Pharmaceutical processes typically take place on the scale of gallons or liters, even tens of thousands of liters. So why would anyone worry about one little microliter? Because in some situations, deviations as small as one microliter can mean the difference between success and failure.
Modern pharmaceutical research, quality control and testing laboratories increasingly rely on small volumes to identify drug candidates, perform toxicity screening or dosage testing, and to determine whether pharmaceutical products are within tolerance and can be released for public consumption.
A microliter is one-millionth of a liter. The technical definition is one cubic millimeter, a tiny cube that is only one millimeter on each edge – roughly the size of a medium-sized salt crystal. Such a small object can easily escape notice by the naked eye.
Then there are the physical properties. One microliter of liquid is a very small droplet, weighing only about one one-thousandth of a gram. Left unnoticed on a laboratory countertop, a droplet this small will evaporate completely in a very short amount of time.
High Performance Liquid Chromatography (HPLC) is a common analytical technique in pharmaceutical laboratories, and every HPLC analysis begins with “injection” of a small measured sample. Injector loops are used to measure out these small samples. Injection volumes of 20 microliters are common and loops as small as five microliters are sometimes used. Since the analytical results are directly proportional to the injected volume, an injection variation of plus or minus one microliter will yield a result that is too high or too low by five percent or more. The consequence of either variation is erroneous data.
Gas Chromatography (GC) is another analytical technique where the results depend on a highly accurate sample volume, and GC injections are often even smaller than those used for HPLC. Five microliter liquid injections are common in GC methods, and it’s not unheard of to have volumes as small as one microliter. Because of the proportional nature of this method, inaccuracy of just one microliter could produce a result that is in error by 20 percent or more.
Serial Dilution is another common laboratory technique that is increasingly becoming miniaturized. Serial dilution preparations in high density formats such as 96-well and 384-well microtiter plates are common. Working volumes in these wells are about 200 microliters and 50 microliters, respectively. That’s total volume, so the sample volume in a ten-fold dilution can be as small as five microliters. In this case, an under-delivery of one microliter in the first dilution causes an 18 percent error, and the error grows with each subsequent dilution to reach 64 percent by the fifth dilution in a series. Because the serial dilution method is frequently used in all types of pharmaceutical laboratories, and is often used for dose response testing and other volume-critical studies, these collateral errors caused by the initial one microliter error can be quite serious.
Polymerase Chain Reaction (PCR) is a technique for identification of nucleic acids. Because of the extreme sensitivity of this method, reagent expense, and the desire to perform testing in high density formats, methods have been developed that use single-microliter quantities of liquid materials. A common assay may call for 5.5 microliters of pure water and one microliter each of primers and master mix. Failure to deliver either of the one microliter aliquots would result in a false negative and completely invalidate the test. The strong growth of PCR assays in pharmaceutical labs and PCR-based genetic testing in clinical settings, and the constantly developing automation of PCR methods suggest that the number of failures of PCR tests of all types will only increase in the future.
From these examples we see that an error of just one microliter can cause errors that range from moderate (5 percent error in the results) to severe (20 and 64 percent errors in quantitative results). A one microliter error can even produce false negative results in qualitative testing. These sorts of errors are troublesome enough from the perspective of analytical quality in any laboratory, but are all the more severe in the strictly regulated pharmaceutical world.
Out of Specification Test Results (OOS) can and do result when quantitative and qualitative results are inaccurate. A “false failure” occurs when a drug product is actually in tolerance, but analytical error produces results that are outside established tolerances. In all cases an OOS investigation is required to prove that the drug is pure, safe and effective. These investigations consume time and money, and while the investigation is being completed, the product is held suspect. This can result in late shipments, production disruptions, and other hidden costs, as well as a reduction in good will and market credibility. Improper handling of OOS results is a leading cause of FDA citations.1
Improper Release Decisions (False Pass): Worse than a false failure is a false pass situation. This occurs when through analytical error a troubled product tests within tolerance and is released for distribution, even though the product actually is outside established tolerances. These sorts of errors expose a company to liability when and if the error is eventually detected. In extreme cases improperly released products may be consumed before the problem is detected, thus generating a potential for product liability actions by consumers.
Poor Reproducibility is always a concern in pharmaceutical testing and has ramifications ranging from inability to properly validate and transfer new methods, to falling out of compliance with existing methods. Poor reproducibility means that the same analysis on the same samples produces different results. This is a serious problem as evidenced by recent FDA notices.
FDA’s Quality System Requirements mandate that when equipment is found to be significantly out of calibration (and as we have shown even one microliter can be significant in some situations) remedial action must be performed. Remedial actions must ensure that there are no adverse impacts to process and finished products, and process and method validation studies, and that device or drug applications are within specified tolerances.3 Obviously, it is beneficial to avoid questions in these areas.
While one microliter may seem like an insignificantly small amount of liquid, there are a number of critical pharmaceutical processes where deviations of one microliter can produce large errors in results. These errors can be quite large, and the possible consequences cannot be taken lightly.
Given the trend towards handling smaller volumes in life sciences laboratories, methods to accurately measure such small volumes quickly and efficiently were needed. New measurement technologies like ratiometric photometry, have been developed that provide pharmaceutical labs with this capability.
Knowing when and where to employ these new technologies is a matter of analysis and judgment. The good news is that the latest quality systems standards regulations provide solid guidance for their application. This removes the guesswork and subjective judgment and gives the pharmaceutical laboratorian a clear direction to assure process and product quality.
George Rodrigues, PhD, is Senior Scientific Manager at Artel, a leading innovator in liquid delivery quality assurance. Rodrigues is responsible for developing and delivering communications and consulting programs designed to maximize laboratory quality and productivity through science-based management of liquid delivery. Rodrigues earned his BS in Chemical Engineering at the U.C. Berkeley, and a PhD in Chemical Engineering at the University of Wisconsin.