Direct Analysis of DBS: Emerging and Desirable Technologies

Keywords::

• automation
• direct analysis
• direct desorption
• dried blood spot
• dried matrix spot
• drug development
• hematocrit
• internal standard
• regulated bioanalysis

Anyone with an interest in following current trends in bioanalysis over the last few years will have noticed the growing interest in dried blood spots (DBS) and subsequently, dried matrix spots (DMS). It is not difficult to understand the reasons for the widespread, contemporary interest in the foregoing topics. The use of DBS offers numerous advantages over the traditional wet plasma sampling techniques that have been used in quantitative bioanalysis for many years to support drug–development studies [1–4]. DBS sampling uses less blood, which enables juvenile toxicology and pediatric studies, and using fewer animals is of course a very welcome ethical advantage. Eliminating the use of satellite animals also improves data quality, and further money saving comes from the reduced shipping and storage costs.

Are there any disadvantages? While pharmaceutical companies and drug-development organizations appreciate the above advantages, switching to DMS analysis adds complexity for the bioanalyst. Specifically, extraction is more complex, sensitivity is lower and ion suppression is higher than for wet plasma analysis. The established generic manual extraction technique for analyzing DMS samples involves punching a disk from the center of the DMS; transferring the disk to a tube and adding an extraction solvent containing internal standard; shaking the sample for ∼2 h; centrifuging the sample; transferring the supernatant to a fresh tube; then analyzing the sample using LC–MS/MS [1]. While this is still a relatively straightforward extraction process, any additional complexity compared with protein precipitation is, of course, undesirable in a high-sample-throughput environment. Indeed, some resistance to accepting the technique has been observed for this reason alone by those who fail to appreciate ‘the bigger picture’. So, this is a problem we need to overcome if we want to maximize the important ethical, financial and data quality advantages that DMS sampling undoubtedly offers.

A potential solution is the use of direct analysis techniques that could eliminate manual extraction steps completely. If we could also offer additional advantages over manual extraction, such as higher sensitivity, enhanced internal standard performance and online sample dilution, DMS analysis should become as attractive to the bioanalyst as it is to the rest of the drug-development organization. A number of emerging options exist that can potentially aid bioanalytical efficiency. We can summarize these into the following three categories: automated DMS analysis, direct elution and direct desorption.

Automated DMS analysis

Automated DMS analysis uses a semi- or fully automated ‘card-punching’ instrument [101–103], possibly in tandem with a liquid-handling robot [104] to relieve some of the manual burden of DMS analysis. The advantage this offers is that such instruments are already available and use exactly the same workflow that bioanalysts are accustomed to. The use of such instruments can clearly be of value in high-throughput environments, but they do not provide the seamless workflow on offer from direct elution and direct analysis. These emerging techniques offer significant additional advantages that are worth the additional resource involved in their development.

Direct analysis: direct elution & direct desorption

The term direct analysis encompasses both direct elution and direct desorption techniques, and for the purposes of this article describes any technique where we are eliminating the manual extraction steps involved in sample extraction, including punching the DMS. The ideal workflow is simple: DMS samples are loaded onto the direct analysis instrument and from here the extraction, separation and detection of the analyte are completely automated. Clearly, the key step here is how our compound of interest is extracted from the matrix and the paper substrate. Direct elution describes a number of similar techniques where the analyte of interest is extracted through contact with a suitable extraction solvent. One of the advantages to direct elution is that a liquid eluate is produced that can be separated and detected using LC–MS/MS – a technique that is readily available in the bioanalytical laboratory and is familiar to, and accepted by, bioanalysts and regulatory authorities. Techniques are emerging from CAMAG (TLC–MS and DBS–MS 500) [5,6,105], Spark Holland [106], Prolab (SCAP) [107] and Advion (liquid extraction surface analysis and liquid microjunction surface sampling probe utilizing the TriVersa NanoMate^®) [7,108] that encompass a variety of approaches and varying levels of automation and additional functionality.

For the purposes of this article the term direct desorption (or ambient ionization surface sampling) describes a host of techniques [8,9] that use a non-liquid elution basis for analyte extraction and do not produce a liquid extract. Such techniques include thermal desorption (DART, LDTD), laser desorption (LAESI), and gas jet desorption (desorption electrospray ionization [DESI]).

One of the potential advantages of direct desorption is the elimination of LC separation, and this is also one of the challenges that must be overcome if it is to be introduced to a regulated bioanalytical environment. Eliminating LC would be a welcome and significant simplification of the bioanalytical workflow. However, in practice removing LC can result in poor sensitivity due to ion suppression and reduced selectivity, and risks assay interference via metabolite decomposition (e.g., N-oxides and glucuronides) into parent compounds during MS ionization. Perhaps different types of separation techniques than those commonly found in the bioanalytical laboratory (e.g., ion mobility, high-field asymmetric waveform ion mobility or some novel approach) could assist in this regard. Whatever the solution may end up being, it is clear that it will be a significant change to current bioanalytical workflows, and this means a significant input of resources (financial and time) is required to develop these techniques and underpin the underlying science behind them to ensure that they are acceptable for regulatory approval. However, this is not the only barrier to the practical application of direct desorption. Extensive investigation has yet to identify a direct desorption technique that offers adequate sensitivity across a range of representative pharmaceutical small-molecule compounds. DMS analysis using direct analysis in real time (DART) on the widely used paper substrate format [109,110] has demonstrated considerably lower sensitivity than that currently achievable from manual DMS extraction. DESI [10], and in particular paper spray [11], have demonstrated sensitivity close to, and in some cases exceeding that for manual DMS extraction for some compounds, but has also been equally disappointing for others. Significant improvements to DART sensitivity have been reported when blood spots have been applied to a non-paper substrate (a glass slide), but unfortunately such an application is unlikely to be compatible with our simple and robust operational drug-development workflow in practice. A paper-type substrate that allows the DART gas jet to permeate the full depth of the DMS (rather than just the surface) could enhance the sensitivity on offer.

The potential advantages on offer from direct desorption are significant, and the elimination of LC in particular would transform the way bioanalysis is performed, and the environments where analytical instrumentation could be used for nonregulated applications (e.g., doctors surgeries). However, the challenges described above clearly show that the use of direct desorption for supporting regulated drug-development studies has to overcome a series of barriers, and thus, currently, can only be considered a long-term future goal.

In contrast, because of its compatibility with the existing bioanalytical LC–MS/MS workflow, using direct elution in practice is a realistic short-term goal. In addition, direct elution coupled to LC–MS/MS has also been shown to offer significant increases in assay sensitivity compared with manual extraction across a range of compounds without compromising chromatographic performance [5]. Initial work has also shown that there may be some promise in utilizing direct elution without HPLC [5]. This highlights that direct elution is a flexible technique that should be compatible with both currently used detectors and those in the future that may make other direct desorption techniques a realistic option.

Direct elution performance, additional functionality & future perspective

High-throughput DMS quantitative bioanalysis via direct elution is an obtainable short-term goal, while practical use of direct desorption is likely to be some years away. Compatibility with LC–MS/MS is the main barrier to introducing a direct desorption technique compatible with regulated high-throughput bioanalysis, but many other issues exist that need to be addressed before any direct analysis technique can be introduced.

Robust automation is a challenge, due to the flexible nature of the cardboard surround of 85 × 53 mm four-spot substrate cards [109,110] favored by clinicians and bioanalysts. Significant progress has been made in this area and the latest prototype instruments utilize highly efficient/robust card handling with suitable error handling and capacity for up to 500 DMS cards [105]. Another automation-related issue is the imprecise positioning of spots on the cards. An automated instrument requires a visual recognition system, most likely a camera or scanner, to accurately locate the position of each spot so that extraction from the center of the sample can be carried out. Intelligent software is also required that can identify ‘bad samples’ – those that fall outside predetermined size and shape parameters.

Robust and reproducible performance is of extremely high importance for any bioanalytical technique. Chromatographic performance measured over many hundreds of consecutive DBS and dried plasma spot samples has shown that direct elution performs at least as well as the conventional manual extraction technique [GSK, Unpublished Data]. The relatively dirty extracts (compared with manual extraction) being introduced into the HPLC columns and detectors were of some concern, leading many to believe that more elaborate guard columns and/or trapping column techniques may be required. However, initial studies have shown that the technique is robust without any further complexity to the analytical technique, which is of course favorable. Modifications to direct elution instrumentation have already been proven to effectively eliminate carryover between samples [6].

Internal standard (IS) performance is an area in which the current manual extraction method can be improved on, and developing a new technique provides an opportunity to do this. Adding the IS as part of the extraction solvent means that the IS is not fully integrated into the matrix components and substrate of the sample prior to the extraction process and, therefore, does not correct for any variability during the extraction process. Ideally, the IS would be added to the matrix before being spotted onto the substrate at the site of sampling, or the substrate would be pretreated with IS to ensure that the IS is integrated with the sample and extracted with the analyte. Unfortunately, both these options are logistically not feasible when dealing with multiple studies, study centers and compounds if the cost and procedural simplifications on offer from DMS are to be kept intact. A suitable and logistically feasible compromise would be to apply IS to DMS in the analytical laboratory, prior to extraction. This would improve integration of IS to the sample compared with currently used manual extraction techniques as long as the IS is given sufficient time prior to extraction for it to bind to matrix components and the paper substrate. Spray technologies are being integrated into fully automated direct elution prototypes and initial results have shown that IS reproducibility is at least as good as that obtained using manual extraction [GSK, Unpublished Data].

A useable direct analysis solution also needs to be able to accommodate sample dilution requirements. The ideal solution would be to use detectors that have a much larger linear dynamic range (a minimum of 5–6 orders of magnitude) than what we currently use (typically 3–4 orders of magnitude). Analytical methods could then be developed with larger ranges that would encompass the higher concentration samples sometimes encountered in early stage drug development studies. Unfortunately, such detectors are not currently available, and are not likely be in the short-term future. A reasonably simple alternative is available for direct elution whereby the liquid eluate following extraction could be diverted and diluted with matrix-matched solvent before being redirected to the HPLC column and/or detector.

Hematocrit

An area that is currently under scrutiny is how DBS sample hematocrit affects assay bias [12]. Current thinking is that this issue needs to be addressed before practical application of DBS analysis can progress to the next level, and any direct analysis technique needs to be compatible with this solution. The overall assay bias caused by changes in hematocrit is made up of three components; change to the area of the spot, variable recovery and suppression. The current strategy is that recovery and ion suppression bias can be eliminated by optimizing the extraction and HPLC methods. The ideal solution for controlling the area bias would be the use of a substrate that behaves independently of hematocrit, which would enable the current workflow to remain unchanged.

If such a substrate fails to be identified, the alternative could be to elute/extract the whole DBS sample rather than a central punch or sampling area, which would eliminate the effect of any area variability. Unfortunately, this would necessitate a change to our current workflow whereby an accurate volume of blood would have to be dispensed in the clinic. Numerous options are being explored that should make this possible if required, and the latest research has shown that direct elution is compatible with whole spot elution without loss of sensitivity or chromatographic performance [GSK, Unpublished Data].

Acknowledgements

The author would like to thank Neil Spooner and Philip Denniff (GlaxoSmithKline R&D, Ware, UK), and Babur Chowdhry and Frank Pullen (University of Greenwich, UK) for their valuable input into the topics discussed in this article.

Financial & competing interests disclosure

The author is an employee of GlaxoSmithKline (GSK) Research and Development and is eligible for GSK stock options and has stock ownership. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript.