The Rogues Gallery

Newsweek on Translational Research

Sharon Begley, the science editor for Newsweek, wrote a recent commentary on translational research – research which is designed to take basic science knowledge and apply it to clinical practice – curing patients. Her thesis is that the current system of academic research and NIH funding slows down translational research, preventing or delaying the development of new medical treatments.

Orac at Respectful Insolence and Tim Kreider at Science-Based Medicine have already written nice take-downs of Begley’s reasoning. I will not repeat their points. But I will summarize them and then follow with an anecdote of my own to reinforce their points.

Begley’s criticism of the current system is essentially that exciting new science is academically valued, while translational research to see if treatments work is less valued. As a result high impact journals favor the former. Also researchers are likely to delay the publication of useful clinical findings until they have done enough background scientific research to get their results into a more prestigious journal, which is important to their career and getting funding.

There are many weaknesses in Begley’s argument. Kreider points out that she relies entirely on the reports of four scientists, without back up from statistics or without getting the other side of the story from journal editors or NIH grant funders. Researchers often think they are being treated unfairly by the system – that their work is not appreciated enough, and the system is too resistant to their ideas. This is just human nature, and an observation subject to confirmation bias.

Orac also points out that Begley misunderstands the nature of research journals – that different journals have different niches. Some journals publish basic science, others publish clinical findings, some publish both. A particular high impact journal may not publish a purely clinical piece because that is not its niche.

This is not to say that the system is perfectly balanced. There is a longstanding known bias toward publishing positive studies over negative – and there is currently pressure on journal editors to balance this better. The “impact factor” of journals is the ratio of their publications to citations of their publications – the more other people refer to their studies, the higher impact they have. This system certainly favors novel research over replication. It also favors doing the basic science (looking at mechanisms) to back up clinical research.

But I, like Orac and Tim, disagree with Begley that this is a bad thing. She seems to be falling for the assumption that the more translational research the quicker cures will be found.

There are many reasons to doubt this assumption, however. The first is the realization that most published clinical studies are wrong in their conclusions (as per the work of John Ioannidis). This is a matter of statistics, considering that most new ideas in science are wrong. It takes time for the literature to work itself out.

Having your basic science ducks in a row also helps keep clinical research on target. As Ioannidis points out, the prior probability is a key factor in determining how much of the clinical research will ultimately be correct or misleading. Prior probability is derived from basic research.

What this means is that we need a proper balance between basic science, translational research, and pure clinical research. Shifting the balance towards translational and clinical research will not necessary bring faster cures, and it may slow it down. This is where my anecdote comes into play.

I am involved with clinical research into treatments for amyotrophic lateral sclerosis (ALS or Lou Gehrig’s Disease). This is a degenerative motor neuron disease that leads to weakness and in most cased eventually to death, within an average of 2-3 years. Sufferers and their families are understandably anxious for a cure.

ALS is also proving to be a very complex set of diseases (it’s not a single pathological disease, but a collection of diseases with the same common outcome). The more we learn about it the more complex it becomes – each answer leads to more questions.

Over the last two decades there has been a huge push for more clinical research. Some of this push came from patient groups who managed to gather significant funding. They granted some of this funding to researchers to do clinical research – studies that had the potential to find a cure. What they did was use their funding to shift the balance of research from basic science to clinical studies.

The result of this was many clinical trials – a few of which I happily participated in – but no cures. There is one drug approved for ALS currently – riluzole, which was the result of pharmaceutical industry research. All the other drugs tested to date have been depressingly negative.

Sure – we learned a lot along the way. And there is a reasonable case to be made for just screening substances to see if they are of use in a terminal and incurable disease. And of course basic science research has continued all this time, with important advances in our knowledge.

But in retrospect it seems that we simply did not understand ALS well enough to judge the prior probability of any particular treatment working, and this lead to a long string of negative studies. Perhaps what is needed now is more emphasis on the basic science to understand ALS better to help guide future clinical research.

In other words – there are finite resources for medical research – in funding but also in the researchers themselves. We want to strategically allocate these resources to maximize return. It is naive to think that simply shifting resources to clinical research will do this, and that scientists pursuing basic science are just having fun and pushing their careers.

The ALS story, in my opinion, is a cautionary tale. The optimal balance of basic to clinical research is likely achieved by scientists following their interests, chasing down interesting ideas, and the collective wisdom of the scientific community. We should make sure the system is not being artificially forced by counterproductive academic and publication biases. But likewise, we should not use politics and funding to force the system either.