The Concern of development of AI in Medicine

Having begun a course on the Potential of AI in medicine (please see the ‘Online Courses’ page) this week, I thought that it would be apt to have a look at the issues and limitations surrounding the development of AI in medicine.

AI (artificial intelligence) and ML (machine learning) is currently providing untapped and rapidly growing sources of data for patient benefit, including the potential of improving diagnostic accuracy, more reliably predicting prognosis, targeting treatments, and increasing the operational efficiency of health systems. But, surrounding these developments is a large amount of concern as, at the moment, they don’t have defined guidelines and rarely undergo the same degree of scrutiny as other medical interventions, such as those in pharmacology. Current policy and ethical guidelines for AI technology are lagging behind the progress AI has made in the health care field.

So, what regulation is their currently on health related algorithms? Firstly, developers should ensure that their algorithm complies with the Medical Device Regulation, which until 2010 did not regulate independent software products. This covers any product which claims to have a medical nature, including providing diagnostic information, making recommendations for treatment, or providing risk predictions of disease. Next they must consult the Medicines and Healthcare products Regulatory Agency who have published guidance for developers that covers the regulation in greater detail. If an algorithm falls within the remit of this regulation, the developer must then seek regulatory approval or accreditation in the form of a ‘CE’ mark before marketing it. The developer must ensure that the device meets the relevant essential requirements before applying the CE mark. These requirements include:

  • Benefits to the patient shall outweigh any risks.

  • Manufacture and design shall take account of the generally acknowledged gold standard.

  • Devices shall achieve the performance intended by the manufacturer.

  • Software must be validated according to the gold standard, taking into account the principles of development lifecycle, risk management, validation, and verification.

  • Confirmation of conformity must be based on clinical data; evaluation of these data must follow a defined and methodologically sound procedure.

In addition manufacturers are required to have post market surveillance provision to review experience gained from device use and to apply any necessary corrective actions.

Furthermore, the use of ML/AI algorithms might be regulated indirectly by other legislation or regulatory agencies. The highest profile additional legislative framework to be aware of might be the European Union’s General Data Protection Regulations. Others include the United Kingdom’s Care Quality Commission who are tasked with monitoring compliance with NHS Digital’s Clinical risk management standards; a contractual requirement placed on developers engaging in service provision to the UK’s health service.

However, this isn’t considered enough. Through answering 20 critical questions, spanning issues of transparency, reproducibility, ethics, and effectiveness, the BMJ aimed to identify where further work is needed to build consensus on what constitutes acceptable practice. Encouraging patients, clinicians, academics, and all manner of healthcare decision makers to ask the challenging questions raised by the BMJ will contribute to the development of safe and effective ML/AI based tools in healthcare. Developing a definitive framework for how to undertake effective and ethical research in ML/AI will involve many challenges, and the challenges facing us with regard to AI in medicine are new and different. These challenges include finding common terminology (where key terms partly or fully overlap in meaning), balancing the need for robust empirical evidence of effectiveness without stifling innovation, identifying how best to manage the many open questions regarding best practices of development and communication of results, the role of different venues of communication and reporting, simultaneously providing sufficiently detailed advice to produce actionable guidance for non-experts, and balancing the need for transparency against the risk of undermining intellectual property rights. Addressing these challenges of transparency, reproducibility, ethics, and effectiveness are important in delivering health benefits from ML/AI.

Further work is also required to identify themes of algorithmic bias and unfairness while developing mitigations to address these; to reduce brittleness and improve generalisability, and to develop methods for improved interpretability of machine learning predictions. If these goals can be achieved, the benefits for patients are likely to be transformational.

Regulation that balances the pace of innovation with the potential for harm, alongside thoughtful post-market surveillance, is required to ensure that patients are not exposed to dangerous interventions nor deprived of access to beneficial innovations. Mechanisms to enable direct comparisons of AI systems must be developed, including the use of independent, local and representative test sets. Developers of AI algorithms must be vigilant to potential dangers, including dataset shift, accidental fitting of confounders, unintended discriminatory bias, the challenges of generalisation to new populations, and the unintended negative consequences of new algorithms on health outcomes.

Other issues include the fact that data is balkanized along organizational boundaries, severely constraining the ability to provide services to patients across a care continuum within one organization or across organizations. Another issue is the lack of ability to execute the successful use of AI on front line medicine; firstly simply adding AI applications to a fragmented system will not create sustainable change, secondly, most healthcare organizations lack the data infrastructure required to collect the data needed to optimally train algorithms.

There are so many other issues which I don’t have time to talk in detail about now, including who to blame when the machine makes a mistake, that ends in death; or while one may share photos on Facebook to their family, would you feel confident for AI to screen those posts for depression or maybe even suicide? Finally, in 2015 Jeremy Hunt, former UK Health Secretary, said that ‘we have the chance to make NHS patients the most powerful patients in the world – and we should leap at the opportunity.’ But this seems blind to reality, for example what about those patients who find technology a burden, who don’t have the latest iPhone and a data plan?

Therefore, if developed in an appropriate manner, AI and ML in medicine will have revolutionary impact; but regulation is required to ensure that these developments are generalised across the NHS, so as not to leave anyone behind. Furthermore, they should be maximised in terms of transparency, reproducibility, ethics, and effectiveness.

Mental Health Issues with One Shared Cause Part 2

Last week, I discussed how mental health conditions could all have one cause; one which can be inherited in a risk called the ‘p factor’.

But, there is more to the story than just genetics, environmental and psychological factors also play a key role. For example, child abuse, drug and alcohol abuse, and a traumatic experience such as being placed in warfare can also be a key contributor to susceptibility of developing a mental health condition.

Plomin and his colleagues have now attempted to quantify the genetic component of the p factor. From tests on 7000 pairs of twins – they estimate it’s heritability is approximately 55%. This means that genetic differences explain over half of the variation between people’s general susceptibility to metal health problems. The study also showed the stability of the p factor across someone’s lifetime.

Studies have now shown where these genes work. In 2018, Gandal and his colleague Dan Geschwind showed, from 7000 post mortems, that they control activity at the synapse. Another study by Hammerschlag backs this up. Her team investigated more than 7000 sets of genes with variants linked to five common mental health conditions, ‘almost all of these gene sets play a role in the synapse’ she says. So, to put it simply, the p factor effects communication between brain cells.

In a more recent study Maxime Taquet, from Oxford University, and his colleagues found that there is a ‘vulnerability network’ in the brains of children at high genetic risk of developing mental health conditions. They found differences in three key areas when comparing these to children with low genetic susceptibility: a structure called the default network (active whilst the brain is at rest), a structure involved in planning and control, and the part of the brain that processes vision. But how does having a brain with differences in these features influence an individual’s psychology?

Caspi and Moffitt (who have carried out a similar test to the one described above and found differences in a brain circuit crucial for monitoring and processing information so that it can be used in higher cortical functions such as regulating emotions, thoughts and behaviours) think that a high p factor probably manifests as a combination of disordered thinking, difficulties in regulating emotions and a tendency towards negative feelings.

So what has been done as a result of these breakthroughs so far?

Already, many drugs have been used for treatment of a variety of disorders; ‘antipsychotics, for example, have not only been used for psychosis, but also in mania, delirium, agitation and other conditions’ said Tova Fuller, of the University of California. The p factor makes sense of these ‘transdiagnostic’ therapies, but it isn’t ideal, as these drugs weren’t created with the p factor in mind.

But Gandal says that ‘if we can figure out the biology of the p factor, then it might be possible to target the mechanisms involved and develop therapies that work better across disorders. These could be given to a large number of patients, rather than treating each person based on their specific pattern of symptoms’.

Other than drugs, cognitive behavioural therapy (CBT) has also had transdiagnostic value. Currently, there are separate therapeutic guidelines for specific conditions. However the p factor may help clinicians to develop a one-size-fits-all approach to CBT called the common elements treatment to ensure that more people globally get the treatment that they need. Moffit doesn’t however disregard the need for specialists that help to treat, for example, schizophrenia.

Moffit also thinks that the existence of the p factor should shift a prompt from treating conditions themselves to treating the often distressing symptoms that people feel, and providing people with the tools to help themselves when issues arise.

Plomin also thinks that there is no boundary between mental health conditions, saying that he believes that ‘diagnostic guidelines are mostly a myth.’ He thinks that it ‘implies there are mentally ill versus the “normals”, really we’re all somewhere along a continuum.’

Mental Health Issues with One Shared Cause Part 1

Mental health conditions are disorders that affect your mood, thinking and behavior. They include anything from depression and phobias to anorexia and schizophrenia. In the UK, one in four people experience a form of mental health condition each year. Currently, if suffering, you may seek help from a medical professional, and they will give you a diagnosis from one of the hundreds of conditions listed in psychiatry’s classification bible: ‘Diagnostic and Statistical Manual of Mental Disorders’. Note that they are not easy to diagnose, as symptoms come out differently in different patients. You would then begin treatment tailored to your condition.

The question that has been raised however is that of whether this the correct approach?

Neuroscientist Anke Hammerschlag at Vrije University Amsterdam said that ‘for millenia, we’ve put all these pyschiatric conditions in seperate corners. But maybe that’s not how it works biologically.’ There is growing evidence to support this view – many mental health problems appear to share an underlying cause, something researchers now call the ‘p factor’. This could revolutionise treatment, putting more focus on symptoms rather than labeling conditions, and offering more general treatments.

‘P factor’ is tied closely to one of the most famous concepts in psychology. More than a century ago, British psychologist Charles Spearman noted that children’s performance in one sort of mental task, for example verbal fluency was correlated to their mental skill in other areas, such as mathematical reasoning, spatial manipulation and logic. So generally, those who are good in one area, tend to be good at another, and those who struggle in one area, tend to struggle in others. Using a statistical tool called factor analysis, Spearman showed that this is because these different mental abilities are all linked by overarching cognitive capacity, which he named the general intelligence, or g factor.

Applying the same logic to mental health diagnoses provided the first hints that something similar might be going on. There are a wide range of mental health conditions that manifest with different behavioural and psychological symptoms. Like cognitive skills, they cluster in individuals, either at the same time or one after another. So, in 2012, Benjamin Lahey at the University of Chicago and his colleagues analysed information on such diagnoses among 30,000 people studied over three years. Using factor analysis they found that the observed patterns of illness were best explained by a general tendency towards mental health conditions. Avshalom Caspi and Terrie Moffi at King’s College London got the same result the following year. They used information from 1000 people who had been tracked for four decades since their birth in the early 1970s. It was Caspi and Moffit who coined the term ‘p factor’ as a broad susceptibility to mental health problems. ‘Once you have any given mental disorder, it increases the liklihood that you’ll have multiple other kinds of disorders’, says Caspi.

It has been long known that these conditions are highly heritable. For example, twin studies have shown that the heritability of schizophrenia is 80%, and major depression is 45%. But, having a parent or sibling diagnosed with a given condition doesn’t just increase the chance that you will get that condition, but it also increases the chance that you will be diagnosed with a different condition. For example, if your parent has schizophrenia, your risk of developing bipolar disorder doubles. This makes sense if you don’t just inherit the risk for one kind of condition, but a more generalised risk: the p factor.

The real genetic advancements came with the introduction of SNP chips, which allow scientists to use a small DNA sample  to scan someone’s genome and discover which genetic variants they carry. These chips have found more than 10 million single-nucleotide polymorphisms (a substitution of a single nucleotide that occurs at a specific position in the genome, where each variation is present at a level of more than 1% in the population). This was used to identify the p factor a few years before Moffitt and Caspi coined the term. In 2009, the International Schizophrenia Consortium used SNP chips to genetically analyse 3000 people with the condition. The analysis found that schizophrenia was linked to thousands of variants, each having a small effect. Intriguingly, these same variants also increased the risk of bipolar disorder.

In 2013, an international group called Psychiatric Genetics Consortium analysed genomic data from more than 30,000 people diagnosed with conditions including bipolar disorder, major depression and schizophrenia. Again, genetic risk variants cut across the traditional diagnostic boundaries of psychiatry. This can be contrasted with neurological conditions such as Alzheimer’s and Parkinson’s which have little or nothing in common with each other or with psychiatric conditions; this was found by a 2018 study on 265,000 people with 2 psychiatric and neurological conditions.

Thus, whilst for neurological disorders, it is easy to divide people into two groups: those who carry the risk variant, and those who don’t, with mental health conditions, it isn’t that easy; few people carry a small amount of risk variants, and few people carry a lot, but most people fall somewhere between. Plomin says that ‘There’s no break point at which the number of variants suddenly leads to a diagnosable psychiatric disorder.’

Rejection-Proof Skin Made on Pigs

Currently, two problems face with regard to organ donation: a shortage of human donor organs; and people who get transplants have to take medicine for the rest of their lives to suppress the immune system from rejecting the new organ. The need for organs is dire. Each day, 20 people die waiting for an organ transplant. More than 113,000 people in the United States are currently waiting for one, while only 36,528 transplants were performed in 2018, according to government data. Every year, the waiting list grows, greatly outpacing the number of available organs. For decades, researchers looked to animal donors as a way to ease this chronic shortage, but transplants from animals have often failed. At least three teams have added human genes to pigs to try to solve these problems, by stopping rejection by a recipient’s immune system.

Lijin Zou at the First Affiliated Hospital of Nanchang University of China and his colleagues have created pigs in which they have added eight human genes to reduce the chance of a donor organ being rejected, and removed the three pig genes that trigger organ rejection. They then transplanted skin from these pigs to monkeys. The skin survived for 25 days without the monkeys without needing any immune system suppressing drugs.

The pigs’ skin, which looks remarkably similar to humans’ skin and is referred to as Xeno-Skin, will be transplanted by surgeons at Massachusetts General Hospital to a small group of burn victims in an attempt to speed up the healing process. It’s the first experiment approved by the U.S. Food and Drug Administration to use animal tissue in humans, a necessary step toward someday transferring entire organs grown in animals to people who need them – a process known as xenotransplantation.

Xeno-Skin, developed by Boston-based biotech company XenoTherapeutics, shows promise. So far, one patient has received the genetically engineered pig skin graft, and five more burn victims are slated to receive it. The grafts are meant to be temporary and will be removed once the patients’ own skin has grown back. Doctors involved in the trial say the donor tissue appears to be healing as well as a human skin graft, which was transplanted next to the pig skin for side-by-side comparison. The process also hasn’t caused negative reactions like provoking an immune response or transmitting animal viruses, two major issues in xenotransplantation. ‘We’re trying to replicate exactly the same mechanisms that are used in the standard of care, or the gold standard treatment, for severe and extensive burns,’ said Paul Holzer, CEO of XenoTherapeutics.

Interest in this type of approach has surged in the past decade as advances such as CRISPR gene editing have made it feasible to make extensive changes to the genomes of animals.

In December, Luhan Yang at biotech firms Qihan Bio in China and eGenesis in the US reported that her team had created pigs with nine added human genes and dozens of deleted pig genes; these include the three deleted by Zou’s team, and also many viral genes known as porcine endogenous retroviruses (PERVs). Porcine endogenous retroviruses (PERVs) represent a particular risk for xenotransplantation using pig cells, tissues or organs. PERVs are integrated in the genome of all pig strains and can be released as particles that infect human cells. If PERVs start infecting human cells after a transplant, there is a risk that they may cause cancer says John Coffin at Tufts University School of Medicine in Boston. But this is still better than the outcomes if people don’t get the transplant.

Image result for Rejection-Proof Skin Made on Pigs

Predicting Heart Attack Risk

A group of London doctors are using artificial intelligence to predict which patients with chest pains are at the greatest risk of a life threatening heart attack.

In the largest study of its kind, funded by British Heart Foundation and published in the journal Circulation, researchers took routine CMR scans from more than 1,000 patients attending St Bartholomew’s Hospital and the Royal Free Hospital and used a new automated artificial intelligence technique to analyse the images. By doing this, the teams were able to precisely and instantaneously quantify the blood flow to the heart muscle and deliver the measurements to the medical teams treating the patients. By comparing the AI-generated blood flow results with the health outcomes of each patient, the team found that the patients with reduced blood flow were more likely to have adverse health outcomes including death, heart attack, stroke and heart failure. The AI technique was therefore shown for the first time to be able to predict which patients might die or suffer major adverse events, better than a doctor could on their own with traditional approaches.

This ensures that patients in need of treatment receive medication to improve blood flow, or undergo procedures, such as stenting, to open blocked heart valves.

One of the doctors at Barts Health, Kristopher Knott said that ‘adverse events were significantly higher when blood flow [is] low’, therefore ‘as poor blood flow is treatable, these better predictions ultimately lead to better patient care.’

Not only can this technology help with the prediction of heart attacks, but it can also help with the prediction of heart disease – the leading global cause of death. One of the most common symptoms of this disease is reduced blood flow and, although non-invasive blood flow assessments are available, including Cardiovascular Magnetic Resonance (CMR) imaging, up until now, the scan images have been incredibly difficult to analyse in a manner precise enough to deliver a prognosis or recommend treatment.

As Professor James Moon (from Barts and UCL) said: ‘artificial intelligence is moving out of computer labs and into the real world of healthcare.’ AI is becoming more commonly used in day to day healthcare, and I can only see this continuing, along with a rise in AI healthcare scans from home.

Perfusion mapThis is a myocardial blood flow ‘perfusion map’ created and analysed using AI showing an area of the heart receiving a reduced blood supply (arrow) and putting the patient at risk of heart attacks and other adverse events.

Covid-19

Coronaviruses (CoV) are a large family of viruses that cause illness ranging from the common cold to more severe diseases such as Middle East Respiratory Syndrome (MERS-CoV) and Severe Acute Respiratory Syndrome (SARS-CoV). On 31st December 2019, a novel coronavirus (COVID-19) was identified in Wuhan, China. Since Coronaviruses are zoonotic, it is believed that this was caused by an animal sold at one of Wuhan’s markets.

The symptoms of Coronavirus are that of a fever, with coughing and shortness of breath. The severity of these symptoms can range from very mild to severe, even death. Although understanding of this disease continues to grow, most people with severe illness have been of an older age or had other significant existing medical conditions. This is similar to what is seen in people who have severe infections with other respiratory illnesses, such as influenza.

It appears to be spreading from person to person among those in close contact. It may be spread by respiratory droplets released when someone with the virus coughs or sneezes. It’s not known if a person can catch the virus by touching a surface that an infected person has touched, and then putting his or her hand to the mouth. But, on 26th January, China’s health minister Ma Xiaowei said that the virus can spread before the person experiences the symptoms. Robin Thompson from Oxford University said that this would ‘explain why the virus is spreading quicker the SARS’.

The scale of the outbreak will depend on how quickly and easily the virus is passed between people. Using data from 18th January, it appears that, on average, each person infected with the virus passes onto between 1.5 and 3.5 other people, according to analysis by Natsuko Imai and her colleagues at Imperial College London. Another study by Shi Zhao and his colleagues at the Chinese University in Hong Kong estimates that the average person can pass it onto 3 to 5 people.

In comparing this virus to pneumonia and SARS (infecting 8000 people in 2003, when a global outbreak began). So far, the fatality rate seems to be lower, with a death rate of 2.8% based on the reported cases so far, compared to 9.6% for SARS; but, it is still too early to be sure just how dangerous the virus is. Thompson says that ‘SARS took several months to cause a thousand cases, this has caused [almost] 3000 cases in 3 weeks’. SARS stopped in 2004 and since then, no cases have been found; it was stopped by isolating those infected and screening air travel passengers. This is more difficult with a more infectious virus.

The worrying thing is if the virus mutates to become more contagious or deadly. As yet, there is no evidence that the virus has mutated. However, the WHO has declared that it is a world health emergency. It has now infected over 69 000 people and has killed over 1 500 people. Almost all of the deaths to date have been in China, but Europe had its first death in Paris last week, and there are many more cases across Europe, America, and Australia.

Due to these figures, health authorities, especially in China, have taken many unprecedented measures in an attempt to stop the virus from spreading. Wuhan has been placed in lockdown, public transport stopped, the airport closed and the use of personal motor vehicles banned. Similar measures have been taken in several other cities with millions of cases. The Chinese government has also temporarily banned the sale of wildlife in markets or restaurants, as it is thought that the virus passed from bats to people, or possibly snakes or minks. These animals were all on sale at Wuhan’s Huanan seafood market, where the first infections were reported.

Action has also been taken in the UK: they have introduced advanced monitoring at airports with direct flights from China. A team of public health experts has been established in Heathrow to support anyone travelling in from China who feels unwell. These hubs will bring in rotational teams of 7 clinicians, working in shifts, who will be on hand to support patients on arrival. This is in addition to medical staff who are already permanently in place at all UK airports and the advice issued to all UK airports for people travelling to and from China. The government has also issued clinical guidance for the detection and diagnosis coronavirus, and infection prevention and control. The Chief Medical Officer, Medical Director at PHE and Medical Director at NHSE have issued advice via a CAS (Central Alerting System) alert to frontline staff to increase awareness of the situation and any actions to take. Most people who develop symptoms will get them after leaving the airport and so the priority is providing UK residents and travellers with the latest information to make sure they know what to do if they experience symptoms, and the NHS and PHE have an established plan to respond to someone who becomes unwell. China has also introduced port-of-exit screening so people already exhibiting symptoms are not allowed to leave the country.

Therefore, although the UK, and other richer countries seem to be on top of things, preparing isolation units in hospitals, this isn’t the case world wide. Furthermore Jennifer Nuzzo at the Johns Hopkins Bloomberg School of Public Health in Maryland said that ‘no country is fully prepared’, and US senator Chris Murphy said that some countries ‘aren’t taking this seriously enough’. Mark Woolhouse at the University of Edinburgh said that ‘of the coronavirus is worse than swine flu, it will be horrendously difficult to handle’. This is to the extent that ‘we won’t be able to control it, and it will have to run its course.’ Overall, we are unsure about the true strength of the virus. Many pharmaceutical companies are working on a vaccine, but Nuzzo thinks that efforts should focus on preparing the communities to cope with the virus rather than trying to halt its spread.

Image result for coronavirus l

The Old Operating Theatre

Yesterday, I visited the Old Operating Theatre in London. It began as an operating theatre in 1822 (making it the oldest surviving surgical theatre in Europe), predating anaesthetics and antiseptics. It is housed in the roof of what used to be a church, with a glass ceiling to maximise to light.

Hospitals at that time were a place for the poor – those rich enough would have private healthcare in their house, perhaps carrying out surgery on their kitchen table. At St. Thomas’, by the 18th century, there were nineteen wards, including the four ‘foul wards’ for those with venereal diseases. There was strict segregation between male and female, and venereal and non-venereal patients. Furthermore, a shortage of beds meant that patients frequently had to share. The types of medical condition that the hospital treated were very restrictive, so patients often lied about their circumstances, such as women lying about pregnancy, or they wouldn’t be admitted. Despite the shortage of beds, patients often stayed for many weeks, but after three months, you were discharged no matter what.

The ‘Rule of Patients’ were visually reinforced so ‘swearing, assaulting other patients, stealing, gambling, excessive drinking and “immodest” behaviour’ were forbidden, with patients being discharged if they refused to comply. In addition, the patients would be forced to work in the hospital, as far as their illness would allow.

There were also issues at this time with ‘quacks’ – those who pretend to have knowledge or qualifications but don’t. Those who rejected Orthodox medicine would often visit these people. This was until 1858 when a medical act set up the ‘Medical Register’ of qualified doctors. An example of the issue with ‘quacks’ can be seen in Lionel Lockyer, who claimed that his pills contained sunbeams. They actually contained antimony, so severe damage was caused.

At this time, many herbal remedies were being used: Marshmallow was used to sooth inflamed areas; mint was used for common colds and disturbances in the gastro-intestinal system; poppies were used as a sedative and pain relief; rosemary was used as a diuretic; sinapis (mustard) was used as a stimulant; thyme was used for digestive complaints; willow bark was used as a pain killer. This is only to name a few. On top of this, there were some herbal remedies which were more individually targeted at specific diseases, such as sage was targeted at typhoid. Some preventative remedies were also used: an orange would have been a precious gift in the stocking of a Victorian child as it prevented scurvy. Limes were proposed by John Woodall in 1779 to prevent scurvy – when he issued them as standard dietary provision to sailors scurvy disappeared almost overnight.

Other than herbal medicine, alchemy began to take off. Paracelsus was the main person in charge of this movement. His medicines largely involved metals such as antimony, lead and mercury. However, alchemy processes also involved plants; so the process anticipates modern pharmacology. Specific chemical medicines could be used to target specific problems. The main piece of equipment at this time was the alembic – usually made out of glass or pottery, it has a round body (circubit) and a spout angled downwards with a receiving vessel attached to it. The material would be heated in the body of the alembic and vaporise. They would then condense in the cooler tube and drip into the receiver.

Image result for alembic

The only problem was that Paracelsus thought that nothing was toxic if you use the correct dose. Before Paracelsus, the main belief of disease was that of the humours (blood, yellow bile, black bile, phlegm). If one was in excess, then this is why disease occurs. Paracelsus suggested that disease could be a living entity of itself that has come from outside the body to attack it – so, his thoughts anticipated the later understanding of viral and bacterial diseases.

On October 16th 1846 anaesthesia was developed, but this operating theatre opened in 1822 – for the fist 24 years patients, with no anaesthetic, would be held down to the operating table by men, whilst the surgeon did his work. There were some plants which could stimulate sleep and were sometimes used, including: henbane; nightshade (still used to day as a pre-medication given to patients prior to gas anaesthesia; it enlarges and dilates the pupil to improve access to the eye for ophthalmic surgery as well); cocaine; mandrake; opium; bryony and foxglove. These all contain Atropine, but this is toxic. Other non-plant anaesthetics were also used such as mercury and carbolic acid – both of which are highly toxic.

Germ theory was only discovered by Louis Pasteur between 1860 and 1864. Before then, surgery was carried out paying no heed to disease. In the theatre, hundreds of medical students would watch on in a small space, coughing and sneezing. Everyone would wear their normal clothes, whilst the surgeon might where a gown with other patients blood splattered on it. No one would wash their hands, or wear gloves, and after the surgery, instruments would be wiped, but not cleaned before being placed in a container lined with silk – a prime place for germs be caught and stay. The bandage applied to the patient might already have someone else’s blood on it as well. Therefore, after the surgery, a patient faced a large risk of infection.

To perform a leg amputation, the surgeon would slice through the skin with a knife, cut through the bone with a saw – but not in one clean cut, but in multiple as the saw gets stuck. They would then pull out loose bits of bone with a piece of equipment resembling pliers, and sew up the patient. The blood vessels in the leg would be sealed with heated oil, tar, or sewn up, but would often come loose.

Overall, it’s a fascinating place to visit, if you want to learn more about the history of medicine and more specifically surgery.

Oxitec’s Mosquitoes

It seems to be becoming a recurring theme on this blog – attempting to stop illness before the symptoms come to light. We have looked at technology that can do this, and doctors searching for issues such as loneliness which could lead to illness, but recently, I read about another way which this can be done – stopping the sources of disease, one of which is insects, and in this article, mosquitoes.

This Brazil based trial, lead by Oxitec, noticed that, in African forests a few thousand years ago, the female Aedes aegypti drank the blood of many species, which ended up including humans, to the extent that they (a subspecies of the Aedes aegypti) required humans for survival, or rather for survival of their eggs. In the 15th century, slave ships then carried them to America. From there, they reached every tropical region, as they require temperatures over 20 degrees Celsius for survival. Nowadays, these mosquitoes are developing resistance to the pesticides that we previously relied on to control them, and they can spread diseases such as dengue, yellow fever, Zika, and chikungunya.

Oxitec have come up with an idea which would mean a quick plummet of thousands of mosquitoes around the world. They propose, and have indeed tried, to genetically modify male Aedes aegypti mosquitoes so that they would express a fluorescent trait and a ‘late-acting lethality’ trait, which means that most of their offspring die as larvae, with the entomological SIDS gene, and do not survive to adulthood to reproduce. Repeated releases of many millions or billions of GM males, vastly outnumbering the wild male mosquito population, are intended to reduce the total adult population of mosquitoes over time.

The first trial reduced the numbers of Aedes aegypti in the city of Jacobina, Brazil by at least 70%. When the release of the mutated mosquitoes stopped, the numbers of Aedes aegypti rebounded. Reading some articles on the success in the Cayman Islands, it seems that it wasn’t all that successfull. “To date all the measures recorded have shown no significant reduction in the abundance of Aedes aegypti in the release area” (MRCU scientist, 4th April 2017). However Oxitec claims that this was too early on in the trial, and in fact it ended up with an 80% reduction.

Some newspapers have latched onto the fact that these GM mosquitoes may be harder to kill. However, since they massively reduce mosquito populations, the headlines in these papers are slightly misleading (such as ‘Deadly super mosquitoes that are even tougher’ – the Sun). Furthermore, the average mosquito lifespan is only 2 months. But are they successfully interbreeding, resulting in a constantly increasing GM super-mosquito population? This is uncertain.

It seems obvious that genetic engineering could help us with the prevention of disease in the future. But, as seen, genetic engineering doesn’t come without risks. For example, in the US, pollen from trial plots of glyphosate-resistant GM bentgrass for use on golf courses spread to wild bentgrass in 2003. This produced herbicide-resistant wild grass. Furthermore, earlier this year, dairy cattle supposed to have only a tiny DNA change to make them hornless, stopping the painful methods to dehorn cows, also gave them a gene for antibiotic resistance. Therefore, do the risks outweigh the benefits?

To me, it seems like they do, after all, these risks can be minimised and potentially wiped out all together – it therefore requires more thought and research however. But, surely it is worth trying to prevent diseases like Zika, which can cause serious birth defects. Furthermore, judging by the results of the initial trial, it seems to have worked.

Off-Label Cancer Drug Testing

When standard treatments for cancer fail, doctors can prescribe drugs that haven’t been approved for that particular cancer type. The purpose of this is not only to provide patients who haven’t improved on one treatment, another treatment, but also reduce the massive cost of many treatments – in 2013, the Netherlands spent nearly €733 million (£626 million) on cancer treatment and drugs. Edwin Cuppen, director of Hartwig Medical Foundation, said that ‘We can’t afford to give these new, expensive drugs to everyone’.

Although still in trials at the moment, mostly in the Netherlands, the way that it works is that once a drug has been approved for one purpose, doctors can prescribe it for another. This can be potentially dangerous, and is carried out on an ad-hoc basis, therefore, often insurance companies won’t pay out for off-label use. For example, in the 1980s, in the US, some heart drugs were used off label, but this resulted in 50,000 premature deaths.Therefore, the question that lies is, is it worth the risk?

Emile Voest, a molecular oncology and immunology researcher affiliated with the Cancer Institute in Amsterdam, and his colleagues have set up a more reasoned method for distribution of off-label drugs. This involves sequencing the whole genome of tumors in people for whom standard treatment has failed. This information is then reviewed to pinpoint other drugs that could potentially help them, via finding similar biomarkers. For example, several drugs that have been approved to treat breast cancers have been seen to produce excessive amounts of HER2 (a growth factor). Therefore, other drugs that produce lots of HER2 might also work treating breast cancer.

One somewhat unique aspect of the program is that hospitals and research centers send fresh frozen biopsy samples, not formalin-fixed paraffin-embedded samples (FFPE), which are traditionally how biopsies are preserved. Researchers determined that DNA extracted from FFPE biopsies was not consistently high enough quality. The CPCT team is currently receiving around 100 samples per month and growing. On average, Cuppen said, 70-80% of the fresh-frozen biopsies can be successfully sequenced, with the main limitation being the percentage of tumor cells in the biopsy. At least 30 percent tumor content is needed in order to obtain enough tumor DNA for whole-genome sequencing to be informative.

In the Netherlands, they have been trialing this with the help of hospitals, charities and pharmaceutical companies, which are donating drugs free of charge. Other countries are now taking on this trial, including Canada, Denmark and Italy – the overall number of patients enrolled in this trial now exceeds 1000. In the trial, someone with a particular cancer is given a drug that could potentially help them. Similar patients are then assigned the same drug until eight patients have tried it. If at least one patient benefits from the trial, more patients are enrolled onto the trial of that drug. If no one benefits from the trial, the cohort is stopped. This is because it would be awful to give patients who don’t have long to live drugs that won’t help them and could have bad side effects.

The researchers also plan to implement RNA sequencing and want to test whether analyzing circulating tumor DNA from plasma could serve as a good proxy for sequencing DNA from the tumors themselves. Analyzing ctDNA (Circulating tumor DNA; tumor-derived fragmented DNA in the bloodstream that is not associated with cells) could be advantageous for patients from whom a metastatic tissue sample is difficult to come by, but they first need to make sure that the same mutations found in even distant metastases can be identified in the ctDNA.

Although the trial still requires proper randomised controls to reach final conclusions,  from the first trial, of 215 people, about a third had some benefit, and one or two had complete remission. As of now, this trial has lead to some health insurers in the Netherlands agreeing to pay for off-label use, despite not being approved by the European Medicines Agency and it therefore seems to have been a success and could be considered to be worthy of continuing.

 

Parkinson’s Disease

On Wednesday 22nd January, I attended a talk by Samuel Deutsch entitled ‘Predicting Parkinson’s in our Sleep’.

He began by talking generally about the field of neurology – the study and treatment of disorders of the Central and Peripheral Nervous system. It therefore includes a wide range of disorders including psychological, and psychiatric disorders. But it is a complicated field as often the disorders’ causes aren’t clear cut. For example phantom limb syndrome, with 80-100% incidence, has no clear cause. Some people believe that it is due to a cortical re-organisation, while others believe that it is due to nervous pathways being hijacked. Furthermore, neurological diseases rarely have a cure – it is more about management symptoms; there is no ‘one pill treats all approach’. So, in the case of phantom limb syndrome, therapies include mirror therapy, stimulation therapy, and virtual/augmented reality. Furthermore, with our now ageing population, ‘the brain is dying before the body does’, and neurological diseases are therefore on the rise.

Image result for mirror therapy for phantom limb pain"Image result for virtual reality for phantom limb pain"

He then went on to speak about Parkinson’s, a motor-system disorder which includes cognitive, homeostatic and psychological issues. It is usually diagnosed in the 60s, and in 2018, 145 000 people were diagnosed; but this is expected to rise by 20%, by 2025. It is caused by a build up of alpha-synuclein in the synapses, which results in the death of neurones in the basal ganglia of the brain.

Parkinson’s has many symptoms, the main four being a tremor, rigidity (how well limbs move), bradykinesia (global slowness of movement), and postural instability. Alone, any one of these symptoms doesn’t mean Parkinson’s, for example, tremors could just be an essential tremor, which is very common. Other results of this disorder include psychiatric symptoms, dementia, cognitive impairment, and other motor problems. However it is extremely difficult to diagnose as it doesn’t occur the same in everyone; it could be cognitive first and motor later, or vice versa.

The treatments of Parkinson’s are wide-ranging. There are some drugs which can be used, such as: levodopa (produces more dopamine), DA agonists, MAO-B inhibitor and COMT inhibitor. All of these aim to increase the amount of dopamine in the nervous system, and therefore cocaine can have the same effect – however this isn’t recommended. Dopamine disregulation can occur as a side-effect to any of these drugs. This results in an inability to control urges, such as gambling. Therapies can also help, such as exercise, diet, SALT (speech and language therapy) and occupational therapy. We even discussed the possible future of stem cells in the treatment of Parkinson’s, helping to regrow cells in the basal ganglia. Although this wouldn’t cure the disease as the build up of alpha-synuclein in the synapses, would continue to kill cells in the basal ganglia.

He then went on to talk about RBD, which is a REM sleeping disorder when you act out your dreams while you are asleep. Usually you are paralysed during REM sleep, but with this disorder you aren’t. This can be extremely dangerous as, in 2010, an man who lived in Oregan with RBD beat up his wife while he was asleep, but he had no idea what he was doing. He was then sentenced to prison. RBD is usually diagnosed in your 50s.

The link between these two diseases is that on autopsy, 98% of people with RBD had a build up of alpha-synuclein. Furthermore, roughly 50% of patients with RBD have Parkinson’s. So, Mr Deutsch and his team at the John Radcliffe are trying to draw links between the two diseases. Since RBD develops 10 years earlier, they are trying to use this as a predictor of Parkinson’s, as one of the most important parts to treatment of Parkinson’s is catching it early.

Mr Deutsch currently has a cohort of 1 000 patients with Parkinson’s, 300 healthy patients, 150 relatives of people with Parkinson’s, and 300 RBD patients. His work is attempting to find whether there are cognitive phenotypes in prodromal Parkinson’s Disease – this is easy, and cheap to screen. This shows heterozygous profiles, and can be compared to RBD results for the same test.

If a link is found, anti-inflammatory or anti-oxidant promoting diets have been found to reduce microgial activation in rats (whether this is reliable is questionable, as it has only currently been tested on rats). Furthermore, exenatide, a glucagon-like peptide-1 receptor agonist that is used for treatment of diabetes, has been found to slow Parkinson’s Disease.

Another research project that occurred earlier this year (2020) across vast expanses of the population of China has found that the genes: LRRK2, PARKN, SNCA, PARK7, PINK1 and GBA have a link to Parkinson’s. Therefore, you can tell people at birth that they are at risk to developing Parkinson’s, and ensure that they have a good diet, and exercise regime to reduce the chances of getting it in the future. Specific genes being affected also points towards the future of treatments of Parkinson’s, in changing the expression of genes – however this isn’t currently possible.

He concluded by reinstating the fact that there is currently no cure to Parkinson’s, and that his work, and the work of doctor’s generally isn’t always about curing diseases – it is about making lives better by alleviating pain.