Non-viral in vivo immune gene therapy of cancer: combined strategies for treatment of systemic disease
- First Online:
- Cite this article as:
- Tangney, M., Casey, G., Larkin, J.O. et al. Cancer Immunol Immunother (2006) 55: 1443. doi:10.1007/s00262-006-0169-z
- 75 Views
Many patients with various types of cancers have already by the time of presentation, micrometastases in their tissues and are left after treatment in a minimal residual disease state [Am J Gastroenterol 95(12), 2000]. To prevent tumour recurrence these patients require a systemic based therapy, but current modalities are limited by toxicity or lack of efficacy. We have previously reported that immune reactivity to the primary tumour is an important regulator of micrometastases and determinant of prognosis. This suggests that recruitment of specific anti-tumour mechanisms within the primary tumour could be used advantageously for tumour control as either primary or neo-adjuvant treatments. Recently, we have focused on methods of stimulating immune eradication of solid tumours and minimal residual disease using gene therapy approaches. Gene therapy is now a realistic prospect and a number of delivery approaches have been explored, including the use of viral and non-viral vectors. Non-viral vectors have received significant attention since, in spite of their relative delivery inefficiency, they may be safer and have greater potential for delivery of larger genetic units. By in vivo electroporation of the primary tumour with plasmid expressing GM-CSF and B7-1, we aim to stimulate immune eradication of the treated tumour and associated metastases. In this symposium report, we describe an effective gene based approach for cancer immunotherapy by inducing cytokine and immune co-stimulatory molecule expression by the growing cells of the primary tumour using a plasmid electroporation gene delivery strategy. We discuss the potential for enhancement of this therapy by its application as a neoadjuvant to surgical excision and by its use in combination with suppressor T cell depletion.
Antigen presenting cell
Cytotoxic T lymphocytes
Lymphokine activated killer cell
Minimal residual disease
Tumour associated antigen
Tumour infiltrating lymphocyte
Regulatory T cell
In patients with cancer, most therapeutic failures and deaths are due to metastatic disease which is evident at presentation or becomes manifest at varying intervals after treatment. Dissemination of tumour cells may have already occurred at presentation and cures achieved only by forestalling tumour growth at systemic level. Current systemic therapies are limited by poor responsiveness and toxicity, and consequently, there is continued interest in the development of biological based therapies that specifically target tumourigenic mechanisms. In general, therapeutic efficacy is mainly confined to the setting of minimal residual disease with little in the way of responses in overt metastasis with the exception of a minority of patients with disseminated breast or prostatic cancers [24, 45]. There is, however, optimism from recent advances in molecular biology and biotechnology that targeting of metastases may be achieved by gene, immune and anti-angiogenic based therapies with little collateral injury.
Recently the relationship between anti-tumour immune responses and tumour progression and prognosis is becoming clearer [24, 35]. Although most cancers occur in patients with a globally intact immune system and continue to progress despite immunologic responses, evidence for the existence of immune control of tumour growth has come to light over the years with findings such as more aggressive tumour growth with earlier metastatic development and worse survival prospects in athymic animals or immunosuppressed patients [19, 22]. The argument in favour of the importance or potential of immune responses to tumour was not resolved by earlier attempts at immunotherapies resulting in meagre clinical responses to cytokine and LAK cell therapies and to vaccination strategies with adjuvants, tumour cells and antigens. A number of simultaneous discoveries have helped to clarify matters in favour of the protagonists of anti-tumour immunity: (i) immune reactivity within the primary tumour is associated with the presence of CD8+ immune effector and memory cells, clearance of systemic micrometastases and an improved prognosis [24, 35]; (ii) the finding that Treg cells preferentially home to tumour sites and locally accumulate with increasing tumour burden may explain the failure of local tumour containment while simultaneously an effective concomitant systemic immunity to minimal disease is preserved ; (iii) the ability to manipulate the immune system at tumour level with rational local gene based therapies gives hope that both host and tumour specific therapies will evolve.
Immune reactivity within the primary tumour is associated with clearance of systemic micrometastases
There is strong evidence that for many cancers, the oncological potential of the minimal residual disease or micrometastases can be altered by changes in either the global or tumour specific immune functions [28, 49]. Induction or acquisition of immune suppressed states leads to early recurrence and rapid progression of cancers. By contrast, in experimental models, abrogation of immune tolerance to tumours or vaccination-based strategies has been shown to control minimal residual disease states. The goal of our studies is to achieve an immune mediated systemic control of cancer using an in vivo gene therapy approach localised to the primary tumour. Such an approach could be used to develop both a host and tumour specific therapy by promoting immune reactivity to heterogeneous populations of tumour cells.
Failure of immune responses to destroy malignant cells
A number of interactions between cancer cells and host immune cells in the tumour microenvironment create a network that may promote tumour growth, protect the tumour from immune attack and attenuate immunotherapeutic efficacy . Also, an inhibited intra-tumoural immune effector mechanism not only fails to eliminate the primary tumour, but also selects for immune escape variant tumour cells, a process termed ‘immunoediting’ . Therefore, weak tumour specific immune response is not simply due to a passive process whereby adaptive immunity is shielded from detection of TAAs, but involves active tolerising processes during the evolution and growth of the tumour.
The cellular basis of immune responses against malignancies broadly involves tumour antigen presentation at the local lymph node, with consequent recruitment of tumour specific effector lymphocytes to the tumour region. Mechanisms of tumour immune evasion involve impaired antigen presentation through weakly expressed surface antigens, immune selective pressure for tumour cells which lack expression of MHC class I molecules, impaired proliferation within the tumour of both APC and lymphocytes, and the apoptotic killing of tumour infiltrating lymphocytes by factors elaborated by the cancer cell, by the Fas counterattack and/or by Treg [3, 29, 39]. Even when specific effector T cells are recruited to the tumour, they are subject to depletion because of absence of synchronous co-stimulation with T cell receptor engagement. Most tumours do not express the immune co-stimulatory molecules B7-1/B7-2 and T cell receptor engagement without co-stimulation is thought to induce a local T cell anergy [17, 18, 49]. Furthermore, CD4+ T cells are known to cross-compete for antigenic complexes on the surface of APC, inhibiting activation of other potentially reactive T cells of the same and differing specificities . This process may represent a physiologically important mechanism for influencing the quality and quantity of CD4+ T cell responses.
The net result of all the immune evasive mechanisms is an inadequate effector to target cell ratio within the growing tumour mass for cancer cell killing, while anti-tumour cellular and humoural responses are measurable at systemic level. This is thought to explain the paradox that solid tumours grow in the presence of a strong immune reaction but after resection of the primary tumour, minimal residual disease is contained when the immune effector targeting of single cell or micro cluster disease is feasible.
Regulatory T cells
Treg preferentially home to tumour sites and locally accumulate with increasing tumour burden . This finding is now thought to be the most important explanation for the failure of local tumour containment while simultaneously an effective concomitant systemic immunity to minimal disease is preserved. Treg have a key role in the maintenance of immune tolerance to both self and foreign antigens. In particular, a minor population (10%) of CD4+ T cells, which co-expresses the IL-2R chain (CD25), has been described to be crucial for the control of auto-reactive T cells in vivo . Functional and molecular characterisation of Treg cells has only recently been made possible by the association of cell markers, such as CD25, CTLA-4, GITR, and FoxP3 gene product, with immune regulation . Several studies have shown that elevated numbers of CD4+CD25+ Tregs can be found systemically, within the tumour and within malignant effusions in advanced cancer patients with various forms of malignancies, and that high Treg frequencies are associated with reduced survival .
The exact mechanism by which CD25+ Treg cells exert their suppressive effects remains unknown. CD25+CD4+ Tregs may act in a cell contact-dependent manner by competing directly for stimulatory ligands on the APC, by reducing the levels of essential growth factors such as IL-2, or by directly transmitting an as-yet uncharacterized negative signal. A bystander suppression may also be achieved through the secretion of immune inhibitory cytokines such as IL-10 and TGF-β. It has been postulated that CD4+CD25+ immunoregulatory cells may be engaged in continuously up-regulating the activation thresholds of other T cells, thereby avoiding effective generation of tumour immunity while inhibiting autoimmunity [4, 16]. Thus, breaking immunological tolerance may allow effective induction of tumour immunity. The potential role of CD25+ Treg removal for the induction of tumour rejection has been demonstrated [4, 34, 47, 48]. Deletion of CD25+ Treg cells leads to the activation of otherwise silent tumour-specific CD8+ cells as well as tumour-non-specific CD4+ CD8+ NK-like effector cells [40, 42]. In addition, recent publications have documented that CD25+ Treg cells may contribute to the control of memory CD8+ T cell responses . These findings suggest that inhibition of CD25+ Treg cell action might have a beneficial effect on the induction of anti-tumour immunity when combined with different strategies of vaccination or immunogene therapies. A recent study has provided the first clinical evidence that in vivo elimination of Tregs is capable of enhancing the magnitude of vaccine-mediated, tumour specific T cell responses in humans .
We have noted that murine tumours greater than 150 mm3 are not curable by our immunogene therapy. The mechanism of immune down regulation during late enlargement of the tumour (high tumour burden) is in part mediated by increased numbers of antigen specific Treg. It is likely that some of the immune recruitment to the tumour by immunotherapy would be suppressed by this Treg mediated suppressive network. Yu et al.  have shown a selective accumulation of Treg cells inside murine tumours, increasing with and permitting tumour growth while a concomitant immunity remained evident at systemic levels. The removal of these immunoregulatory T cells unmasks natural tumour immunosurveillance and this strategy has been experimentally shown to potentiate anti-tumour vaccines or immunotherapy of cancers . Yu et al.  illustrated that the suppression of anti-tumour immunity by Treg occurs predominantly at the tumour site, and that local reversal of suppression, even at a late stage of tumour development, can be an effective treatment for well-established cancers. In the murine fibrosarcoma model used by Yu et al. antibody mediated CD4+ cell depletion during the effector phase, rather than priming phase, successfully enhanced anti-tumour immunity. Furthermore, local depletion of Treg cells inside the tumour resulted in a change of cytokine milieu and led to the eradication of well-established highly aggressive tumours and the development of long-term anti-tumour memory. Onizuka et al.  demonstrated that a single i.v. administration of anti-CD25 monoclonal antibody 4 days prior to tumour induction caused the regression of various leukaemias, myelomas and sarcomas from different mouse strains. Kinetic analysis showed that the administration of antibody later than 2 days after tumour inoculation caused no tumour regression, irrespective of depletion of Treg.
Immune therapy of cancer
The development ab initio or the augmentation of similar anti-tumour immune responses has been used as a cancer therapy in experimental clinical settings. To date, such immunotherapeutic strategies have been relatively ineffective for larger tumours, are limited by toxicity and by the expense and difficulties involved with patient individualised large-scale production of therapeutics. Systemic delivery of cytokines such as IL-2 is toxic and expensive with little efficacy except for a small percentage of patients with either renal cancers or malignant melanomas. Cell based immunotherapies are currently limited by toxicity, low therapeutic efficacy and by the logistics involved in large scale culture and production of patient specific cells for reinfusion. Therapies examined to date include LAK cells for T cell independent tumour cell killing, the large scale growth of TIL to improve the effector target ratio of tumour specific T cells, the ex vivo production and culture of tumour antigen primed DC for tumour specific immune activation, the ex vivo fusion of tumour and DC to facilitate delivery and presentation of weakly expressed tumour antigens. Another strategy involving the ex vivo genetic engineering of patient specific tumour cells to produce cytokines for vaccination is also limited by efficacy and similar production difficulties while viral or cell based vaccines for single defined antigens such as MAGE1 or CEA are ineffective in antigen negative or undifferentiated tumour cell variants . Leukaemia and melanoma cells modified ex vivo to express B7 or other co-stimulatory molecules have also been used successfully as whole cell vaccines [8, 11, 21]. The graft versus leukaemia effects are due mainly to recognition of minor histocompatability (H) antigens . Targeting of specific minor H antigens has recently been shown to constitute an effective therapy for both leukaemias and solid tumours .
In situ gene therapy is now a realistic prospect and involves the delivery of genetic information into the nucleus of a cell to facilitate the production of a therapeutic protein. In the context of cancer, this may involve correction or silencing a relevant mutant gene or engineered production of therapeutics such as cytokines or immunostimulatory peptides. Clinical successes have been achieved with viral gene therapy and a p53 coding adenovirus has been licensed as the first commercially available gene therapy . However, there are still significant difficulties to be overcome before achieving an efficient and safe gene medicine. These include overcoming biological barriers to targeted delivery and ensuring an appropriate level and duration of expression. A number of delivery approaches have been explored, including the use of viral and non-viral vectors. Although developments of viral vectors are promising, their use is still problematic because their intrinsic immunogenicity restricts their repeated usage, their delivery capability is restricted to relatively small amounts of DNA and they have a small risk of random integration with oncogene activation with consequent leukaemia [13, 25]. Conditionally replicating viruses restricted to cells with specific gene deficits are also in development but may be limited by immunogenicity and tumour cell heterogeneity. Recently non-viral vectors have received significant attention since, in spite of their relative inefficiency, they are safer, less immunogenic and have greater potential for delivery of larger genetic units.
Immunogene therapy: in vivo electroporation mediated delivery of plasmid coding for GM-CSF and B7-1
We hypothesised that transfection of tumours with a CpG rich plasmid coding for the combination of the cytokine GM-CSF and the co-stimulatory molecule B7-1 (CD80) would induce an effective anti-tumour response by augmentating and focussing the immune effector system at tumour mass level. We transfected growing murine tumours by means of in vivo electroporation utilising intratumoural needle electrodes after injection of plasmid DNA to tumour.
Unmethylated CpG dinucleotide motifs found in DNA of bacterial origin, including plasmids, induce immune responses via the stimulation of TLR receptors on leukocytes . CpG sequences stimulate Th1-biased pro-inflammatory immune responses characterized by cytokine release and activation of both the innate and adaptive immune systems. While these effects can often limit the utility of gene therapy efforts, in applications such as vaccination and cancer immunotherapy plasmid DNA CpG-mediated stimulation of the immune system may be beneficial.
GM-CSF is a cytokine that acts as a critical factor for development and differentiation of macrophages and DC, the most potent APCs. DCs are felt to be critical to the activation of T cell responses, and GM-CSF can act to recruit and mature DCs locally where antigen can be efficiently taken up and then presented to T cells in the draining lymph nodes to generate a systemic, tumour-specific response [3, 12, 26]. Experiments using genetically modified tumour cell vaccines revealed the potent ability of this cytokine to enhance anti-tumour immunity . A recent phase I/II trial showed a dose dependant efficacy of GM-CSF in the treatment of non-small cell lung carcinoma with significantly prolonged survival in patients receiving vaccines secreting GM-CSF at more than 40 ng/24 h per 106 cells . Other groups have reported efficacy in various tumour models for treatment by immune peptides such as interleukin-2 (IL–2) and GM-CSF either delivered directly or by tumour cell gene transfection [3, 15, 43].
The B7-1 co-stimulatory molecule is usually expressed on the surface of professional APCs, such as DC or B cells, initially binding to CD28 on CD4+ (helper) T cells, and subsequently acting as a second signal following antigen-TCR engagement [17, 18]. Tumour cells usually lack expression of CD80 and when T cell signalling occurs in the absence of co-stimulation, T cells may become clonally anergic . B7-1 transduced tumour cells are expected to present both the antigen (TCR receptor) and the co-stimulatory (CD28-mediated) signals to CD8+ CTL simultaneously, leading to efficient activation of CTLs without requiring the assistance of CD4+ helper T cells. Transfection with B7-1 has already resulted in tumour cell rejection in several tumour models [5–7, 37, 44] and the anti-tumour immunity induced by B7-1 transduced tumour cells has been found to depend on the inherent antigenicity of the tumour .
Plasmid DNA is a relatively safe alternative to viral vectors for cancer gene therapy studies. The toxicity is generally very low, and large-scale production is relatively easy. However, a major obstacle that has prevented the widespread application of plasmid DNA is its relative inefficiency in gene transduction. Electroporation has long been used to effectively transport molecules, including DNA, into living cells in vitro . When used in conjunction with DNA plasmids, it greatly enhances the local transfection efficiency over plasmid injection alone, or encapsulated delivery . Our in vivo electro-gene therapy protocol appears to be safe and non-toxic . All mice remain healthy throughout the course of experiments and there have been no treatment related deaths. Similarly we have found that electroporation when used clinically in the treatment of solid tumours with bleomycin (electrochemotherapy) is safe and non-traumatic, and can be applied under local or general anaesthetic, depending on anatomical location.
Responses to treatment of primary tumour
Electroporation in vivo of a plasmid both GM-CSF and B7-1 genes significantly reduces the average growth rate of tumours, with 60% treated tumours completely and permanently regressing. This average 60% cure rate is consistent throughout several treatment groups. Tumour plasmid injection in the absence of electroporation has no significant effect on tumour growth, nor does electroporation alone. The growth rate of non-responding treated tumours remains the same as that of gene free plasmid treated tumours. Tumours that received both GM-CSF and B7-1 had a significantly reduced tumour growth compared with equivalent gene quantities of either GM-CSF or B7-1 alone. Consistently, in all experiments, we found that in 40% of mice the responses of the primary tumours to immunogene therapy were partial and temporary. Although the cause of this local therapeutic failure is not evident, results from a JBS hepatic tumour model suggest that even in mice where the primary tumour was unresponsive, there was at least some degree of gene expression, which indeed did induce a concomitant immune response that was effective against micrometastases.
The GM-CSF produced by the tumours did not induce systemic toxicity. This is in keeping with the observations of Heller et al.  who found that the level of GM-CSF resulting from delivered gene, while increased within the tumour, was not measurable in peripheral blood and did not induce systemic toxicity. It is also noteworthy that on observation of the mice for greater than 3 months after eradication of the tumours, there is no evidence of autoimmune disease suggesting that in this model, at least, immune clearance of tumours can be achieved while maintaining autoimmune control.
Cure of systemic metastases
This intratumoural in vivo gene therapy also resulted in systemic responses that are inhibitory of metastatic growth in mice. When ‘cured’ animals were rechallenged to assess for an anti-tumoural immunological response, all mice challenged with the same tumour type remained tumour free to 100 days, whereas 100% of heterologous tumour challenged mice succumbed to their disease. This demonstrates that treatment induced an antigen-specific immune response giving resistance to the same tumour cell line. With previously fibrosarcoma ‘cured’ mice, the novel colon tumours were seen to grow at the same rate as in naïve mice.
To further demonstrate the systemic nature of responses to tumours arising from this treatment, effective cure of established distal metastases was achieved following treatment of the primary s.c. tumour in mice. To mimic an MRD tumour model, groups of mice received s.c. JBS tumours, followed 7 days later by hepatic JBS tumours. S.c. tumours were treated as usual. Results showed the absence of liver tumours in all mice where s.c. tumour regression occurred. In mice where the s.c. tumour did not regress, liver tumours were either absent or liver weight statistically significantly less than animals that did not receive GM-CSF/B7-1 genes. Therefore, clearance of the primary tumour mass by the gene therapy does not appear to be a prerequisite for control of distal tumours given that murine liver tumours regressed even in those cases where the treated primary tumours did not respond.
Adoptive transfer of anti-tumour lymphocytes as a therapy
When splenocytes derived from ‘cured’ animals were adoptively transferred into naïve mice, which were subsequently challenged, we found specific resistance to tumour establishment. We then set out to assess the effects of transfer of splenocytes derived from ‘cured’ animals to mice bearing established, palpable, vascularised tumours. Regression of the initial tumour occurred in 50% mice receiving ‘cured’ splenocytes, indicating transfer of efficient immunological recognition and cytotoxicity capable of clearing the vascularised tumour masses. No effect on primary tumour growth was observed in mice receiving naïve splenocytes. Therefore, adoptive transfer of specific tumouricidal lymphocytes derived from immunogene treated tumour bearing hosts may also have potential as a therapy.
Clearance of the primary tumour is not a prerequisite for control of metastases
In our murine experiments, some tumour types examined and larger fibrosarcoma tumours proved unresponsive to immunogene therapy. Our clinical findings on the association between intratumoural inflammatory reaction, clearance of micrometastases and prognosis, indicate that the presence of a systemic anti-tumour reactivity confers a survival advantage after removal of the primary tumour. It is probable that our gene therapy would be effective against MRD when used in a neoadjuvant setting in the primary cancer as the disseminated micrometastases are likely to share the same antigen spectrum as the primary tumour. The combination of immune gene therapy and traditional surgical excision could prove to be an effective therapy of some human cancers.
Our results in murine models hold immediate implications for treatment of patients given that (1) there is regression of the primary, treated tumour, and (2) perhaps more importantly, systemic protection is afforded against distal micrometastases. Unlike previously described systems, clinical development of this treatment would obviate the necessity to obtain tumour tissue from each patient or for large-scale culture of cells. It would also be more efficient and less toxic than systems involving sustained delivery of pharmaceutical cytokines that are subject to immune neutralising, hypersensitivity and systemic inflammatory responses. The production and purification costs of plasmid DNA are negligible and EP immunogene therapy of tumours is safe. We have demonstrated the clinical efficacy and safety of EP as part of our electrochemotherapy programme, and are currently developing electrodes to facilitate endoscopic and radiologic based treatment of internal malignancies.
Immunogene therapies may yield most success when used in combination with surgery and/or suppressor cell deletion. It is possible that gene therapy induced anti-tumour effector responses may be attenuated by Treg within the growing cancer and that immunogene therapy alone might be inadequate in larger tumours. We are at present investigating the combination of Treg cell depletion with our immunogene therapy. In a clinical setting, approaches other than antibody-mediated depletion of Treg cells may be preferable, such as the recently reported recombinant IL-2 diphtheria toxin conjugate .
Given that in situ immune therapy is individual to the host and the tumour at any stage of its development, treatment of tumours with immunogene therapy as a neoadjuvant to surgery or electrochemotherapy may prove effective in patients. We aim to bring to clinical reality this immune based gene therapy that may be applied to a diversity of cancer types.
Mark Tangney and Gerald C. O’Sullivan are funded by the Health Research Board of Ireland.