoma cancer patients, has reverted chemoresistance in B cell lymphoma cell lines to chemosensitivity [33]. Chemosensitization of tumor cells was due to downregulation of TNF-alpha secretion, but not to downmodulation of either the MDR-1 or bcl-2 gene products. Also, Her2-neu downregulation by mAb Herceptin increased tumor cell sensitivity to cisplatin by decreasing DNA repair activity following cisplatin-induced DNA damage [62, 115].

  2. b.

    In several clinical trials, IT was followed by salvage CT [4, 6, 103, 148] (Table 4). Patients treated with this combination therapy showed higher clinical response rates as compared to historical controls treated with CT alone, although larger randomized and carefully controlled trials must be conducted to convincingly demonstrate beneficial effects of combination therapies. It is not known whether in the trials mentioned above IT “conditioned” the tumor to destruction by CT as shown for combinations of mAb and CT [33, 62, 115]. Gabrilovich [53] suggests that the anti-tumor effects of IT followed by CT are exerted independently by the two therapies and synergistic effects of this combination therapy may be dependent on optimal timing and scheduling of the two therapies. Specifically, CT may need to be started quickly after the administration of IT as anti-tumor immune responses generated by IT can not be sustained for a long period of time in cancer patients [53]. On the other hand, studies in tumor-bearing experimental animals have shown that delaying CT after IT increases the anti-tumor efficacy of this combined treatment, evidently through inhibition of vaccine-induced regulatory T cells by the chemotherapeutic drug [28] (Table 3).

The therapeutic induction of lymphopenia has raised considerable interest in the context of adoptive lymphocyte transfer therapies and vaccination of melanoma patients [100]. Transient lymphopenia is thought to enhance the efficiency of these therapies by activating homeostatic mechanisms that stimulate the tumor-reactive effector T cells and by counteracting tumor-induced suppression by mechanisms such as regulatory T cells or other mechanisms. Lymphodepletion also enhances T cell homing into tumor beds and intra-tumoral proliferation of effector cells [42, 44] (Table 5).

In an animal model, synergism between CT and adoptive lymphocyte IT was dependent on the involvement of the tumor microenvironment [152] (Table 5). Thus, treating well established tumors expressing low levels of antigen with a chemotherapeutic drug caused sufficient release of antigen to sensitize stromal cells for destruction by adoptively transferred cytotoxic T cells (CTL), resulting in tumor growth inhibition.

In summary, the demonstration of statistically significant survival enhancement in cancer patients treated in randomized phase III trials with mAb and CT vs patients treated with either therapy alone, raises great expectations for combination therapies consisting of active specific IT or adoptive lymphocyte IT with CT, as suggested by studies in experimental animals.