Prof. Ulrich Dirnagl, a stroke researcher at Charité, received the prestigious Thomas Willis Lecture Award from the American Heart Association in February 2016. Prof. Dirnagl, who is a member of the BIH Faculty (along with all teachers at Charité - Universitätsmedizin Berlin and senior MDC scientists), primarily focuses on understanding the mechanisms of stroke related damage and developing novel therapies from this knowledge. Frustrated by the lack of reproducibility of research in his field and a disappointing track record of taking bench finding to efficient therapies in patients, he is also interested in improving the quality of preclinical research. We put four questions to the prize winner.
Prof. Dirnagl, you won this year’s Thomas Willis Lecture Award. What does the distinction mean for your research, and for you personally?
The Willis Lecture Award does not recognize a specific research result, but rather the winner’s contribution to advancing the field. I cannot think of another prize that is as highly regarded in my field. Those who have won the award over the years form a kind of Hall of Fame of stroke researchers. So it’s an amazing feeling to be part of that; I’m extremely happy about it!
One of your main concerns is improving the predictive power of basic clinical research. Could you give us an idea of where the challenges currently lie?
The replication crisis is a big topic at the moment. It turns out that a lot of laboratory findings – and often the spectacular ones – cannot be replicated. The pharmaceutical industry was the first to notice this, as it often cannot reproduce the results published by academic biomedicine. We, too, can treat many diseases very successfully in our animal models, but the therapies are rarely successful in patients. This is known as the translational roadblock, which may in part be the result of a lack of robustness of the preclinical research on which the clinical studies were based. Various research fields are now running large-scale replication studies, with rather sobering results.
As well as your scientific research into stroke pathophysiology, you are also active in the fledgling field of meta-research. Can you tell us a bit about that?
Meta-research is the scientific – i.e. also quantitative – study of research practice. The aim is to make research as efficient and reproducible as possible. It turns out that considerable room for improvement exists in the basic and preclinical fields of biomedicine in particular. Such improvements include reducing bias, increasing statistical power, and optimizing publishing practices. At the same time, my seminars on experimental design, statistics and good scientific practice for medical students and PhD candidates at Charité and in various international graduate programs in and around Berlin show me that we also have an enormous amount of catching-up to do in terms of methodological expertise. From my contact with students I can gather that they need more guidance and support in their research. And this is where the challenges lie: we have to make our research more resilient and more predictive. This is especially true in cases where researchers have to go from laboratory findings to clinical proof-of-concept studies, which are designed to establish the fundamental existence of disease mechanisms and the feasibility of new therapies.
At BIH, we also want to achieve outstanding quality in preclinical research so that as well as producing original and spectacular findings, we can also make our results more reproducible and more robust. What measures do you think will secure the most effective improvements in this regard?
One example would be to provide tools and services for researchers such as an electronic lab journal to help scientists plan and analyze experimental studies, establish quality assurance measures, or ensure that they comply with reporting guidelines. Such measures could also include support for publishing “negative” results, or a fund for replicating particularly important results. Decisions on researchers’ appointments, funding and career progression should also not be solely based on the impact factor of their publications. This will also be of key importance. Furthermore, we could use new indicators, in some cases those that have yet to be developed – such as quality of results and their significance for clinical application – to create incentives for more quality-awareness in publishing. These are just a few ideas. In any case, the implementation of these measures will have to be scientifically monitored so that we can check if they are effective and develop more and better approaches. We could then offer them as blueprints to other research institutes. Within the context of the international discussion on reducing waste and increasing value, BIH could assume a leading global position and thereby optimize the value of its research.