Editing cancer with the immune system

Immunologists have long suspected that the immune system plays an important and intricate role in combating cancer.

Hypotheses about immune cells patrolling the body to keep cancer at bay were proposed decades ago to explain events such as tumours that suddenly stop growing or even go away, undetected dormant tumours transferred to organ recipients following a transplant, and immune-depleted subjects being more susceptible to cancer.

However, hard laboratory-based evidence for this idea was scant and the idea of cancer ‘immunosurveillance’ remained controversial. In 2001, Professor Robert Schreiber of the Washington University School of Medicine in St Louis and colleagues proposed a new elaboration of the model called ‘cancer immunoediting’.

In this model, three outcomes are possible following a cancerous insult – elimination, equilibrium or escape of the tumour – with all influenced and perhaps modulated by immune function.

It was also around this time that Professor Mark Smyth, who heads the cancer immunology program at the Peter MacCallum Cancer Centre in Melbourne, started to collaborate with Schrieber on some mouse models.

The ultimate application of their research was targeted at developing a way to tweak the immune system in favour of elimination or at least equilibrium, while preventing escape of the tumour to grow out and fulfil its malignant potential.

At the 2008 Australasian Society for Immunology (ASI) meeting in Canberra this week, Smyth will present his team’s latest results on the research, published in Nature late last year. He will also present new data from a different model that validates their primary findings.

Smyth’s interest in this area grew out of earlier work he was doing with Professor Joe Trapani at Peter Mac on cytotoxic lymphocyte mechanisms to protect against viral infections.

“I have always been very interested in investigating basic biological responses using mice, so I started to analyse some of the knockouts we had procured for the cell-death/apoptosis research for the importance of these same pathways in tumour development,” Smyth says.

“The basic idea is that the immune system has an ongoing relationship with tumours as they develop and is part of the extrinsic regulation of tumour growth and spread.

“Most of us are familiar with the gatekeepers of the genome like the p53 gene and other tumour suppressors, as well as defined oncogenic changes that can occur. These are all genetic fundamentals that underlie cancer development and they are all intrinsic to the cancer cell.

“But then there are all these other processes that determine whether a lesion becomes clinically evident, and the immune system is one of those, and we have now spent a lot of time trying to define this role in mouse models.”

Immune pathways had been looked at previously in terms of infection, Smyth says, but not to any great depth in relation to tumour control, and certainly not in longitudinal models. “We thought that these earlier systems were not very representative of what really happens in cancer,” he says.

---PB--- Immune pathways and tumour control

Such studies usually involved bolus doses of tumour injected into transgenic mice and an immediate assessment of tumour development. So, by looking at some carcinogen-induced models as well as tumour suppressor-loss model systems, and by taking some of these models out to 100 days following treatment, Smyth started to find that some of these immune-related pathways were also important in tumour development.

The Nature study used experimental mice treated with the strong carcinogen, methylcholanthrene (MCA), to investigate Smyth’s initial findings further.

“When a group of animals is treated with a carcinogen, most develop tumours fairly quickly. These lesions generally grow out over an eight- to 14-week period and the animals have to be sacrificed because the tumours get so big.

“What we noticed early on was that some of the mice treated with MCA had what looked like tumours – small obtruding masses – on the injected side of the animal. However, these masses did not progress or grow, and if you watched those mice over a couple of hundred days the lesions just disappeared.”

Closer examination of the masses revealed aberrant fibroblasts and cells with a very low proliferative index, suggesting cells that were neither growing nor dying – the ‘tumours’ were in a state of dormancy.

“We started to wonder about the immune system in this particular group of experimental mice in which there was clearly some sort of tumour activity happening that was maintained for a long time.”

In a key experiment, the researchers injected mice with a low dose of the MCA carcinogen. A small percentage of the animals – 15 to 20 per cent – quickly developed tumours and so were excluded from further analysis.

The mice remaining were divided into two groups: half were given a control antibody and the other half a mixture of antibodies that would together knock out interferon gamma activity and T cells, crucial cell pathways and cell-types involved in the immune response.

In these immune-depleted animals, about 50 per cent developed full-blown tumours at the site of the small MCA-induced masses, compared to only a few of the control mice.

“It was pretty clear that an immune mechanism was somehow preventing the initial lesions from growing further and progressing, sort of like an extrinsic tumour suppressor mechanism,” Smyth says.

Similar results were also seen in mice left for up to 400 days after the MCA before treating with either the control or immune-depleting antibodies (the initial batch of experiments used 200 days).

Even after this time, Smyth and Schreiber found some tumours growing out, suggesting that this cancer equilibrium phase could be maintained for a long time, but that malignant potential remained in those animals, he says.

---PB--- Immune editing mechanisms

More recent experiments looking at the possible mechanisms underlying this immune editing of cancer are starting to reveal a pattern of particular immune pathways and cell types, although it is too early in this work for Smyth to be more specific and the findings are unpublished. They do confirm, however, that adaptive immunity is key to this cancer equilibrium phenomenon, he says.

“We have knocked out a few of the innate immunity pathways or molecules and do not see the same result, even though these factors are known to be important in the initial protective response against tumours.”

Smyth and Schreiber are now trying to recapitulate the cancer equilibrium evidence in other models. Their primary piece of data so far is from a long study on transgenic mice that are heterozygous for p53.

“Left for 750 days, about 60 per cent of these animals develop tumours naturally, but there was also a whole bunch in our large cohort that did not develop tumours. So, we took these and treated them with antibody as in the earlier experiments.

“We monitored these mice for another 200 days, which is really getting to the end of the animal lifespan, (so it becomes sort of representative of human ageing in cancer), and found similar numbers of mice dying in both groups.”

However, autopsies revealed that far more of the animals whose immune system had been depressed had formed tumours compared to the controls (1/26 in the control antibody group compared to 6/25 in the immune-depleted group).

“So it looks like there is a malignant potential in those immuno-compromised animals that is being held back by the immune system.”

For these findings to become of clinical importance, Smyth says the researchers need to understand the control mechanism at work, which they don’t yet. Exactly how the immune system is holding these tumours at bay remains to be established and Smyth will continue to use his animal model systems to try and find out some of the many questions remaining.

“For example, what are the anti-proliferative mechanisms – why are they not growing? Which immune cell types and players are involved; what exactly are the immune cells seeing in the cancer cells; and why are the lesions neither progressing to cancer nor completely resolving?”

---PB--- The tumour neighbourhood

The immune system thus seems to have an ongoing and multi-tasking role in cancer, and what emerges is a tumour that to some degree has been shaped by the immune pathways. Smyth is most intrigued by the nature of these changes happening intrinsically in the tumour cell and the subsequent activation response of the immune system.

“What is the link between these changes and then how are they represented by the tumour to its neighbourhood? As one approach, we would like to map the genetic changes in the aberrant fibroblasts introduced by the carcinogen and then try and understand how that relates to what the immune system is doing.

“Can we see a pattern emerging, such as a particular change in p53, for example, being mimicked by an immune response against a particular molecule or pathway? This could teach us something very valuable that could then be applied to treat cancer or prevent a tumour developing.”

Using this phenomenon to ‘control’ rather than cure cancer is really an entirely new idea in cancer treatment. Smyth and his colleagues see this information being applied in combination treatments – developing reagents to target specific components in the immune system together with established cytotoxic therapies to target the tumour cell.

Smyth thinks that prevention of cancer development will be particularly important here. “The lesions we are seeing clinically could very well be the ones that have already developed escape mechanisms, and we have no idea whether we could perhaps push the tumour cells to go back into that equilibrium phase,” he says. “

“We just can’t answer that yet, but we would really like to.”

Biography:

Mark Smyth has studied natural and passive immunity to cancer throughout his research career, publishing over 200 articles in leading peer-reviewed journals. After completeing his PhD on monoclonal antibodies in cancer and transplantation at the University of Melbourne in 1988, Smyth took up a post-doctoral position at the National Cancer Institute in the US, where he cemented a lasting interest and expertise in lymphocyte-mediated cytotoxicity.

In 1991, Smyth returned to the Austin Research Institute in Melbourne to set up his own group, looking at the biological relevance of lymphocyte granule proteins, death ligands and cytokines in the immune control of various cancers. A project designed to more specifically target T cells to cancer cells by genetic engineering of their surface with monoclonal antibodies originated in his group in 1993. He went on to assess the anti-tumour activity and function of these gene-engineered T cells in pre-clinical mouse models of cancer.

The new millennium saw Smyth relocate to the Peter MacCallum Cancer Centre in Melbourne. Here, he began to translate his research into the clinic, particularly the development of breast cancer-specific T cells. For his work on innate immune surveillance of cancer he received the 2002 William B. Coley Medal from the Cancer Research Institute and the 2007 Brupbacher Foundation Prize.