The occurrence of septic shock: difficulty in predicting the outcomes

Dr Manuela Ferrario - San Raffaele Scientific Institute, Milan, Italy

Shockomics is a European project with the proposal to study multiscale factors in acute heart failure in shock patients. Shock patients is a condition where the patient is characterised by a low perfusion but at the moment the current therapy tries to target the symptoms but not the cause of the pathology. So the need to understand this pathology, the pathways of inflammation involved in this pathology, is essential for a targeted therapy to the root cause of this pathology. Our project involves several clinical units and we are collecting data in Geneva, in Barcelona and in Brussels. We will connect haemodynamic data but also blood samples to perform a multiscale analysis in these patients in order to get information from transcriptomics, metabolomics and proteomics. So the idea is to understand the inflammatory pathway in this condition.

How can studying the plasma metabolome help in understanding which patients are at risk of septic shock?

It’s very important because we know that the metabolism is the output of a very complex system and in order to understand the entire inflammatory pathway it’s important to start from the base of the pyramid, so the genome, transcriptome, but also to reach the upper of the pyramid, the metabolomes. It’s very important and promising in order to find biomarkers that really stratify the patients according to their severity but also in order to identify of patients so to find the target therapy for those patients with those biomarkers.

How long have you been working on this project?

It started in 2013 and it will end in 2017. It’s a four year European project.

Can you discuss some of the new statistical approaches that may be useful for the clinical implementation of metabolomics?

It’s very important to tackle several problems from a statistical point of view. One of these is that you have few patients, or at least less patients, than the number of metabolites, well none, for example. So it’s important to identify the most important metabolites involved in inflammatory pathways. But at the same time if you want to integrate metabolites in this, let’s say, huge view of the pathologists, the integration of metabolites with proteomics, transcriptomics, it’s very important from a statistical point of view to find the right approach in order to integrate all these measurements. I think that these two topics are very interesting and challenging for the next future.

Have you started to identify any biomarkers in your project?

We started in a sub-study in an available biobank to help this study which is, I think, the only one, only biobank available with septic and sepsis, septic shock patients. We found some metabolites that are very important and even in a very small subset highlight some metabolites that are coherent with other published papers. So these metabolites, at least currently, are a strong predictor of mortality and associated to mortality.