As we are all acutely aware of, within regulated bioanalysis, we have a requirement for the production of rugged and reliable quantitative methodologies. This pertains to our analytes, typically drugs or drug candidates and select metabolites, in a given biological matrix of a given species. These complete methods include three elements that marry together as a final entity to produce high-performance data capable of swiftly swimming through a full method validation. Constituting the analytical endpoint, we have the chromatography and the subsequent mass spectrometric detection, but prior to these, we have the sample preparation, always involving some means of extraction of the analytes from the matrix, making it a most potent avenue for attaining great selectivity. Or, with certain decisions, to avoid any meaningful selectivity. The broad bracket of extractive options encompassing such different extents of selectivity is the methodological aspect dwelt on here.
Beyond discovery-focused liquid chromatography–mass spectrometry (LC–MS) workflows, and particularly since the arrival of triple quadrupole mass spectrometry as the top-notch mode of detection a few decades ago, there has been a widespread strong preference for the option of rapid and straightforward sample preparation over longer, more laborious techniques.
There exists a cliché that I still hear sometimes, but not as frequently as around 20 years ago: “quick and dirty,” which reflects the zeitgeist just alluded to, from the time of the popular industry onset of tandem mass spectrometry. The phrase, of course, pertains mainly to the sample preparation; a direct reference to the route of protein precipitation for plasma samples—in essence, the very mindset leading to dilute-and-shoot whenever possible, and minimal sample processing with minimal direct costs and time required.
However, although this approach may be fine for certain analytical domains, can we afford to have an aversion to looking any further than “quick and dirty” in regulated bioanalytical LC–MS? It’s not that I am determined by any means to adhere to a “laborious and clean” outlook, and speed is, of course, a desirable feature. Rather, the crux of the matter is the cleanliness, the selectivity, the tip-top instrumental operation, and ultimately the high-performance data obtained when taking the trouble to prepare and extract samples in appropriate ways. I also wish that latter phrase would simply roll off the tongue a little easier.
When there are oodles of sensitivity and an ideal internal standard of practically identical physicochemical properties as the assigned analyte, performance may well be tickety-boo with a simple, non-selective sample extraction. Tiny injection volumes, hence minimal signal drift, no significant matrix effect, and all bias values in calibrants and quality controls (QCs) tight around the ideal. However, when faced with the need to inject greater amounts on-column, we are prompted to consider an alternative, more involved extraction, like liquid-liquid (LLE), or the analogous and easily automatable technique of supported-liquid extraction (SLE), typically great for small molecules once properly optimized. Then there is generally the most selective and tunable option and, crucially in this new era, by far the most applicable to biologics, solid-phase extraction (SPE).
The approach of SPE is least amenable to the principle of “quick and dirty”, but with a vast array of sorbent chemistry flavors and combinations to choose from, together with the associated wash and elution procedure offering a great deal of fruitful optimization, it will bring the user as close to the solution domain as possible in terms of interferent elimination.
Where quantitative issues remain post-extraction, it must be borne in mind that with less selectivity in the chosen sample extraction, there is accordingly more emphasis on the liquid chromatography and mass spectrometry to provide the necessary discriminating power. A good marriage of selectivity, between extraction and LC–MS, is there to be found.
Lastly, I often imagine that in certain contexts, such as regulated bioanalysis, the goal of cost-saving through inexpensive and non-selective sample extractions is a misnomer. There are inescapable running cost implications of “dirty” samples, instrumental fouling of the interface and beyond, leading to increased downtime and critical part replacements. Interwoven are batch failures due to phenomena such as signal drift, and simple blockages of various flow paths, leading to repeat analysis—a very expensive and unattractive scenario.
That concludes my salute to the sample preparative domain, underrated but marvellously potent as can be perceived if we look beyond the basic.
Disclaimer
The opinions expressed are solely my own and do not express the views or opinions of my employer.