The Rise and Fall of Modern Medicine

by James Le Fanu

  The following year, in February 2001, the two leading science journals, Nature and Science, each published a complete version of that ‘most wondrous map ever produced’ as a large, multi-coloured poster displaying the full complement of around 20,000 human genes.2 It was, as Science observed, ‘an awe-inspiring sight’. Indeed, it was awesome twice over. Back in the 1950s, when Francis Crick and James Watson described the structure of the double helix, they had no detailed knowledge of a single gene, what it is or what it does. Now, thanks to the techniques of the New Genetics, those involved in the Human Genome Project had, in just over a decade, successfully culled from the 3 billion chemicals of DNA strung out along those intertwining strands the hard currency of each of the 20,000 or so genes that determine who we are.

  The human genome map, like Thomas Jefferson’s map of the United States, portrays the major features of the genetic landscape with great precision. Whereas, in the decade prior to its completion, it had taken the best part of seven years to find the defective gene responsible for the common lung disorder cystic fibrosis, now anyone could locate it on that multi-coloured poster in as many seconds. Here too at a glance you can pick out the gene for the hormone insulin, which controls the level of sugar in the blood, or the gene for haemoglobin which transports oxygen to the tissues. To be sure, the functions of many of those genes remained obscure, but now, knowing their precise location and the sequence of which they are composed, it would be only a matter of time before they too would be known. ‘Today will be recorded as one of the most significant dates in history,’ insisted one of the major architects of the HGP, Dr Michael Dexter of the Wellcome Trust. ‘Just as Copernicus changed our understanding of the solar system and man’s place within it, so knowledge of the human genome will change how we see ourselves and our relationships to others.’

  The director of the Project, Francis Collins, spelled out the implications for the future of medicine.3 Within the next few years, he anticipated, scientists would have identified the ‘five to ten’ genes involved in common diseases. By the year 2010 there would, he anticipated, be predictive tests that would inform healthy people of their risk of subsequently developing numerous serious conditions such as diabetes, Alzheimer’s and several forms of cancer. This would, he anticipated, transform the whole process of pharmaceutical innovation, which would focus on developing ‘gene-based designer drugs’. For those with cancer it would even be possible to categorise the genes that had gone awry and target specific therapies tailored to the individual patient, thus minimising the hazard of side-effects. ‘This is a time of dramatic change,’ Francis Collins wrote, where before long most family doctors would ‘become practitioners of gene-based medicine capable of advising their patients on how to enhance their chances of staying well’.

  These optimistic predictions conceal a most ambitious change in direction that needs emphasising. The two major research priorities in the twenty or so years leading up to the Human Genome Project (as outlined in ‘The Brave New World of the New Genetics’) were, first, to identify the genes for those hormones and proteins whose deficiency could be corrected by the techniques of biotechnology – insulin for diabetes, growth hormone for dwarfism, clotting factors in haemophilia, and so on. The second priority was the protracted search for the genes involved in one or other of the genetic disorders due to a defect in a single gene – cystic fibrosis, Huntington’s disease, Duchenne’s muscular dystrophy. The completion of the Human Genome Project would of course enormously simplify this process, resulting in the identification of the genetic mutations responsible for more than 1,000 of these single gene disorders. They are, however, all extremely rare, while the practicalities of doing something about them – whether preventing them by prenatal screening or curing them with gene therapy – remain intractable for the reasons already described.

  The crux then of Francis Collins’s ambitious prospectus might best be described as the ‘relevance imperative’, where genetics would have to transcend its concern with these (relatively rare) single gene disorders to become of general relevance to the mainstream by proffering solutions to the common disorders that afflict tens of millions – diabetes, arthritis, cancer and so on. And even more ambitiously, he suggested the Project would usher in the era of ‘personalised genomics’ where it was possible to imagine a situation where everyone might have their genome sequenced, and the knowledge so generated could not only predict the disorders they might encounter but determine the most appropriate treatments for them to receive.

  This might seem rather over-ambitious. It had taken the best part of a decade and $3 billion to sequence that first human genome; what chance then that ‘personalised genomics’ would ever become a viable proposition? But, as ever, it is a mistake to underestimate the pace of technological innovation, and no more so than here, as over the next ten years the speed of genome sequencing increased 50,000-fold, telescoping that decade down to a few days – and all for a few thousand dollars.4

  The New Genetics, liberated now to investigate the intricacies of the human genome by techniques faster and cheaper by several orders of magnitude, has proved immensely productive: typically a dozen or more groups of scientists working in vast citadels of research routinely generate millions of megabytes of basic biological data every week.5 It has, in short, become Big Science, whose sequencing projects now extend to hundreds of other forms of life, traversing the full range of its complexity from the very simplest bacteria and viruses to the worm, fly, mouse, chimpanzee and many others. Biomedical research has as a result become the major game in town, whose funding, doubling and doubling again to more than $100 billion a year, now dwarfs that of all the other sciences combined.

  The tenth anniversary of the completion of the Human Genome Project in 2010 provided an obvious opportunity to reflect on not just the progress so far but the difficulties encountered and how they might be overcome. Francis Collins, in an article in Nature, ‘Has the Revolution Arrived?’, affirmed it had – with the one important caveat that ‘the consequences for the practice of medicine have so far been modest’.6

  This caveat was spelled out rather more emphatically elsewhere in the journal, with the observation that the notion of the tailoring of treatments based on the knowledge of a person’s genetic endowment was ‘not on the horizon’ – while the prospect of those gene-based designer drugs for common illnesses ‘no longer seems a foregone conclusion’.7,8 The Lancet echoed this scepticism, observing how ‘so far the benefits from the Human Genome Project are scarce’, while noting that ‘therapeutic returns from the substantial investment in genomics are badly needed’.9 Nicholas Wade, science correspondent for the New York Times, noted pessimistically how ‘after ten years of effort’ scientists were ‘almost back to square one in knowing where to look for the [genetic] basis of common diseases’.10 ‘The mountain has laboured’, Steve Jones, Professor of Genetics at London’s University College, observed tartly, ‘and brought forth a mouse.’11

  So what had happened? The official view emphasises the astonishing progress in the speed and cost of sequencing and the prodigious quantities of biological data so generated, but concedes that the practicalities of what those 20,000 human genes actually do, and how they interact together, have proved vastly more complex than originally supposed. This is undoubtedly true – as is its necessary corollary, that it will take much longer to ‘work it all out’ before there can be any hope of reaping all those therapeutic benefits that Francis Collins so confidently predicted.

  But the more substantial – and indeed apparently insurmountable – difficulty is the quite unexpected discovery that 95 per cent of the genetic heritability in the predisposition to common illnesses such as diabetes or arthritis is ‘missing’.12 This sounds very serious and clearly requires clarification.

  Prior to the completion of the Human Genome Project the quest for the genetic defect in those single gene disorders such as sickle cell anaemia or cystic fibrosis was relatively straightforward, thoug
h extremely arduous. Compare the sequence of, say, the haemoglobin genes in those with and without sickle cell disease and it is possible to infer that any difference in the arrangement of the CGAT letters is likely to be the genetic cause. But the situation with common disorders such as diabetes is much more difficult. They are, to start with, ‘polygenic’; that is, the genetic predisposition lies not in a single gene but in the interaction of many genes and the proteins that they code for.

  The possibility of pinpointing what those genes might be derives from the observation that while, by definition, we humans all share an identical complement of genes, the comparison of one genome with another reveals that 10 million or so of those 3 billion CGAT letters will differ between individuals. These single-letter differences (known as Single Nucleotide Polymorphisms or SNPs) might then serve as a reference point or signpost of genetic variation between individuals and thus might account for why some people are predisposed to have diabetes and others are not.

  The relevant investigations pursuing this possibility (known as Genome Wide Association Studies, or GWAS) are massive undertakings, where blood samples from tens of thousands of people are processed through sophisticated sequencing machines that can identify half a million or so of these variant SNPs in a single swoop, the findings being stored and analysed in powerful supercomputers.13

  Some sense of the scale of this enterprise is illustrated by a study involving the collaboration of fifty groups of scientists who examined the SNP genetic variants at 500,000 different positions in the genomes of 17,000 individuals to identify those that might be implicated in seven common diseases – arthritis, raised blood pressure, Crohn’s, heart disease, manic depression and diabetes. The cumulative biological data generated by 600 of these GWAS studies have identified dozens of genetic ‘loci’ – forty involved in determining height, almost thirty for Crohn’s, twenty for obesity and diabetes, and so on. But when it came to adding up the contribution of each, it soon became clear that the genetic basis for these disorders remained as elusive as ever.14,15,16,17

  Thus, it is possible to estimate from twin and family studies that genes contribute between 80 and 90 per cent of the difference between the tall and the short. But the net contribution of the forty or so ‘height genes’ identified by the GWAS add up to less than 5 per cent of that ‘heritability’. So too with diabetes, where the two dozen implicated genes explain less than one part in twenty of its inheritance liability. Put another way, there could be as many as 800 different genes (potentially many more) contributing to these common disorders – each with a tiny predictive value. ‘Our chances of being born with a predisposition to a common illness are not represented by the roll of a single die,’ observed Professor of Genetics Steve Jones, ‘but a gamble involving huge numbers of cards. People, rather than drawing one fatal error, lose life’s poker game in complicated and unpredictable ways. So many small cards can be shuffled that everyone fails in their own private fashion.’18

  It is scarcely necessary to spell out the significance of what Nature subsequently described as ‘The Case of the Missing Heritability’. For the best part of thirty years The New Genetics, as the driving force of medical research, was predicated on the assumption that finding the genetic cause of disease would open the way to new and much more effective treatment. But when it takes the interaction of 800 or more genes (780 of which might as yet remain unknown) to predispose to (say) diabetes, the notion that it might be possible to develop targeted gene-based therapies is indeed ‘no longer a foregone conclusion’. And so too the promise of ‘personalised genomics’ that it might be possible to tailor treatments on the basis of a person’s genetic endowment to maximise their effectiveness, no matter how rapid or cheap the sequencing process might eventually become.19 ‘It is now pretty clear this talk about personalised risk profiles for most common diseases and a whole flood of new drugs targets is wishful thinking,’ remarks David Goldstein of Duke University.

  The New Genetics can scarcely be allowed to fail, as it has become so massive an enterprise, employing legions of scientists, the cost of whose projects routinely run into tens of millions of pounds. So while the findings of those genome-wide association studies may, as Science tactfully puts it, ‘not have broken any floodgates of understanding’, there is no shortage of similar projects organised on semi-industrial lines, capable of generating gigabytes of biological data.20 Still, that ‘missing heritability’ must eventually influence public perceptions as to what genetics can reasonably be expected to achieve. The biology of life is complex, billions upon billions of times more complex than the inanimate physical world. And part of that complexity is that genes for the most part do not act as discreet independent entities with specific properties – that would make them suitable targets for those gene-based therapies. Rather they contribute to a network of interactions with quite different functions in different tissues.

  This is not to suggest that the Human Genome Project was futile. On the contrary it must in time be seen as one of the most influential achievements of the twentieth century – if not quite for the reasons anticipated. Rather we have here, as the historian of science Evelyn Fox Keller puts it:

  One of those rare and wonderful moments when success teaches us humility . . . we lulled ourselves into believing that in discovering the basis for genetic information we have found the ‘secret of life’. We were confident that if we could only decode the message in the sequence of chemicals we would understand the ‘programme’ that makes an organism what it is. But now there is at least a tacit acknowledgment of how large that gap between genetic ‘information’ and biological meaning really is.21

  There is no ready explanation for that gap between ‘genetic information’ and ‘biological meaning’, but the two most important (if hardly acknowledged) findings of the many genome projects on diverse forms of life hint at the deep inscrutability of the relationship between those genetic instructions and the structures to which they give rise. Those projects were predicated on the entirely reasonable assumption that spelling out the full gene sequences of worm, fly, mouse, man and many others must, to a greater or lesser extent, account for those particularities of form and attribute that so readily distinguish one form of life from another. But, on the contrary, the situation has turned out to be virtually the reverse of that predicted, with a near equivalence of a (very modest) 20,000 genes across the whole range of organismic complexity, from a millimetre-long worm to ourselves.22

  Next comes the astonishing revelation of the interchangeability of the regulatory genes, where, for example, the gene that orchestrates the formation of the fly’s compound eye does the same job for our very different camera-type eye, and so on.23 There is, in short, nothing in the genomes of fly and man to account for why a fly should have two wings, four legs and a dot-sized brain and we should have two arms, two legs and a mind capable of understanding the origins of the universe. The genetic instructions must be there, of course, for otherwise the diverse forms of life would not reproduce themselves with such fidelity from generation to generation. But we have moved, in the light of these extraordinary findings, from supposing those instructions are at least knowable in principle to recognizing that we have no conception what they might be.

  It might seem pointless to enquire why this might be so, but the explanation must lie at least in part in the simple elegance of the two intertwining strands of the double helix, which for so long has held out the promise that it might be possible to understand ‘the secret of life’. That simple elegance cannot be because the double helix is simple, but because it has to be simple, if it is to copy the genetic material every time the cell divides.24 And that obligation to be simple requires the double helix to condense within the one-dimensional sequence of the CGAT chemicals of the genes strung out along the two intertwining strands the billion-fold biological complexities of the three-dimensional forms and attributes that so readily distinguish one form of life from another – flies from worms from frogs f
rom humans. This is not to deny the substantial contribution of genetics, but it cannot conceal ‘the higher truth’ that the question of what might conjure the richness of biological meaning from the monotony of that genetic information would seem, in the current state of knowledge, insoluble.

  3

  BIG PHARMA RULES

  The respective contributions of technical innovation and the pharmaceutical industry to the medical expansionism of the last decade might seem to run in parallel with doctors not just doing more but prescribing (vastly) more – in Britain an additional 300 million prescriptions a year in just ten years. This therapeutic enthusiasm has, as noted, proved enormously beneficial to the drug companies, doubling their annual revenues from $400 billion to $800 billion – with just a single drug such as the cholesterol-lowering Lipitor earning its manufacturer Pfizer a staggering $12 billion a year, or almost half Google’s entire annual revenue.

  Those parallel trends can, however, be deceptive. No one would dispute the value of such technical innovations as coronary angioplasty in promptly relieving the anginal symptoms of heart disease, but the benefits of those 300 million additional prescriptions are much more difficult to quantify. Here the suspicion, hinted at in ‘Looking to the Future’, that the drug companies may have orchestrated this massive upswing in drug prescribing to their advantage would seem to be confirmed by the publication within the past ten years of a shelf-ful of highly critical books with such self-explanatory titles as On the Take; Our Daily Meds: How the Pharmaceutical Companies Transformed Themselves into Slick Marketing Machines and Hooked the Nation on Prescription Drugs; Overdosed America; Overtreated: How Too Much Medicine is Making Us Sicker and Poorer; Selling Sickness: How the World’s Biggest Pharmaceutical Companies Are Turning Us All Into Patients; and, the most influential of all, The Truth About the Drug Companies, How They Deceive Us and What to Do About it by Dr Marcia Angell, the chief editor of the most prestigious of all US medical journals, the New England Journal of Medicine.1,2