Bioanalysis: LC-MS-MS, Sample Prep and Dried Blood Spot Analysis

Bioanalysis uses a variety of separation techniques to analyse samples ranging from plasma and urine to dried blood spots. Participants in this Technology Forum are Ling Bei, Patrik Appelblad, and Dave Lentz of EMD Millipore; Nadine Boudreau of PharmaNet Canada; Diab Elmashni, Jeff Zonderman, and Simon Szwandt of Thermo Fisher Scientific; and Debadeep Bhattacharya of Waters Corporation.

Bioanalysis uses liquid chromatography–tandem mass spectrometry (LC–MS-MS) as the main separation and quantification technique, and typically batch sizes can consist of up to several hundred sample, standard, and QC injections. What should you do to ensure that the chromatographic separation remains consistent throughout the run?

Bei, Appelblad, and Lentz: Bioanalysis samples (saliva, urine, whole-blood, plasma, serum, CSF, tissue, and so forth) may contain water, ionic salts, an assortment of proteins and lipids, anticoagulants and stabilizers, as well as many other substances which can interact temporarily or permanently with stationary phases, causing compromised LC column lifetime. This can result in gradual deterioration in its performance over its lifetime, throughout a long run, or even rapid failure in the first few injections due to pore blockage by the adsorption of large protein molecules or precipitation of sample components in the mobile phase. Appropriate sample prep, including filtration, protein precipitation, and possibly solid-phase extraction (SPE) clean-up steps help a great deal. Automated column switching for these clean-up steps can save time and make them more effective. Guard columns can help a great deal, as long as what they collect doesn’t just get washed onto the column anyway in washout steps. Long gradient reequilibration times are often necessary to make baselines stable enough for the next injection. Cleaning or “blank” gradient runs or wash steps between injections are routinely used as well. In addition to these considerations, bioanalytical methods often push the limits of detection of the instrumentation used. All these factors make it very necessary for bioanalysts to design methods thoughtfully, to prepare for the worst and hope for the best. These methods should contain self-monitoring steps that will periodically document system performance so that when results are reviewed after long unattended runs, they can be sure that the data they’ve got were generated by a system in control and can be trusted. Frequent injections of standard concentration profiles to check linearity will give some assurance of continued suitability. Repeated blank injections or negative spikes will reveal carryover issues. Replicate injections of samples as well as standards will help test reproducibility and overall precision.

What can most dramatically help ensure consistency with your method over time is by first establishing appropriate goals for it and then taking a step back to see the bigger picture and prioritize those goals and assess their practicality. Choosing the right chromatographic selectivity needed is key. Assays should be robust, using no more separation power than needed. If the detection sensitivity is sufficient, you can use a larger stationary phase particle size, which is less prone to matrix effects. The smaller the particles used in the column, the more prone the column will be to clogging, so a 3- or 5-µm material will be more appropriate. There must be enough flexibility built into the method to allow for slight changes in sensitivity in detection and to accommodate differences among samples. Also, choose the right column dimensions. Bioanalyses require rugged methods that are stable over long series of samples, for days, weeks or years — one that doesn’t depend on several critical method parameters with tight working ranges to work. A little thought up front will pay off later in a lot of peace of mind so you can trust your results.

Boudreau: During method development, the best chromatographic conditions are determined to have the most robust method possible. Analytical column and mobile phase selections are determined according to the analyte of interest and the matrix. Also, during validation the robustness of the method is also verified. This is done by injecting several samples that will represent a study sample analysis run. Moreover, during validation since numerous samples are injected in each run this will give a good idea also on the robustness of the same. Furthermore, extra cleaning steps could be performed before each run to prevent any deterioration of the columns (like cleaning of column and peak tubing ) before each run. Of course, the extraction procedure can have a great impact also on chromatography. Having the cleanest possible samples is an asset. This will increase lifetime of analytical columns.

During study sample analysis, system suitability samples (SST) are injected at the beginning of a batch run. This will permit monitoring of the sensitivity, selectivity, peak response, retention time and peak shape. The SST samples are included in each run and are verified by a qualified individual.

Elmashni: You should ensure you have set up your gradient to complete and equilibrate the column after each injection before the next run is started. Also you should compare the periodic QC injections to ensure the retention time, peak capacity, and column plate count are consistent. When you begin to see a large variation in those three items from QC injection to QC injection then you know you are losing some of your chromatographic efficiency.

Bhattacharya: LC–MS–MS technology, owing to its sensitivity and speed of analysis, is the chosen technique in bioanalysis. However, achieving the desired sensitivity and resolution within a short time has been a challenge. The growing demand of optimizing the lower limit of quantification (LLOQ) of a drug molecule to a lower amount and the ability to analyze peptide drugs and biomarkers makes it more challenging for the LC–MS–MS instruments.

Bioanalysis involves the analysis of extracts of plasma and urine from different species, and these biological fluids contain many endogenous components. How important are the sample preparation techniques that you use to maintaining the life of your high performance liquid chromatography (HPLC) columns?

Bei, Appelblad, and Lentz: As was mentioned before, there are many things to consider in dealing with biosamples when the primary goal is to maximize column lifetime, and there are many things one can do to achieve that. But isn’t it always really a tradeoff between the time and cost of sample clean-up vs the cost of the HPLC column? It all depends on the goal of the method. Some users even choose to inject very dirty samples and forgo sample preparation completely, just accepting very short column lifetimes of a few dozen injections. They have decided this is the optimum use of their resources and it works for them. Others may develop very rigorous sample preparation protocols we just discussed that will ensure column longevity. But the truth is that all HPLC columns have finite lifetimes. The only ones that last forever are the ones that are never used. Although they are highly technical products with many years of research and development behind them and are the separation workhorses of the system, they are, after all, consumable items. Nobody expects tyres to last the whole lifetime of a car. With the high quality of modern columns, analysts who buy good quality columns can be very sure that the next replacement column they buy will perform just as the previous one did when it was new. The good news is that even those who choose to accept reduced column lifetimes still have alternatives to that in the modern age of HPLC. For example, silica monoliths are much more forgiving of dirty samples. Their useful lifetime is also not only much longer than that of particulate columns but is more consistent throughout it because there is no particle bed to shift with each injection. New hydrophilic interaction chromatography (HILIC) materials can be used to remove detergents, unlike most traditional SPE materials. So there are new choices available to challenge old paradigms with more freedom to give back to users to find the solutions that work best for them.

Boudreau: It is crucial to select the extraction method to obtain the cleanest samples. Extracted samples should be similar as much as possible to neat solutions. The cleaner the extract is, the longer the column will last and the chromatography will remain consistent. The newest technologies in LC–MS–MS allow us to achieve very good sensitivity, offering analysts the choice to dilute the samples by a higher factor, improving the cleanliness of the extracted samples. Moreover, new SPE sorbents and protein precipitation filtration plates are in direct relation with clean samples.

Zonderman: The sample preparation method directly impacts the extent of how clean a sample arrives to the ultrahigh-pressure liquid chromatography (UHPLC) column and is very important to the lifetime of the column — especially as more and more UHPLC analyses are used with sub-2-µm particles. These columns require a cleaner sample extract to maintain an optimal column lifetime. Many people have moved away from a protein crash to liquid–liquid or SPE to address this issue. Unfortunately, this can require extra effort and add cost. Instead, we recommend implementing an automated online sample preparation.

Bhattacharya: Matrix interference is one of the major effects faced in the pharmaceutical world and is caused by endogenous and exogenous materials, phospholipids that are not resolved from analytes, inefficient chromatographic conditions, and so forth. Matrix effects can affect the ionization process of an analyte during an LC–MS experiment, resulting in batch failures and irreproducibility. In addition, the presence of endogenous and exogenous materials also affects the lifetime of the column that is used in the LC system.

In the world of bioanalysis encompassing pharmaceutical companies to CROs, there is a constant demand for reaching a lower concentration to quantify a given drug molecule or candidate. In addition, the growing market of peptide therapeutics and biomarkers also requires an efficient method for removing matrix effects in the sample so that a proper LC–MS analysis of the peptide can be performed. Such a demand of removing or quantifying matrix effects not only challenges the capabilities of the LC–MS systems, but also calls for an efficient sample preparation method.

There is great interest in the bioanalysis of dried blood spots from samples of nonclinical and human studies because of cheaper transport costs, simpler sample preparation, and so forth. What is or will be the impact on bioanalytical method development and validation as well as its application to routine analysis?

Bei, Appelblad, and Lentz: Dried blood spot (DBS) analysis has been around for almost a half century. It was a tremendous advance over traditional venous blood collection in terms of portability, stability, and safety. It requires far less sample (a drop or two of blood) and is much more accepted and convenient in developing countries, allowing far greater numbers of samples to be taken and assayed, dramatically increasing population data in disease studies. DBS analysis has integrated well into the bioanalytical workflow, providing low cost, fast quantitative results, in areas such as paediatric studies and preclinical rodent studies where only small volume samples can be obtained, and Phase II/III studies, where collection and storage of samples are limited to ambient conditions. Stability studies of DBS samples in extreme environments have proven its reliability in inhibiting degradation of its components over long periods of time.

Internal standards allow fully quantitative methods to be developed from them. Analysis of the dried spots can even now be fully automated with recent innovations in high performance thin-layer chromatography (HPTLC) technology. These methods validate very well and have proven to be robust and accurate. Instrumentation has even been developed to connect them directly to mass spectrometers, without the need for a chromatographic column separation.

These very interesting applications will continue to have a positive impact on future routine bioanalytical method development as they illustrate the possibilities that open up when old assumptions are challenged. Sample storage can be much less of a bottleneck now. Dramatic increases in throughput are then possible. Breaking away from the “everything in solution” paradigm has allowed DBS techniques to develop beyond the restrictions of conventional wisdom about classical analyses. The fact that DBS methods validate well also shows that reliable quantification is possible under conditions previously ignored. There are also implications for LC development — that the HPLC column isn’t always a critical component in routine quantitative bioanalysis. This may help to encourage development of new column technologies that take advantage of this, to concentrate its contribution on more targeted resolution tasks, freed from the total burden of separation.

Boudreau: Though being of very high interest for sample storage and handling, the arrival of DBS will bring new challenges for method development. Being able to achieve the requested LLOQ with lower blood volume will be a challenge for some methods. Moreover, additional validation tests will need to be added to cover all aspects of sample handling, shipment, storage, and processing (for example, blood temperature, punch size). Finally, because no studies have been audited yet by any regulatory agencies, new questions will probably be raised and will need to be addressed in future DBS studies.

Szwandt: Dried blood spots are nothing new to routine analysis, having been used for many years in paediatric medicine. Recently there has been a significant push within the pharma industry to assess and develop dried blood spot analysis (and indeed urine and plasma spots too).

Utilizing dried blood spots has significant advantages over the more traditional liquid plasma sampling. Firstly it requires much lower sample volumes. This is particularly advantageous when looking at animal studies and paediatric studies. Sampling is generally much easier, less invasive and requires less training when compared to venipuncture. Perhaps the biggest driving force in the move to DBS is the potential for very large cost savings. The savings include sampling, processing of the sample (for example, centrifugation), storage of the sample (–20 °C or –70 °C), and shipment of the sample (usually on dry ice). There is also the cost associated with training technicians and clinicians.

There is the potential to generate better toxicokinetic (TK) and pharmacokinetic (PK) data. This can be exemplified by the use of single animals for TK studies rather than one animal per sample. It also benefits the significant drive in the 3 Rs: reduction, replacement, and refinement in the use of animals.

When it comes to processing the samples typically it can be easier to extract the sample, leading to cleaner extracts when compared with typical plasma crashes. Some of the advantages can also be seen as disadvantages. Low sample volume can lead to problems when considering inhaled products or very low dose treatments, and the required limits of quantification may be difficult to achieve. This is also particularly important for the MIST guidelines for metabolites in safety testing. There is also the question of sample stability while the sample is drying on the card. The clinical lab conditions may not be suitable (humidity, temperature, and so forth). In sample analysis many labs have questioned the potential for “carry over” related to the punching of cards, and furthermore there is a lack of accepted automation for sample preparation.

There are many challenges still facing the bioanalytical community as it continues to push forward with DBS. Firstly, all historic PK and TK data are typically from plasma samples, and does not necessarily reflect the true systemic kinetics of whole blood. This single fact is a significant hurdle that has to be overcome. All DBS analysis has to be validated in parallel with classical liquid (plasma) sampling. Sensitivity, particularly with inhaled and or very low dose products, will continue to prove challenging. Here systemic concentrations are extremely low and in some case alternative analytical techniques have been employed, including nanospray. While validating, the effects of treatments to the DBS cards have to be considered. These additives and preservatives may potentially cause analytical issues including suppression.

Most importantly perhaps is the need for recognition and acceptance in the bioanalytical community and especially by regulatory authorities (FDA) of the transition from plasma-based TK and PK data to blood-based data.

Bhattacharya: Use of dried blood spots (DBS) has gained tremendous interest in the bioanalysis community in the recent past owing to its low volume use, which allows fewer animal compromises during the phase II development of a drug. In addition, DBS offers potential advantages in transport costs of samples, serial versus composite sampling, paediatric sampling, and so forth. However, some of the aspects that should be considered for validation of DBS methods include extra matrix effects from card; shipping of QCs; hematocrit issues with clinical samples; compound and metabolite stability on card; effect of different blood spot volumes; and accuracy of spot size cutting.

If you are interested in participating in any upcoming Technology Forums please contact The LCGC editorial team for more information.