Introduction

Cancer results from a succession of genetic mutations that result in activation of oncogenes or inactivation of tumour suppressor genes. These changes can occur both early in the process of malignant transformation, and later, as the tumour becomes invasive. The success in the development of the technology for the transfer of genetic material into mammalian cells has raised the possibility of treating cancer at a molecular level. Despite initial enthusiasm, gene therapy has still not become a standard treatment modality for cancer. In this chapter we will review the approaches that have been attempted and consider why gene therapy is still an experimental approach.

Principles of gene therapy

Several strategies have been developed that involve the insertion of genetic material into cancer cells or immune cells involved in tumour cell kill. The success of these approaches depends on the ability to deliver the genetic material to the target cells. The transfer of genetic material to a cell is termed ‘transduction’ and the delivery systems used to transfer genes to target cells are called ‘vectors’. Transduction of adequate amounts of genetic material into tumour cells represents one of the most challenging areas of gene therapy, and vector technology is one of the most important areas of current research.

For a vector to be practical for everyday clinical use, it must be:

1. Easy to manufacture

2. Specific to tumour cells or host cells that may benefit from modification

3. Efficient at transducing genetic material

4. […]

Introduction

Management options available for women at high lifetime risk of breast cancer due to their family history, or carriage of a mutation in BRCA1/2 (which confer a lifetime risk of breast cancer of 85% (Ford et al., 1994; Wooster et al., 1994) are limited. Screening with mammography or even magnetic resonance imaging (MRI) is one option, and this can be combined with entering trials of chemo-prevention. However, many women are now seriously considering or undertaking prophylactic mastectomy if found to be mutation carriers for BRCA1 or BRCA2. The efficacy of surgical procedures to reduce the risk of breast cancer is controversial (Goodnight et al., 1984; Zeigler and Kroll, 1991), although it would appear that the residual risk of breast cancer is dependent upon the amount of remaining breast tissue following the surgical procedure. Recent work suggests that more women than previously are considering prophylactic mastectomy (JW et al., 1996; Lynch et al., 1997) and that protocols should be in place to deal with these requests. It has been suggested that surgery will increase life expectancy in BRCA1 or BRCA2 mutation carriers (Schrag et al., 1997). A recent study by Hartmann et al., (1999) has demonstrated that women with a high risk of breast cancer can significantly reduce the incidence of the disease with prophylactic surgery. However, the level of reduction in those at highest risk (BRCA1/BRCA2 carriers) is still unclear.

Despite advances in surgery and chemotherapy, the overall prognosis for ovarian cancer remains poor. It has improved little over the last 30 years. The best way of improving outcome may be to detect the condition at an early stage through screening the population at risk. The high incidence of disease in those with a strong family history of ovarian cancer makes them particularly amenable to this strategy.

Why should we screen?

There are estimated to be approximately 50 000 women in the UK who have a significant family history of ovarian cancer with two or more affected close relatives. These women have an approximately ten-fold increased risk compared with the general population. This translates to an average lifetime risk of developing ovarian cancer of 15%.

The prognosis for ovarian cancer is generally poor, with an overall 5-year survival of about 30%. Seventy per cent of women are diagnosed with stage III or IV disease, with 5-year survivals of 15–20% and less than 5% respectively (Teneriello and Park, 1995). The lack of symptoms of early ovarian cancer results in women frequently presenting with advanced disease. This is due to the location of the ovaries within the peritoneal cavity, which results in minimal local irritation or interference with vital structures until ovarian enlargement is considerable, or metastasis occurs. Initial symptoms may be so vague that multiple consultations with a GP may occur before a gynaecological referral is initiated.

Introduction

The volume has covered aspects of the development of therapeutic cancer vaccine strategies against a variety of molecular targets and diseases with a strong bias to the generation of specific cell-mediated (CTL) responses. This brief overview will consider some common lessons with respect to: (1) target molecules; (2) delivery systems; and (3) evaluation methodology relevant to success of immunotherapy for cancer. A summary of the rationale, optimism, limitations and further keys for development for the various cancer vaccine approaches outlined in this volume is given in Table 14.1.

Target molecules

Viral targets

When there is an established viral aetiology for particular malignancies such as HPV with cancer of the uterine cervix or EBV with nasopharyngeal carcinoma, virally encoded tumour-associated molecules offer exogenous cancer vaccine targets where there is unlikely to be immunological tolerance at the immune repertoire level. However, prevention will always be better than cure, so immunization to reduce infection is likely to be more efficacious and cost-effective than immuno-therapeutic approaches. This is clearly shown by the example of the association of hepatitis B virus with hepatocellular carcinoma, where classical prophylactic vaccination programmes have dramatically influenced the incidence of the cancer in at risk populations. A similar strategy for the high risk papillomaviruses associated with cervical neoplasia is also planned.

Overall, the implementation of worldwide immunization against viruses such as HPV, where the malignant disease is a late complication of the viral infection, may be difficult.

Introduction

The ideal cancer vaccine should stimulate cytotoxic T cells (CTL), helper T cells and antibodies. The CTLs will efficiently kill all tumour cells expressing target antigen and MHC. Helper T cells will help in the production of CTLs but will also migrate to tissues expressing the target antigen. Once they have localized within the tissues they will release the cytotoxic cytokines (TNFβ, IFN γ) and recruit nonspecific effector cells such as macrophages. Both of these cytotoxic effects will result in tumour cell death of antigen positive or negative cells. They are therefore synergistic with CTL killing. T helper cells can also recruit natural killer (NK) cells that will kill any tumour cells that have lost MHC expression. As this is a common mechanism for tumours to evade CTL killing it is an important component of any immune response induced by a cancer vaccine. The potential of antibody responses to contribute to antitumour effects is less clear. The ‘type 1’ T cells that help in the activation of CTLs can also help in the production of specific subclasses of antibodies (IgG2a in mice and IgG1 in humans). These antibodies will kill any tumour cell expressing target antigen by antibody-dependent cellular cytotoxicity that is mediated by Fc receptor expressing leucocytes, including NK cells. T helper cell recruitment of NK cells into tumour tissues is therefore also essential for antibodymediated tumour killing.

Introduction

While a widely efficacious tumour vaccine is not yet available, a great deal of progress has been made in the development of effective cancer vaccines. Vaccines designed to treat patients with metastatic cancer have shown the first evidence of efficacy in the clinic. In this chapter, we will focus on efforts in which recombinant poxviruses have been used in the clinical and preclinical treatment of cancer. This work is based on a ‘reductionistic’ approach which has made possible an understanding of the interactions between the immune system and tumour cells on a molecular level. The thrust of this work comes from observations, discussed at greater length below, that infection of a tumour-bearing animal with a recombinant poxvirus encoding a tumour-associated antigen can result in tumour destruction and prolong the survival of the animal.

While viruses are demonstrably immunogenic, tumour cells have notoriously poorimmunogenicity.The reasons for this apparent lack of immunogenicity (as discussed in Chapter 1) may be that cancer antigens are generally not presented to the immune system in a micro-environment that favours the activation of immune cells. Although no single known mechanism can explain poor tumour immunogenicity in all experimental models studied, the molecular bases can be separated conceptually into four distinct groupings: (1) lack of expression of co-stimulatory molecules; (2) production of immuno-inhibitory substances; (3) poor antigen processing and presentation; and (4) variability in the expression of antigen by tumours.

Introduction

Vaccination strategies for the treatment of human cancer depend on the existence of tumour antigens which are able to elicit specific immune responses in the tumour-bearing host. The specific recognition of antigens by the immune system is accomplished by two targeting systems: CD4+ and CD8+ T lymphocytes recognize processed antigens presented on MHC class II and class I molecules, respectively, while B lymphocytes produce antibody molecules that bind specifically to unprocessed antigens. The analysis of humoral and cellular immune responses in cancer patients had indicated for a long time that cancer-specific antigens do indeed exist and are recognized by the immune system of the tumour-bearing host. However, the molecular nature of these antigens remained unclear until cloning techniques were developed that used established cytotoxic T lymphocyte (CTL) clones or circulating antibodies as probes for screening of tumour-derived expression libraries. The CTL approach and the antigens identified by it are reviewed elsewhere in this book (Chapter 11). This chapter is intended to give an introduction to the serological approach, to summarize the current status of antigens identified and to provide a perspective for the use of these antigens for cancer immunotherapy.

Rationale for using the antibody repertoire of cancer patients for the identification of tumour antigens

A variety of in vitro studies and animal tumour models demonstrated that CTLs are the protagonists of an effective cytotoxic antitumoural immune response and motivated the search for antigens recognized by CD8+ T lymphocytes.

Introduction

As early as the turn of the century, Paul Erhlich suggested that ‘aberrant germs’ (tumours) occurred at a high frequency in all humans but were kept in check by the immune system. Developments in understanding of the protective roles of antibodies and phagocytes in infectious disease in the early years of the century led to attempts to stimulate the immune system to reject tumours. The New York surgeon, Coley, used bacterial vaccines to cause a ‘commotion in the blood’ and occasional regressions following treatment or occurring spontaneously were taken as evidence of an effective immune response.

Early experimental work demonstrated that transplanted (allogeneic) tumours usually regressed. However, it was soon realized that this was a consequence of the genetic disparity of host and tumour and was revealing immune responses to foreign tissue transplants, not tumour antigens. However, what these early studies did show was that a strong immune response could prevent the growth of a tumour and cure the animal.

Immune surveillance

In the 1950s, Burnett and Thomas restated Erhlich's idea as the theory of ‘immune surveillance’. It was proposed that the immune system was able to recognize abnormal cells, which were destroyed before they could develop into a tumour. Since tumours do develop in many individuals it was also suggested that the immune system played a role in delaying growth or causing regression of established tumours.

Introduction

Epstein–Barr virus (EBV) is a member of the Herpes family of viruses and infects the human population worldwide. In the 35 years since its discovery, EBV has attracted ever-increasing attention from medical, molecular biological, virological and epidemiological viewpoints. It is one of the most efficient cellular growthtransforming viruses known and yet, following natural primary infection during childhood, it generally coexists within its host completely asymptomatically and, in common with other herpesviruses, establishes a persistent infection which is maintained lifelong. In Western communities about 85–90% of all adults carry the virus, whereas in developing countries the infection level approaches 100% by the age of two. At any given time, about 20% of virus-positive individuals shed infectious virus in saliva which is believed to form the primary route of transmission. However, under certain circumstances its pathogenic potential is unleashed and EBV is associated with a wide spectrum of clinical conditions, many of which are malignant.

Clinical significance of EBV

In developed countries primary infection is often delayed for several years. Following first time infection during adolescence or young adulthood, clinical infectious mononucleosis (IM) develops in about half of the instances. In the United States alone, it is estimated that there are about 100000 new cases per year. However, despite the significant morbidity which is directly attributable to the aetiological role of EBV in the causation of IM, it is more the virus' associations with various human cancers which are the prime considerations behind the need for vaccination and/or therapy.

Introduction

There are tantalizing glimpses of the intrinsic ability of the human immune system to control emerging cancer cells. This is seen most clearly in EBV-associated B-cell tumours which emerge in patients undergoing immunosuppression following organ transplantation. Humoral immune responses against tumour antigens of melanoma and of epithelial malignancies have also been detected in patients. However, when patients have cancer, the tumour cells have obviously evaded any immunological weapons. Vaccination as a treatment has to be able to activate a defeated immune system, in a situation where tumour antigens may still be present, and where the patient may be debilitated by disease or treatment.

Tumours of B lymphocytes are attractive for study because of the availability of cells, and because of an increasing understanding of the array of potential target molecules expressed at the cell surface. There is also a clinical need to develop new treatments for the many categories for which chemotherapy has not improved survival. However, these tumours exist despite exposure to the full power of the immune system. If we can succeed in activating an antitumour response against B-cell tumours, therefore, we should be in a strong position to attack other cancers which are less exposed. Molecular technology is facilitating vaccine development in three ways: first, by revealing novel tumour antigens; second, by aiding our understanding of immune mechanisms; and third, by facilitating new delivery systems to mobilize those mechanisms.

This review will focus on malignancies arising from B lymphocytes.

Introduction

The Poxviridae are a family of double-stranded DNA viruses that characteristically replicate in the cytoplasm of infected cells and, by encoding a wide range of hostindependent enzymes, are able to express gene products and replicate DNA, even in enucleated cells. They are further subclassified but only two members of the subfamily Chordopoxviridae (poxviruses of vertebrates), molluscum contagiosum virus and variola (smallpox), are specifically human viruses. Infection with poxviruses of other species can occur with cowpox, vaccinia and monkeypox, sometimes resulting in localized outbreaks. In addition, skin infections from inadvertent local inoculation with orf (a virus of sheep and goats), pseudocowpox (producing milker's nodules), tannapox and yabapoxvirus occur. The most notorious member of this family – variola virus – was historically the cause of widespread pandemics of smallpox. It is estimated that 400000 people died each year in Europe at the turn of the eighteenth century with a consequent major impact on communities. Clinically, it has been recognized that immunity was lifelong following exposure to variola, resulting in the practice of variolation (direct inoculation of infected scabs from patients into healthy recipients). Although immunity to variola and a significant reduction in mortality was achieved, a substantial morbidity and mortality associated with the procedure remained. This provided the background to the classical experiments of Jenner, who used direct inoculation of cowpox to induce protective immunity to experimental exposure to variola.

Introduction

The realistic prospect of developing a vaccine for the treatment or prevention of colon cancer derives from recent advances in molecular biology and a better understanding of the host–tumour immune response. Although the concept of vaccination is not new, the concept that T-cell responses are critical to achieving tumour rejection, coupled with the identification of novel tumour antigens that contain T-cell-specific epitopes, has led to new strategies in the application of tumour vaccines.

Colon cancer

Cancer of the colon or rectum is one of the most common neoplasms in the Western world. The highest incidence is reported in the United States, but rates are especially high in Australia, New Zealand and areas of northern and western Europe. Estimates suggest that over 131000 new cases are diagnosed annually in the United States and that about 5% of the population will develop colorectal cancer before the age of 75 years. When diagnosed in its earliest stages colorectal cancer is highly curable with surgical treatment. However, once the disease has spread to regional lymph nodes or elsewhere, the cure rates decrease dramatically. Mortality from advanced disease has remained largely unchanged over the last 50 years with over 56000 deaths annually in the United States. Recent advances in colorectal cancer research have focused on understanding the genetic changes associated with disease progression, defining methods for earlier diagnosis, and improving therapy for advanced disease.

Introduction

During recent years there have been major strides in our understanding of both the molecular basis of malignant disease and the cellular and molecular immunological response thereto. Nevertheless, despite these exciting developments, arguably the most successful immunotherapeutic approach for the treatment of any malignancy is the relatively unsophisticated approach, involving the intravesical administration of BCG in transitional cell carcinoma of the bladder. In this chapter we briefly describe the natural history of bladder cancer, the conventional treatments for this malignancy, the effects of BCG on the immune system and bladder cancer in vitro and in vivo, and speculate on its possible mode of action. Finally, we comment on the problems associated with this treatment and highlight current and future developments aimed at further improving the antitumour potential of BCG and other mycobacteria. For other excellent reviews in this area we would recommend Jackson and James, O'Donnell and DeWolf, Martinez-Pineiro and Martinez-Pineiro, Lamm, and Kurth.

Bladder cancer as a disease

Bladder cancer is the fourth most common cancer among men and the eighth most common among women. The annual incidence of bladder cancer in the USA is 54000 patients and the total number is currently estimated at 300000 (O'Donnell, personal communication). This results in over 11000 deaths a year. Females are relatively spared with a ratio of 1 to 3, possibly due to the different occupations and environments to which they are exposed.

Introduction

Spontaneous regressions of human tumours have been reported in different types of cancer, especially in melanoma and renal cell carcinoma, but also in other types of cancer, such as nonsmall cell lung cancer, bladder carcinoma and breast cancer. These observations suggest the interaction of the immune system with antigenic determinants presented by the tumour. Early attempts to activate the immune system against tumour growth were based on observations made in cultured melanoma cells which were shown to be lysed by autologous CD8+ T lymphocytes in vitro. The clinical translation of this observation was applied to single patients with metastatic melanoma, who received irradiated autologous tumour cells as a vaccine. Two patients (SK-29 and MZ-2) with recurrent metastatic melanoma have been observed by our group since 1978 and 1982, respectively. Both patients received intradermal immunization with irradiated autologous tumour cells for an extended period of time. Complete regression of tumour manifestations was documented after prolonged immunization with autologous tumour cells. The patients have remained free of disease until now for 19 and 14 years, respectively. Based on these favourable clinical evolutions, a systematic search was initiated to identify and characterize the cancer antigens and immune effector mechanisms mediating these tumour regressions in vivo.

Human tumour antigens recognized by the immune system

Cancer testis antigens

Cytotoxic T lymphocytes (CTL) have been isolated first from melanoma patients that effectively lyse melanoma cells in vitro. Antigenic peptides have been found to be presented by MHC class I and II molecules.