UK: +44 (0)1223 264428
USA: +1 (919) 901 0909
by Dan Haworth
Healthcare systems are overburdened and short of cash, patients don’t want to wait days or weeks for their blood results to come back, and front line medical staff would rather treat their patients in one consultation.
This is great news for Point of Care (PoC) diagnostics innovators, particularly molecular-based systems which have the potential to compete with centralised lab analysers in terms of analytical performance. They promise ease of use and low cost – so why isn’t every GP surgery or Physician Office Lab (PoL) offering these tests?
There are several reasons, but one is that molecular testing is either too expensive or the ‘Platform’ system does not provide the ‘test menu’ to justify the initial outlay. GPs and PoL want one true ‘Platform’ – a one-system-fits-all disease targets, rather than a one-trick pony which only tests a few targets from a specific sample matrix.
Whilst the industry is working on new technologies to take PoC to the next level, at Cambridge Design Partnership we are looking at novel methods to implement sample preparation in a low cost, user-friendly way. We take the complete system view of the device to be developed, and work closely with assay scientists to define the fundamental biochemistry requirements needed to develop a robust product. At CDP, we have the ability to leverage technologies across different market sectors and are always on the lookout for innovative ways to overcome the challenge.
Implementing a molecular assay can be simplified to three stages:
1. Sample preparation
2. Amplification of nucleic acids
With amplification technologies such as Polymerase Chain Reaction (PCR) or isothermal amplification well established, detection becomes less of a challenge as the target DNA (or RNA) is replicated to billions of copies.
The challenge that is frequently overlooked when device developers embark on a new project is how to achieve effective sample preparation. Successfully processing a range of sample types, simply and at low cost is inherently a consumable-based activity – and this is very challenging.
THE IDEAL SAMPLE PREP PROCESS
The ‘ideal’ sample preparation process would be to add raw sample directly to the biochemical reagents, which then amplify to generate a detectable signal. This may not always be possible as many samples are not in liquid format (e.g. swabs), and even samples which are, will rarely have the right chemical composition. For example, the pH or salinity may be incorrect for the reaction to proceed anywhere near optimal speed and performance.
Three sample preparation considerations are necessary for molecular assays:
1. Pathogen lysis
2. Overcoming matrix inhibition
3. Target titre
The key challenges associated with each are examined below from a biochemistry-implementation perspective. Look out for future articles which examine Sample Preparation from a user-interaction perspective and how this informs the product design.
This involves breaking down the cell wall to expose the nucleic acids to the biochemistry. Whilst some pathogens can be easily lysed (e.g. by moderate heating), others require aggressive methods such as mechanical bead-beating, extreme pH or exposure to nasty hydrogen bond disrupting (chaotropic) agents.
In ‘ideal’ sample preparation, the chemicals used in lysis are carried with the sample into the reaction chamber. Lysing chemicals may destroy the reaction environment and inhibit the reaction so sample dilution or neutralisation may be the only way to regain performance, but this adds complexity to the device – and also reduces the concentration of the target.
OVERCOMING MATRIX INHIBITION
Target DNA can be present in a range of different sample matrices such as biological fluids (urine, blood, saliva) or synthetic carrier media such as swab elution buffers.
Physical properties such as high viscosity (e.g. sputum) can make samples difficult to process and mix; dark samples (blood) may pose challenges for certain detection methods such as fluorescence. If the sample contains auto-fluorescent compounds (e.g Urobilin in urine) this may first need to be removed.
The matrix will also contain a ‘soup’ of biological inhibitors that reduce reaction performance and alter the chemical environment (e.g. by pH or salinity) or deactivate the amplification reactants. High concentrations of non-target DNA and DNases also pose significant threat to reaction performance.
As with the lysis chemical approach, sample dilution may also be effective to reduce the effects of inhibition but at the cost of reducing target titre.
The minimum volume of sample needed to achieve the desired test sensitivity is driven by the analytical Limit of Detection (LoD) of the assay. Of course DNA needs to exist in the sample in the first place, but due to the presence of inhibitors, a single copy of DNA may not be enough to initiate a reaction. The LoD may be, say, 10 copies.
If the titre is inherently low, one option is to increase the reaction volume, but larger reaction volumes normally mean higher costs (more reagents to lyophilise, more plastic waste etc.).
A second option is to carry out a concentration step, and this can be achieved by capturing the DNA on a solid such as the surface of a magnetic bead, washing to remove the non-analyte components followed by elution from the surface prior to mixing with biological reagents. This method developed by Boom et al, can be applied universally across many different sample matrices. However, the consumable devices tend to be very expensive, bulky cassette formats as they need to carry out complex fluid handling operations such as pumping, valving, metering, etc.
Effective sample preparation is a fundamental consideration when designing a PoC device which is high performance, easily useable, costing ‘nothing’ and which is capable of testing many different sample types.
Minimal sample preparation runs the risk of not offering the wide test menu and so limiting the ‘Platform’ appeal that users demand, but universal sample preparation is complex and expensive to implement at the point-of-care.
Look out for our next Blog – Sample Preparation from a User Perspective. To talk more about our PoC diagnostics development capabilities please contact Dan Haworth at firstname.lastname@example.org.
1Journal of Clinical Microbiology, Mar. 1990, p. 495-503
Louise explains the innovation opportunities offered by patient-centred-design.
13 January 2020
Karla shows us how modelling renal vasculature can help better understand CKD.
22 November 2019
Stay up to date with all our work and our latest news by signing up to our newsletter.