11. I, ROBOT (SURGEON)

Boston, 7 November 2005

On a cold, bright January morning I am walking south across Westminster Bridge to St Thomas’ Hospital, an institution founded by Augustinian monks in the twelfth century. Today it occupies a white-tiled modernist complex memorably described in Pevsner’s The Buildings of England as ‘monster cubes’.¹ Pevsner rather liked the bold development, but when it appeared on the Thames skyline in the 1970s some of the neighbours were less keen. MPs and peers in the Palace of Westminster, whose agreeable river terrace overlooks the hospital, vented their indignation in Parliament, describing it variously as an ‘icebox’, a ‘monstrosity’, and a ‘travesty of architecture’. Whatever its aesthetic merits, St Thomas’ has a proud tradition of innovation: I am there to observe a procedure generally regarded as the greatest advance in cardiac surgery since the turn of the millennium – and one that can be performed without a surgeon.

The patient is a man in his eighties with aortic stenosis, a narrowed valve which is restricting outflow from the left ventricle into the aorta. His heart struggles to pump sufficient blood through the reduced aperture, and the muscle of the affected ventricle has thickened as the organ tries to compensate. If left unchecked, this will eventually lead to heart failure. For a healthier patient the solution would be simple: an operation to remove the diseased valve and replace it with a prosthesis. But the man’s age and a long list of other medical conditions make open-heart surgery out of the question. Happily, for the last few years another option has been available for such high-risk patients: transcatheter aortic valve implantation, known as TAVI for short.

When I arrive in the cath lab, wearing a heavy lead gown to protect me from X-rays, the patient is already lying on the table. He will remain awake throughout the procedure, receiving only a sedative and a powerful analgesic to render it painless. I am shown the valve to be implanted, three leaflets fashioned from bovine pericardium fixed inside a collapsible metal stent. After being soaked in saline it is crimped on to a balloon catheter, squeezing it from the size and shape of a tube of lipstick to a long, thin object like a pencil.

The consultant cardiologist, Bernard Prendergast, has already threaded a guidewire through an incision in the patient’s groin, entering the femoral artery and then the aorta, until the tip of the wire has arrived at the diseased aortic valve. The catheter, with its precious cargo, is then placed over the guidewire and pushed gently up the aorta. When it reaches the upper part of the vessel we can track its progress on one of the large X-ray screens above the table. We watch intently as the metal stent describes a slow curve around the aortic arch before coming to rest just above the heart. There is a pause as the team checks everything is ready, while on the screen the silhouette of the furled valve oscillates gently as it is buffeted by pulses of high-pressure arterial blood. When Dr Prendergast is satisfied that the catheter is precisely aligned with the aortic valve he presses a button to inflate the tiny balloon. As it expands it forces the metal stent outwards and back to its normal diameter, and on the X-ray monitor it suddenly snaps into position, firmly anchored at the top of the ventricle. For a second or two the patient becomes agitated as the balloon obstructs the aorta and stops the flow of blood to his brain; but as soon as it is deflated he becomes calm again.

Dr Prendergast and his colleagues peer at the monitors to check the positioning of the device. In a conventional operation the diseased valve would be excised before the prosthesis is sewn in; during a TAVI procedure the old valve is left untouched and the new one simply placed inside it. This makes correct placement vital, since unless the device fits snugly in the annulus there may be a leak around its edge. The X-ray picture shows that the new valve is securely anchored and moving in unison with the heart. Satisfied that everything has gone according to plan, Dr Prendergast removes the catheter and announces the good news in a voice that is probably audible on the other side of the river. Just minutes after being given a new heart valve, the patient raises an arm from under the drapes and shakes the cardiologist’s hand warmly. The entire procedure has taken less than an hour.

TAVI is the latest example of the interventional cardiologist’s encroachment on the surgeon’s territory – and, according to many experts, this is what the future will look like. Though available for little more than a decade, it is already having a dramatic impact on surgical practice: in Germany the majority of aortic valve replacements, more than 10,000 a year, are now performed using the catheter rather than the scalpel.² In the UK the figure is much lower, since the procedure is still significantly more expensive than surgery; this is largely down to the cost of the valve itself, which can be as much as £20,000 for a single device. But as the manufacturers recoup their initial outlay on research and development it is likely to become more affordable – and its advantages are numerous. Early results suggest that it is every bit as effective as open-heart surgery, without many of surgery’s undesirable aspects: the large chest incision, the heart-lung machine, the long post-operative recovery.

While it has only recently become a clinical reality, the essential idea of TAVI was first suggested more than half a century ago. In 1965 Hywel Davies, a cardiologist at Guy’s Hospital in London, was mulling over the problem of aortic regurgitation, in which blood flows backwards from the aorta into the heart. He was looking fora short-term therapy for patients too sick for immediate surgery – something which would allow them to recover for a few days or weeks, until they were strong enough to undergo an operation. He hit upon the idea of a temporary device which could be inserted through a blood vessel, and designed a simple artificial valve resembling a conical parachute. Because it was made from fabric it could be collapsed and mounted on to a catheter; in experiments on dogs Davies showed that it was easy to introduce it via the femoral artery and lodge it in the descending aorta. It was inserted with the top of the ‘parachute’ uppermost, so that any backwards flow would be caught by its inside surface like air hitting the underside of a real parachute canopy. As the fabric filled with blood it would balloon outwards, sealing the vessel and stopping most of the anomalous blood flow.³

This was a truly imaginative suggestion, made at a time when catheter therapies had barely been conceived of, let alone tested. But Davies found that his prototype tended to provoke blood clots, and he was never able to use it on a patient. The idea was forgotten, and another two decades passed before anybody considered anything similar. That moment came in 1988 when a trainee cardiologist from Denmark, Henning Rud Andersen, was at a conference in Arizona attending a lecture about coronary artery stenting. It was the first he had heard of the technique, which at the time had been used in only a few dozen patients, and as he sat in the auditorium he had a thought which at first he dismissed as ridiculous: why not make a bigger stent, put a valve in the middle of it, and implant it into the heart via a catheter? On reflection he realised that this was not such an absurd idea, and when he returned home to Aarhus he visited a local butcher to buy a supply of pig hearts. Working in a poky room in the basement of his hospital with basic tools obtained from a local DIY warehouse, Andersen constructed his first experimental prototypes. He began by cutting out the aortic valves from the pig hearts, mounted each inside a home-made metal lattice, then compressed the whole contraption around a balloon.⁴

Within a few months Andersen was ready to test the device in animals, and on 1 May 1989 he implanted the first in a pig.⁵ It thrived with its prosthesis, and Andersen assumed that his colleagues would be excited by his work’s obvious clinical potential. But nobody was prepared to take the concept seriously – folding up a valve and then unfurling it inside the heart seemed wilfully eccentric – and it took him several years to find a journal willing to publish his research. When one finally did in 1992 his paper was largely ignored, and crucially none of the major biotechnology firms showed any interest in developing the device. Andersen’s ‘crazy’ idea worked; but it still sank without trace.

Thinking that nothing would ever come of it, Andersen sold the patent for his device and moved on to other projects. And then at the turn of the century there was a sudden explosion of interest in the idea of transcatheter valve implantation. In 2000 an interventional cardiologist in London, Philipp Bonhoeffer, replaced the diseased pulmonary valve of a twelve-year-old boy using a valve taken from a cow’s jugular vein which had been mounted in a stent and put in position using a balloon catheter.⁶ In his report, Bonhoeffer suggested that the technique might be used to replace other cardiac valves – and in France another cardiologist was indeed already working on doing the same for the aortic valve. Alain Cribier had been developing novel catheter therapies for years; it was his company that bought Andersen’s patent in 1995, and he had persisted with the idea even after one potential investor told him that TAVI was ‘the most stupid project ever heard of’.⁷ Eventually he managed to raise the necessary funds for development and long-term testing, and by 2000 had a working prototype. Rather than use an entire valve cut from a dead heart, as Andersen had, Cribier employed a bioprosthetic one: three leaflets made from glutaraldehyde-treated bovine pericardium, mounted in a collapsible stainless-steel stent. Prototypes were implanted in sheep to test their durability: after two and a half years, during which they opened and closed more than 100 million times, the valves still worked perfectly.⁸

Cribier was ready to test the device in humans, but his first patient could not be anybody eligible for conventional surgical valve replacement, which is safe and highly effective: to test an unproven new procedure on such a patient would be to expose them to unnecessary risk. In early 2002 he was introduced to a fifty-seven-year-old man who was, in surgical terms, a hopeless case. He had catastrophic aortic stenosis which had so weakened his heart that with each stroke it could pump less than a quarter of the normal volume of blood; in addition, the blood vessels of his extremities were ravaged by atherosclerosis, and he had chronic pancreatitis and lung cancer. Several surgeons had declined to operate on him, and his referral to Cribier’s clinic in Rouen was a final roll of the dice. An initial attempt to open the stenotic valve using a simple balloon catheter failed, and a week after this treatment Cribier recorded in his notes that his patient was near death, with his heart barely functioning. The man’s family agreed that an experimental treatment was preferable to none at all, and on 16 April he became the first person to receive a new aortic valve without open-heart surgery.

Over the next couple of days the patient’s condition improved dramatically: he was able to get out of bed, and the signs of heart failure began to retreat. But shortly afterwards complications arose, most seriously a deterioration in the condition of the blood vessels in his right leg, which had to be amputated ten weeks later. Infection set in, and four months after the operation he died.⁹ He had not lived long – nobody expected him to – but the episode had proved the feasibility of the approach, with clear short-term benefit to the patient. Andersen’s animal experiments thirteen years earlier had been received with indifference; when Cribier presented a video of the operation to colleagues they sat in stupefied silence, realising that they were watching something that would change the nature of heart surgery.¹⁰

At first Cribier was only permitted to perform TAVI on desperately sick patients who were too great a risk for surgery. Predictably, many died within weeks, but there were also some astonishing successes: a handful of patients lived for more than five years after being declared untreatable. These results attracted the interest of the large medical device companies, and huge amounts of money were poured into improving the technology. Clinical trials were set up to investigate whether the procedure was as good as its proponents claimed, and in 2010 a large international study concluded that for those too sick for open-heart surgery,fn1 TAVI was far superior to any other treatment.¹¹ Better still, even if a patient was well enough to go under the knife, it made no difference to their survival prospects if they opted for TAVI instead.¹²

When surgeons and cardiologists overcame their initial scepticism about TAVI they quickly realised that it opened up a vista of exciting new surgical possibilities. In 2012 Lars SØndergaard, a medic from Copenhagen, became the first to apply the transcatheter approach to the mitral valve, implanting a prosthesis in an eighty-six-year-old patient.¹³ As well as replacing diseased valves it is now also possible to repair them, using clever imitations of the techniques used by surgeons. If the leaflets no longer close properly, a cardiologist can pin them together with tiny plastic clips,¹⁴ or remodel the entire valve by implanting a stiff piece of plastic around its circumference.¹⁵ The technology is still in its infancy, but many experts believe that this will eventually become the default option for valvular disease, making surgery increasingly rare.

While TAVI is impressive, there is one even more spectacular example of the capabilities of the catheter. Paediatric cardiologists at a few specialist centres have recently started using it to break the last taboo of heart surgery – operating on an unborn child. Little more than a century ago most surgeons believed that the adult heart was out of bounds; the notion of foetal cardiac surgery would have been inconceivable, since it is only in the last thirty years that specialists have been able to diagnose foetal heart conditions with any degree of confidence. Although antenatal scans using ultrasound became commonplace in the 1970s, doctors had no idea how to interpret the grainy black-and-white images of the tiny beating heart. It was ‘like trying to read a foreign language without a dictionary’,¹⁶ in the words of Lindsey Allan, a paediatric cardiologist who in the early 1980s was working at Guy’s Hospital in London. Despite this, she not only devised a method of visualising the most important features of the foetal heart,¹⁷ but also showed that it was possible to differentiate between normal and abnormal anatomy, and to identify specific congenital malformations well before birth.¹⁸

Allan was also involved in the first attempts to correct congenital abnormalities in an unborn child. Her team at Guy’s treated a series of babies with aortic stenosis: shortly after birth they inflated a balloon catheter inside the valve, in the hope that slightly widening it would improve the blood flow and alleviate the condition. Sadly all died, having already sustained irreversible damage to the heart muscle while in the womb. Reasoning that the only hope was to intervene earlier, Allan and her colleagues decided in 1989 to try the procedure in utero. Under local anaesthetic, a stiff hollow needle was passed through the mother’s abdominal wall, into the uterus and through the chest of the foetus, directly into the left ventricle of its heart. This terrifying undertaking would have been inconceivable only twenty years earlier, but ultrasound images allowed the doctors to see the structures through which they were passing the needle. Once the needle was in the left ventricle a guidewire was passed through it and across the narrowed aortic valve; the needle was withdrawn, and a balloon catheter threaded on to the wire and into the heart. During the first procedure the valve could not be reached, and the foetus died the next day. The second attempt was more successful: the balloon was inflated inside the valve, dilating it as intended. The pregnancy continued to term, and a repeat procedure was performed after birth, but the child’s heart muscle was already compromised and it died after five weeks.¹⁹ These first attempts ended in failure – but later technical improvements made it a viable method of treating valve problems in pregnancy, successfully extending the lives of babies with the condition. Then, around the turn of the millennium, doctors started to apply the technique to an altogether more formidable problem.

Nowhere is the progress of cardiac surgery more stunning than in the field of congenital heart disease. Malformations of the heart are the most common form of birth defect, with as many as 5 per cent of all babies born with some sort of cardiac anomaly – though most of these will cause no serious, lasting problems.²⁰ The organ is especially prone to abnormal development in the womb, with a myriad of possible ways in which its structures can be distorted or transposed. Some we have already encountered, but there are dozens of others, some vanishingly rare. Over several decades, specialists have managed to find ways of taming most; but one which remains a significant challenge to even the best surgeon is hypoplastic left heart syndrome (HLHS), in which the entire left side of the heart fails to develop properly. The ventricle and aorta are much smaller than they should be, and the mitral valve is either absent or undersized. Until the early 1980s this was a lethal defect which inevitably killed babies within days of birth, but a sequence of complex palliative operations now makes it possible for many to live into adulthood.fn2

Because the left ventricle is incapable of propelling oxygenated blood into the body, babies born with HLHS can only survive if there is some communication between the pulmonary and systemic circulations, allowing the right ventricle to pump blood both to the lungs and to the rest of the body. Until they close a few days after birth, the ductus arteriosus (between the aorta and pulmonary artery) and the foramen ovale (between the left and right atria) provide routes for oxygenated blood to travel back into the pulmonary circulation. Some children with HLHS also have an atrial septal defect, a persistent hole in the tissue between the atria of the heart which improves their chances of survival by increasing the amount of oxygenated blood which reaches the sole functioning pumping chamber. When surgeons realised that an ASD conferred a survival benefit, they began to create one artificially in babies born with an intact septum, usually a few hours after birth. But it was already too late: elevated blood pressure was causing permanent damage to the delicate vessels of the lung while the child was still in the womb.

The logical – albeit risky – response was to intervene even earlier. So in 2000 a team at Boston Children’s Hospital adopted a new procedure to create an artificial septal defect during the final trimester of pregnancy: they would deliberately create one heart defect in order to treat another. A needle was passed through the wall of the uterus and into the baby’s heart, and a balloon catheter used to create a hole between the left and right atria.²¹ This reduced the pressures in the pulmonary circulation and hence limited the damage to the lungs; but the tissues of a growing foetus have a remarkable ability to repair themselves, and the artificially created hole would often heal within a few weeks. Cardiologists needed to find a way of keeping it open until birth, when surgeons would be able to perform a more comprehensive repair.

In September 2005 a couple from Virginia, Angela and Jay VanDerwerken, visited their local hospital for a routine antenatal scan. They were devastated to learn that their unborn child had hypoplastic left heart syndrome, and the prognosis was poor. The ultrasound pictures revealed an intact septum, making it likely that even before birth her lungs would be damaged beyond repair. They were told that they could either terminate the pregnancy or accept that their daughter would have to undergo open-heart surgery within hours of her birth, with only a 20 per cent chance that she would survive. Dismayed by this gloomy outlook, the VanDerwerkens returned home, where Angela researched the condition online. Although few hospitals offered any treatment for HLHS, she found several references to the Boston Fetal Cardiac Intervention Program, the team of doctors who had pioneered the use of the balloon catheter during pregnancy.

They arranged an appointment with Wayne Tworetzky, the director of foetal cardiology at Boston Children’s Hospital, who performed a scan and confirmed that the condition was treatable. A greying, softly spoken South African, Tworetzky explained that his team had recently developed a new procedure, but that it had never been tested on a patient. It would mean not just making a hole in the septum, but also inserting a device to prevent it from closing. The VanDerwerkens had few qualms about accepting the opportunity: the alternatives gave their daughter a negligible chance of life.²²

The procedure took place at Brigham and Women’s Hospital on 7 November 2005, thirty weeks into the pregnancy, in a crowded operating theatre. Sixteen doctors took part, with expertise drawn from a range of specialisms: cardiologists, surgeons, and no fewer than four anaesthetists – two to look after the mother, and two for her unborn child. Mother and child needed to be completely immobilised during a delicate procedure lasting several hours, so both were given a general anaesthetic. The team watched on the screen of an ultrasound scanner as a thin needle was guided through the wall of the uterus, then the foetus’s chest and finally into her heart – an object the size of a grape. A guidewire was placed in the cardiac chambers, and then a tiny balloon catheter was inserted and used to create an opening in the atrial septum. This had all been done before on other patients; but now the cardiologists added a refinement. The balloon was withdrawn, then returned to the heart, this time loaded with a 2.5 millimetre stent which was set in the opening between the left and right atria.²³ There was a charged silence as the balloon was inflated to expand the stent; then, as the team saw on the monitor that blood was flowing freely through the aperture, the room erupted in cheers.

Grace VanDerwerken was born in early January after a normal labour, and shortly afterwards underwent open-heart surgery, the first of three operations she would need in infancy. After a fortnight she was allowed home, her healthy pink complexion proving that the interventions had succeeded in producing a functional circulation. But just when she seemed to be out of danger, Grace died suddenly at the age of thirty-six days – not as a consequence of the surgery, but from a rare arrhythmia, a complication of HLHS that occurs in just 5 per cent of cases. This was the cruellest luck, when she had seemingly overcome the grim odds against her. Her death was a tragic loss, but her parents’ courage had brought about a new era in foetal surgery. Grace was the first child to have a device implanted in her heart while still in the womb – a landmark procedure which has since been performed successfully on many other foetuses diagnosed with HLHS.

Grace’s story is a dispatch from a new frontier of medicine, a world of microscopic precision and miraculous engineering in which life-threatening heart conditions are treated without recourse to the scalpel. But this is not the only novel method of eradicating the trauma of open-heart surgery. In 1983 the French surgeon Alain Carpentier made a whimsical attempt to imagine the operating theatre of the future. It reads like something out of Buck Rogers, complete with lasers to open the patient’s chest and a mysterious electrical device which renders patients unconscious without any need for anaesthetics. In Carpentier’s mega-hospital of the year 2050, automation has done away with the large surgical teams of the twentieth century. A single surgeon performs a valve repair assisted only by a machine, a robot which fetches instruments and helps with the simpler parts of the procedure.²⁴ Nothing dates faster than a vision of the future, however, and several of Carpentier’s ideas already look like relics of the 1980s. But he did correctly anticipate the use of email to share scans and medical records, and made one other prediction that was surprisingly accurate: fifteen years later he himself performed open-heart surgery assisted by a robot.

If you’re now picturing a chrome-plated android endowed with artificial intelligence, superhuman dexterity and impeccable clinical judgment, you may be disappointed. Surgical robots are not the autonomous machines of science fiction: they cannot think for themselves, but are passive devices which can only carry out instructions given by a human. Instruments held by the robot are controlled by a surgeon sitting at a console a short distance from the operating table – the device is, in effect, an extension of the surgeon’s own hands. The rationale for creating such a machine was not to automate the process of surgery, but to make it possible to operate accurately through an incision too small to admit the human hand.

Surgical robotics is the natural evolution of an idea born in the late nineteenth century, but which began to have a significant clinical impact only thirty years ago. In the 1890s the German surgeon Georg Kelling became preoccupied with the dangers of abdominal surgery. He wanted to find a way of treating patients without making a large and risky incision, and at a medical conference in 1901 demonstrated a new technique which he called ‘koelioscopie’. He made two tiny holes in the abdominal wall of a live dog, through one of which he inserted a cystoscope, a magnifying instrument like a small telescope. Air was then pumped through the second incision to inflate the cavity, giving a clear view of the abdominal organs and creating enough space to manipulate surgical instruments.²⁵ This was the first demonstration of laparoscopic or ‘keyhole’ surgery, a technique which only became common practice many decades later when technological developments – notably the invention of flexible optical fibres – made it a genuine alternative to conventional surgery. The advantages of laparoscopy were soon apparent: in the hands of a competent surgeon the smaller incisions reduced operative risk, allowed the patient to recover more quickly and reduced post-operative pain. In the 1980s there was a great deal of interest in these minimally invasive techniques, and it was not long before surgeons began to ponder their application in operations on the heart and lungs.

The principal aim in using such an approach was to avoid sternotomy, the usual method of opening the chest for open-heart procedures. This entails cutting through the skin and the tissue underneath it before splitting the breastbone with a pneumatic saw; metal retractors are then used to pull the ribcage apart and give the surgeon a good view of the heart and surrounding vessels. Although most people recover from a sternotomy without significant pain, it leaves an unsightly scar down the centre of the chest; the incision may take months to heal; and it can be risky to repeat the procedure in the event that further surgery is needed. There were good reasons for looking for an alternative.

In 1996 two surgeons from the Cleveland Clinic, Delos Cosgrove and José Navia, performed a series of mitral valve repairs using a simple, minimally invasive, technique: rather than opening the chest, they tilted their patient slightly and made an incision between two ribs on his right side. This small aperture allowed just enough space to admit instruments and allow them to open the heart chambers.²⁶ Others adopted the approach, often using a video camera to provide a better view of what they were doing inside the heart. Though the method reduced pain and bleeding, the limitations of working through such small incisions were clear. The instruments were unwieldy – typically twice as long as normal – and even the best high-definition camera gave only a flat two-dimensional image of the complex 3D structures they were trying to cut and sew.

The fundamental challenge – that of performing intricate manipulations in a restricted space – was hardly new. Industry had been using robotics since the 1970s to assemble complex electronics or handle dangerous chemicals, applications where a high degree of precision was required. Smaller desktop robots subsequently became a familiar sight in medical laboratories, so it required little imagination to realise that the technology might be adapted for the operating theatre.²⁷ In 1988 a robot was used for the first time to assist with a prostate operation, and other applications followed, but only a decade later was one used in cardiac surgery.

On 7 May 1998 Alain Carpentier performed the first robotic heart operation, using a prototype machine with two arms designed to replicate the full range of motion of the human hand. The procedure began with a small incision on the right side of the patient’s chest, creating an aperture just 6 by 4 centimetres. Cardiopulmonary bypass was initiated, and once the heart had been taken out of the circulation the right atrium was opened. The camera and robotic instruments were then lowered into the cardiac chambers and Carpentier sat down at a computer console a few feet away from the table. The patient had a large atrial septal defect and an associated aneurysm of the septum; as a 3D image of these features appeared on the video screen, Carpentier had the eerie sense that he was actually inside the organ and inspecting it at close quarters. He slipped his fingers into the motion-sensitive rings controlling the instruments. One of the robotic ‘hands’ held forceps, the other a pair of scissors, a needle or a device for tying knots, as required. As he worked, a computer continuously analysed his movements and eradicated any tremor: even the best surgeon’s hands tremble from time to time. In a procedure lasting several hours, Carpentier was able to cut out the aneurysm and place a patch of pericardium across the septal defect to repair it. The operation was completed without a hitch.²⁸

A few weeks later, Carpentier’s friend and collaborator Friedrich-Wilhelm Mohr applied the technology to mitral valve surgery, performing complex repairs through three small incisions (known as ‘ports’) made between his patients’ ribs. Mohr used a voice-activated robot capable of responding to simple verbal commands; it could also memorise the structures of the heart and move to particular positions automatically. The system proved so effective that he was able to complete the entire operation with only the assistance of a single scrub nurse – fulfilling the prophecy made by Carpentier more than a decade earlier.²⁹ But the new surgical tool passed its most exacting test later in the year, when both Mohr³⁰ and Carpentier³¹ showed that it was possible to perform even coronary artery bypass grafting – a notoriously fiddly procedure – using a robotic assistant.

Although the early results of robotic CABG were excellent, doubts were soon raised about a machine’s ability to suture the tiny coronary arteries with as much accuracy as the human hand. Devices have now been developed to speed up this process: instead of sutures, tiny metal clips like staples are punched through the edges of the blood vessels.³² This is still a new procedure, but current research suggests that – in the short to medium term, at least – the outcome of robotic surgery is just as good as that of the old hand-sewn method.³³ Similarly, after thousands of operations on mitral valve repair the results are no worse than would be expected from a human surgeon. Patients undergoing robotic surgery tend to have a shorter stay in hospital, return to work earlier and experience less pain.³⁴

So is the robot the future of heart surgery? Alas, probably not. When surgeons adopt a new procedure, one question matters above all others: will this save more lives than existing techniques? Robotic surgery may allow patients to spend less time in a hospital bed, but it is no better at keeping them alive than an old-fashioned surgeon. And there are other barriers to its general adoption: robots are hugely expensive to buy and maintain; surgeons require extra training and experience before they become adept in their use; and operations take longer when a robot is involved. One surgeon put it like this: ‘Cardiologists have used innovation to make their jobs easier. We’re managing to make our jobs more difficult.’ The robotic surgeon will supplant its human counterpart only if and when it can do a better job.

The desire to reduce the effects of surgery by using minimally invasive methods has led some surgeons to explore another, rather disconcerting, possibility. In October 1998 surgeons at Guven Hospital in the Turkish capital Ankara began to perform heart surgery on patients who remained wide awake. Instead of a general anaesthetic they were given a spinal injection to numb the entire chest area while the surgeons completed a coronary artery bypass graft on the beating heart. The rationale behind this strange regression to an earlier medical age is twofold: to eliminate the risks inherent to anaesthesia, and to enable constant assessment of the patient’s neurological condition, a useful indicator of rare but dangerous complications.

Today, perhaps the most adventurous pioneer of conscious cardiac surgery is Vivek Jawali, from the Wockhardt Heart Institute in Bangalore, who in 2004 began performing a range of procedures, including CABG, valve replacement and even open-heart repair of congenital defects, without general anaesthetic or mechanical ventilation. In several of his patients a phenomenon previously unknown to medicine was observed: after being attached to the heart-lung machine they spontaneously – and for unknown reasons – stopped breathing. In normal circumstances this would have been an emergency, but because their blood was being oxygenated mechanically there was no cause for alarm.³⁵

Conscious cardiac surgery must count among the most surreal experiences anybody has ever had at the hands of the medical profession – so, just for a moment, imagine it from the patient’s point of view. You’re lying on an operating table, wide awake. Your breastbone has been sawn through to reveal your internal organs – though somebody has considerately hung a surgical drape in front of your face so that you can’t catch a glance of your own heart and lungs. A surgeon has not only stopped your heart but cut it open, and at this very moment is rummaging inside it with a forceps and a needle. Weirdest of all, although you can reply to the anaesthetist’s questions you haven’t needed to breathe for the last hour; the sensation is strangely liberating, like a dream in which the action takes place underwater yet inexplicably you don’t drown. You may find the idea distasteful, even repellent, but most patients who have been through the experience report no discomfort or unpleasant memories of the procedure.

Conscious heart surgery is an intriguing idea which has been taken up with enthusiasm in a few centres in India, China and Japan, but is only ever likely to be useful for a small subset of patients for whom a general anaesthetic is clinically undesirable. Like keyhole surgery and robotics, it amounts to an eye-catching innovation rather than a revolution. Indeed, most recent developments in cardiac surgery fall into this category, with little sense of anything radically new on the horizon, of any breakthrough to compare with TAVI. When interventional cardiologists and surgeons reflect on the future of their field the contrast between their responses is surprising. The cardiologists brim with enthusiasm, and cannot wait to explain how catheter interventions might develop in the coming years. Surgeons have less to say, and say it without any great excitement: many agree that there is little remaining scope for innovation. They now have a tried and tested repertoire of operations to treat all the commonly encountered heart conditions. Most are so effective that there is little need to develop new ones; instead, efforts are concentrated on refining existing methods, reducing mortality and the incidence of complications.

Professional sportspeople like to talk about the ‘aggregation of marginal gains’, the insight that achieving tiny improvements in many different areas will eventually add up to a substantial competitive advantage. The philosophy was pioneered and popularised by Dave Brailsford, the coach under whose guidance the British track cycling team dominated successive Olympics from 2008 onwards. In his determination to improve performance Brailsford studied every conceivable aspect of his cyclists’ lives, from the aerodynamics of their riding positions to the bedding they slept on – even bringing in a surgeon to teach them to wash their hands properly as a safeguard against illness. Cardiac surgery is now following a similar path, looking at how outcomes can be improved by comparatively minor changes to operative technique, anaesthesia, medication and follow-up care. Fifty years ago a typical research paper might describe a novel operation, including pictures and a case report or two. Today it’s more likely to read like an actuarial report, with reams of data subjected to rigorous statistical analysis. This is important work, but it does not exactly quicken the pulse.

It would be wrong, however, to imply that innovation has ground to a halt. Much of the most exciting contemporary research focuses on the greatest, most fundamental cardiac question of all: what can the surgeon do about the failing heart? Christiaan Barnard believed he had the answer, and half a century later transplantation remains the gold standard of care for patients in irreversible heart failure once drugs have ceased to be effective. It’s an excellent operation, too, with patients surviving an average of fifteen years. But it will never be the panacea that many predicted, because there just aren’t enough donor hearts to go round. In 2007 there were 88 people on the UK waiting list for a heart transplant; by 2016 there were 249, even though the number of operations performed rose sharply.³⁶ Fewer than half of patients on the waiting list live long enough to receive a new organ, and some spend a year or more in intensive care before finally undergoing surgery. The lack of available hearts is critical, and there is little that can be done about it: fewer healthy young people die than they used to, a development which is hardly to be regretted.

With nowhere near enough organs available, surgeons have had to think laterally. One new strategy is the use of organs from so-called ‘marginal’ donors: those aged over fifty or who show some signs of disease. In the 1970s most hearts came from accident victims in their teens or early twenties; at one major European transplant centre the average age of a donor organ now stands at fifty-five, and surgeons routinely transplant hearts taken from donors in their sixties.³⁷ Another solution is the use of technology to preserve donor hearts, enabling them to be transported long distances. Conventionally, the organ is packed in ice within minutes of its ceasing to contract, a measure that preserves its tissues for up to four hours. In 2006 a new medical device called the TransMedics Organ Care System appeared on the market, a portable pump-oxygenator the size of a small suitcase which circulates warm blood through the heart while it is in transit. Nicknamed the ‘heart in a box’, it was adopted by Harefield Hospital in 2011 to enable the retrieval of geographically distant donor organs, since it allows a heart to be kept beating and perfused with blood for more than eight hours – long enough to get between any two points in the British Isles.³⁸

Desperation has also led surgeons to challenge one of the oldest orthodoxies about transplantation. For many years the only suitable donor was believed to be a brain-dead patient whose heart was still beating; if its contractions had ceased, ischaemia (interruption of the blood supply) would cause permanent damage to its tissues within minutes of death. This remained the accepted wisdom until 2006, when Stephen Large, a transplant surgeon at Papworth Hospital, proved that a stopped heart could in fact be restarted, and continue to function, up to forty minutes after the death of the patient.³⁹ The discovery suggested an entirely new source of donor organs, although it also raised complex new ethical questions: what criteria should be used to determine death? How soon after the heart had stopped beating was it acceptable to remove it? And how would consent be obtained from family members during the brief window when the organ remained viable? Only after these had been discussed and new protocols agreed was the approach put into clinical practice.

The first modern transplant from a non-beating-heart donor was performed in mid-2014 by surgeons at St Vincent’s Hospital in Sydney, Australia.fn3 The organ came from a donor whose life support was withdrawn at the behest of the family. Half an hour after it had stopped contracting it was withdrawn and attached to the Organ Care System. Fresh blood was now pumped through the organ and as oxygen reached the myocardium it began to beat once more – and continued to do so for four hours as it was transported to Sydney by road.⁴⁰ There it was implanted into the chest of Michelle Gribilar, a fifty-seven-year-old who had been desperately ill with congenital heart failure. Less than a month later she was well enough to go home, and was soon able to walk long distances without fatigue.

This method was used to perform several transplants, but it had significant drawbacks, in particular the fact that the donor’s own blood was used to perfuse the heart while in the Organ Care device – an unsatisfactory arrangement, since at the point of death the blood is full of toxic chemicals produced as a response to stress and injury, which may harm the organ.⁴¹ To get around this problem Stephen Large and his team modified the procedure for Europe’s first non-beating-heart transplant, which took place at Papworth in February 2015. When the donor had been declared dead the heart was restarted while still in the body, with the brain deliberately excluded from the circulation to avoid catecholamine storm – the dramatic release of hormones such as adrenaline which would otherwise flood the body during brain death. Once the heart had been assessed and stabilised it was stopped again, removed from the body and attached to the Organ Care System ready for transplant. The first recipient, a sixty-year-old former boxer called Huseyin Ulucan, did well, but the second was an even more spectacular success. The patient was a man who at twenty-three was dying from heart failure and had spent two months in the hospital’s critical care unit. Eight weeks after his transplant his doctors were taken aback when they received a video of him ‘recuperating’. Stephen Large was astonished: ‘I remember seeing this guy in intensive care [and] thinking, when’s it going to happen, the cardiac arrest? Two, three weeks? And now here he is, doing motocross and water-skiing on one leg.’⁴²

By returning to non-beating-heart donation it is estimated that it may be possible to increase the number of available organs by as much as 20 per cent.⁴³ But this will only help in the short term, and I was surprised to find some transplant surgeons yearning for the demise of their own speciality. One said, ‘I want to see the end of heart transplantation, or transplantation in general. Relying on the organs of somebody who’s died is a strange position to be in. We find ourselves complaining that there isn’t enough death!’ André Simon, the surgeon in charge of the heart transplant programme at Harefield Hospital, suggests that transplantation will continue to play a role, but only for a small number of carefully selected patients. VADs are now improving at such a rate, he believes, that they will eventually be the answer in most cases, and we may one day even have a reliable and permanent artificial heart.⁴⁴

Indeed, a new generation of total artificial hearts is now in development. For decades the only models available were pneumatically powered, a cumbersome arrangement which left patients permanently tethered to an air pump. The development of VADs with tiny rotary electrical motors provided another option, and several companies are now working on artificial hearts based on the same technology. In addition to being much smaller and more efficient than the pneumatic pumps, these devices are far more durable, since the rotors that impel the blood are suspended magnetically and are not subject to the wear and tear caused by friction.⁴⁵ Animal trials have shown promising results, but none has yet been implanted in a patient.

Another type of total artificial heart has, however, recently been tested in humans. Alain Carpentier, who in his ninth decade continues to innovate, has collaborated with engineers from the French aeronautical firm Airbus to design a pulsatile, hydraulically powered device whose unique feature is the use of bioprosthetic materials similar to those used in replacement valves. Unlike earlier artificial hearts, its design mimics the shape of the natural organ; the internal surfaces are lined with preserved bovine pericardial tissue, a biological surface far kinder to the red blood cells than the polymers previously used. The blood is propelled by membranes, which pulsate back and forth under alternating pressure from silicone oil pumped by an electric motor. The device is also the first to regulate its own rate automatically, detecting whether the patient is at rest or exercising. Since the motor is housed inside the body, the patient merely needs to wear a battery on a belt rather than carry a bulky drive unit.⁴⁶ Carpentier’s artificial heart was first implanted in December 2013; although the first four patients have since died – two following component failures – the results were encouraging, and a larger clinical trial is now underway.⁴⁷

One drawback to the artificial heart still leads many surgeons to dismiss the entire concept out of hand: the price tag. These high-precision devices cost in excess of £100,000 each, and no healthcare service in the world, publicly or privately funded, could afford to provide them to everybody in need of one. Where else to look, then? Some believe that the cure for heart failure will not be provided by a machine, but by a farmyard animal. Researchers realised decades ago that the supply of transplant donors was never likely to meet demand, and began to look for an alternative source of organs. In the 1960s the American surgeon Keith Reemtsma had had some success transplanting chimpanzee kidneys into patients, and although the first attempts by James Hardy and Christiaan Barnard to use non-human hearts all ended in failure the idea seemed credible. If the difficulties could be overcome, xenotransplantation – the transplantation of organs from another species – had one huge advantage: in theory, donor animals could be farmed to produce an almost unlimited supply of organs.

Unfortunately the task proved far more challenging than anybody had imagined. The main obstacle is a phenomenon known as hyperacute rejection. In the 1960s the British surgeon Roy Calne began a series of animal experiments in which he transplanted goat kidneys into dogs. He found that within a matter of minutes the organs died, ‘turning port red, maroon, purple, and then black before my eyes’.⁴⁸ The immune system quickly recognises cells from another individual as alien, and begins to attack them – but when the tissue comes from a different species the assault is far more aggressive. In the 1980s, when research into xenotransplantation began to gather pace, many investigators believed that hyperacute rejection could be mitigated if the donor organs came from animals closely related to humans, since the immune system reacts more violently to tissues which are genetically dissimilar. Baboons and chimpanzees were the obvious candidates, but these animals take a decade to reach maturity, which makes breeding impractical. Moreover, the heart of even the largest chimp is too small to sustain the average human’s circulation. In the mid-1980s David Cooper, a British surgeon then working in Oklahoma, became interested in the possibilities of xenotransplantation. Having abandoned the idea of using primate organs he turned his attention to pigs. Unlikely as it may seem, pigs have hearts very similar to ours; they are also easy to breed, and take only a year to reach full size.

Unfortunately, pig hearts also present a far greater immunological challenge. In his early experiments Cooper found that when a monkey heart was transplanted into a baboon it might function for a few days or weeks before rejection set in. When a pig’s organ was used instead, it would last only a few minutes before being destroyed by the baboon’s immune system. It transpired that primates, including humans, produce antibodies against pig tissue – an immune response that is not innate, but which develops in the first months of life. Why we should acquire immunity against pigs was mysterious, since antibodies are usually produced against pathogens that the body has recognised as a threat. Why should babies’ immune systems learn to attack pig tissue months before they are able to eat solids, let alone pork?

After many months, Cooper’s research group discovered that this immune response is not directed specifically at pig tissue, but at a molecule found on the surface of porcine cells. Galactose-α-1,3-galactose is a sugar that plays a role in the biochemistry of all mammals, with the notable exception of primates such as humans. It is also present in the cell membranes of the microbes which colonise our gut shortly after birth, a fact that explains why the human immune system produces antibodies against it. In the first few months of life the body learns to recognise the sugar as a feature of non-human cells, and will attack any structure containing it.⁴⁹

Efforts to get rid of these antibodies stalled until 1992, when a small biotechnology firm in Cambridge, Imutran, unveiled a major breakthrough. Her name was Astrid, and she was a genetically modified, or transgenic, sow. Her DNA had been altered so that her cells produced human decay-accelerating factor (hDAF), a protein specific to human tissues.⁵⁰ It was hoped that this would fool the immune system into recognising porcine organs as human, and thus prevent rejection; indeed, in 1995 Imutran announced that surgeons at Papworth intended to perform the first pig-to-human heart transplants the following year. The era of xenotransplantation seemed to be just months away.⁵¹

The prediction turned out to be horribly premature. Researchers soon found that the immunological barriers were even more complex than they at first appeared. And then experts in another field, microbiology, broached a subject which cast doubt on the entire enterprise: trans-species infection. Transplanting a pig organ into a human might also risk introducing pig viruses into the patient’s bloodstream, with unknown consequences. Many diseases we now think of as human – including HIV, Ebola and measles – originated in animals before crossing the species barrier, so there appeared to be a genuine danger of creating new types of infection. Most concerning of all were a group of organisms called porcine endogenous retroviruses (PERVs), viruses which seemed impossible to eliminate, since they are actually part of the pig’s genome. Laboratory research published in 1997 showed that PERVs were capable of infecting human cells – in a test tube, at any rate.⁵² Nobody knew whether this would also happen in a living body, or what the consequences would be; such setbacks caused many to lose faith in a project which only recently had looked so promising.fn4

Several companies withdrew their funding for xenotransplantation, and the media lost interest. Ironically, though, it was only after their work had fallen out of the public eye that researchers started to make significant progress with the problem of rejection. In February 1997 a team at the Roslin Institute in Edinburgh introduced the world to Dolly, an eight-month-old sheep who instantly became the most famous ungulate on the planet. She had been cloned from a mammary gland cell taken from another sheep – the first time that specialised adult cells had been used to clone a mammal. Her birth was a watershed in genetic engineering, since it demonstrated that scientists had attained a new degree of sophistication in manipulating DNA. One way of making pig tissue compatible with the human body is to alter its biochemistry – and the best way of doing that is to tinker with the pig’s genome. Dolly’s creation told researchers that such a thing might be possible.

On Christmas Day 2001, five healthy piglets were born at a research facility in Blacksburg, Virginia. They looked perfectly normal, but in one respect they were profoundly unlike any pig that had lived before. Their cells lacked a molecule called α1,3-galactosyltransferase, an enzyme responsible for the synthesis of the galactose sugar which provokes hyperacute rejection when detected by the human immune system. Scientists from the University of Pittsburgh had succeeded in ‘knocking out’ the gene that triggers the production of the enzyme, so that the pigs’ cells were incapable of manufacturing the sugar.⁵³ Creating these so-called ‘Gal knockout’ pigs was an important step towards a rejection-free xenograft, and David Cooper’s team in Boston subsequently transplanted several Gal knockout pig hearts into baboons. Two were given normal pig hearts, and hyperacute rejection killed them both within twenty minutes. But the Gal knockout organs caused no such reaction, and one animal lived for six months with its new heart – a dramatic improvement on previous attempts at xenotransplantation.⁵⁴

This was only the first piece in a complicated jigsaw, however: the animals eventually succumbed to rejection, indicating that there remained other biochemical processes to be mastered. Research established that other parts of the immune system were also implicated, and that the blood coagulation systems of pigs and primates are fundamentally incompatible.⁵⁵ In 2009 a group in Germany bred triple-transgenic pigs to overcome several rejection mechanisms simultaneously, a development which produced the best results so far.⁵⁶ One baboon lived for a year with a porcine heart implanted in parallel with its own – the experiment was testing for rejection rather than assessing whether the transplanted organ could sustain the circulation on its own.⁵⁷

Research is still identifying further immunological mechanisms which need to be overcome before xenotransplantation can be applied to humans. The transgenic pigs currently being used have as many as six genetic manipulations to alter their biochemistry, and more may be necessary. When baboons are given porcine organs they require immunosuppressive drugs, as do patients who receive a conventional transplant, but the ultimate goal is to produce a pig heart that the human body cannot distinguish from its own tissue. Clinical trials may finally be on the horizon, although they will not be given the go-ahead until researchers can demonstrate consistent long-term survival in primates with a xenograft which actually sustains the circulation.⁵⁸

Research into xenotransplantation – not just for hearts, but kidneys, pancreatic cells and livers – is now going on in several countries including China, South Korea and the US. David Cooper, who has worked in the field for over thirty years, believes that it will not be long before we can reap the benefits: ‘You’re not going to have to get up in the middle of the night to fly somewhere to get a heart; it won’t come from somebody who’s brain-dead; you’ll be able to do the transplant the next morning on the routine operating list. The patient won’t be in an intensive-care unit for three months waiting for a donor – we’ll be able to say, we’ll transplant him tomorrow.’⁵⁹ It’s an enticing prospect, but there aren’t many surgeons who share his optimism. Several cite Norman Shumway, the father of transplantation, who used to joke that ‘Xenografting is the future of transplantation, and always will be.’⁶⁰ In addition to the practical difficulties, there are significant ethical and psychological questions which still need to be answered. Many patients may have religious or moral objections to receiving a heart from a pig, or even see it as a threat to their humanity. And so far, with clinical use still some way off, doctors have barely attempted to gauge public opinion on the subject.⁶¹

Surveying the landscape of possible treatments, there is surprisingly little consensus on the most promising option for heart failure: some proponents of VADs are dismissive of the total artificial heart, while transplant enthusiasts suggest that mechanical devices will always remain too expensive for general use. And there is one still more tantalising notion: that we will one day be able to engineer spare parts for the heart, or even an entire organ, in the laboratory.

Growing tissues outside the body was first proposed in 1897 by Leo Loeb, a German pathologist working in Chicago. While investigating wound healing in the skin of guinea pigs, he succeeded in transplanting cells extracted from a guinea pig foetus and then embedded in a nutrient medium, a mixture of agar and blood serum.⁶² Finding that they continued to grow while isolated from the blood and neighbouring tissue, Loeb suggested that by using similar methods it might be possible to cultivate tissues outside their usual biological environment. By the end of the next decade growing cells in vitro (from the Latin, ‘in glass’) had become an indispensable laboratory technique, but it was not until much later that anybody contemplated using it to build specific structures to repair the body. In the 1980s surgeons began to fabricate artificial skin for burns patients, seeding sheets of collagen or polymer with specialised cells in the hope that they would multiply and form a skin-like protective layer.⁶³ But researchers had loftier ambitions, and a new field – tissue engineering – began to emerge. In an influential article published in 1993, two of its pioneers, Joseph Vacanti and Robert Langer, defined this nascent discipline as applying ‘the principles of biology and engineering to the development of functional substitutes for damaged tissue’.⁶⁴

High on the list of priorities for tissue engineers was the creation of artificial blood vessels, which would have applications across the full range of surgical specialisms. In 1999 surgeons in Tokyo performed a remarkable operation in which they gave a four-year-old girl a new artery grown from cells taken from elsewhere in her body. She had been born with a rare congenital defect which had completely obliterated the right branch of her pulmonary artery, the vessel conveying blood to the right lung. A short section of vein was excised from her leg, and cells from its inside wall were removed in the laboratory. They were then left to multiply in a bioreactor, a vessel which bathed them in a warm nutrient broth, simulating conditions inside the body. After eight weeks they had increased in number to more than 12 million, and were used to seed the inside of a polymer tube which functioned as a scaffold for the new vessel. The tissue was allowed to continue growing for ten days, and then the graft was transplanted. During a straightforward procedure the blocked artery was removed, and the tissue-engineered vessel sutured in its place. Two months later the polymer scaffold around the tissue, designed to break down inside the body, had completely dissolved, leaving only new tissue which would – it was hoped – grow with the patient.⁶⁵

The limitation of this technique was that the tissue used to engineer the new artery could only come from another blood vessel, since the cells which make up the lining of these vessels – the endothelium – are highly specialised. But at the turn of the millennium a new world of possibility opened up when researchers gained a powerful new tool: stem cell technology. In contrast to the endothelial cells in the blood vessels (for instance), stem cells are not specialised to one function but have the potential to develop into many different tissue types. One type of stem cell is found in growing embryos, and another in parts of the adult body, including the bone marrow (where they generate the cells of the blood and immune system) and skin. In 1998 James Thomson, a biologist at the University of Wisconsin, succeeded in isolating stem cells from human embryos and growing them in the laboratory,⁶⁶ resulting in a wealth of new research into how cells differentiate and possible new therapies. But an arguably even more important breakthrough came nine years later. Shinya Yamanaka, a researcher at Kyoto University, showed that it was possible to genetically ‘reprogram’ skin cells and convert them into stem cells.⁶⁷ The implications were enormous. In theory it would now be possible to harvest mature, specialised cells from a patient, reprogram them as stem cells, then choose which type of tissue they would become.

One possible application of this discovery is in the treatment of heart attacks. When the cardiac muscle is damaged by an interruption in its blood supply the body does little to repair it: scar tissue forms, but few new muscle cells, or myocytes, appear. Existing methods of treatment – whether drugs, CABG or stenting – do nothing to restore the tissue which has been injured, so an effective means of replacing lost myocardium would be a major therapeutic breakthrough. To encourage the growth of new muscle, scientists first tried extracting adult stem cells from the patient’s bone marrow and injecting them into the coronary arteries, in the hope that some would adhere to the myocardium and convert into myocytes – but the results were disappointing.fn5

Rather than engineer new tissue in situ, Sanjay Sinha, a cardiologist at the University of Cambridge, is attempting to grow a ‘patch’ of artificial myocardium in the laboratory for later implantation in the operating theatre. His technique starts with undifferentiated stem cells, which are then encouraged to develop into several types of specialised cell: not just myocytes, but also smooth muscle tissue and vascular cells. These are then seeded on to a scaffold made from collagen, a tough protein found in connective tissue. The presence of several different cell types means that when they have had time to proliferate, the new tissue will develop its own blood supply. Clinical trials are still some years away, but Sinha hopes that one day it will be possible to repair a damaged heart by sewing one of these myocardial patches over areas of muscle scarred by a heart attack.⁶⁸

Using advanced tissue-engineering techniques, researchers have already succeeded in replicating structures more complex than myocardium, including the creation of replacement valves from the patient’s own tissue. This can be done by harvesting cells from elsewhere in the body (usually the blood vessels) and breeding them in a bioreactor, before seeding them on to a biodegradable polymer scaffold designed in the shape of a valve. Once the cells are in place they are allowed to proliferate before implantation, after which the scaffold melts away, leaving nothing but new tissue.⁶⁹ The one major disadvantage of this approach is that each valve has to be tailor-made for a specific patient, a process which takes weeks. In the last couple of years a group in Berlin has refined the process by tissue-engineering a valve and then stripping it of cellular material, leaving behind just the extracellular matrix, the structure which holds the cells in position. The end result is therefore not quite a valve, but a skeleton on which the body lays down new tissue.⁷⁰ Valves manufactured in this way can be implanted, via catheter, in anybody; moreover, unlike conventional prosthetic devices, if the recipient is a child the new valve should grow with them.⁷¹

If it is possible to tissue-engineer a valve, then why not an entire heart? For many researchers this has come to be the ultimate prize, and the idea is not necessarily as fanciful as it first appears. In 2008 a team led by Doris Taylor, a scientist at the University of Minnesota, announced the creation of the world’s first bioartificial heart. They began by pumping detergents through hearts excised from rats. This removed all the cellular tissue from them, leaving a ghostly heart-shaped skeleton of extracellular matrix and connective fibre, which was used as a scaffold on to which cardiac or blood-vessel cells were seeded. The organ was then cultured in a bioreactor to encourage cell multiplication, with blood constantly perfused through the coronary arteries. After four days it was possible to see the new tissue contracting, and after a week the heart was even capable of pumping blood – though only 2 per cent of its normal volume.⁷²

This was a brilliant achievement, but scaling the procedure up to generate a human-sized heart is made far more difficult by the much greater number of cells required. Surgeons in Heidelberg have since applied similar techniques to generate a human-sized cardiac scaffold covered in living tissue. The original heart came from a pig, and after it had been decellularised it was populated with human vascular cells and cardiac cells harvested from a newborn rat.fn6 After ten days the walls of the organ had become lined with new myocardium which even showed signs of electrical activity.⁷³ As a proof of concept the experiment was a success, though after three weeks of culture the organ could neither contract nor pump blood.

Growing tissues and organs in a bioreactor is a laborious business, but recent improvements in 3D printing offer the tantalising possibility of manufacturing a new heart rapidly and to order. 3D printers work by breaking down a three-dimensional object into a series of thin two-dimensional ‘slices’, which are laid down one on top of another. The technology has already been employed to manufacture complex engineering components out of metal or plastic, but it is now being used to generate tissues in the laboratory. The process is a bit like making a three-dimensional colour photocopy, as the finished item is an exact replica of an original. To make an aortic valve, for instance, researchers at Cornell University took a pig’s valve and X-rayed it in a high-resolution CT scanner. This gave them a precise map of its internal structure which could be used as a template. Using the data from the scan, the printer extruded thin jets of a hydrogel, a water-absorbent polymer which mimics natural tissue, gradually building up a duplicate of the pig valve layer by layer. This scaffold could then be seeded with living cells and incubated in the normal way.⁷⁴

Pushing the technology further, Adam Feinberg, a materials scientist at Carnegie Mellon University in Pittsburgh, recently succeeded in fabricating the first anatomically accurate 3D-printed heart. He used the heart of a chick embryo, chosen because its complex internal anatomy made it particularly difficult to replicate. The tiny organ, just 2.5 millimetres in diameter, was mapped microscopically to provide a template for the printer, which was set up so that it produced a replica ten times life size. This facsimile was made of hydrogel and contained no tissue, but it did show a remarkable fidelity to the original organ.⁷⁵ Since then, Feinberg has used natural proteins such as fibrin and collagen to 3D-print hearts; his group is now aiming to include living cardiac cells in the hydrogels extruded by the device so that it is able to print living tissue and so construct a viable organ.⁷⁶ For many researchers in this field a fully tissue-engineered heart is the ultimate prize, but even those involved acknowledge that their goal is a long way off. When I asked one whether he could see it being a realistic alternative to transplantation, he laughed before suggesting that it would take another forty years to perfect.

We are, therefore, left with several competing visions of the future. Within a few decades it is possible that we will be breeding transgenic pigs in vast sterile farms and harvesting their hearts to implant in sick patients. Or that new organs will be 3D-printed to order in factories, before being dispatched in drones to wherever they are needed. Or maybe an unexpected breakthrough in energy technology will make it possible to develop a fully implantable, permanent mechanical heart. More mundanely, it’s also conceivable that advances in drug therapy and preventative care will render such dramatic interventions largely unnecessary. A few retired medics can still recall wards full of patients suffering from polio, tuberculosis or scarlet fever; perhaps heart disease will go the same way as these illnesses, which have now virtually disappeared in the West.

Most cardiac surgeons are too cautious to forecast which, if any, of these scenarios is most likely to come to pass. Perhaps they realise that, as a breed, when they do venture to make predictions they often turn out not to be very good at it. In 1910 Harry Sherman, one of the first surgeons to suture a beating heart, gave a talk at a meeting of the Medical Society of the State of California. He told his colleagues: ‘I do not see how the majority of heart conditions which come to the physician can look to surgery for relief. A too large heart cannot be shaved down to a proper size, nor can a new heart muscle be made.’⁷⁷ If he was being unduly pessimistic, a report published in 2004 by the World Health Organisation went to the other extreme, predicting implausibly that by 2020 ‘nanosurgeons’, microscopic robots, would be able to crawl through arteries to scrape away fatty deposits and make minor repairs.⁷⁸

Whatever the future holds, it is worth reflecting on how much has been achieved in so little time. Speaking in 1902, six years after Ludwig Rehn became the first person to perform cardiac surgery, Harry Sherman remarked that ‘The road to the heart is only two or three centimetres in a direct line, but it has taken surgery nearly 2,400 years to travel it.’⁷⁹ Overcoming centuries of cultural and medical prejudice required a degree of courage and vision still difficult to appreciate today. Even after that first step had been taken, another fifty years elapsed before surgeons began to make any real progress; and then in a dizzying period of three decades they learned how to open the heart, repair and even replace it. In most fields, an era of such fundamental discoveries happens only once – if at all – and it is unlikely that cardiac surgeons will ever again captivate the world as Christiaan Barnard and his colleagues did in 1967. But the history of heart surgery is littered with breakthroughs nobody saw coming, and as long as there are surgeons of talent and imagination, and a determination to do better for their patients, there is every chance that they will continue to surprise us.