Issue link: https://www.e-digitaleditions.com/i/1171020
16 Pharmaceutical Technology REGULATORY SOURCEBOOK SEPTEMBER 2019 P h a r mTe c h . c o m Analytical Procedure Validation scribes that solutions and drug substances are considered homogeneous samples. For analyti- cal procedures of drug products, the components of the drug product or the drug product may be mixed until homogeneous. Using individual drug product units should absolutely be avoided as it will bring the variability within the batch to the evaluation of the analytical procedures. However, for some analytical procedures the sam- ple preparation of a drug product unit (e.g., tablet, blister etc.) will be part of the method (e.g., dissolu- tion). In these cases, at the repeatability precision level it will not be possible to separate all sources of analytical variability from product variability. In the following examples, the authors detail a reversed-phase HPLC assay method to quantify a process-related solvent in an API, a cascade impac- tion method for an inhaled device, and an auto- mated content method for a tablet. Example 1: Reversed-phase HPLC assay method to quantify a process-related solvent in an API A reversed-phase gradient HPLC external standard- ization assay method is used to quantify a process- related solvent in an API. A reproducibility exercise was performed instead of an IP study during the technology transfer from an R&D site to the manufacturing site. Reproducibility was demonstrated by measuring the solvent from six preparations of two API batches in six independent runs—two analysts, instruments and columns on each of two sites, giving a total of four different ana- lysts, instruments and columns, used across six differ- ent days. An additional advantage of this design over that seen in Table I is that it enables more levels of the intermediate precision factors to be explored. Factors were selected that were considered most likely to impact method performance and combined for the study as shown in Table II. Given the sim- plicity of the chromatography—a single component assay with the solvent peak eluting isocratically at 100% aqueous conditions—it was not thought that evaluation of the individual factors on their own was likely to be needed (instrument and column factors were deemed low risk) and thus only three of the eight possible combinations of instrument/ column/analyst were run for each site resulting in imbalance of factor levels within each site. It was understood that potential variability existed in the external standardization that could be impacted by different analysts. As such, two analysts were used at each site. Had it been desirable to estimate the ef- fects independently, choosing a half-fraction design in the three factors (four of the eight combinations) would have been preferable. Six independent runs provide five degrees of freedom at the IP level, which allows the effects of random events on the precision of the analytical procedure to be established for routine operation of the method. Key variables known to impact the precision and accuracy of the solvent content results are incorporated into the design. Repeat- Table II: Reproducibility design for Example 1. Condition Factor: Site 1 2 Analyst 1 1 2 3 3 4 Instrument 1 2 2 3 4 4 Column 1 2 1 3 4 3 Independent Run 1 2 3 4 5 6 Within Run Two preparations per batch per independent run