‘A hundred years ago you’d have looked at all the infectious diseases and said, “Oh my God, we’ve got to have a cure for each one – there’s TB in the lung, and in the bone, and then there’s this and that . . .” But it turns out they share a great deal of commonality, and if we hit them with antibiotics we can more or less eradicate, at least for a time, infectious disease due to bacteria. Now the fact is, bacteria are much, much more genetically complex and heterogeneous and hardy and resourceful in the evolutionary sense than cancer cells.’
The task for drug developers, Evan believes, is to find the commonality of cancer – the ‘mission critical’ mutation without which no tumour can survive. Not everyone agrees with this analysis; most cancer researchers are still backing the idea of targeted therapy tailored to the individual patient’s tumour characteristics. But whatever the perspective, tumour suppressors are obvious candidates for investigation, and much of the effort of academic researchers and their counterparts in the pharmaceutical industry is focused on repairing or enhancing the body’s natural capacity to single out and eliminate rogue cells.
People have high hopes and many imaginative ideas for p53-based therapies, though the journey from the lab to the patient’s bedside is often frustratingly slow. Ironically, as the explosive speed of technological advance makes it ever easier and quicker for scientists to develop potential new drugs, the rules and regulations governing the process get ever more tight: it typically takes a decade or more for a promising new therapy to be approved for use on patients. Very many prototypes never make it that far; drug development is, by its very nature, a painstaking process of trial and error, but even the ‘failures’ teach valuable lessons along the way.
VIRUSES AS DRUGS
The first person to try p53-based therapy in humans was Jack Roth, who in 1996 recruited to his study nine patients with inoperable lung cancer whose tumours were no longer responding to conventional therapy. In a pleasing twist to the story, Roth’s therapy made a virtue of the pernicious properties of viruses – the fact that the only way they can survive and reproduce is to invade the living cells of the host organism and hijack the machinery of replication. Using genetic engineering, he and his colleagues converted a common virus into a vehicle for transporting good copies of p53 into cells where the gene is dysfunctional. This engineered virus they injected direct into the patients’ tumours and found, to their gratification, that the strategy worked: the p53 gene was successfully transferred to the tumour cells; it switched on to produce healthy protein, and the patients suffered no significant side effects.
However, the viral vector, or delivery vehicle, proved poor at evading the sentries of the immune system, and in subsequent prototypes the scientists coated the virus with a substance to give it a better chance of killing tumour cells before being wiped out itself by the immune system. They also changed the delivery vehicle from a retrovirus to an adenovirus – the type that causes the common cold and other respiratory infections.
Therapies based on this design have been tested now in thousands of patients in clinical trials mostly in the US and China. They have proved effective, especially when used in conjunction with conventional chemo- or radiotherapy. Patients also need to be carefully selected for their suitability, since the treatment works better under some conditions than others. It tends to be most effective, for example, in tumour cells in which existing wild-type p53 protein is trapped by over-expression of its natural controller Mdm2; or when mutant p53 protein is produced at such low levels in the cancer cells that it cannot overwhelm the wild-type protein produced by the gene therapy. (You will remember that in some cases where a person has a mutant and a wild-type copy of the p53 gene, the mutant protein is powerful enough to knock out the function of the wild-type protein – the so-called ‘dominant-negative’ effect. Someone with such a powerful mutant will not be a good candidate for the gene-transfer therapy.)
Roth’s pioneering work in the mid-1990s led to the development of two trademarked products, Advexin in the US and Gendicine in China. As well as being injected directly into the tumour, these can be administered by injection into an artery or vein like chemotherapy, and have been tested in a number of tumour types including lung, liver, and head and neck with varying degrees of success. They seem to be effective also, used alone, in preventing early lesions in the mouth from turning malignant. In 2003, Gendicine was approved by the Chinese regulatory authorities for use in the clinic. The first gene-therapy product to receive official approval anywhere in the world, it is used today, in conjunction with radiation, to treat patients with head and neck cancer in China.
In 2007, Zhang Shanwen of Beijing Cancer Hospital, who chaired the clinical trials of Gendicine, gave an indication of its effectiveness. At a conference in China, he presented data from a trial in which 26 patients were treated with gene therapy plus radiation and 27 controls were given radiotherapy alone. Seventeen of the 26 patients who received the combined therapy were still alive five years later, of whom 16 remained completely tumour-free. Of the 27 controls given radiotherapy alone, 14 were still alive five years later, 10 of them tumour-free.
However, despite some remarkable individual success stories and despite being almost identical to Gendicine, the US product, Advexin, has had a rocky ride. The quest to get this product – considered an ‘orphan drug’ because of its limited market potential as a therapy primarily for head and neck cancer – into the clinic has been enormously expensive. When the US Food and Drug Administration (FDA) declined approval in September 2008 because there was not enough evidence of its effectiveness, the manufacturer, Introgen Therapeutics Inc. of Houston, went bankrupt. Today, Vivante, a small company that rose from the ashes of Introgen and was itself acquired in 2010 by the Swiss-based giant Lonza, holds the licence for Advexin and continues the quest for approval from regulatory authorities in the US and Europe. Meanwhile, the Chinese manufacturer of Gendicine is also seeking FDA approval for its product in the US and in India.
In the early 1990s, Frank McCormick at the University of California, San Francisco, began developing a therapy that uses the common-cold virus in a very different way. He had observed that cancer cells and adenoviruses share some important characteristics, one of which is that in order to stay alive they need p53 to be out of action. Here was a trait he could exploit. But it needed a good deal of engineering to ensure that the virus would target and kill only cancer cells and not cause more widespread infection. Essentially, McCormick removed the mechanism by which the virus itself normally knocks out p53 when it enters our bodies. This meant that his engineered virus could survive only in cells which already had no functioning p53 – that is, cancer cells. In these the virus grows and multiplies until the cells literally burst. However, if the engineered virus invades cells with functioning p53 – i.e. non-malignant cells – it withers and dies because it no longer has the machinery to knock out the tumour suppressor. The process by which the cancer cells burst is known as oncolysis, and part of the beauty of McCormick’s mechanism as a therapy is that engineered virus particles spilling from the burst cells can infect and destroy neighbouring cancer cells in the same way, but pose no threat to normal cells in the body.
In 1992 McCormick co-founded Onyx Pharmaceuticals Inc. to develop his idea, and in 1996 the therapeutic agent ONYX-015 entered clinical trials in the US – the first engineered oncolytic virus ever to be tested in humans. Those early trials, first on patients with head and neck cancer and then on those with a variety of other tumour types, looked good. The gene-therapy community was riding high. Then came a body blow.
In 1999, 18-year-old Jesse Gelsinger, who had enrolled voluntarily in a clinical trial at the University of Pennsylvania, died suddenly of multiple organ failure after his immune system over-reacted catastrophically to the agent he was given. The product under trial was an engineered adenovirus carrying a gene to correct the serious but rare liver disorder Gelsinger had been born with that leads to the build-up of ammonia in
the bloodstream. After his death, the FDA temporarily suspended clinical trials of all gene therapy and subsequently tightened the rules on safety precautions. These were difficult times for pharmaceutical companies developing such agents, and in 2003 Onyx sold the licence for ONYX-015 to the Chinese company Shenzhen Si Biono Gene Technologies Ltd.
In the meantime, China itself had been developing an oncolytic agent very similar to ONYX-015. Oncorine, manufactured by Shanghai Sunway Biotech Ltd, was the first engineered oncolytic virus worldwide to reach the medicine cabinet when it was approved by the Chinese regulatory authorities in 2005. This approval raised the spirits of the depressed gene-therapy field. Today Oncorine is used in China in conjunction with chemotherapy (as an alternative to Gendicine) to treat tumours of the head and neck, and the data from trials suggest it is roughly twice as effective as chemotherapy alone. The goal of the Chinese companies is still to obtain approval for ONYX-015 or Oncorine for use in the US and Europe.
However, the propensity of the viral vector to be detected and wiped out by the patient’s immune system before it can deliver its cargo to the cancer cells remains a major challenge for scientists working to refine gene therapy. Another challenge is to find ways of reaching the scattered metastases with these drugs, for it is these secondary tumours that tend to kill the patient with cancer.
SMALL MOLECULES KICK-START STRESS RESPONSE
Other new strategies for treatment being explored start with the fact that in very many cancers p53 is not mutant, but the normal protein is inactivated by some other mechanism. In cervical cancer, for example, around 90 per cent of cases are caused by infection with human papilloma virus (HPV), a sexually transmitted disease that can also cause genital warts. Scientists at the US National Cancer Institute discovered in 1990 that one of the viral genes in cells infected with HPV produces a protein called E6 that completely stymies the action of p53. ‘What happens,’ explained Karen Vousden, one of the NCI team at the time, ‘is that E6 binds to p53 along with some other proteins. The end result is that the p53 protein is degraded very rapidly – it’s just broken up into little bits – so the cell never manages to make any p53 protein that’s functional. It’s as though there isn’t any p53 at all.’ Preventing infection with HPV in the first place was the obvious solution here, and a vaccine capable of doing just that was approved for the market in autumn 2005.
But the HPV story is an unusual one. More typically the normal p53 protein is prevented from carrying out its functions by abnormalities elsewhere in the tumour-suppression pathway, such as over-zealous behaviour on the part of its controller, Mdm2. This, you will remember from Chapter 13, is the gene switched on by p53 that produces a protein that, in its turn, binds to p53 protein and marks it up for destruction. This dance of death between p53 and Mdm2 goes on in an endless cycle, taking about 20 minutes to complete each time. In this way, Mdm2 ensures that the enormously powerful p53 – with its ability to kill cells or stop them dividing – is kept in check until needed. Once researchers began to understand this feedback mechanism, they figured that if they could release p53 from the clutches of Mdm2 by blocking the interaction between the two proteins, they should be able to reactivate normal p53 in cells where it was abnormally restrained – that is, cancer cells. In such cells, they reasoned, p53 is like a loaded gun primed to go off, but with the trigger jammed; the challenge was to find something that would release the trigger.
Working in his Dundee lab, David Lane became, in the late 1990s, the first person to manage to do this, with a tiny molecule that plugged the docking site between p53 and Mdm2. Because of the awkward shapes of proteins in general and the flexibility of p53 in particular, this was a huge technical challenge, Lane commented when I spoke to him for this book. ‘As tough as the search for antiretroviral drugs for HIV?’ I queried. ‘HIV is a very instructive example, actually,’ he replied. ‘When I was growing up as a young microbiologist, I was told there would never be a drug to treat a virus; it would be impossible, because the viruses are so close to the host and they use the host machinery. I was also told you’ll never get a drug that inhibits a protein-protein interaction. And the immune system will never have a role in helping to clear cancer cells. So you know, you get told these things with absolute certainty by people . . . And of course they’re always wrong!’ he laughed.
Today there are at least half-a-dozen drugs in development that disrupt the bond between p53 and Mdm2. The earliest and best known is Nutlin, produced by the pharmaceutical giant Hoffmann-La Roche since 2004 (the drug derives its name from the company’s research institute in Nutley, New Jersey, where it was developed). But the initial excitement generated by Roche’s success soon gave way to serious concern. The first reports of Nutlin described experiments with cells and tissues growing on gels in Petri dishes in the lab. But a mouse experiment reported in 2006 that uncoupled p53 from its controller had a catastrophic outcome that gave everyone working with this strategy pause for thought.
Having engineered a mouse with the gene for the p53 controller Mdm2 knocked out, scientists in Gerard Evan’s lab gave the animal a drug to switch on the tumour suppressor. With no controller, the p53 protein went into overdrive in cells throughout the body, resulting in mass, generalised apoptosis – effectively a mouse that melted. In fact, the mouse with no Mdm2 at all was not a relevant model for Nutlin and other such drugs, which are designed to uncouple p53 from its controller only transiently. However, it did raise questions about how to limit the destructive activity of p53 to the tumour sites. This is something drug developers have worked hard on, and modern versions of the drug show very little activity beyond their target cells.
In the lab, Nutlin has inhibited the growth of cells taken from a wide range of cancers, including colon, lung, breast, skin and blood, and it has shown activity in animal models too. But the results have been puzzlingly inconsistent, says Lane, who works closely with the Roche team on the continuing development of Nutlin. In recent experiments with acute myeloid leukaemia (AML), for example, they treated cancer cells from a number of different patients in Petri dishes; they found that while all the cells went into growth arrest at the same low dose of the drug and some committed suicide, it took 10 to 20 times the ‘normal’ dose to induce apoptosis in others. What was going on? No one is certain yet, though a reasonable hypothesis is that the balance between pro- and anti-suicide proteins active in a cell at the time of treatment affects its sensitivity. Until they understand fully the forces at work, the drug developers will be unable to say exactly which patients with AML should be treated with Nutlin and at what dosages to induce the desired effect, cell death, says Lane.
Meanwhile, researchers are investigating the use of Nutlin in combination with conventional chemo- and radiotherapy, and have found it to be more effective than using either Nutlin or conventional therapy alone in a number of tumour types. One objective of combination therapy is to harness the synergy between the different agents in order to be able to reduce the dosage of the conventional drugs – and thus their distressing side effects for patients – without reducing effectiveness. This is a pressing need in the case of sarcomas, which include bone cancers and are among the most common cancers in children, and here Nutlin looks promising. In 2011, scientists trying to kill sarcoma cells in the lab found they could reduce the amount of some conventional drugs by a factor of 10 when they used them in combination with Nutlin and still achieve the same or a better result as when they used the chemotherapy drugs alone.
Breaking the bond between p53 and its controller Mdm2 is such an attractive option for drug developers that a number of big international pharmaceutical companies, including Merck and Sanofi, and several smaller ones are in the race to get a drug of this kind into the clinic. Until they can do so, however, they still need definitive answers to the vital questions: what effect do drugs of this nature have, if any, in normal cells and exactly how toxic might they be?
Galina Selivanova at the Karolinska Institute in Stockholm is
working on a drug of this design which she has named RITA. She points out that in order to kill cells, it is generally not enough for p53 simply to be present; to become active, the tumour suppressor needs to receive clear signals that the cell is under stress – signals that are likely to be strongest in cancer cells. ‘My hope is that if you have an Mdm2 inhibitor which is not too strong – maybe it’s enough to release just some p53 from Mdm2 – it will not have very drastic effects in normal cells. But in tumour cells, where you have all these signals which are activating p53, it will kill.’
MENDING THE MUTANT
When she left Russia in 1992 with a PhD in bacterial genetics from Moscow University, Selivanova intended to spend just three months of summer gaining experience with work on higher organisms before returning home. However, she joined the lab of Klas Wiman at the Karolinska Institute, discovered p53 and never went back. ‘It was so exciting from the start,’ she told me when I met her at a mutant p53 meeting in Toronto. ‘p53 is unbelievably interesting. Everything you do opens new questions, new perspectives.’ She joined the p53 community just as people were beginning to think seriously about translation – how they might use the wealth of knowledge they had accumulated to improve the treatment of people with cancer. It was a topic with personal significance: Selivanova had seen her own mother die of a brain tumour, and she soon found herself drawn into the quest.
Besides her own work with RITA, she and Wiman have worked together on another drug, known as PRIMA-1, that is turning out to be one of the most exciting p53-based therapies in development. The drug is designed to work in cancer cells where p53 is mutant and the protein it produces misshapen so that it cannot bind to DNA, as it should, in order to switch on other genes. PRIMA-1 is able to restore the mutant protein to its normal shape, and serendipity played a large part in its discovery. In 1995, the two scientists were studying small scraps of protein called peptides, looking for ones that could regulate the activity of p53. They were intrigued to discover one peptide that was able to activate both normal and mutant p53.
p53 Page 23