Features

A new, dispassionate analysis is needed of all the available data on breast cancer screening programmes and their effectiveness, says an Oxford University researcher in an article today in the BMJ medical journal.

Although it saves lives, the benefits and harms of breast cancer screening are perhaps more evenly balanced than many of us might naively imagine.

This has led to increasingly heated debates among specialists in the field who take one side or the other, says Professor Klim McPherson in his BMJ article. These polarised arguments are not helping women to make an informed decision about screening, he argues.

It can be a very difficult thing to assess the benefits of cancer screening programmes – something OxSciBlog has looked at before.

You have to be clear that picking up breast cancer early on through screening leads to improved outcomes (fewer deaths) for those women. You also have to be wary of false positives, or treating cases where perhaps there was no need – the type of cancer involved may not have progressed, for example. This is known as over-diagnosis and over-treatment, and can potentially be harmful.

The BMJ’s press release today also highlights the number of women that need to take part in a screening programme before lives are saved:

'A recent US report on screening for breast cancer estimated that the mortality reductions attributable to breast screening are 15% for women aged 39–49, 14% for those aged 50–59, and 32% for those aged 60–69. Worse still, estimated numbers of women needed to be invited to a US screening programme in order to save one life are high. For the younger group it is nearly 2,000 while in those aged 60–69 it is still nearly 400. In the UK, the figure is around 1,600 for women aged 40–55.'

Klim McPherson, visiting professor of public health epidemiology at Oxford’s Nuffield Department of Obstetrics and Gynaecology, tells OxSciBlog that: ‘There has been a debate about whether the actual benefit of mammography in earlier diagnosis justifies the risk of over-diagnosis and over-treatment it may cause.

‘The evidence on breast screening programmes is that the chance of saving a woman's life is small but real, and the chance of unnecessary anxiety and treatment is also low and real – and not properly discussed. The uncertainty is in the relative size of these two risks for a woman.’

This uncertainty does need addressing, he says, and it may be time to look again at the NHS screening programme thoroughly.

The burden of breast cancer is unremitting and we must do anything we can to contain it, he is clear. But screening for a progressive disease is justified only if earlier diagnosis and treatment improve disease progression enough to justify the harms of screening.

‘Breast cancer screening is no panacea, is expensive and it is irresponsible not to properly assess it, so that women can be better informed,’ says Professor McPherson.

‘The problem is that the debate between cynics and enthusiasts is going nowhere and getting more and more heated. It needs to be properly resolved by, for example, the National Institute for Health and Clinical Excellence (NICE) using all the available data dispassionately.’

In the last of a series of articles marking the 10th anniversary of the first draft human genome, OxSciBlog talks to Professor George Ebers about what we know now about the genetics involved in multiple sclerosis.

Since the sequencing of the human genome, our understanding has greatly increased of the subtle interplay between our genes, our lifestyles, and all that we encounter and are exposed to as we grow up and develop.

We know more about the interactions between our genes, the intricate control systems within the body that switch them on and off, and have some evidence of the way outside factors can link to our genes and make us more susceptible to some conditions.

The idea that we are at the mercy of our genes, that all we are is hard-written into our DNA – if it ever truly held sway – has long gone. It is clear that genetic and lifestyle or ‘environmental’ factors are both important in many complex diseases, from heart disease to obesity and mental health conditions.

Multiple sclerosis, the complex and devastating neurological disease which leads to progressive loss of function in the nervous system, is one condition which offers a clear illustration of this.

Genes linked to MS
Professor George Ebers of the Department for Clinical Neurology says: ‘The sequencing of the human genome has provided a vital tool for understanding the interactions between genes and the environment, how “nature” and “nurture” are bound up together, in many human diseases such as multiple sclerosis.

Over the past 15 years and more, George Ebers has amassed a large database of genetic and lifestyle information from Canadian patients with multiple sclerosis and their families. This has proved to be very powerful in answering many questions about MS.

It’s been known for a long time that multiple sclerosis is associated with the major histocompatibility complex (MHC), a big player in the human immune system. Work by George Ebers and colleagues, though, has demonstrated that the picture is more complicated than originally thought.

They pinpointed the single strongest genetic effect in MS to a variation in one gene called HLA-DRB1 that is involved in the MHC. This is a highly variable genetic region, but one variant that may be present, called HLA-DRB1*15, is strongly linked to susceptibility to the condition.

Unfortunately it’s not quite as simple as having one or two copies of this genetic variant increases your risk of MS by a certain amount. Other variations present in this tight cluster of genes appear to modify further a person’s risk of developing MS. They could increase or decrease the risk associated with HLA-DRB1*15. Further genetic variations in this region appear to affect the severity of the condition where it develops.

It’s a complex picture of genes influencing the effects of other genes – a phenomenon called epistasis. You can’t just consider the effects of each gene individually in isolation.

Getting more sun
But there’s yet more to it than just genes. Where you live in the world has an effect. The further you are from the equator, the greater is your risk of MS. That is, there are more cases in Edinburgh than Marseilles and more in Sydney than North Queensland.

The evidence is accruing that the amount of sunlight people get influences the number of cases of MS that are seen. Sunlight helps the body produce vitamin D, and those populations with lower levels of the vitamin do have more cases of MS. This and other potential health benefits has led to suggestions in Scotland (where rates of MS are the highest in the world) of widely distributing vitamin D supplements.

The group has also identified a modest ‘month of birth effect’, where babies born in May have slightly higher risk of MS than those born in November. One possible reason? Mothers who are pregnant over winter months see less sun and have lower vitamin D levels.

Last year, George Ebers, Julian Knight and colleagues made a direct connection between these observations and the genetics of MS. They showed that vitamin D binds to a section of DNA next to the important genetic region for MS – HLA-DRB1 – and alters the amount of protein that’s produced from the gene.

Here is a route by which genes and the environment (the amount of sunlight a person experiences and their vitamin D levels) interact, and where they are both known to influence MS risk. How this process increases MS risk is unknown, but the finding suddenly allows new studies to investigate.

Enter epigenetics
However, George Ebers thinks this is only one way in which genes and environmental factors interact in multiple sclerosis. ‘Epigenetics – an emerging field that moves beyond the sequence of the DNA code alone to consider other changes in the chemical and physical structure of our genes – is turning out to be the most important of these gene-environment interactions,’ he says.

There are a number of processes occurring on a DNA molecule that can switch genes on and off, controlling their effect. These modifications can be added or removed, so can be temporary, and do not involve changes in DNA sequence. Jane Mellor has described many of these processes previously on OxSciBlog and the effects they can have in disease, in growth and development and across generations.

Epigenetics adds yet another level of complexity, and geneticists are increasingly moving away from considering DNA sequence alone to include these additional chemical markers on the DNA.

Epigenetics is likely to be behind what are some otherwise surprising observations in multiple sclerosis. Data from George’s Canadian study has revealed that mothers tend to contribute more to disease development than fathers. Comparing where nieces and nephews are affected and where there’s also an uncle or aunt with the condition shows this effect. The parent connecting the niece/nephew through the family tree to the uncle/aunt is found to be more likely the mother than the father.

His team are now looking at how the genetic region important for MS might be chemically modified, and how these epigenetic markers get passed down the generations. Such epigenetic changes to the DNA might occur in response to environmental factors to predispose people to disease, perhaps they could occur in the womb for example. Even more confusingly, the genetic trait mysteriously disappears after two or three generations. An article by Michael Gross in a recent issue of Oxford Today explains it like this:

'These epigenetic markers can remain on a gene long enough to be passed on to the next generation or two, but they don't last forever. Studying large families in North America descended from one European and one Native American parent - which is of particular interest, as the MS susceptibility trait is virtually unheard of among Native Americans - Ebers' group found that the disease hit two generations with the predicted pattern, but then disappeared in the third.

Ebers says his group has now shown three ways in which the environment interacts with the genes to influence MS risk. 'Firstly, gene expression in the main genetic region for MS risk is regulated by vitamin D. Secondly, the month of birth effect maps to the same genetic region, and finally we have shown that epigenetic modifications determining MS risk take place in this same region.'

These findings, says Ebers, confirm that the 'genetic cause' that scientists were hunting for so long, really reflects environment-gene interaction, an epigenetic pattern that is passed on for one or two generations but then tends to disappear. 'Multiple sclerosis is the first common disorder with an important epigenetic mechanism', Ebers concludes.'

To investigate the epigenetic changes that are having an effect on MS will be a challenge, however. New experimental techniques will be required to identify and pinpoint those extra chemical or structural modifications to the DNA that are important in disease, says George Ebers.

‘We can recognise clues from genetic epidemiological studies that epigenetics is playing a role in disease, and it is possible to identify disease-specific epigenetic markers. But the pressing challenge in this field is to come up with efficient and reliable sequencing methods for detecting the key epigenetic changes.’

He is confident this will happen. ‘We can expect many new insights from epigenetics in the coming years, in areas from cancer to multiple sclerosis, continuing to build on the knowledge we have gained from the human genomics studies of the last ten years,’ George Ebers says.

‘My hopes are very high and I believe the understanding of epigenetics will represent and embody the most important contributions genetics has to offer. Epigenetics is the interface between genetics and the environment and its understanding holds the key to preventing many common diseases.'

OxSciBlog has been building up to the 10th anniversary of the first draft human genome on Saturday with a series of articles looking at how different areas of research have been affected by knowing our DNA code.

We talk to Dr Deborah Gill of the Nuffield Department of Clinical Laboratory Sciences who, with Dr Steve Hyde, runs a group hoping to develop a gene therapy for cystic fibrosis.

Cystic fibrosis [CF] affects over 8,500 people in the UK and is the most common life-threatening single-gene disorder, says Deborah Gill. It occurs when two copies of a faulty gene are inherited together from a child’s parents. Around 1 in 25 people carry a faulty copy of the CF gene, and if two carriers have a child there’s a 1 in 4 chance of the child having cystic fibrosis.

The loss of this one gene affects many of the internal organs, particularly the lungs and gut where mucus begins to build up. Children tend to be diagnosed with the condition when they fail to thrive as normal and it becomes clear they are having difficulty absorbing food from the gut. Gradually, the patient’s lungs will become clogged with mucus and despite daily physiotherapy, the lungs become progressively more damaged.

‘There is no cure for cystic fibrosis,’ says Deborah Gill. ‘In addition to medication and nutritional supplements, the lung problems are managed using daily physiotherapy and antibiotics to both treat and avoid infections. Half of patients with cystic fibrosis currently live into their thirties.’

While those born today with cystic fibrosis are likely to live longer thanks to improvements in treatment, the number also hides a lot of teenage deaths from the condition, she explains.

Breathing it in
‘When the gene for cystic fibrosis was discovered in 1989, very quickly people began to think about gene therapy as a prospect,’ says Deborah. The aim was to replace the faulty or malfunctioning gene with a working copy wherever it is needed in the body. ‘Since then we’ve come up against every problem there is in gene therapy. It’s the delivery of the gene that’s the tricky bit.’

Although a functioning CF gene is absent throughout the body in cystic fibrosis patients, it is in the lungs where it is a real problem. ‘The lung is our focus for gene therapy as it is the cause of premature death in cystic fibrosis,’ explains Deborah. ‘The lung is essentially open to the atmosphere, so you should be able to simply breathe in a gene therapy. We want to capitalise on the existence of nebulisers for diseases like asthma and use this as a way of introducing a functioning version of the gene in an aerosol to patients.'

‘It’s a golden opportunity to get the gene in an inhaler and allow patients to just breathe it in,’ she says.

Gene therapies that have been investigated have largely fallen into two camps: those that have used types of viruses to get the genetic material into the cells where it is wanted and those that have used capsules made up of fat molecules or lipids.

Any gene therapy wanting to reach the cells in the lung lining will need repeated doses as these cells get renewed and replaced, but methods using current viruses are poor for continuing to get gene expression dose after dose.

Deborah Gill’s group uses lipid particles for their aerosol gene delivery approach. The lipids encase a circle of DNA called a plasmid. The plasmid includes a full version of the CF gene that isn’t functioning in cystic fibrosis patients. They use an aerosol droplet size that gets the gene where it is needed in the airways of the lungs.

Engineering for expression
While other groups have tinkered with the lipid capsules to try and get better results from gene therapy, Deborah Gill and colleagues have worked on engineering the plasmid DNA itself.

‘It’s possible to manipulate plasmid DNA in quite specific and sophisticated ways. It’s a method which is just beginning to be exploited,’ says Deborah. ‘We’ve been able to design the plasmids specifically to get long-term expression of the gene in the lungs.’

Expression describes the process where proteins are made from a gene. The CF gene delivered on the plasmid needs to be expressed continually if the gene therapy is to help relieve chronic lung symptoms in cystic fibrosis patients.

In the beginning, to control expression of the CF gene in their plasmid, the group used a standard DNA sequence, or promoter, taken from a virus. This is an approach that’s been used many times in gene therapy, and it works. You get large amounts of protein made from the CF gene – just not for very long. It turns out that the human body is more subtle than that.

The group found that if you use a general-purpose human promoter sequence in your plasmid instead, you get less CF protein made, but that production is sustained over much longer time periods. It fits much better with the body’s complex control systems that govern gene expression.

The group has also shown that the exact DNA sequence of the plasmid is important too, and needs to be carefully designed at all points. Human DNA tends to have a chemical group – a methyl group – added to the DNA everywhere a ‘C’ in the DNA code appears next to a ‘G’. An un-methylated CG group tends to be recognised as DNA from a virus or bacteria and destroyed. The plasmid DNA in gene therapy is un-methylated, but needs to hang around.

By carefully making sure their plasmid is free of all ‘CG’s in its code, the group has made sure they get expression of the CF gene for longer in the lungs and with less inflammation.

It’s through careful manipulations like this, that the group have fine-tuned their gene therapy approach for cystic fibrosis.

They have now begun clinical trials of their aerosol gene therapy in cystic fibrosis patients. These first studies in a handful of patients are designed to check for safety and to get the dose of the gene therapy right. They are not yet looking for CF gene expression to treat the condition, which will be the next step. But Deborah says she is ‘encouraged’ by the early results.

Prospects for gene therapy
The gene for cystic fibrosis was discovered over 20 years ago, yet we still don’t have a successful gene therapy.

‘It was massively overhyped at the time,’ Deborah admits. ‘As soon as the gene was discovered, many labs started working on a therapy for cystic fibrosis. There was a lot of repetition and this perhaps did the field a disservice. It seemed to many as if gene therapy didn’t work and the new biotechnology companies that had formed couldn’t make a return on people’s investments.’

‘But everyone was trying something new,’ she adds. ‘People hadn’t done gene therapy before. When this started in the early 1990s, no one had a way of making plasmid DNA suitable to give to patients. No methods had been published and there were no companies offering DNA sequences for sale. It was OK for cloning small pieces of DNA in the lab, but not for scaling up to make DNA like a pharmaceutical.’

That’s changed. ‘We are there now,’ she explains. ‘There are companies who can manufacture DNA on a large scale. It just takes a long time.’

In her view, the human genome project added a great deal to the tools available for her work in gene therapy. Although Deborah says she doesn’t get ‘excited’ by genomics – she says it appeals to those who like scale! – she believes the masses of information genomics has produced has ‘provided all the tools to make things really easy’.

‘It has revolutionised gene therapy on a support basis,’ she says. ‘There is great potential for exploiting that information and we haven’t scratched the surface yet.’

And while gene therapy has been slow to fulfil the initial hype and excitement, Deborah sees things changing rapidly in the field.

‘The discovery of micro RNAs – a delicate system where RNA molecules modulate gene expression – gives us a subtle and universal new way to manipulate protein expression,’ she adds. ‘The RNAs act like a rheostat for gene expression – continuously turning it up and down. This has become a really big new area for gene therapy in the last 2-3 years. We are nowhere near seeing just how this will affect the field yet.'

‘But the challenge of delivering the appropriate genetic material to the right part of the body remains the same for all. Still, we are getting a wider array of targets and tools to customise gene therapy for particular diseases.’

Deborah says, ‘I would like to see a world where we have a basket of different viral and plasmid approaches for gene delivery, where we know how to deliver to any organ such as the liver, or heart, and where we can stitch all these different elements together to treat a specific disease. Gradually this is all coming together.’

Streptococcus meets Saving Private Ryan

Humans are the casualties of a bacterial war for survival, new research shows.

The story comes from our friends at the Wellcome Trust who highlight work by Sam Brown of Oxford University's Department of Zoology published in Current Biology today.

Our bodies contain a wide range of bacteria which mostly do us no harm. But now and then a bacterium will evolve properties which are potentially deadly to its human host. The big question for evolutionary theory is, 'why?', when killing off its host could mean the bacterium is killing itself.

Sam explains: 'For many microbes, living in harmony with their host is the best option, so why do some suddenly turn nasty? Sometimes the answer is obvious – for example, the cold virus makes its host sneeze, helping it spread wider. But for other bacteria and viruses, which do not normally cause disease, the reason isn't at all clear.'

The team of Oxford and US scientists modelled in mice how the bacterium Streptococcus pneumoniae interacts with other bacteria, showing that competition for space can cause deadlier forms of bacteria to evolve.

S. pneumoniae usually exists in the nasal passage, where it sits quietly: as many as two in five people in some countries will carry the bug without being aware of it. When S. pneumoniae is forced to share space with Haemophilus influenzae, another common, ordinarily asymptomatic bacterium, the two battle for space.

But H. influenzae has an extra trick up its sleeve, calling on our immune system to help get rid of its competitor by recruiting white blood cells called neutrophils, which surround and attack the S. pneumoniae bacteria.

However, calling in an 'immune strike' on its own position has some unintended consequences.

For, in scenes of microscopic carnage reminiscent of the opening of Saving Private Ryan, whilst S. pneumoniae benefits, as H. influenzae are less able to withstand the immune attacks, when a sufficient amount of H. influenzae are present, the more virulent, armoured strain of S. pneumoniae out-competes its rivals: its thick sugar coating enables it to escape attack from the neutrophils, but, unfortunately for us, this coating also makes it more deadly when it enters the human blood stream.

'Creating a new armour is costly to S. pneumoniae in terms of the energy expended to make it, but it means the bacterium wins the battle with H. influenzae,' Sam tells us.

'However, it also means that if S. pneumoniae enters the blood stream, the immune system is unable to stop its rampant progress. Our bodies are unable to cope and the armoured bug could pay the ultimate price: death to its host and death to itself.'

Dr Sam Brown is a Wellcome Trust Research Career Development Fellow based at Oxford's Department of Zoology.

Thanks to Craig Brierley at Wellcome. 

There were 451 million clinical cases of Plasmodium falciparum malaria globally in 2007, according to new research by the Malaria Atlas Project [MAP].

The new study was led by Simon Hay of Oxford University's Department of Zoology and is published today in PLoS Medicine.

This estimate of how many people become ill because of the disease (which kills about 1m people a year) is almost double the previous one provided by the World Health Organization (it estimated 247m for 2007) and highlights the difficulties of tracking a global disease.

MAP, which is mainly funded by the Wellcome Trust, set out to tackle some of the uncertainties surrounding the number of cases of malaria gobally: estimates for the disease are made particularly difficult because in places where it is endemic diagnosis is often inaccurate and national reporting of cases incomplete.

The researchers used a recently published map of modern-day malaria risk and more advanced statistical techniques that better describe uncertainty. They found that there could have been between 349-552m clinical cases of P. falciparum worldwide in 2007 and came up with a combined estimate of 451m cases.

They also discovered that more than half of the estimated malaria burden and its associated uncertainty was contributed by India, Nigeria, the Democratic Republic of Congo, and Myanmar (Burma).

'The uncertainty in our knowledge of the true malaria burden in a mere four countries, confounds our ability to assess progress in relation to international development targets at the global level. It is clear that we urgently need an increased focus on reliably enumerating the clinical burden of malaria in these nations,' Simon told us.

'The divergence in our estimates and those of the World Health Organization is greatest in Asia and acute in India. We have sought to explore on a country by country basis how these differences arise, the relative uncertainty in the alternative burden estimation approaches and the potential insights that could be gained by hybridising the two.'

Oxford University's Bob Snow, who leads the MAP group in Kenya, said: 'Our estimates for P. falciparum malaria alone are almost twice those provided by the WHO, which include both P. falciparum and P. vivax malaria.'

Bob added: 'Getting the numbers right is fundamental to reporting on success or otherwise of increased donor funding. A valid question remains about whether agencies charged with the responsibility of supporting the delivery of malaria interventions should be the same ones expected to report progress.'