Novelty over impact is medical science’s fatal flaw.
What we can do to realise the full potential of the future of medicine
We have optimised medical research for the wrong metric, and it is sabotaging our own success.
It is now, more than ever, imperative that we do something about this. Medical science has come under attack, and the reality is we have made it easy for them.
The reproducibility Achilles’ heel.
I just read the other day that the reproducibility crisis in science was used as a justification to cut NIH funding. It’s hard to make a sound counterargument.
The primacy of publishing novel findings in high-impact publications as a metric for academic success blots out any concern about the lack of reproducibility. Yet reproducibility is not a new problem. It’s been a topic of discussion for decades. When a system has a simple positive feedback loop that promotes novelty over everything else, the system becomes very difficult to change.
The long-overdue shift to treating biology, not disease
Making use of the exploding ability to characterise all of the molecules that our cells are producing, has led to an exponential expansion of our understanding of the complexity of biology. This has led many to suggest that we are on the verge of a new era where we abandon old definitions of disease to focus on the characteristics of biology.
However promising that might sound, we have been constrained by arbitrary definitions of disease for too long. Our approach to health and disease has calcified into a system optimised for novelty that resists change.
We have known for some time that processes and even the cells that orchestrate those processes are the same throughout the body. Take fibrosis, for example. It is a critical feature in many different diseases.
Developing anti-fibrotic therapies is certainly happening. Yet the challenges are that fibrosis is a complex process and likely needs to be attacked in multiple ways, and that anti-fibrotic therapies often have side effects that are hard to tolerate.
This highlights the challenge. Even though fibrosis is a process involved in many different diseases, the concept of fibrosis is itself a bucket of underlying biological processes. For example, here are the known fibrotic processes in the liver:
It is likely that each one of these processes can also be broken down into a similar diagram. Many, if not all, of these molecular components are involved in other disease processes.
The future could be that we go to the doctor, he or she does not just prescribe an anti-fibrotic. Instead, he or she draws blood or collects samples, and our molecular processes are characterised. Then we get a personalised treatment regimen of multiple medicines and even a modified version of our own cells that tunes up the processes that have gone awry.
Aside from oncology, the practice of medicine remains bound to existing disease definitions which lump people together into artificially defined groups. The focus on novelty, not impact, has not helped resolve what is an unbelievable translational gap of 17 years from research findings to clinical implementation.
What is holding this shift back?
It is the optimisation for novelty. Researchers who find new molecules or pathways are rewarded.
What we need is more work to understand how these processes all connect together.
More work to know how we can embrace the complexity of what we know and develop therapies that nudge our underlying biological processes back to a healthy state.
This is how we get to treating biology and not disease.
For example, a basket trial where, based on molecular profiling, a custom anti-fibrotic and anti-inflammatory therapy regimen is delivered. We also need increased efforts to understand the molecular interactions and what they mean to developing new therapeutic strategies.
I should disclose at this point that I am an adviser for the company TopMD, which is creating pathway biomarkers by incorporating the complexities of biological processes into a topological map. It is analyses like this that will help enable a new paradigm of treating biology and not disease.
Crisis as an opportunity
The diminution in the valued position of science is very troubling. It’s troubling not just for the US, but the entire world. Many of the scientists I work with are the leaders in their field in Europe, this is important for them too. The drastic cut in science funding is a symptom of a much deeper disease. Simply restoring the funding in the US would be a missed opportunity.
Restoring the funding does not address the underlying problem - the incentive to be novel, not confirmatory, not synergistic.
Novelty is the wrong performance indicator.
Maybe at one point, when our understanding of biology was much simpler and there were low-hanging target molecules, a novel finding would equate with a rapid pace of new therapy development. That is no longer the case. We need to shift from a scientific mindset that values novelty to one that values impact.
The system, at least in the US, has experienced a shock, and when there is a shock it is a good moment to consider making changes. The system right now is less capable of resisting change. The damage is done, now is the time to rebuild better, but how?
Valuing collaboration
Europe has a long history of valuing collaboration over individual performance. Probably because the amount of funding available for medical research in Europe is a fraction of that in the US, even with the cuts to the NIH. EU funding bodies put a premium on large collaborative projects and student exchanges, both of which are ways of doing more with less by banking on the added value of combining different disciplines and different perspectives.
Does it work?
The evaluation of the one of the bigger programs that the US has sought to emulate, Innovative Health Initiative (IHI) suggests that it does.
The IHI is a public-private partnership that has funded more than 220 consortium projects. I know from experience that these projects are a fantastic opportunity to experience all of the good and all of the bad of large-scale collaborations.
When you look at field-normalised citation impact, all of the very first IMI call projects were above the field-normalised norm (the IHI was call the IMI at that time). In later calls some projects were 5x the norm. The U-BIOPRED project, a IMI first call topic, was a project which was 3x the norm is a project I helped develop. That project formed a lot of my thinking about how to make a consortium project highly interactive. This metric is, of course, anchored in publications which are biased towards novelty; nonetheless, it validates the potential of the collaboration through a big consortium project approach.
The reality is that many of these types of collaborations are still constrained by the dominant novelty focus of science. The novelty focus fuels silo thinking even when you are supposed to be collaborating.
Many of the IHI projects have been focused on improving our ability to have a finer-grained understanding of the biology of health and disease. This positions them well to drive the transition from treating disease to treating biology. What is more important that most do not realise is that consortium projects are where stakeholders come together to break the bottlenecks that are holding back a field. It is this innovation implementation focus that makes consortium projects our best hope for rebuilding the degrading trust in biomedical science. One of the best ways to deal with the whims of politics is to focus on delivering unmistakable value. Unmistakable value defangs politics.
Now is the time to make a change.
The opportunity we have before us is to move beyond the shock that the defunding of biomedical science in the US has created and push forward with more funding of collaborative projects that move us toward a paradigm where we treat biology and not disease. Such a move is likely to open up many new ways to treat disease.
It is also likely to open up deeper personalisation of medicine, which with the new forms of treatment has become more possible than ever before. A really exciting opportunity is that it will make the process of intervening in the pre-disease phase possible.
It is inherent that we understand the biology that leads to a disease in order to intercept a disease before it manifests. Otherwise what are you treating? Here focusing on the biology enables us to treat to biomarkers before anyone has to suffer from a disease.
The reality is that the slow pace of translating medical research into meaningful impact has put medical science at risk. It is not that medicine has stagnated. No, indeed there are new forms of treatment. There is also a whole lot of potential that goes unfulfilled.
Potential to deliver the right medicine to the right patient at the right time.
Potential to deliver treatment regimens that achieve nearly 100% effectiveness.
Potential to interrupt disease processes before they produce disease.
At the root of a lot of lost potential is the inability to shift to treating biology, not disease. This stagnation was inevitable. When we optimise for novelty and not impact, the system will be weighted towards producing novel insights that are not reproducible.
Consortium science as the solution.
To break free of this stagnation, more emphasis should be placed on consortium science. When we work in a consortium project, it is almost a given that we will share techniques and it’s a great format for reproducing models and results. Most importantly, consortium projects endow us with collective agency.
Agency to solve the problems that arise in the delivery of a project.
Agency to create the critical mass that can have the influence needed to shift paradigms.
Agency to drive forward the impact we are all waiting for.
This means that NIH funding should be restored with an emphasis on funding consortia that break the bottlenecks holding back innovation. Even in the US, there is a good track record of this with the Cell Atlas projects and, of course, the Human Genome Project.
It also means that rewards in science should be focused on what has been done to move towards impact, not novelty.
On an individual level, it means choosing to invest time and effort in big consortium projects. It means grounding even your most basic science work in impact.
This newsletter is about the type of big projects we need to make medical science even better than it was before. Show you support by subscribing.
Great read-- I look forward to the change from oversimplistic classifications of disease. I hope that payers and regulatory will be open to intervening on biology and health status.
Very nicely written. Enjoyed reading