BioPharm October eBook: Best Practices 2018

Issue link:

Contents of this Issue


Page 23 of 27

24 BioPharm International eBook October 2018 At this stage, a risk analysis should be conducted to identify and rank parameters that are likely to impact method performance and, therefore, conformance to the ATP. Risk assessments are typically itera- tive throughout the lifecycle of a method. They are performed dur- ing method development to iden- tify method parameters that present highest risk as well as the end of method development to ensure that any potential risks associated with the developed method are identi- fied, as appropriately, and mitigated. Risk analysis is also performed prior to method transfer or during a prod- uct change (e.g., route of adminis- tration, formulation, or process). The risk assessment performed during method development should be focused on identifying factors that could impact method linearity, precision, accuracy, signal-to-noise, and specificity, in addition to other method parameters. The risk analy- sis should include materials needed for method execution, equipment, operators, and various method ele- ments. Using chromatography as an example, such elements would likely include column selection, column temperature, mobile phases, run time, flow rate, and gradient. Each of these risk-analysis components are assigned a risk weight; those deemed to have medium-to-high risk would require a mitigation plan. The mitigation plan is typically comprised of non-experimental fixes for materials, operators, etc., and experimental fixes for method elements (i.e., column tempera- ture, gradient, flow rate) that are addressed through a DoE approach. Fu r t her r isk assessment w i l l also be performed when transi- tioning the method from develop- ment to the commercial stage. For this assessment, focus is placed on method ruggedness or more spe- cifically on the sources of reagents, laboratory practices, the environ- ment, test ing c ycle t imes, a nd equipment availability. DESIGN OF EXPERIMENT Once an appropriate risk analysis has been completed, it is time to design the experiments. For this, one should consider the following steps: • B e m i nd f u l of t he met ho d requirements per ATP such as repeatability, precision, accuracy, limit of detection (LOD)/limit of quantitation (LOQ), linearity, and resolution. • Def ine the range of concen- trations to be measured by the method, along with solution matrix of the analyte. • Obtain/prepare reference stan- dards for bias/accuracy evalua- tion. • Def ine d isc rete steps in t he method, previously identified in early scouting studies. • Identify responses that address method purpose. • Identify an error control plan. • Design the experimental matrix and sampling plan. • Run study. • Analyze results and determine settings and processing condi- tions that result in desired preci- sion, accuracy, and linearity. • Document design space of the method and upper and lower limits for critical parameters. • Ru n con f i r mator y st udy for method settings/limits. With these steps, it is important to be careful in defining the range of concentrations, as this will be important in defining the method design space. If the design space is too restrictive, the method will be limited. There are many options for the experimental matrix in the DoE study. For small studies in which only two or three factors are being evaluated, a full factorial design is appropriate. For studies with five or more variables, then alternate desig ns should be explored, as detailed in Figure 2. Analysis of variance (ANOVA)/ analysis of covariance (ANCOVA) and range analysis are the basis of the statistical analysis methods for response surface designs. Range analysis is frequently employed to establish each factor's effect and identify the best level for the differ- ent factors. For a factor, the range of means is the difference of the maximum and minimum means. In the design space of a method, the factor (or variable) with the Biopharma Laboratory Best Practices Analytical Methods Figure 2. Experimental design approach based on number of variables.

Articles in this issue

Links on this page

Archives of this issue

view archives of BioPharm - BioPharm October eBook: Best Practices 2018