The Chairman (Dr D. J. P. Hare, F.I.A.): Welcome to the Institute and Faculty’s Spring Lecture.

We are absolutely thrilled to have with us Professor Robert Winston and a wide-ranging, diverse audience. One of the important things that we try to do within the Institute and Faculty is to move actuarial science forward, and to hold events like this that help to stimulate debate and broaden our minds, in the hope that we are helping the public interest by advancing actuarial themes and other topics related to the work and the expertise that we have as actuaries.

This is one of the largest audiences we have ever had, which is a testament to the quality of the speaker that we have.

A few words of introduction: Professor Robert Winston is an eminent speaker. He is a professor at Imperial College London. He is also a professor at the Institute of Reproductive and Developmental Biology, aiming to improve human transplantation.

Robert has had a distinguished career in medicine, in broadcasting and also in the House of Lords. If you want to discuss the Care Bill with him afterwards, he will be happy to discuss the amendments that he has put forward. That is something about which we have been in the press this week, with some research that we have been doing.

Robert, in his spare time, is passionate about classical music, vintage cars and fine wine, as well as a certain football team who performed reasonably well in the League and have a big game at the weekend.

I am going to stop now and say thank you very much for coming, Robert. We look forward to your lecture.

Prof R. Winston: I am going to try a number of different themes out on you, largely because I find actuaries such frightening people.

The Rift Valley in Kenya is where we believe hominids first came to be. Certainly, the earliest pre-human remains are in this area, the eastern part of Africa.

Wherever you look through 360°, there is no sign now of human habitation, so it was quite surprising when there was a buzzing from my pocket. I put this skull under my arm and took my mobile phone out (which does not work in London but was working brilliantly in the middle of the savannah) and it was Richard Dawkins saying that he was objecting to people teaching The Creation in schools and would I sign a letter to The Times about evolution. I said that I would not because I was posing for a photograph about evolution with these skulls as he was talking to me.

In my hand I had the skull of Australopithecus afarensis, who, when they dug her up, was called Lucy because the palaeontologists had the radio on and it was playing The Beatles singing “Lucy in the Sky with Diamonds”. The brain capacity of that creature three and a half million years ago was about 400 ml.

Homo erectus has a brain capacity of around 900 ml. So, in about 1 million years, which in evolutionary time is very short, there has been life on the planet for nearly 4,000 million years – the brain has doubled. That is quite interesting for obvious reasons. The modern human brain is about 1,420 ml, according to my colleagues in Cambridge who have done an extensive meta-analysis in various papers. It is a little bigger in men than it is in women, on average. Women have a brain size that is about 70 ml less. You gentlemen are looking rather pleased with yourselves at that thought. Actually, that is simply because we males are not prepared to ask directions to where we are going. It is not related to intelligence.

I have drawn a timeline here that should represent 100,000 years, which is the time Homo sapiens has been on earth. The intriguing issue is this: there is another timeline which, to the same scale, is about 400 years. Given that it took 1 million years for humans to put a stick on the hand axe so they could throw it and kill at a distance, what has happened in the last 400? Well, in 1599, Shakespeare wrote Hamlet. Then, in 1609, Galileo saw four moons going round Jupiter and was able to show that Copernicus’s theory that we were not a geocentric universe was correct. By 1800, we had Trevithick, who built a steam engine which goes under its own power up a lane in Camborne in Cornwall. The engineering is good enough for the engine to really go, a kind of Puffing Billy machine. The engineering was not perfect because the steering was so bad that it went into a ditch. They could not get it out of the ditch. They went to the nearest pub and toasted themselves on brandy and roast goose. Then they were so drunk that they could not put the fire out afterwards! That is an extraordinary development, as you can see, in a very short time since the Renaissance.

Then about 26 years after Trevithick, we have the opening movement of Opus 131, the C sharp minor quartet by Beethoven. It is an extraordinary piece of music. In fact, we could just sit here and listen to it because it tells us so much about the human mind. This astonishing isolated man writes an iconic piece of music like this that almost changes the mould of chamber music. So it is not just science and engineering that is changing. It is our cultural activity, the whole way we think.

That is what I want to talk to you about. Since that was composed, we can now make portable computing, we have landed on the moon and we can make living organisms in the laboratory. What is happening is remarkable, which is why I am reticent because the idea of predicting where we will be in any aspect of human affairs now is increasingly difficult.

In 1599, Shakespeare might have leant over the parapet of London Bridge, as it was then, spat into the water, looked at the South Bank and said, “This view is not going to change in my lifetime. I have built the Globe Theatre”. Actually, he would have been wrong about that because in 1613 they produced Henry VIII, with the latest technology cannon, set light to the thatch and the whole thing burnt down. But the principle of having a predictable universe, a predictable life, was there to a large extent. Only complete accidents happened.

It is fair to say that he could not possibly have predicted where we would have been culturally or scientifically now.

It makes life particularly demanding when you are trying to work out where you are going to be. So I have finished with the futility of prediction. I will be touching on where I see the future of healthcare, genomics and genetics, whether the human genome is producing what was anticipated, or whether in fact there is something else going on that is more intriguing. I want to talk about our relationship with public health because that is going to be of growing importance.

One of the issues will be the darkness of mental illness and the promises of regenerative medicine, in which people are increasingly heavily involved. I probably will not give myself time to talk much about digital technology; but the digital age and engineering are making a massive difference, particularly with old people. It is going to make a huge difference in how we manage healthcare in the future. We might want to talk about that, perhaps, afterwards.

Finally, and most importantly, is public engagement, which is something that is not usually factored in; and, together with that, must come a better understanding of social sciences. It is intriguing that, while there is a scientific adviser in every department of government, there is not a good social scientist at the top level.

I want to start off with my great-great-grandfather. My great-great-grandfather lived in Kent and did not have any means of transport. So if he went from place to place, he walked. My great-grandfather was rich enough to have a horse and cart. My grandfather in the early 1900s had a mushroom-coloured Ford Prefect, which I remember very well. If you look at the catalogue for the Ford Prefect in the 1960s, the mushroom colour is not in the catalogue. We called it a mushroom colour because there was not “mushroom” inside it!

What is interesting about this is that it is a good example of how we cannot predict how our technology will interface with other aspects of human activity. The intriguing issue now, of course, is that all of us in this room are not only threatened by climate change, which is one issue, but we are also threatened by pandemic infection. The reason for that is we do not travel by a mushroom-coloured Ford Prefect; we travel by aircraft. We travel by aircraft round the world in a way that means that within a few hours, we are exposed to pathogens for which our immune system does not have any defence. That is extraordinary. It is quite interesting to consider how the stockpiling, for example, of anti-viral drugs by the government, which was reasonable to do at the time, turned out to be misplaced because (a) they were not used and (b) it probably paid far more for them than it needed to do. But we, as scientists, were telling the government to do that. Now the government is getting lambasted for doing it, which is fairly ridiculous. It is easy to be wise after the event. Certainly, it is a good example of how different areas of human ingenuity affect ways that we are finding difficult to predict.

In the year 2000, “Ladies and gentlemen, the President of the United States accompanied by Dr Francis Collins and Dr J. Craig Venter” was the announcement of the first sequence of the human genome.

Francis Collins did much of the public sequencing and Craig Venter, who was from the private sector, sequenced his own genome using a different technique. I am going to paraphrase what President Clinton said. It is a good example of how, apparently, a biological development of major importance is taken up in such a big way that it is announced in the media on the major channels with a presidential broadcast in Washington and the Prime Minister of Britain also in the broadcast. The statements they made, together with the statements from the Wellcome Trust about the sequence of the genome, are interesting because they affect some of the things about which you are thinking.

Essentially, it is probably worth listening to what was said by a few people. I am going to play you a clip. I want to show you how science can be tilted by all sorts of different pressures. I have just shown you Clinton talking with Francis Collins on one side and Craig Venter, who is certainly the equal scientifically, on the other side.

I am now going to play you the same broadcast again, but this time it is a broadcast that has been put out by a religious programme. Bear in mind that Francis Collins is a deeply committed Christian and Craig Venter is an atheist. It may seem like a diversion to show you this at all, but you will see why it is not.

“Ladies and gentlemen, the President of the United States is accompanied by Dr Francis Collins”.

“We are here to celebrate the completion of the first survey of the entire human genome. Without a doubt, this is the most important, most wondrous, map ever produced by humankind. Today we are learning the language in which God created life.

When you look at the sequence of the human genome, our own instruction book, these three billion letters that determine all of the biological attributes of a human being, that is a pretty awesome experience. After all, out of this research most of us believe we will arrive at cures for terrible diseases for which we currently do not have much to offer. I think our strongest mandate down through the centuries of the Judaeo-Christian tradition – one of the strongest mandates – is healing.

It is going to make more difference than any other single scientific advance for doctors like me and for medical scientists. It will completely revolutionise the way that people investigate and treat human disease. Instead of wandering around looking for treatments, as happened with antibiotics, or stumbling across them, we will have a map in our hands. When you try to work out what range of illnesses will be influenced by the human genome project, the only one that might not be is traffic accidents”.

Prof Winston: One of the things that I want to suggest to you is that the only thing for which that might be useful is traffic accidents. I say that “tongue in cheek”.

It is interesting to consider what Clinton said. Clinton said that this is the most important map ever drawn by mankind. But the map of Washington, DC is a good deal more useful, so far.

It is also interesting here to see how, for example, in the second clip, the atheist, Craig Venter, is cut out of it completely because it does not suit the story.

Mr Tony Blair, at the other end of the line, who has been scripted by two of the leading scientists in this country, one of whom was the President of the Royal Society at the time (a man whom I deeply respect and a wonderful man) says that this endeavour tells us more about our humanity than any previous human achievement. Yet, what Shakespeare does in 1599 with Hamlet tells us probably more about our humanity than any aspect of the genome sequencing.

The third thing is that the man who funds the British project, Michael Dexter, from the Wellcome Trust, says that this is an invention more important than the invention of the wheel. If so, it is my contention that it has not yet revolved around its axis. I am not saying that it will not, and I am not saying that the genome is not useful, because undoubtedly for scientists it is immensely useful. We can use it in all sorts of ways to understand a whole range of mechanisms better. In terms of changing healthcare, you cannot say it has had a significant impact except in my own field. My field is a rare field, where we can now use the genome to some extent to look at rare diseases that are caused by a single gene defect. There are 6,000 of those diseases, and each one is so rare that only a tiny proportion of people die from them.

The genetic diseases kill children before the age of 3 or 4 usually. The big diseases that kill adults, like diabetes, heart disease, dementia and so on, are not in fact greatly helped so far by the sequencing of the genome. Over the years that I have been interested peripherally in endocrinology and diabetes, I have seen many of my colleagues trying to arrive at a genetic predisposition for diabetes and failing. This common disease is going to be the great scourge in this country with two million people expected to be affected in the next two decades. That is one prediction we can be sure about; a disease that undoubtedly affects every organ in the body.

One of the issues to be aware of is that scientists are likely to get carried away by their enthusiasm, just as all professionals are. We are not to be trusted entirely. We are not objective, although we claim to be objective. Actually, we are as excited and as subjective as anybody. I have published some 320 scientific publications of experiments – many of them are useless. I cannot think of more than a handful of those experiments, probably fewer than 15, where I went into that experiment and the publication without beforehand having an assumption in my mind of what I was going to find and what I was going to prove, because that is the way science is practically done. We do not admit it very often, but it is true.

In the last year data has been published by ENCODE. There is a series of publications and it appears impossible to understand, but we now know that there are something like four million variations that can be seen in sites in the genome that will have an effect on various outcomes, and some of these are not even DNA. Most of them are RNA, for example.

It is going to be significantly more complicated than simply being able to read the genome and then saying that a particular person is going to be susceptible to a specific disease in a real, meaningful way. That is going to be an increasing issue. I will explain in a second where I think we will be going.

It does not stop this kind of nonsense in newspapers. When that data was published, the following day The Sun said we can now have babies to order, because we can use DNA switches to improve the health of babies or change them. This was taken up by the press generally. These switches can control health, physique and looks. There is no evidence that that is true.

However, that impetus has undoubtedly resulted in this kind of endeavour. You can have your genome sequenced for a price. I refer to Carole Cadwalladr in The Observer, who had a five-page spread last year. She had her genome sequenced, and the article was about what she found as a result. First of all, she found that she has the male pattern baldness genome. Well, she could have found that out by looking at her brothers and or her father. That would have done perfectly well. She says in the article that, apparently, she has a sprinter’s power gene, whatever that is. We might forgive her, because the Olympics were going on in London at the time and maybe there was some kind of obsession. I do not believe that Carole Cadwalladr runs, or trains to run. So it would not be very useful to her. She might have late-onset Alzheimer’s because in fact they did sequence an area in the genome, which might give some information about that. But she wisely chose not to look at it, because it is not predictive as far as we know, and it might cause more distress than anything else.

Then she says that she has the mutation for galactosaemia. Finally, we have a disease propensity which might be useful. Galactosaemia is not a pleasant thing to have. The trouble is that that mutation occurs in about one in 16,000 of us, and it is completely unimportant if you have the mutation unless you marry somebody with that mutation. Then your children will have a one-in-four risk of having that disease. The idea of screening for this would be nonsense, because most of the time you would not turn up anything. There is a question here of hanging back. It may, of course, change if we can deal with very big data, but I have some misgivings about that.

Then she said that it was great news because “I do not have a risk for breast cancer because, I have not got the BRCA1 gene”. That is a complete misunderstanding. The breast cancer gene does give you a high probability of breast cancer, but most people with breast cancer do not have that gene. When she does not have that mutation, it reduces her risk of this common disease, which will perhaps affect one-in-nine women, by about 1% or 2%. So it is not as much use as you might think.

There is nobody here that suffers from the waist circumference gene except me. She had apparently had a gene for conscientiousness tested, whatever that is. I was quite intrigued by that so I e-mailed her. I said, “Your editor obviously thinks you are conscientious, because you have a five-page spread. Do you think you are?”. She did not reply to my e-mail!

So where do we go from here? I could go to great lengths in different areas. One of the great hopes of genomics has undoubtedly been drug design. You can look at drug design extensively. Given that one-third of us in this room will develop cancer, having cancer drugs that are going to be associated with our genomics will be useful. The great breakthrough that was announced with huge trumpets and alarums was the fact that one particular patient, after having been given a certain drug, was treated, and all the metastases he had throughout his body just dried up within 14 days. It sounded like an amazing result.

When the figures were looked at more carefully with a proper controlled trial, we see that, in spite of saying that this is the biggest breakthrough in cancer or melanoma treatment for 30 years, the average survival of these patients was still only 5 months. If you think about it, that is not terribly surprising because this drug is an RAF inhibitor, which is a particular mutation that gives rise to a cancer. We evolve at the rate of one generation every 20 years, but the tumour cells in our body are changing constantly, evolving all the time, developing new mutations in a cancer, and there are millions of cells doing that. It is not surprising that those metastases might dry up and then other mutations take over, which are equally damaging so the patient dies anyway. There was a modest cost for this anti-cancer genomic drug. Most of them cost much more because to develop a modern genomic drug might cost around half a billion dollars, with testing many compounds in the process, and about a 13- to 14-year run-in time.

So it is not quite so simple. It may well be that it is not going to be genomics that is going to change this pattern but our understanding of the immune system; for example, where the immune system might be used to kill foreign cells that we have in our bodies in the right sort of way. I think immunotherapy, which is not necessarily going to need genomics much at all, might be increasingly important.

I cannot talk about healthcare without raising another issue. That is mental ill-health. The Mental Health Foundation said in 2007 that mental health presents one of the greatest challenges that current and future generations will face. In health, economic and social terms, the burden created is immense and growing. I could have talked about the Health and Care Bill but I decided not to. We are shockingly inadequate in pretty much every advanced country in how we deal with mental health. Depression, for example, is extremely common. These diseases are massively important economically and to people because they are so common in families. Although there may be a genetic predisposition to depression, it is clearly not a simple predisposition, and so far genomics has not been of any value at all in looking at that.

The issue is this, which is intriguing. You realise how useless we are at making a diagnosis. There are two types of seasonal affective disorder, winter and summer depression.

What is the difference between somebody who has winter depression and somebody who has summer depression, for example? Why is it to do with light? The point is if you take this or any aspect of depression, the phenotype of the disease is completely unknown. It is not defined. When you go to your doctor with diabetes, you go with a symptom. Perhaps you might feel tired. You might be passing urine frequently. You might be wasting away. You may have an infection. But the underlying cause, the phenotype, is definable. With depression you have a symptom and all we can do is treat the symptom, because we do not understand what is going on underneath that. That is a problem. So, definition of a phenotype in mental illness is going to be one of the biggest problems in healthcare.

There are suggestions that we might get there in some way. Unless we do, we are going to be awash with mental ill-health, not just with depression, which is common, but with the more serious diseases that cause severe changes in mental health. In this I would include dementia.

So, what can we do now? For years we have been able to stimulate the brain to understand it. For example, we can stick an electrode in the brain while a patient is on an operating table. They do not even need to be unconscious because the brain does not feel pain. You can work out what areas are working when you stick a needle in. We can also look more passively at electrical activity from the cortex of the brain. Nowadays, we can wear the magic hairnet and map out the brain rather well.

It turns out that electrical recording of the brain is pretty basic and not very discriminatory. In recent years we can look using magnetic resonance imaging. We can see blood flow where areas of the brain are activated so we can map it out quite well. The trouble with this is that it is an artificial arrangement because people sit clamped in a large, hostile machine, which makes a loud noise. Because it takes a long time to get the image you want, you are not seeing blood flow contemporaneous with the stimulus. You are seeing it afterwards. You have to make that judgement. So statistics are difficult.

We can also use positron emission scanning. This involves giving a radioisotope to a person and looking, for example, at metabolism in the brain and therefore working out what the brain is doing where there is metabolic activity. The trouble with that is that resolution is not very high and you cannot repeat it because radioisotopes cannot be given again and again to subjects. Magnetometry is much more benign. We can put magnets over the head and try to map out what is going on and work out why people are thinking or even moving in a particular way, but it does not work very well.

Most recently, we have been able to stick single electrodes into the tiny individual neurons in the brain and work out what the neurons are doing in different areas of the brain. You can do it in humans, although it is difficult to obtain ethical approval, for obvious reasons. It can be done during surgery, for example. But you cannot then do it very accurately. It is mostly done in rats.

The latest development, which is intriguing, is optogenetics.

Optogenetics is a new technique that has been around for only a couple of years. It turns out that we can look at brain function in an extraordinary way. We can modify individual neurons in the cells of the brain. There are about 100 billion neurons in your brain and my brain – probably fewer in my brain, because I am older. Each of those neurons can be modified, or groups of them can be modified, so that they are genetically changed. The genome of that nerve cell will in fact be sensitive to light. There are a number of proteins that will respond to a particular wavelength, purple, green, red, orange and so on.

Some of the neurons, which all have connections, will inhibit certain functions. The ones with long nerve roots will be stimulatory. For example, if you want to work out what the brain is doing, you can target with genetics. You can introduce what will become a dye in individual cells that have the same genetic predisposition, the same function, to the exclusion of the other ones, and then you can shine a light into the head of the animal and you can identify what it does in relation to those things being inhibited or increased. You can see modified behaviour.

Here is one example and the person who worked this out is likely to receive a Nobel Prize. What he has done is take the memory part of a mouse’s brain. He has stained all these cells red and they are consequently sensitive to a particular wavelength of light. He can create a false memory in the mouse. That may seem bizarre to you.

To give you an example, you might put mice in a box where they receive, not a serious electric shock, but one that will inhibit them from going back in again. By stimulating these nerve cells in a particular way, you can create the memory of having been in that box before. In fact, when they see the box they will not go into it even though they have never actually had an electric shock. It is an extraordinary thing. We are able to create false cognition. Memory is just the start, and there will be other things to follow. That may be an intriguing way of trying to define the phenotype of a whole range of diseases, particularly, I believe, in the field of depression, where presumably the reward centres of the brain are not working correctly. Then we can design drugs that are not just blanket drugs that do not work very well but that are targeted to the phenotype that we are interested in. That will be a real possibility in the next decade with techniques like this. It is an example of where science may make a difference.

The problem with genetics can be shown in a Tube map. If you travel from Westminster on the Jubilee Line to Canning Town, you travel about 7 miles. Here is something intriguing. It turns out that in Westminster the average life expectancy of a male is 77 and in Canning Town it is 70. Seven miles makes a 7-year difference.

The average life expectancy in Calton in Glasgow is 54, which is lower than it is in Mali in Africa or in Mozambique in Africa. If you walk a few kilometres to Lenzie it is 82, one of the highest in the world. I think it is only beaten by Israel and one or two other countries. That is extraordinary because the genetics of those two populations is the same. This is not a genetic thing. So what is going on here?

You might think that it is due to smoking, alcohol or obesity. My niece, who is a Cambridge mathematician, looked at that population mathematically. She has done some calculations and she said that when you correct for those factors, it shows that something else is going on. It is not due entirely to those factors that are usually associated with Glasgow, with the dreadful diseases that people suffer from. Something else is happening.

I am willing to bet with you that that something is epigenetic; that it is not the genome, but it is what happens as a response to our environment. That response to our environment might be two generations earlier. That is a very intriguing issue.

Francoise Champagne has looked at both mice and rats in McGill University. It turns out that they all breastfeed in the litter pretty well, but some mothers do not lick or groom their offspring at all well. They deprive them of the usual physical contact after birth that they would expect beyond breastfeeding. It turns out that, when they do this, with the low groomed mice, although their DNA does not change in its letters, some chemicals become attached to one part of the DNA so it becomes methylated. That ultimately changes its function so that when they grow up they have a change in the oestrogen receptor motor region in the brain. This means that these offspring, when they become adults, are more aggressive and are cognitively impaired. They are less able to find their way out of a maze and they are less able to find, for example, a platform under water if they are swimming around and want to be rescued, even though they might have seen the platform before. Normal mice will remember that, but these mice cannot remember so well.

What is intriguing is that this is passed on to the next generation. So what the mother does may affect her offspring as a grandmother. We know that that is increasingly likely in human care as well. For example, we know that boys who smoke at the age of 9, because their sperm is starting to develop in their testicle stem cells, are more likely to have obese children, and that may be passed on to another generation. However, we have not been looking at this for long enough yet.

At the moment I am involved in a big study in Singapore. A intense cohort study on children who were born 4 years ago have been followed since conception. We have followed the mothers from the moment they conceived right through pregnancy, delivery and afterwards.

The researchers are looking at the bacteria they carry, their genes and their epigenetics. We are looking at a whole range of things; and we are starting to find out things that we did not anticipate. This is a good example of where predictions often fall flat. What will be interesting with cognitive testing, done every 6 months with these children, is whether or not there will be a relationship with cognition. I bet that there will be.

Let me show you now – to change the theme slightly – one of my own experiments, which is now very much out of date, but it was at the time regarded as important. It concerns the human embryo of a woman who had lost a previous child at the age of three and a half from one of the rare gene defects. At the time we did not have the genome sequence. At least we had enough of her DNA sequence to know where the defect might be and we were able to identify one single letter that was missing in the three billion base pairs in the human alphabet that made it. That was before the sequencing and with the sequencing it is much more efficient.

The mother said to me that she was adamant that she wanted to try, if at all possible, to avoid another child that might die. She could not face that. But she could not see herself using contraception for the rest of her life. She certainly was not prepared to abstain and she was not prepared to consider an abortion after antenatal diagnosis. She felt that that would be unethical for her.

She asked if there was anything we could do. We worked on this problem for 5 years, devising a technique to drill a hole in the shell of the egg of the human embryo, remove a single cell, knowing that removing one cell in an embryo will not kill it. In fact, many human embryos are deficient with a single cell, or often two cells, missing. We now know from stem cell biology that these cells all have the same potential anyway, so if you lose one cell it is not critical. We made little pipettes in the laboratory. They were made of glass and were very fine, so fine that if you waved them in the air they would just break. However, they looked very robust under the microscope. The embryo was immobilised. Then we made a little hole in the side of the embryo and we sucked out a single cell for genetic analysis.

At the same time, while we were working on making this a safe procedure, by looking at animal models, for example, we also worked on how we might test for the DNA.

We had to invent a way of finding that single letter, which we did, I am glad to say, successfully. We were able to put back inside this particular woman two embryos that we had biopsied in this way and which were apparently normal. It was an act of faith on behalf of the woman.

The two little girls, Lisa and Harriet, had their 25th birthday 2 or 3 weeks ago. The press called them perfect babies, but all we did was to ensure that they did not have that one base pair mutation. This is an example where genomics could work to treat these diseases, but they are rare diseases. Very few people die from these defects. The most common is cystic fibrosis. Now we are improving at dealing with cystic fibrosis, so many of these people live to be adults and give birth to children perfectly happily without massive problems, or with fewer problems than they used to have.

What is intriguing is the report from the journal Science. One of the leading scientists in the world in the field of genetics is saying of our work in Hammersmith that it is now possible to remove a single cell, and we can now typify the DNA. We can run a supercomputer with all the data of the relevant properties of the proteins, and soon we will be able to work out how this will interact with the environment, and then we could make a colour movie to show the mother when we transfer the embryo back into the uterus, and then we will be able to show her her own baby speaking and singing before it is implanted in the womb, when it is still an invisible embryo.

When one of the leading scientists in the world says that of your work, you feel very proud. If you are wise, you do not feel proud for longer than about 6 minutes, because of science delusion.

Beware, as often we talk up these achievements in a way, which makes them completely unrealistic. At the time there was wild enthusiasm for pre-implantation diagnosis, which is now widely used. I wish that I had discussed it with my colleagues and thought about patenting it. I wish I had patented the embryo biopsy, because in my view it has been misused in private clinics, which is something I will come onto in a second because it is relevant to private healthcare.

People began to think that because human embryos are so often abnormal and of course only about 18% of human embryos implant themselves in the womb, we could perhaps improve IVF by taking a single cell away, identifying any defects in that embryo and then not putting that embryo back.

The problem is that so many embryos are mosaic. That is to say, they have different constructs of the DNA in each cell.

For example, an embryo might have three cells, each with a different chromosome 2. A male embryo might have one cell with one copy of the Y chromosome, no Y chromosome in another cell and two Y chromosomes in another. There is massive variety in those individual cells. So this mosaicism is confounding, and therefore you might well predict that the idea of removing a single cell and then analysing it might not be quite as useful as hoped. That does not stop people in a commercial environment from trying it out, particularly when people are desperate to improve their success rate.

Taking a cell away still might tell you when there is a genetic disease, but it will not tell you whether the embryo is capable of being viable. That is the issue. That does not stop it being resolved, and what has happened is this: a good example just published by Dr Voet of Belgium at the Sanger Institute in Cambridge. He has taken single cells and shown, for example, all the genetic sequences and how they are expressing in a single cell in one phase of development and then, a few hours later, in another phase of development. Of course, the gene expression pattern is totally different. What we have here is a massive information overload in just one cell. It will be difficult to compute what is going on because there are literally hundreds of thousands of different genetic influences working here. The idea of looking at a single cell, analysing it, is in my view fantastical.

There are 11 or 12 clinics in London that are charging anything from £2,000 to £5,000 to do this in an attempt to improve an IVF rate. That is one of the issues in the NHS because, as the fragmentation occurs continually, we are going to see this kind of market in healthcare in a range of fields. It happened in IVF early because the NHS does not deal with IVF very well and, when it does, it makes a profit out of it, even though it is not strictly legal to do so. Nobody notices that. Trusts are so short of money that the government has turned a blind eye to it.

This is a real problem. There is a market with the combination of desperate people and scientists who are prepared to try an experimental treatment. What I find extraordinary is that we have a regulatory authority, the Human Fertilisation and Embryology Authority, but it does not stop this from happening. It does not say “This is an experimental treatment; if you want to do this experimental treatment, then you cannot charge for it because it is an experiment. It is not ethical”.

Had we patented embryo biopsy, we could have controlled it. You think of patent in terms of making a profit, but you could have given a licence only to people who were going to use it experimentally or who are going to use it only for things where you knew it might work.

As a result of this burgeoning market, more and more women, in order to try to reduce the number of multiple births so that not too many embryos are put back at the same time, are having their embryos frozen in liquid nitrogen at £250 – maybe £350 – a year; it probably costs more now. It is perhaps worth bearing in mind that the canister contains many embryos and a litre of liquid nitrogen, which might be changed five times a year, and each litre of liquid nitrogen costs a few pence. So where is the cost being passed on to the patient to justify these charges?

You see how that can happen in a privatised system. I do not think that we ever predicted that. In fact, for IVF I did this piece of maths some years ago. I took the staff costs of running a large unit with 2,000 cycles annually, which is what we used to run when I was directing the unit at Hammersmith. The basic staff cost, which is by far your biggest expense, is somewhere around £700. That includes consultation, and everything else. When you add in some of the things that you might want to do, like using a better microscope, and so on, you might add on another £100 or £200.

The last time that an IVF clinic in London was sold, last year, I understand that it went for £400 million. The building would be worth quite a lot of money but the capital equipment might be worth £2 million. A house in Harley Street: what can that be worth? Perhaps £5 million?

I am using IVF here deliberately as a model for why privatisation of healthcare might increase massively, and in a way that we cannot easily control with the kind of reforms that are happening and at a time when a government is strapped for the cash needed. Meanwhile, of course, you can see advertisements on the Tube encouraging people to freeze their eggs. If I advertised on the Tube that I was going to freeze your eggs, the General Medical Council would strike me off. But I can hide behind a clinic. I can own the clinic, and the clinic can advertise on my behalf. The regulatory authority does nothing. One of the issues is that biopsy of embryos, we now know, may cause defects when these animals – in this case, mice – grow up. Humans are not that much different, but you see brain changes in some of these mice that never normally occur and are rather similar to disseminated sclerosis. There is disruption of neurons. We are already seeing that some children, even younger children, who had some of these IVF procedures, may have changes in cognition. For example, when sperm are injected into the egg, there is pretty clear evidence that autism is more common.

By the way, biopsies of the embryo – and there have been several meta-analyses done with randomised control trials – seems to reduce the pregnancy rate by half. That is what the meta-analysis by one of the best groups in Belgium shows. That does not stop it happening. That is an issue, too. Unfortunately, we are not doing enough academic medicine and acting in the findings.

More and more women are recognising that they are ageing. When they were at puberty they had about 300,000 eggs and by the time they reach the menopause 40 years later, they have none left. It is not just one ovulation a month. While you have been listening to me, they have lost, on average, about two or three eggs by the process of cell death, which occurs in all our bodies all the time.

By comparison, the men have made in the same time some 100,000 new sperm, each one genetically unique. So, that inequality is something that we have to work out in our society better to make sure that women are supported through education, higher education and skills training. In my view, late childbirth is not such a bad idea because we now realise in our society that women are more likely to be in a stable position. The question is how we can protect women to make sure that we can do that. This is not the way.

When it came on the market, I did some calculations in my own laboratory and predicted that it would be <10% successful. People were very critical. They said I was just trying to criticise the market. But we have managed to winkle out this information from the regulatory authority. It was not easy to get it out of them. There were 2,200 women treated between 1991 and 2012 who had 20,000 eggs frozen, 253 IVF cycles have been done. Less than 10% have become pregnant and we do not know how many live births there were. We do not know how many abnormal babies there might be in the future, either. That prediction of about 10% is not far short. It is not something that we should be contemplating.

That is why I argue that the issue of prediction affects us biologically, too. However, we get carried away, and it is understandable.

When you look down a microscope and see a whole piece of heart muscle beating at the right time, you think how colossally powerful it is. It is easy to get carried away, given that one-third of us will die of cancer and another third of us will have heart disease. It is obvious that regenerative medicine might be a real issue.

It will be interesting to see how that works out. I do not know whether we can clearly say what will happen. It is not going to be nearly as easy to make organs. We will certainly be able to make tissues. We will be able to make bone marrow. We will probably be able to make sperm and eggs eventually. Whether we will be prepared to fertilise them, I do not know because it looks so dangerous. We will be able to make fat cells. We will be able to remodel tissues like the liver. But making a whole organ will be very complicated.

The idea of using, for example, 3D printing, which has been recently suggested, does not seem to me likely at all.

A paper published about 2 years ago shows that drug design with a genome has really only taken off in cancer treatment. It turns out there are many novel targets that we are looking up in the genome. While this is an increasing market for drug companies, in all the other fields of medicine there has been very little progress, for example, in gastro-intestinal, genito-urinary, immune system drugs and so on. Cancer drugs have taken off but they are not particularly successful in the way that we imagined.

It is relevant – and a digression – to show you how two research councils, The Engineering and Physical Sciences Research Council and the Medical Research Council – I was a member of the latter until recently – has the biggest spend of all. More and more is being spent on health issues. An engineering research council is now spending nearly £80 million on engineering that will deal with older people and what they can do in the home.

How that will pan out in terms of how we manage to protect people remains to be seen. Clearly, it is an important factor.

I mentioned public engagement. One of the real issues is that we are not doing the social science research we need to do. One of the cleverest people of our generation, Steve Jobs, has transformed the world in all sorts of ways. Yet, when he developed pancreatic cancer he does not go to evidence-based medicine, he turns to herbal remedies, with no evidence at all that they are going to work.

What is interesting here is how we can make ourselves more literate, and clearly one of the effects of the digital revolution will be how we can educate the population to protect their own health more effectively. One thing that will be needed is an understanding of why it is we do not readily seek the care that we might expect.

One of the issues is that we still see the science in black and white. A rather famous headline from The Sun says how so many cancers are missed: “14 million dead from a smear scandal”. Of course, it is nonsense. First of all, the statistics do not look at the whole population. But also it implies that, in a cancer test like this, the cancer cells obviously could not be missed by the microscope compared with a normal group of cells. With the way that this is published, any solicitor or bricklayer on the top deck of a London omnibus could say, “Well, I could tell the difference between a cancer and a normal smear”. What is forgotten is that science is not black and white. There are grey areas. That is one of the problems. We always tend to present things in black and white. That is part of the educational process.

I am going to conclude with what we are doing in my laboratory at the moment. One mouse has been genetically modified. Other mice, as a result, have a brown coat. The father has a dominant gene for a black coat so all his offspring should have a black coat but they do not. We have managed to get the genome changed in all four offspring. That is highly efficient. Usually making a transgenic animal – putting a gene into its cells – is highly inefficient (1%–5%).

So far, transgenic technology, which I view as being more important than the genome, has told us a huge amount about genetics, but it has always been very inefficient. However, men are producing sperm all the time. Rather than trying to transfer embryos, which are difficult to get hold of, why not just modify sperm? If we did that we could perhaps end up with sperm that you could introduce to the female and then have a transgenic animal. And it turns out it is highly efficient.

Now, let me give a fatuous example where this might work. The world record for the 1,500 m – the premier Olympic event – was just under 4 minutes 100 years ago, in 1912. The last British person to win the world record was Steve Ovett in three and a half minutes in 1984. Since then we have knocked another 4 seconds off, so in 30 years we have knocked off just 4 seconds. Basically, human performance is, as you might expect, flattening because we cannot fundamentally change our bodies much.

So, suppose we can modify a mouse, for example, so that it runs faster using genes that modify the way that it uses oxygen and how its muscles exercise. In an experiment by Hakimi in Chicago, two mice race each other on a treadmill. One has been modified. The other one is just a normal mouse. The normal mouse gives up after about 200 m. But the modified mouse, which, incidentally, is much slimmer because it uses carbohydrates much more effectively, goes on for 2 hours and it is still running. Can it go on for 3? It loves it, obviously. It can get off any time it wants. Three hours later it is still running. It runs for over 4 hours, amazingly, then has a breather and starts again, and goes on for about 6 hours altogether.

The trouble with making that mouse was that it was very laborious, because it took endless experiments to incorporate the DNA. It was very difficult. But with sperm we might do it. For example, there are two genes: one is stained green and one is pink. You put the two bits of DNA in simultaneously. The pink ones are still going in. They are in the sperm still in the second wave. But what is extraordinary here is this: four live embryos dancing. They have all incorporated a double dose of DNA. So could this mean that if you can modify speed, you could modify cognition or memory or height or beauty or aggression?

We never thought that we would be able to do this. Everybody said that to modify sperm would be impossible. But here it is, just published. The interesting thing about this is that ultimately, I suppose philosophically, if our primal moral argument for our ethics is the sanctity of human life, if we were able to make superhumans simply by artificial insemination, and with a market behind us at a time when the world is in conflict, what price is humanity? What price has human life? That is an intriguing issue because it is one that we do not debate. It is one I suspect we will have to debate fairly soon because we can certainly do this in the pig. It is the same size as the human and it is not more difficult. We do not produce quite as many sperm as pigs do. A pig produces a cupful of semen. I will stop with that thought!

The Chairman: I have learnt more than I was expecting to in this talk. I am grateful for all the information.

Who would like to ask the first question?

Miss K. Lo Dico, F.I.A.: Thank you for an interesting talk. You said that Glaswegian life expectancy between two boroughs showed a discrepancy of 20-odd years.

You said the reason for that was epigenetics, and it is more likely to be social environment.

Do you think it would be more effective to focus on those kinds of quick wins to address social conditions rather than try to address these medical conditions?

Prof Winston: That is what I was trying to imply. I did not spell it out, but you are quite right. That is the message. In the United States, and in Britain and in one or two other places as well, we are brilliant at doing experiments such as I have just explained. This looks at individual mechanisms in a cell in a way that is quite extraordinary. It may have much less impact on healthcare than we may think.

What we are bad at is public health. We cannot explain that 7-mile distance to Canning Town. We cannot explain the journey from Calton to Lenzie. I suppose the big challenge is how we try to work out what we do for public health. It is interesting because it is only in the last couple of governments that we even had a minister for public health.

So far, what we do for public health – mental ill-health is a prime example – is shockingly inadequate. It is extraordinary. Probably some of you will have heard on the radio this week somebody saying that they have travelled to see their child in a mental institution. The child was admitted on a section 400 miles away because there is no bed closer. The parents had to go from Cornwall to Birmingham and then back again. For a person who is mentally ill, that complete lack of any contact with their own family, who should be most supportive, is a very good example of why we have to integrate much better than we seem to be able to do.

Pollution has come up in London recently. It is very clear from our experiments that pollution is way above what it should be. For example, in Exhibition Road outside Imperial College, it is seven times the European limit. There is little doubt that that will have an epigenetic influence. Most people in my field believe that that might have an effect in pregnancy or before conception.

That is why the Singapore study is interesting because now that is being extended to pre-conception. We are able to do that because we are amazingly lucky with a compliant group of Chinese and Malay women who are so interested in what we are doing that they are prepared to submit their children to the indignities that we put them through. We do not hurt them. They see this as an advantage. They receive good care in the process.

Those cohort studies are going to be really important, and sadly we are not doing that kind of study in Britain. Genetic studies, Biobank and so on, are much cruder.

Public health is going to be colossally important; but how we factor that into your field, I do not know. I feel embarrassed about talking about this, because I do not feel that I have given you any useful information. I do not know how you would use this information to predict where you might go, except to be a bit aware that, for example, these people who are peddling their personal genome on the web are probably not doing a great job.

Mr I. J. Kenna, A.I.A.: Why are there continued investigations into the harm done by tobacco and alcohol but no investigations into the possible harm done by various foods, such as Britain’s favourite food, bacon, for example?

Prof Winston: I think there is a slight misconception here. In fact, my colleague, Professor Sir Stephen Bloom at Imperial College, who is one of the world’s leading experts on fat metabolism, is developing a compound which is extraordinary. This might be useful in public health. He has been looking at how the brain understands when it has been sated. It understands when our appetite has been assuaged. He can now mimic that message to the brain with a drug that he has designed. At the moment it is given by injection, and it seems to work. As a drug development, it is worth millions of dollars.

So, work like that is going on and it is important. The work on tobacco may be less than you might imagine. Tobacco research is still done, but not in vast amounts.

Alcohol receives a massive press, but the amount of research that is done is much less than the press would suggest.

For example, huge claims are made about the damaging effects of tobacco and alcohol in pregnancy. But very little research is done. My view is if there is anybody pregnant in this room, do not worry about having a drink of alcohol. It is really not that important. We have become fetishist. We doctors think we know what is best for health.

We all accept that one of the most important health interventions has been the cessation of smoking. Nobody would disagree with that. But the evidence that alcohol in modest amounts is really damaging to, for example, women in pregnancy, is rather dangerous, two and half per cent of babies will be born abnormal. If you are one of those people who have had a drink during pregnancy and you have an abnormal baby, you are convinced that you have done it. To have an advertisement in the antenatal clinic telling you how dangerous it is for your baby for you to smoke or to drink, or to drink coffee, seems to me to be insensitive and dangerous.

We are not good at public health messages, and we are bad at the social science behind those messages.

A member of the audience: Thank you for your presentation. You spoke about how you felt that diabetes was a concern over the next 20 or so years. Is that the type I or type II?

Prof Winston: It is type II. Type II diabetes is the scourge that is nine times more common than type I, which is usually the diabetes that results in a defect in the pancreas itself. Type II is the diabetes which most of us will be likely to have or to have already and is associated with insulin insensitivity. They both have the same effects in terms of organ failure.

Diabetes causes defects in your vessels, in your heart, your brain, the kidneys, your liver, your feet, your skin, your eyes. In every organ you can think of it has a negative effect. The prediction that diabetes in this country might double in the next 15–20 years is likely.

The big problem is that 60% of us are overweight and 30% of us are clinically obese. We have not been able to control our expanding waistlines. There is no question that that puts a massive strain on one’s endocrine system.

The biggest health intervention at the moment would undoubtedly be the control of obesity. The problem is we are thinking very simplistically – in terms of crisps and bacon when it is almost certainly epigenetic, at least to a large degree. There is probably some prenatal experience that babies have that contributes towards that. It is a big problem.

I think we will sort that. For example, our Singapore study is very interesting. In Singapore, we have three racial groups. We have Malays, Chinese and Indians. The biggest single group that we are studying is the Chinese. It turns out that in all three groups the incidence of gestational diabetes, that is abnormal blood sugars in pregnancy, is much higher than was ever predicted. We never realised that was happening. We are able to follow-up those children and see what happens to their waistline. We think they are going to be obese or at least overweight. That is a problem. It is not just a matter of cutting out the foods that make us fat, it is working out how we control what is a much more complicated mechanism that involves the brain, and involves propensities that are not genetic.

It is a fascinating area and there is much research at present. The predictions world-wide are there are or will be about 450 million people with diabetes. Figures are so inaccurate. When I was working with the World Health Organisation in the 1970s as a young scientist, we were being recruited to try to look at contraceptive programmes around the world. I went to India, Bangladesh and other parts of the world on programmes for contraception. We were told that all the predictions suggested that there would be 115 billion people on the planet. Now we are worried about nine billion. It does seem really rather silly, does it not? It is amazing how often those predictions are wrong.

Mr K. B. Donaldson, F.I.A.: You said that pandemics might be more of an immediate risk than, for example, climate change. You have also talked about cancer, diabetes and other diseases. I wondered where you would rank the ineffectiveness of antibiotics among all these terrifying prospects.

Prof Winston: That is a great question but it is also a rather savage question because it turns out that there is a good example of where genetics has not been terribly useful. For example, one of the big scourges is tuberculosis. World-wide it kills about three million people, and now many more people are getting tuberculosis. It was assumed by many of the drug companies – and they worked hard on this – that if you could sequence mycobacterium tuberculosis, the bacterium that causes the disease, you would have a target to attack. So they sequenced it. It was published in Nature, a top journal.

It has been completely useless. It has not helped with the development of any new antibiotic for tuberculosis at all.

Our strategies for new antibiotics turn out to be more difficult. There has not really been a new antibiotic – there have been modifications of old mechanisms. But, fundamentally, there are no new approaches to antibiotic medicine. That is a real problem.

The pandemic risk with bacteria is probably not that great. But with viruses it is much more serious. Viruses are mutating because they are pure DNA. They are mutating in extraordinary ways and may transfer between animals and humans. That was the risk with bird flu. The concern about bird flu was that we might have a pandemic from either pigs or birds. It was pigs in Mexico and chickens in Thailand.

I do not know whether we were lucky or whether it was just an exaggeration of the risk but it is fair to say that virologists are still extremely concerned about that risk, which is probably rising. It turned out that the antiviral we would have given would probably not have worked anyway.

You cannot blame the government for that. Nobody knew. So, to stockpile that drug seemed sensible. The House of Lords Select Committee, on which I was sitting, made the recommendation in good faith, having looked at the evidence produced from the best bacteriologists, not just in Britain but in America. I remember that we phoned Washington at one point, the National Institutes of Health. We had a clear message that we should be stockpiling. We were the only country in Europe that was doing it sufficiently.

It turned out that we wasted that money. That is likely to happen again.

Mr D. C. E. Wilson, F.I.A.: You talked about public engagement and how we need to have sensible debates about some of these issues, and the role of the press is clearly critical in that. I was given an example the other day of a study that said 10% of people have a supergene, which means they are less exposed to the risk of heart disease. That was the headline in the study. When it came out in the press the next day it said 90% of people have a genetic defect that makes them more exposed to heart disease.

If that is the reaction we are going to see in the press, how can we have the debate? How can we make sure people are sufficiently well informed?

Prof Winston: You are right to raise the question. To be fair to the press, the standard of science reporting has immeasurably improved in the last few years. It is a totally different environment now than what it was. We still see ludicrous headlines. I was subject to one of them last week in the Daily Mail about some lecture I had given. It said that I said IVF was going to change genetic modification, which is not what I said at all. The journalist sitting in the audience got it wrong completely.

We are still obsessed with genetics. We think of the gene for intelligence, the gene for waist circumference, the gene for breast cancer. It is so much more complicated, as I showed from the ENCODE data. You might well have changes in the DNA that increase that risk, but that risk is then denied because there are other changes further upstream or downstream, which mean that that risk is not nearly so great. But we do not yet know about it.

The interesting thing for me about the genome is that people are making predictions but they cannot validate the data, because the only way you can validate the data is by looking back at the population and saying, “Look, this was the gene”, or “This was the genetic sequence, and these are the death rates subsequently”.

If you do it prospectively, the problem is that it takes a long time. That is one of the big limitations. We have been unwise in the way we have promoted this. The press, after all, take what we say and, not unreasonably, will try to make it the best story they can.

Even the most hyperbolic newspapers are changing their stance. The thing I remember vividly was when Fukushima went off. Hundreds of people were destroyed in that dreadful accident in Japan. The one newspaper that you would have been worried about reporting that as “We must not have nuclear power in Britain” – which is what happened in Germany – was the Daily Mail. Mike Hanlon, the science reporter, said, “Come on. Let us just be sensible about this. This was a badly designed piece of engineering. There were all sorts of things that we are not doing and things that we can improve. It is not a risk that we are likely to see and we have to see this in a proper context of a much wider picture”.

I thought that was splendid to see in what normally tends to be a sensationalist newspaper – and Paul Dacre (the editor) tends to go for these shock-horror stories – coming out again and again over a period of some weeks with temperate reports, which would not have happened 10 years ago. It is true in medicine as well. Of course, some papers are better than others.

But the problem is every newspaper every week reports a new gene for something. There was one in The Times this morning. I thought about quoting it, because I thought it was quite interesting, but I cannot remember what it was. The appetite thing is interesting. Professor Sir Stephen Bloom’s work is fascinating. It has been a slow process. He has been working on this for quite a long time.

The Chairman: It is interesting that we have talked a lot about modelling. That is one of the things that we, as actuaries, do. Whenever you are modelling you are simplifying the real world. What you have shown us is that the world is even more complicated than we had thought, and it is more complicated than even you and your colleagues were thinking. That is one of the fascinating points as we grapple with the challenge of developing appropriate models.

One of the other things that is dear to us is how you communicate that uncertainty. Many of us are trying to advise firms. How do you find financial security for the firms who provide insurance benefits to people against such a complicated environment?

You have not made the life of the pensions or insurance actuaries any easier, but maybe being armed with more facts and understanding of the uncertainty helps us in the advice that we give, and also encourages us in the research that we do.

The Chairman: I am sure everybody would like to join me in thanking Robert for a fascinating lecture in the usual way.