OUP user menu

Advances in tumour immunotherapy

J. King, J. Waxman, H. Stauss
DOI: http://dx.doi.org/10.1093/qjmed/hcn050 675-683 First published online: 13 May 2008


The clinical goal of tumour immunotherapy is to provide either active or passive immunity against malignancies by harnessing the immune system to target tumours. Although vaccination is an effective strategy to prevent infectious disease, it is less effective in the therapeutic setting for cancer treatment, which might be related to the low immunogenicity of tumour antigens and the reduced immunocompetence of cancer patients. Recent advances in technology have led to the development of passive immunotherapy approaches that utilize the unique specificity of antibodies and T cell receptors to target selected antigens on tumour cells. These approaches are likely to benefit patients and alter the way that clinicians treat malignant disease. In this article we review recent advances in the immunotherapy of cancer, focusing on new strategies to enhance the efficacy of passive immunotherapy with monoclonal antibodies and antigen-specific T cells.


The clinical goal of tumour immunotherapy is to provide either active or passive immunity against malignancies by harnessing the immune system to target tumours. Although vaccination is an effective strategy to prevent infectious disease, it is less effective in the therapeutic setting for cancer treatment, which might be related to the low immunogenicity of tumour antigens and the reduced immunocompetence of cancer patients. Recent advances in technology have led to the development of passive immunotherapy approaches that utilize the unique specificity of antibodies and T cell receptors (TCR) to target selected antigens on tumour cells. These approaches are likely to benefit patients and alter the way that clinicians treat malignant disease. In this article we review recent advances in the immunotherapy of cancer, focusing on new strategies to enhance the efficacy of passive immunotherapy with monoclonal antibodies and antigen-specific T cells, while limiting the toxic side effects of these treatment modalities.

Passive immunotherapy with monoclonal antibodies

A desire to provide patients with less toxic treatments, coupled with an increase in scientific understanding and technology, has led to the identification of tumour antigens which are suitable targets for immunotherapy. The major effector cells of the adaptive immune response have since been directed against these tumour antigens.

The development of hybridoma technology has allowed the production of large quantities of highly specific antibodies licensed for use against a broad spectrum of diseases. Monoclonal antibodies exert their effects via mechanisms which include triggering apoptosis, activating antibody dependent cellular cytotoxicity, blockade of growth factor receptors and the activation of complement.

Clinical success has been seen with passively acquired monoclonal antibodies directed against a number of targets including HER2, CD20, vascular endothelial growth factor (VEGF) and epidermal growth factor receptors (EGFR). This class of cancer therapeutics continued to grow rapidly, with a far less toxic side effect profile than conventional chemotherapy and radiotherapy.


In 1997, rituximab became the first monoclonal antibody to be approved by the US Food and Drug Administration for a cancer indication. It is a chimeric monoclonal antibody which targets the CD20 receptor on B cells, inducing B cell apoptosis and recruiting other immune effector cells to mediate cell lysis. It mediates its effects through mechanisms including antibody-dependent cellular cytotoxicity (ADCC), complement-dependent cytotoxicity (CDC) and induction of apoptosis of CD20+ cells.1,2

In a landmark clinical trial, previously untreated, elderly patients with diffuse large-B-cell lymphoma were randomized to receive CHOP chemotherapy plus rituximab or CHOP alone. The complete response rate and survival was significantly higher in the group that received CHOP plus rituximab than in the group that only received CHOP, without a significant rise in toxicity.3 An international multicentre trial of patients aged 18–64 then followed. The rituximab group had a statistically significant increase in event free and overall survival.4

Interest in the therapeutic applications of monoclonal antibodies soared in the wake of the success seen with rituximab in clinical trials.


Trastuzumab (Herceptin) is a chimeric antibody against HER2—binding of trastuzumab to the HER2 receptor prevents the activation of the receptor's intracellular tyrosine kinase. Trastuzumab has several possible mechanisms of action.5 These include prevention of HER2-receptor dimerization, increased endocytotic destruction of the receptor and antibody dependent cytotoxicity.6 Trastuzumab may also inhibit tumour angiogenesis by both induction of anti-angiogenic factors and repression of pro-angiogenic factors.7,8

Overexpression of HER2 occurs in around 25% of breast cancer and is associated with a poor prognosis.9–11 The results of several pivotal clinical trials with trastuzumab have led to changes in clinical practice in recent years. A randomized trial of patients with metastatic breast cancer compared standard chemotherapy alone and standard chemotherapy plus trastuzumab. The addition of trastuzumab to chemotherapy was associated with a longer time to disease progression and a 20% reduction in the risk of death. Unfortunately, cardiac dysfunction occurred in up to 27% of patients receiving trastuzumab, most commonly when given in association with an anthracycline.12

Two subsequent trials showed the benefits of trastuzumab in early breast cancer treatment. The first reported data from an international, multicenter, randomized trial comparing one year of trastuzumab with observation in patients with HER2-positive breast cancer, who had completed locoregional therapy and at least four cycles of neoadjuvant or adjuvant chemotherapy. At the first planned interim analysis, 127 recurrences of breast cancer, or contralateral breast cancer, or second nonbreast malignant disease, or death were observed in the trastuzumab group and 220 in the controls. Overall survival in the two groups was not significantly different. Severe cardiotoxicity was again reported in 0.5% of the women who were treated with trastuzumab.13 The second paper described the combined results of two trials that compared adjuvant chemotherapy with or without trastuzumab in women with surgically removed HER2-positive breast cancer. There were 133 events in the trastuzumab group and 261 in the control group. The absolute difference in disease-free survival between the trastuzumab group and the control group was 12% at three years. Trastuzumab therapy was associated with a 33% reduction in the risk of death (P = 0.015). Again, the major toxicity seen was a class III or IV congestive heart failure or death from cardiac causes in the trastuzumab group of 2.9–4.1%.14

As a result of these findings, NICE recommended herceptin as a treatment for both early and metastatic breast cancer in August 2006. For women with HER-2 positive breast cancer who previously found themselves in a poor prognosis group, this represented a real breakthrough in terms of translational cancer research.


Bevacizumab is a monoclonal antibody against VEGF. It acts by binding to VEGF, thus preventing the interaction of VEGF with its receptors on the surface of endothelial cells. The interaction of VEGF with its receptors leads to endothelial cell proliferation and new blood vessel formation in in vitro models of angiogenesis.15

Bevacizumab has seen success in a landmark clinical trial in patients with metastatic colorectal cancer. Eight hundred and thirteen patients were randomized to IFL chemotherapy plus bevacizumab or chemotherapy and placebo. The median duration of survival was 20.3 months in the group given bevacizumab, as compared with 15.6 months in the group given placebo. The median duration of progression-free survival, response rate and the median duration of the response were also greater for bevacizumab than placebo. Hypertension was more common during treatment with bevacizumab, but was easily managed.16


Cetuximab is a chimeric monoclonal antibody directed against the ligand binding domain of EGFR, a member of the ErbB family of receptor tyrosine kinases. Its mechanisms of action have yet to be fully elucidated, but are thought to include ADCC.17 EGFR is over-expressed in several epithelial cancers including those of the head and neck, lung, oesophagus and colon, where it is associated with a poor prognosis.18–21 Cetuximab has been shown to have clinical activity in colon cancer.22 A recent multi centred, randomized, Phase III study compared radiotherapy with radiotherapy plus cetuximab in the treatment of patients with locoregionally advanced head and neck cancer. Progression free survival was significantly better in the cetuximab plus radiotherapy group, as was overall survival (49 months vs. 29 months, P = 0.03).23

This represents the latest challenge to the field of oncology and translational cancer research: providing new treatments with an improved side effect profile that confer a survival benefit and are also cost effective. Antibodies licensed for use in oncology are listed in Table 1.

View this table:
Table 1

Monoclonal antibodies licensed for use in oncology

BevacizumabMetastatic colorectal cancerVEGFHumanized
TositumabCD20+ NHLCD20Murine
AlemtuzumabB-CLL, transplant pre-conditioningCD52Humanized
CetuximabColorectal cancer, head and neck cancerEGFRChimeric
TrastuzumabBreast cancerHER-2/erbB2Humanized
GentuzumabAMLCD33Humanized, linked to toxin
AdecatumumabBreast cancerEpCAMHuman

Anti-CTLA-4 (Cytotoxic T lymphocyte-associated antigen 4) antibodies

Cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) is expressed on activated T cells and regulatory T cells (Tregs) and plays a critical role in regulating natural immune responses by reducing the proliferation of activated T cells.24–26 CTLA-4 blockade with anti CTLA-4 antibodies can induce rejection of several types of solid tumour in mice, although less immunogenic tumours such as B16 melanoma were not rejected.27–30 In a murine melanoma model, CTLA-4 blockade in combination with granulocyte macrophage colony stimulating factor (GM-CSF) immunotherapy has been shown to cause tumour regression which correlated with a change in the balance of Tregs and effector T cells in the tumour microenvironment.31

There are two fully human anti CTLA-4 monoclonal antibodies in clinical trials at present: Tremelimumab (Pfizer) and Ipilimumab (Medarex and Bristol-Myers Squibb), which have been developed to stimulate patients’ immune responses against tumours. They bind to the CTLA-4 molecule on Tregs and activated T cells and prevent B7-1 and B7-2 from binding to CTLA4. B7-CTLA4-mediated down regulation of T-cell activation is thus inhibited, while B7-CD28-mediated T-cell activation continues unopposed by CTLA4-mediated inhibition. In addition, CTLA-4 on Tregs is inhibited from binding its target on dendritic cells (DCs). When CTLA-4 binds to B7 molecules on DCs a regulatory pathway is activated, initiated by the enzyme indoleamine 2,3 dioxygenase (IDO). This pathway is considered to be one of the contact dependent effector mechanims of natural Tregs that express CTLA-4.32

In a Phase I/II trial, five out of 30 metastatic melanoma patients treated with ticilimumab (tremelimumab) demonstrated objective clinical responses. Antitumour responses correlated with an increase in interleukin 2 (IL-2) production and a reduction in Tregs and IL-10 secretion.33 Anti-CTLA-4 antibodies have also induced clinical responses in metastatic renal cell carcinoma,34 and in metastatic melanoma and ovarian cancer patients when used in conjunction with GM-CSF expressing tumour vaccines, antigen loaded DCs vaccines and peptide vaccines.35–38 However, Tregs play a crucial role in suppressing autoimmune disease, and it is important to note that clinically significant autoimmune sequelae—including hypophisitis and thyroiditis—have often been associated with clinical responses in these clinical trials.39

Limitations of monoclonal antibody therapy

Antibody therapy has a number of limitations in addition to their cost. Firstly, since a memory response is not generated, repeated antibody infusions are required. Secondly, because antibodies are chimeric or humanized and retain a small murine component, they are themselves potentially immunogenic, which may cause problems with repeated administration. Thirdly, antibodies can only recognize specific proteins which are present on the cell surface, which limits the range of available targets. Intracellular proteins are broken down into a large number of peptides, which are expressed on the cell surface in association with a major histocompatibility (MHC) molecule. These peptide-MHC complexes can be recognized by the TCR, which is able to target a much larger number of intracellular tumour antigens.

T cell based immunotherapy

CD4+ and CD8+ T cells both show anti tumour activity in isolation, but work best in combination, as CD4+ cells are major producers of cytokines which provide ‘help’ for CD8+ T cells.

When the TCR binds its cognate antigen, intracellular signalling results to trigger proliferation and effector functions such as cytokine production and cytotoxicity. CD4+ T cells recognize peptides in association with MHC class II (MHC II) molecules, while CD8+ T cells recognize peptides presented by MHC class I (MHC I) molecules. Since tumour cells are usually MHC II negative, the role of CD8+ T cells in providing anti tumour immunity was originally investigated in much greater detail. The majority of identified tumour antigens to date are presented by MHC I molecules, and loss of MHC expression is a frequent mechanism by which tumours evade recognition by CD8+ T cells.40–42 Although many tumours do not express MHC II molecules, CD4+ cells can still show protection against such tumours.43,44 MHC II positive antigen presenting cells present antigen to CD4+ cells, resulting in the production of cytokines such as interferon gamma, which has been shown to have an inhibitory effect on tumour vasculature.44

A major obstacle to successful T cell immunotherapy is the poor immunogenicity of tumour associated antigens (TAA) which are expressed in normal tissues as well as being over-expressed on tumours. To reduce the risk of autoimmunity, developing T cells which recognize self antigens with high affinity are deleted in the thymus. Peripheral tolerance mechanisms, including Tregs, also exist to delete or render anergic any self reactive T cells which escaped deletion in the thymus.

Adoptive T cell therapy provides an opportunity to directly transfer the specific effectors of immunity, bypassing the obstacles in the host that might prevent the generation of an effective response in vivo. Antigen-specific T cells may be expanded in vitro, and can be selected to be of high avidity and tested for anti tumour activity before transfer into patients. The adoptive transfer of antigen specific T cells was first used to good effect in the treatment of Epstein Barr virus (EBV) and cytomegalovirus (CMV) infection or reactivation occurring in immunosuppressed bone marrow transplant recipients. EBV related post transplant lymphoproliferative disease (PTLD) is a significant cause of morbidity and mortality in transplant recipients. EBV-specific T-cell lines prepared from peripheral blood collected from donors on the day of marrow harvest have been used either prophylactically, in patients thought to be at risk of developing PTLD, or therapeutically with good clinical effect.45–47

It is important to lymphodeplete patients, prior to adoptively transferring cells, in order for those cells to persist in vivo.48–53 This may be to reduce competition with endogenous T cells for antigen presenting cells, or for cytokines required for growth. Another possible explanation is that lymphodepletion results in the clearance of host suppressor cells such as myeloid suppressor cells or Tregs.54,55 Tregs make up ∼5–10% of CD4+ T cells and are characterized by the additional expression of CD25 and the transcription factor Foxp3. They are a subset of T cells which actively suppress T cell activation and prevent autoimmune disease resulting from pathological self-reactivity.56–58 The critical role Tregs play within the immune system is evidenced by the severe autoimmune syndrome that results from a genetic deficiency in Tregs. However, Tregs are also thought to play a role in tumour immunity by suppressing anti-tumour responses.59 This concept is supported by the isolation of tumour specific CD4+ Tregs from tumour infiltrating lymphocytes60–62 and the observation that removal of CD4+/CD25+ regulatory cells resulted in the in vivo eradication of tumours in a murine model.63–65

In a study of melanoma patients with progressive disease refractory to standard therapy, tumour infiltrating lymphocytes (TIL) which recognized melanoma antigens were expanded ex vivo.66 T cells which were reactive against A2 positive melanoma cell lines in vitro were selected. Patients underwent non-myeloablative chemotherapy conditioning before receiving on average 7.8 × 1010 cells and nine doses of IL-2. Six out of 13 patients showed objective clinical responses, with four further patients showing mixed responses. This study demonstrated for the first time that T cells specific for TAAs could persist and be functionally active in vivo. It was extended to include a total of 35 patients, for whom the objective response rate was 51%, including four patients who showed a complete response. In a proportion of patients, tumour regression was accompanied by in vivo expansion and persistence of the adoptively transferred melanoma antigen specific T cells.67,68 However, there are a number of practical limitations with this type of adoptive therapy. It is neither always possible to isolate TIL from all patients, nor all tumour types, and it is also a very labour intensive process requiring the expansion of a specific set of cells for each individual patient, so is therefore unlikely to become a widely available therapy.

TCR gene transfer

In recent years a new strategy has evolved to enable the adoptive transfer of autologous T cells which are specific for TAAs. This technique enables the production of large numbers of autologous, antigen specific cells. It overcomes the problem of central tolerance and bypasses the need for the host to mount the immune response against tumour antigens. A further advantage of this strategy is that the introduced TCR specificity can be directed against poorly immunogenic targets. Using viral vectors, it is now possible to transfer the genes for the alpha and beta chains of TCR into T cells, thus conferring upon them a given tumour antigen specificity.

In a recent Phase I trial, retroviral gene transfer was used to transduce peripheral blood lymphocytes, taken from patients with melanoma, with the genes encoding the alpha and beta chains of a TCR with specificity for a MART 1 peptide presented by HLA-A*0201.66 The 17 patients in the gene transfer study were lymphodepleted prior to receiving autologous T cells transduced with the MART-1 TCR. The engineered T cells persisted and the two patients with the highest levels of circulating anti melanoma T cells showed objective regression of metastatic lesions and remained in remission 18 months after treatment.69

The results of this study prove that retroviral TCR gene transfer can be used to confer anti tumour specificity upon a large number of T cells, and that these T cells can engraft in patients and persist at high levels.

However, TCR gene therapy carries with it other potential safety concerns (Figure 1), the first of which relates to the use of retroviral vectors and the theoretical risk of insertional mutagenesis. The insertion of retroviral vectors containing viral promoter sequences may alter the expression of genes adjacent to the insertion site. In a landmark gene therapy trial, children with X linked severe combined immunodeficiency syndrome (SCID) received stem cells which had been retrovirally transduced with the common gamma chain of the IL-2 receptor, which is deficient in SCID. While T and B cell function was restored in all cases, 4/20 children went on to develop T-cell leukaemia. The transgene had inserted into the intron of the LMO-2 gene, causing LMO-2 upregulation. LMO-2 is required for the regulation of haematopoesis, but is also an oncogene which is aberrantly expressed in acute lymphocytic leukaemia.70–72

Figure 1.

Safety issues relating to retroviral TCR gene transfer. (1) Retroviral particles, containing the genes for a tumour antigen specific T cell receptor (TCR), infect T cells. (2) Viral reverse transcriptase converts viral RNA into DNA. (3) Integrase then allows viral DNA to stably integrate into the host DNA. This is a potential safety risk, since retroviruses frequently integrate near promoter regions of endogenous genes. (4) The 5′ long terminal repeat (LTR) drives transcription of the introduced TCR genes. (5) The 3′ LTR may activate or enhance the expression of cellular genes adjacent to the insertion site. (6) Endogenous TCR chains are transcribed under the control of the endogenous TCR promoter. (7) TCR α and β chains are translated and imported into the endoplasmic reticulum where the α and β chains form heterodimers and assemble with the CD3 complex (not shown), before being transported to the cell surface. (8) Since exogenous α and β chains can mis-pair with the endogenous α and β chains, this could result in the surface expression of TCR of unknown specificity, which are potentially auto-reactive.

However, it appears that the risk of transformation is much lower when transducing mature T cells, where there are no reported cases of insertional mutagenesis.73 A number of strategies are also being investigated to reduce the risk of insertional mutagenesis, including the use of single rather than multiple vectors, self inactivating vectors with deleted U3 long terminal repeat regions, the addition of suicide genes into retroviral vectors to allow the selective elimination of gene modified cells if transformation or side effects occur, as well as the use of lentiviral vectors, which have been shown to integrate near promoter regions of transcriptional units at a lower frequency compared to retroviral vectors.74

A further potential risk of TCR gene transfer is that the α and β chains of the introduced TCR could ‘mispair’ with β and α chains of the endogenous TCR, respectively, resulting in the expression of TCRs of unknown specificity on the cell surface. These combination TCRs could, in theory, be auto-reactive and recognize self antigens with high affinity, since they will not have undergone education and negative selection in the thymus. A number of strategies have recently been employed to address this issue. Hybrid TCRs have been designed to incorporate murine constant regions and human variable regions. These hybrid TCRs show preferential pairing with each other when introduced into human T cells, combined with superior cell surface expression and biological activity.75,76 TCRs have also been engineered to include an additional cysteine in the constant regions of the α and β chains, resulting in the formation of a second disulphide bond between them. T cells transduced with cysteine-modified TCR secreted more cytokine and showed increased antigen specific lysis when co-cultured with specific tumour cell lines, compared with T cells expressing wild type TCR.76,77

While there is a concern that low avidity, TAA specific CTL from the autologous repertoire may not be efficacious, high avidity, self reactive CTL may pose the opposite problem. The majority of targets for tumour immunotherapy are over-expressed self proteins, so there is a risk that targeting TAA may result in autoimmune damage. Several of the monoclonal antibodies licensed for use in cancer patients have demonstrated unexpected side effects, notably the cardiotoxicity associated with trastuzumab and the risk of bowel perforation associated with bevacizumab. In several murine models, as well as in clinical trials, it has been demonstrated that the successful induction of CTL responses against melanoma TAAs (such as melan A or gp100) has been associated with the development of vitiligo.78–83 T cell therapies targeting TAAs with a more ubiquitous distribution have not been studied in the same detail as yet. It remains to be seen whether the morbidity associated with any resultant autoimmune disease outweighs the anti tumour benefit.


The advent of monoclonal antibodies has changed the face of medical oncology in recent years and monoclonal antibodies are now an established tumour immunotherapy tool. These molecules show great promise in terms of their ability to target tumour cells directly, enhance immune responses and can also work synergistically with conventional chemotherapy. Although the promise of antigen specific T cell therapy has yet to be fulfilled, TCR gene transfer is a strategy which makes it possible to produce large numbers of antigen specific cells. While the early clinical data in melanoma is promising, the in vivo data is at present limited to 17 patients, and further work is required to better understand the factors which enabled the TCR transduced cells to mediate tumour regression. Future improvements in cancer immunotherapy may be achieved by combining strategies which take advantage of the immune enhancing activities—such as those directed against CTLA-4—and the direct anti-tumour of activities of monoclonal antibodies against surface molecules on tumour cells and T cells directed against intracellular tumour antigens.

Conflict of interest: None declared.


View Abstract