Is specificity a crucial parameter in a validated, quantitative method ?

As the part on analytical method validation -with the ICH Q2(R2) scientific guideline as the starting point- is quite a significant piece of the course on instrumental analytical techniques (focus: spectroscopy) I’m teaching at the University of Leuven (@KU Leuven: congratulations with the 600 years anniversary !) obviously there are questions on the exam regarding this important item.

“Which of the experimentally derived validation parameters / characteristics are to be considered the most crucial ones?”, is such a question that challenges the theoretical knowledge (from the course) and the expertise of the development and use of a quantitative method, assuming we use the impurities-quantitative checklist of ICH Q2(R2) as the context for this question.

Parameters typically mentioned in the frame of a validated, quantitative (or limit test) method are (non-limitative): linearity, robustness, sensitivity, accuracy, repeatability, method detection limit, specificity, precision, method range, linear range, intermediate precision, method quantification limit, measurement of uncertainty , selectivity, reproducibility, system suitability, ……

If you’re working in this domain and want to reflect for a minute on your answers, pause your reading and try to put the 4 – 5 most important ones on a paper, and then continue reading.

?

Let’s start with the assumption of two knowledgeable people in the analytical chemistry domain who are discussing on a similar method they each in their own lab have developed and validated. What would be the “best” way of comparing the methods and have a conversation on whether both methods are suited for the intended use and if “yes”, which is the more performant one?

As an example, to make it practical, let’s take the following: “a need for a method to detect and quantify Butyl Hydroxy Toluene (BHT, a polymer additive described in respective compendial monographs) in an aqueous drug formulation at a concentration level of 150 μg/L (should be quantifiable)”

Complying to the intended use of the method and hence be a suited method is key in any method development and validation, one should really target and achieve this state of compliance, but in the same way also not doing more than needed (keep in mind it’s not an academic context) with the latter being the pitfall of skilled scientists in a commercial laboratory where productivity and profit are key values, next to delivering the appropriate science and quality.

I want to state that there’s not one right answer to the question as there can be a difference in perception on the 4 – 5 parameters selected for the answer, though some parameters should be part of the answer without any doubt, other ones have arguments pro and con. Isn’t Science so much more valuable when a two-way conversation is happening ?

?

Helpful could be to look at those parameters which are mainly or completely independent of the target species and/or the matrix in which the method should be established.

For example, linearity or better formulated “the relationship between an analytical signal and the respective amount of the species exhibiting/leading to this signal” is for sure an important aspect of the method established and especially of value during the subsequent measurements of samples (interpolation principles). But the outcome of the search towards this relationship is dependent on the techniques applied, the target, the concentration range, … ?

It’s known that some acidic species (e.g. polymer additives) with a certain chain length are linear in a signal-concentration range by LC/MS/MS whereas other chain lengths don’t show a linear relationship anymore (ex. Octadecanoic versus Arachidic acid). It’s also known that an ICP/MS method exhibits a long linear range for most elements (metals, e.g. silicon Si) where a GF-AAS methodology always has a short linear range (that’s technique intrinsic) for silicon and other metallic species. The same is also valid for Ion Chromatography (IC) and flame-AAS where the relationship is rather quadratic, respectively polynomial then linear, again intrinsic to the use of this techniques.

?

Specificity is always a key parameter, defined as “the ability to assess unequivocally an analyte , e.g. in the presence of other impurities, degradation products, matrix etc….”. It’s that parameter ensuring you’re actually reporting the results of species X and not the signals and respective calculated concentration of Y or a sum of X and Y, …. Whatever accuracy and precision the method holds over the entire method range, it’s worthless if the analytical signals are not from the target species and potential interferences, if present, have not been dealt with.

So, referring to the example question, the signal should be the one from and designated to BHT and not (partly) from BHT-OH, a highly similar chemical species, known to be formed in polymers (intentionally, as BHT is an antioxidant). That’s in essence what specificity should guarantee and why it’s of upmost importance in any method development and validation.

?

Having that conversation could deal with? a HPLC/UV versus a GC/MS method for BHT, with the latter method having a retention time (RT), a relative retention time (RRT) and above all mass spectral data and SIM settings with a target and qualifier ions and the former method a retention time, a specific wavelength (lambda) and possibly a relative retention time (RRT). The HPLC/UV could be suited for intended use if the matrix is relatively easy and “clean”, no interferences are to be expected and observed and the peak shape complies to e.g. predefined symmetry criteria. The strength of this methodology could be increased by applying a fluorescence detector (FLD) in series with the UV-detector (or stand-alone) as the excitation and emission wavelenghs established and set increase the specificity.

But always the GC/MS SIM method will be superior, this of course requires a more expensive equipment, an expertise with mass spectrometry and knowledge in this domain versus the more widely used and in general more accessible HPLC/UV systems.

?

To define the best possible technology to develop an analytical method, next to the instrument’s availability in the laboratory, scientific and analytical knowledge and expertise is required, combined with some expertise in specific targets and matrices to proactively understand potential hurdles and pitfalls rather than facing them during the experimental work.

For example: Analyzing “Si” in an aqueous solution (or drug product) can be done in various ways and with more than one technique, e.g. GF-AAS, ICP/OES, ICP-MS, flame-AAS, but each of the mentioned technologies has its own advantages and disadvantages when it comes to specificity for silicon.

?

Next to specificity, of course other parameters are also crucial and necessary to understand the capabilities of the method established, but that’s for a future thought sharing. As is the variety of Internal standards which are preferably also part of a method.

?

References

ICH Q2(R2): Validation of analytical procedures - scientific guideline (June 2024 version)

ICH Q14 : Analytical procedure development - Scientific?guideline (June 2024 version)

applies to new or revised analytical procedures used for release and stability testing of commercial drug substances and products (chemical and biological/biotechnological)

Nomenclature

ICP/MS – Inductively Coupled Plasma Mass Spectrometry

ICP/OES - ?Inductively Coupled Plasma Optical Emission Spectroscopy

GF-AAS – Graphite Furnace Atomic Absorption Spectroscopy

HPLC/UV – High Pressure Liquid Chromatography UltraViolet

LC/MS/MS – Liquid Chromatography tandem MS

#ICHQ2R2 #ICHQ14 #methodvalidation #specificity #methoddevelopment #pharmaceutical #polymeradditive #extractables #leachables

Emily Brown Aryo Nikopour Shiri Hechter Stacie Mink Alpa Patel Martell Winters Matt Cushing Michelle Lee Matthew R. Jorgensen, PhD, DABT Matthias Onghena Jonas De Vits Johan Jaubin Fran?ois B. James Mullis Mike Hodgson Carola Damen