The low-down on the Bristol GDNF trial
Original article: Parkinson disease and growth factors – is GDNF good enough? Nat Rev Neurol: April 4, 2019.
Two Parkinson’s experts comment on the results of the recent GDNF trial, and discuss possible explanations for why the trial may have failed to meet its primary endpoint.
Why is it important?
It is important to understand what the GDNF trial failure means, and identifying ways in which a future trial should be designed to definitively assess its disease modifying potential.
The potential of GDNF to promote the survival of dopaminergic neurons in Parkinson’s was first identified in the 1990s, when the first sets of promising findings emerged in animal models. The first translation into a clinical trial in people with Parkinson’s delivered GDNF into the fluid filled spaces of the brain, and did not succeed. Subsequently, an open label study delivering it directly into the brain’s movement centre which degenerates in Parkinson’s showed the potential for improvement and neuronal sprouting.
Why was another trial of GDNF even needed? In an open label study, all participating patients and involved clinicians know what treatment is being given. This leaves it open to a placebo effect – that is, improvements associated just with receiving any treatment and being involved in a trial. This “anticipatory” or psychological element is real and large, especially in people with Parkinson’s, for reasons related to dopamine.
The recent double-blind, randomized, placebo-controlled trial of GDNF run in Bristol, UK, showed no benefit on the UPDRS scale assessing movement, which was its so called primary endpoint: the way in which researchers define, before the trial even starts, the criterion for success. However, neuronal improvement was seen using neuroimaging and some benefits in the open label 40 week extension.
Why did the trial fail? Several factors could be at play, and at the moment it is not possible to tell whether any of these is at play here or whether GDNF is simply ineffective.
The patients recruited into the trial may have been too advanced in their disease course for GDNF to have an effect. The dose of GDNF may not have been high enough for the neuronal benefit seen on neuroimaging to translate to tangible movement improvement. The neuroimaging results however do need to be understood better, because in addition to picking up changes in dopaminergic cells, this method also picks up changes in other types of cells. Investigating how these changes map onto each individual patient’s response could be useful. Another difficulty was caused by the abnormally large placebo response. Finally, it is not clear whether enough of the target brain area was actually covered. This may well have been overestimated with the conventional imaging technique used, raising the possibility that better results may be achieved with different or greater coverage.
Some of the many open questions raised in connection to this GDNF trial may be answered by an open label gene therapy study of GDNF which is still ongoing, and is following people over 5 years. But an open label trial can’t in itself answer definitively whether any therapeutic is truly effective, as it doesn’t contain a placebo arm. In a future double blind placebo controlled trial of GDNF, it will be essential to incorporate learning from the previous work and iron out all of these open issues.
Original article: Parkinson disease and growth factors – is GDNF good enough? Nat Rev Neurology: April 4, 2019. Kirkeby, A., & Barker, R. A.