Variability and biases introduced through purification are limiting the translation of EV research into clinical diagnostics and therapeutics. So why purify if you don't have to? 

It is becoming increasingly accepted that the commonly used techniques to purify EVs, do not result in samples which consist of just EVs. Analytical techniques used to measure EVs which are not specific in nature, will therefore measure both EVs as well as contiaminents from the purification process and will not discriminate between these particle populations. This will influence measured concentrations and also limit the ability to link clinical observations to measured data. Moreover, absolute data becomes impossible to recover and any observations that are seen become relative in nature only.

 It has been shown that when purifying blood plasma by ultracentrifugation, that the majority of detectable particles within a sample are in fact not EVs and are instead lipoproteins (Sódar et al 2016). These contaminants make it very difficult to relate experimental or clinical changes, to measured data. The situation is made worse when looking for rare events or sub-populations of EVs, among the milieu of EVs present in blood plasma. Techniques which are non-specific in nature are confounded firstly by contaminants and secondly by EVs that are not the specific EVs they wish to study. In summary non-specific techniques are looking for changes and trends in the specific needle lost among more needles, lost in the haystack.  The ExoView platform bypasses these challenges by only measuring specific EVs which exhibit specific surface antigen.  Variability in data analysis is inherently improved because now only sample variability and instrument variability need to be considered, given that purification variability is removed from the analytical workflow.


The issues associated with sample purification are commonly cited as being the biggest limiting factor in the progression of EV based platforms, into clinical diagnostics and therapeutics. The ideal analytical tool would by-pass sample purification and be able to fully characterize extracellular vesicles direct from biological samples, with high specificity, in low volumes with adequate detection sensitivities to be clinically relevant. 

The ExoView platform is such a tool, it provides phenotyping, sizing and concentration measurements direct from biological fluids therefore bypassing arduous purification protocols which bias the data generated.

For those working in therapeutics, it is desirable to produce highly pure samples of vesicles. In this case, purification is required. However, because most analytical techniques do not have the ability to measure vesicles in both un-purified and purified samples, it can be difficult to assess the efficiency of the purification techniques used. The ExoView platform provides exactly this capability and provides label-free measurements on final product and can be used as a tool that can monitor the efficiency of the purification techniques used in the manufacturing process.

The ExoView platform works with as little as 50µL of sample and captures EVs directly from cell culture medium or plasma onto a functionalized array


Each spot on the array can be functionalized with different antibodies, peptides or aptomers. Once bound the concentration and size or antigen positive EVs can be measured in parallel providing the ideal platform for streamlined exosome research and biomarker discovery. 



As Coumens et al (2017) describe, the methods used in sample collection and isolation can influence significantly the downstream analysis of vesicles in terms of their function, content and physical form/concentration. The paper goes on to describe the main limitations associated with all commonly used techniques to purify (summarized below), making the point that the choice of purification is the single most important determinant of the results of subsequent analysis. The study goes on to explain that no technique that will isolate extracellular vesicles only and that soluble proteins, protein aggregates, lipoproteins and cell organelles and viruses are candidates for co-purification. Techniques which are unable to discriminate those things which co-purify and report them alongside EVs (techniques such as NTA and TRPS) are therefore subject to errors in data interpretation.