Sir Mark Caulfield | NIHR Barts Biomedical Research Centre, Barts Health NHS Trust, London; Queen Mary, University of London, UK
Citation: EMJ. 2025;10[3]:42-48. https://doi.org/10.33590/emj/DGCY5495
Starting at the beginning of your career, what first drew you into the world of genomics, and what about the field continues to fascinate you today?
I trained in medicine, and I’ve been a doctor for 41 years. What I observed as I began to train, particularly as I became more senior in my career, was that, although we had many good medicines to treat high blood pressure, we still had a residual unmet need. So, I thought about this, and I wondered whether, if we understood better the genetic architecture of blood pressure, because we know it’s a mixture of genetic variation and environment, we could identify new pathways that might unveil new therapies, which we could bring forward, test in trials, and then perhaps give to patients to address that unmet need. I started working on that in 1989, and in 1994, I discovered that the angiotensinogen gene was a gene for high blood pressure, and that’s since been validated. I published that finding in a journal, and that kick-started my career. It fitted with another group that had already published work suggesting the gene was implicated. And it turns out that 20 years later, I rediscovered the mechanism in another paper that I had published.
For many clinicians, the rapidly evolving genomics landscape can feel overwhelming. Given your deep clinical and research experience, how do you think we can more effectively bridge the gap between frontline healthcare professionals and the practical benefits of genomics in patient care?
We are entering an era of precision healthcare, where genomics and other -omic tests, which are measures of your RNA, proteins, metabolic signatures, and what life does to your DNA, that is, epigenetics, are becoming relevant to diagnosis and ensuring patients get the right treatment the first time. Also, at present, we do not have sufficient genomics and genetics training in undergraduate medical degrees, so people graduate with limited knowledge, probably confined to rare diseases. Here are some examples of work that I’ve done to help illustrate the point.
In a study of 76,000 whole genomes, we showed 99.4% of us have at least one genetic feature that means a medicine could be ineffective, or even harmful. Twenty-five percent of us have four of these gene-drug interactions, meaning we are likely to experience side effects. What is the relevance of this to practice? If 6.5% of UK hospital admissions are due to adverse drug reactions, and most of us have at least one of these features, our genetic makeup could be fuelling these adverse responses. If we knew this before prescribing medicines, particularly those with known adverse event potential, we could avoid harm. Therefore, it is very likely that, at some point, health systems in various countries around the world will adopt some form of genetic testing to understand the risk of somebody having an adverse reaction if they’re given a certain medicine.
Even during COVID-19, we studied 7,491 intensive care patients and 44,000 others, and we found 23 regions of the DNA linked to severe responses to COVID-19. One of these, tyrosine kinase 2, is the target of baricitinib, originally indicated for inflammatory arthritis.1 In February 2022, a paper was published showing that baricitinib reduced mortality and length of stay in the intensive therapy unit by 13% in addition to dexamethasone and tocilizumab.2 So that’s an example of genomically primed therapy. And if there’s one, there will be others.
Another important consideration is diversity. Typically, the most studied population is of White European ancestry, which means there are populations that are vastly underrepresented. In a study of 65,000 British Bangladeshi and Pakistani individuals, we found that 57% could not activate clopidogrel, a medicine commonly used after heart attacks or following narrowing of the coronary arteries.3 Thirty percent of the White European people were also not able to activate the medicine. We used electronic health records to show that those unable to activate clopidogrel had a higher risk of further heart attacks after being given the drug. This shows how real and important this issue is, making sure patients get the right medicine for the right benefit with modern genomics.
So, what I think is needed is undergraduate strengthening of education in genomics. There is a national master’s programme in genomic medicine, but not everyone has the time to do that. What we will have to do is to continually upskill the workforce in all health systems across the world to be able to adopt what might be not even a genomic, but a multi-omic future, where we may not just measure the genome, we might measure what life has done to your genome, the epigenetics, the RNA, the protein, and the metabolic signatures, all of which build up a life course picture of how you and your body interact with the world. This is definitely the future.
You received a knighthood for your leadership of the 100,000 Genomes Project, which was a landmark study in genomics. What have been the most impactful conclusions from the data so far, and what benefits have patients seen as a result?
On the 5th of July 2013, when Genomics England, London, UK, was formed, we were given the mission of the 100,000 Genomes Project. I was given four letters, which you can find on the World Wide Web. One was on rare disease, one on cancer, one on infection, and one on big data and how to store it. They all said roughly the same thing: here are some things you could look at because they might be tractable to benefit in healthcare. Nobody has ever done this before. The first thing we did was to travel across the world to centres which said they were doing this type of thing, and we learned that although lots of people said they were doing this, they weren’t really doing it as an end-to-end pipeline. For example, innovations in sequencing meant you could sequence at scale, so reading the genetic make-up was no longer the issue. The issue was, how did you analyse it and make it meaningful for patients?
We worked with one of the top sequencing companies in the world, knowing that if they were doing this alongside us, they would learn, invent, and do their utmost to ensure we delivered the best possible quality product. And that’s what we did. After that, we built an analytical pipeline to try to analyse the genomes.
One major output from this was the National Genomic Test Directory. In 2018, with my colleague Dame Sue Hill, NHS England, UK, we created a new genomic medicine service which was national. And to do that, my team and I created a national test directory to drive regional equity of access to genomic testing for 57 million people across the UK. We also negotiated 500,000 whole genomes for use in rare disease and cancer, making them available to patients through this test directory. That test directory is not just a directory for when to perform whole genome sequencing, because whole genome sequencing isn’t needed for every single genomic disorder; it’s a comprehensive system for genomic testing in the UK.
In 2019, I was asked by then Chief Medical Officer, Sally Davis, to look at genomic testing in children. With a large task and finish group, we proposed that all babies and children admitted to neonatal or paediatric intensive care with an unexplained diagnosis should receive genomic testing with a whole genome. We started that service a few years ago, and they are diagnosing about 42–47% of the children with a genomic diagnosis now. The second thing we recommended is that Genomics England and the NHS should test the role of a whole genome sequence in newborn babies, since around one in 17 may have a rare inherited disease. The goal was to see if knowing about these conditions at birth could allow us to intervene, avoid harm, or reduce disability. As a result, we created a website called RX-Genes where you can plug in the name of any genomic disorder, and it will tell you whether it’s treatable or not and what the treatment is.
We have learned a lot because the programme has sequenced 12,000 babies already, finding treatable diagnoses at the rate of one in 250 births. That doesn’t mean their lives will be entirely normal, but it could reduce their disability or screen them for something. For example, one baby’s genome showed mutations in the RB1 gene, a gene in the retina, and if mutated, can lead to early-life tumours of the back of the eye, which can be treated with radiotherapy or removal of the eye. At 6 weeks, an eye exam under anaesthetic detected two areas that had retinoblastoma in the retina, which were treated with laser therapy. If the child makes it to their sixth birthday without having major problems, then the child may retain sight in both eyes and avoid removal of the affected eye.
Another example involved a family whose first baby passed away after 4 months in intensive care. We allowed parents to enrol their deceased child’s DNA in the project to help with future reproductive choices. When the mother became pregnant again, she wanted to know if the second child might be affected. Sequencing showed the first child had a mutation in transcobalamin 2, a protein that helps vitamin B12 enter cells. Case reports suggested that high-dose B12 could bypass the defect. The second child, tested within a week, had the same mutation but received weekly high-dose B12 and has had a normal life course.
By identifying those at risk, we can provide treatment early, reduce disability, improve quality of life, save healthcare costs, and potentially prevent disease. So that’s what this programme sets out to do, and the government has allocated another 650 million GBP to do that over the next few years.
Throughout these projects, we learned that involving patients and families with lived experience is essential because the questions they want answered are not always the ones we think are most important. Their voice was central to the 100,000 Genomes Project and the newborn project, and it is very important to me. When we faced challenges with our analysis pipeline or other setbacks, we met with patients to explain the issues. This maintained a strong connection to their views, and many became ambassadors for the project while continuing to advocate for patients.
Are there any current research projects happening as a result of the 100,000 Genomes Project that you’re particularly excited about? What are their goals, and what impact do you hope they’ll have on patients in the future?
Firstly, when we designed the project, we took great care to design a life course picture of the people we enrolled and their health records. Everyone in the programme is followed longitudinally over their life course, so as they age, there may be new diseases or exposures, and we have their consent to study any condition they encounter. This allows us to work on any disease or exposure that arises during their life. Because of this longitudinal approach, we can show them the impact of specific mutational signatures in cancer on a person’s lifespan.
For example, in the paper we published in 2024 on cancer landscape, we showed that over 5 or 6 years, certain mutational signatures could indicate a higher likelihood of recurrence, increased illness, and sadly, a higher risk of death. This also has implications for clinical trials. Trials are typically conducted over a short period, but if we have all of the cancer registry data, such as the systemic anti-cancer therapy registry, we can look for long latency signals of benefit that may take months or years to emerge, but wouldn’t be captured in the usual clinical trial. The project is a gift that keeps giving.
We can also explore polygenic risk scores, where an aggregate of all the variants that affect your risk of having a common disease is calculated. This could allow someone in their 20s to be told they’re at a risk of developing breast cancer, and are recommended intensive breast screening. Common diseases are a mixture of lifestyle and genomics, not simply genomics. But genomics may enable a new era of preventive precision medicine, allowing health systems to anticipate and reduce future disease risk.
This is what we’re calling now a National Genomic Research Library. Everyone who has a whole genome sequenced in the NHS is offered consent to be a research participant, and over 90 percent of them say yes. This creates a longitudinally followed cohort, essentially a library of information. This library functions as a reading library, not a lending library. People come and work on the data without taking it away. In the era of AI, this scale of data can add value by improving diagnoses and finding answers we do not yet know, such as rare disease diagnoses or identifying genetic variants that predict adverse drug reactions. We set this up as a life course repository for Genomic Health, and the participants are fully informed and very involved. They’ve actually published papers on how to be on our participant panel. They have also published a book of poetry describing their lived experience of rare disease and cancer. They are incredibly committed people who are all very participatory in the programme.
Integrating genomic testing into routine NHS practice required a major transformation. What were the biggest clinical, operational, or ethical hurdles, and how were they overcome?
The challenge of organising the NHS is that it is like an oil tanker. It takes a while to turn it round and change its course. This was taken care of by Dame Sue Hill. I was involved, but she very much did this. She created the seven lab hubs, which were openly tendered for as part of an NHS procurement process in accordance with government rules, and they competed to be lab hubs. The aim was to have seven lab hubs, and it encouraged people to establish partnerships. If one part of the country had a specific set of skills but not the full range, they could pair with another region to cover all the necessary expertise.
Not every lab hub does every test. Certain tests are done in specific lab hubs, and that’s to focus on quality and excellence, and they’re all in National Quality Assurance schemes, and are monitored. It is hard to secure accreditation, and it’s also hard to maintain it, but the NHS lab hubs do that. Every diagnostic test returned to a patient, including whole genome sequencing, goes through a UK quality assurance process. We built the first accredited whole genome sequencing end-to-end pipeline, from DNA to sequence to analysis to patient.
Then there is the ethical side. When we began, the Ethics Committee asked, “You are doing whole genomes, so you will inevitably find things. How will you handle that?” If I have your complete blueprint for life and analyse it, I will probably discover things that could be useful to you, but you might not want to know them. My view is that if I knew something about you and withheld it, that would be unethical. In Genomics England, we treat your DNA as always belonging to you, even if you consent for us to use it. It is your unique blueprint, unlike anyone else’s in the world. You are, in a sense, unique in your 3.3 billion letters. You did not know you were a billionaire,but you are, just not in cash.
The important thing here is that we said to people, if we find something, we need to tell you, and they said they would want to know, but when we asked about specific examples, some did not want certain information. So, we created a list of findings we would return, focusing on those with profound impacts. If I knew you had a high hereditary risk of prostate cancer and did not tell you, you could develop it early in life without the chance for prevention. To me, that is not ethical. Ethics works both ways: it is about protecting people from harm and ensuring they have the opportunity to act on important information.
There is also the ethical question of sequencing newborn babies. They are at the very start of life, and some wonder if this could label them in ways that cause harm. My argument is that if one in 190 babies has a treatable condition and we do not tell their parents, that is far more harmful. Most people agree when you explain it.
Because we involve the public and participants throughout, we can go to them with these questions. In fact, when we asked if they were comfortable with us collecting all their health data, from GPs, from every registry, they surprised us. We expected caution, but they said, collect everything you can. They wanted us to use the resource fully, to find problems early, and to make the work as useful as possible. In many ways, they were less conservative than we were.
Has a lack of public understanding about genomics ever posed challenges during your work? How have you approached communicating complex concepts to patients or the wider public?
When you do a programme that’s groundbreaking and at the forefront, everybody thinks it’s going to be marvellous, and everybody assumes you are only doing it because you already know how. As I shared with you, we were building the aircraft as we were flying it. Traditionally, that’s not what you want to do. But of course, we did everything we could to make it as safe and as fast as possible.
In the early days, we faced challenges from patients. Some would say, I’ve been enrolled in the study for a year, I haven’t got any results, I’m really upset, etc. We were very careful in how we handled that. We spoke directly to people who felt that way and walked them through the process. When we explained what we were doing with their genetic makeup, they often said, I completely see why that would take a year or more.
Overwhelmingly, the people in the project stuck with the project. Every participant has the right to withdraw at any stage, but most stayed. Some did withdraw. Some got their diagnosis and said, I’ve got my answer, and I don’t want to be in this anymore. That is fine. We can keep and use their data for research up until the point they withdraw, but after that, no further data is collected. We respect that; it’s the right thing to do.
What is the greatest challenge currently seen in healthcare that you think could be solved with the use of genomic research?
I think one in two of us will have cancer, and I think that genomics, possibly multi-omics, might allow us, in some cancers, not all, to choose the right therapy the first time. I think the big challenge, and one that is being actively investigated, is to be able to detect cancer early. Most of the time, when we find cancer, it’s already at an advanced stage. But now, using cell-free DNA that leaks out of dead or dying tumour cells into the circulation, we can measure in a blood sample that you have tumour DNA circulating around. That molecular signature from the tumour DNA may even tell us what organ the cancer is in, and can even detect tumours that are not easy to spot on imaging at a very early stage.
We are already using these measures for lung cancer, and they’re allowing us to detect this early. If early detection of cancer works, then we will be in for an entire paradigm shift in cancer treatment, and we will be able to deploy therapies earlier. Of course, we will need a series of studies in early-phase cancers to understand how therapies that have been tested in late-stage disease can be used safely and effectively, because these treatments are not without side effects. So, discovering the cancer earlier may lead to it being removed, it may lead to it being treated with radiotherapy and chemotherapy, or a combination, and monitored afterwards using cell-free DNA to detect recurrence. There is a very large study called the NHS Galleri trial involving 140,000 people, looking at cell-free DNA as a multi-cancer test. It may not work for every cancer we’re testing, but if it works for three or four, it could be the first annual test starting from around your 35th birthday. That would have the potential to completely change the world of cancer care, just through genomics.
How does the UK’s approach to genomic medicine compare to other countries? What lessons can we draw from global efforts, and are there potential opportunities for collaboration and even larger-scale projects?
When I was at Genomics England, we interacted with many programmes around the world. It’s fair to say that there have been some attempts at national initiatives, but there’s been nothing so tightly connected to a healthcare system, free at the point of delivery, as this one, which has led to a transformation that is NHS-wide and regionally equitable. There is no other project that has done that. There are other projects that have sequenced more whole genomes. But what I would put to you is that it is not the number of genomes that you sequence that matters; it’s what you do with the information you get. Having sequenced a million whole genomes is not that important if there is no benefit to humanity.
We tried to make our benefits directly realisable within the UK NHS, and that’s the reason why there is a National Genomic Medicine Service. That wasn’t an instruction from the government when they awarded us the 100,000 Genomes Project. We’re one of the very few countries with a public system that is deploying whole genomes for direct healthcare. Clinicians and scientists working in the NHS can order from a national test directory for specific conditions.
So, we are in the vanguard. There are others doing similar initiatives, and they will eventually catch up. The nature of the UK and the NHS provides a unique foundational health system to do this in. You can do it at the level of a nation here, whereas in other health systems, this is more difficult because they have regional or other things that get in their way. Regional barriers or even laws can make a national-level programme very challenging.
Another issue is that we do not have equitable access to genomics worldwide. Many underserved communities in the global south, like Africa, India, Latin America, and Southeast Asia, have yet to feel the benefits of genomics. But that is coming, and there is a recognition in the global genomics community of the need to address this.
Projects of this scale, UK Biobank, Our Future Health, and All of Us in the USA, galvanise a community behind a programme or a project. People then spend their time doing research and innovating. The people actually leading the project don’t have to do all the work. They just have to assemble a global coalition of intellectuals to do that. And of course, there are things like the Global Alliance for Genomics in healthcare, which is designed to develop, co-develop, and co-create methods, approaches, and standards for doing genomics worldwide. This means that research from around the world becomes portable and usable within programmes here in the UK. That is the benefit of a global community working together and pooling its intellect for the good of humanity.
