There is a growing body of evidence which links Parkinson's to the gastrointestinal tract, opening potential new avenues for treatment.

Efforts to solve thorny problems in medicine are increasingly turning to big data, often by pooling together databases from individual studies and searching for associations and trends within these. This is a particularly helpful strategy especially if one is looking for associations between events that are relatively infrequent. A novel approach to big data in a healthcare context is mining insurance claims databases as a means of answering such questions. Smaller individual studies have revealed intriguing patterns between inflammatory gut conditions such as Crohn’s which increases the risk for Parkinson’s by almost a third. A recent study interrogated a database of 170 million individuals, and succeeded in replicating this rate of increased risk.

Virtual repurposing
If Parkinson’s and IBD share a common disease network, then might drugs used to target one, help combat the other? In particular, could repurposing one of the drugs used to treat IBD, then be used to reduce the risk for Parkinson’s? Circumventing all the difficulties, costs and risks of large scale clinical trials, a 'virtual repurposing' trial was performed simply by analysing this large dataset. The long term outcomes in people with IBD who had received the therapy were compared to those who hadn’t, demonstrating that the therapy, in this case anti-TNFa therapy, could potentially reduce the risk for Parkinson’s by up to 78%.   Read more about this study here.

However, big data approaches and virtual trials come with health warnings. Among the issues to consider are observations of associations which do not necessarily imply causation, ensuring the any potential factor systematically occurs prior to the outcome, and interpreting findings in the absence of crucial data which a planned clinical trial would of course obtain. In the context of this interesting work on anti-TNFa, factors such as smoking could not be factored in, although it is known that it is associated with a lower risk for Parkinson’s. Gender differences, with benefits favouring males more than females, can’t be readily understood, but merit further research.

Big data can only take us so far, but they are useful in generating hypotheses and opening up new questions. For example, the anti-TNFa antibodies that are effective in IBD do not actually cross into the brain. Might IBD render the blood brain barrier more leaky, thus allowing the passage of these molecules after all? Moreover, what is the link with LRRK2, which appears to be a common genetic signature between gut inflammation and Parkinson’s? While the quest for LRRK2 inhibitors is ongoing, findings from big data can help pinpoint complementary approaches which may offer significant benefit to particular groups of patients in the context of personalised medicine.

Original article: Inflammatory bowel disease increases the risk of Parkinson’s disease: a Danish nationwide cohort study 1977–2014.

Original research: Peter I, Dubinsky M, Bressman S, et al. (2018). Anti–tumor necrosis factor therapy and incidence of Parkinson disease among patients with inflammatory bowel disease. JAMA Neurol. doi:10.1001/jamaneurol
.2018.0605

See also Editorial:
Olsen, AL, Riise, T, Brundin, P. (2018). Discovering new benefits from old drugs with big data - Promise for Parkinson disease. JAMA Neurol. doi:10.1001/jamaneurol.2018.0345