Analytical Methods Validation: Design and Execution

By Cliff Nilsen

Validation is a multistep process with USP regulatory guidelines at each step Editor’s Note: This is the first in a series of articles on analytical method validation. The next article will finish defining each of the other method validation components. Method validation is the process of demonstrating through laboratory studies that an analytical method is suitable for its intended use. All analytical methods should be validated, the degree of validation being dictated by the method’s intended use. Methods that are published as part of official monographs, such as in the United States Pharmacopeia (USP), are generally considered valid for applications such as release testing of entities like active drug substances, excipients, drug intermediates such as blends and granulations, and finished drug products. The user need only demonstrate that the official method is suitable for that particular application. While this series of articles looks at traditional method validation parameters, it is mainly devoted to presenting newer Six Sigma statistical methods to provide more scientifically sound method validation results. Method validation guidelines are published by the USP as well as by the International Conference on Harmonisation (ICH), with each defining the required components of an acceptable method validation. The USP goes a step further, defining the validation components that are required for each type of analytical procedure, such as release testing, stability methods, and limits tests. The USP divides methods into three categories: Category I: Analytical methods used for the quantitative analysis of major components of bulk drug substances or active ingredients (including preservatives) in finished drug products. Category II: Analytical methods used to determine impurities in bulk drug substances or degradation compounds in finished drug products. Category III: Analytical methods used to determine performance characteristics, such as dissolution and drug release. The first step is to write a method validation protocol that includes what will be done, how it will be done, and acceptance criteria. Validation Parameters The design of an analytical method validation is driven by the category into which the method fits. The first step is to write a method validation protocol that includes what will be done, how it will be done, and acceptance criteria. The protocol must be written and approved prior to commencement of any validation work. A company should ensure that that one of the approvers is its quality control unit. Typical validation parameters for each category are shown in Table 1 (see p. 31). The validation parameters for a particular method type, as tabulated above, are as follows. Stability indication refers to a process whereby it is demonstrated that, after degradation of the test article, the analytical assay method is capable of producing reliable results for its active substances and preservatives, if applicable, without interference from degradation products. This can be accomplished by performing forced degradation studies on the test article and then assaying it for the active ingredient(s), showing that the active ingredient assay is free of interference from degradation products. For ultraviolet (UV)-active moieties (generally the case), this is done by high performance liquid chromatography (HPLC), using a photodiode array (PDA) detector. The diode array detector produces an entire UV spectrum continuously across a chromatographic peak. A comparison of spectral data at points along a peak (e.g., up slope, apex, and down slope) is used to determine whether or not a peak is pure (no interference), or impure (other components under the peak of interest). The degree of peak purity is used to determine whether or not the peak for the analyte of interest represents that analyte only, or the analyte plus some other compound that is hiding under the peak. PDA detector data can be interpreted several ways. One common method is the use of spectral overlays, which compares an overlay of the spectra at the up slope, apex, and down slope of a peak to look for differences. The overlay for a pure peak will appear as a single spectrum. Another method involves the use of ratiograms, which are plots of wavelength ratios as a function of time. Each peak appears as a square wave. A pure peak will have a perfectly straight horizontal component that is parallel to the baseline. The other technique is the use of purity parameter or purity factor, a mathematical formula used to judge whether or not a peak is pure. Each instrument manufacturer has its own formula. The diode array detector, although widely used, is only useful for HPLC procedures with UV-active materials. Materials that are not UV-active, or those that are assayed by other techniques such as gas chromatography, need to be treated somewhat differently. One approach is to assay the degraded material with two or more different chromatography columns, each having different polarities or functionalities. For example, if one were performing an HPLC procedure with a refractive index detector, using a C18 column, the degraded sample could be assayed using alternate columns such as a phenyl column or a cyano column. If the results obtained on each column for the degraded active are equivalent, there is a high probability that the peak of interest is free of other peak interference. With gas chromatography, this approach can be extended to the use of different detectors such as flame ionization and thermal conductivity, in combination with columns of different polarities (Carbowax versus SE-30, for example). Because flame ionization and thermal conductivity detectors have widely different responses to different compounds, if the assays are equivalent—with different columns and different detectors for an active ingredient after forced degradation—then the peak of interest is most certainly unadulterated by any degradation product. As with all peak purity evaluations, data is limited by instrument sensitivity. Therefore, it is a good idea to spike your matrix with a low level of a suitable impurity to see if the analytical system used can discriminate between the adulterated and unadulterated sample. Stability indication (part of selectivity/specificity) is done using several methods, including forced chemical and physical degradation. The approach is dictated by whether or not authentic samples of impurities and degradation products are available.