Design Space Calculation by In Silico Robustness Simulation with Modeling Error Propagation in QbD Framework of RP-HPLC Method Development

Abstract

Robustness of chromatographic methods is a key element in method validation that should be ensured within the method development stage. Several approaches in the literature were proposed to incorporate robustness in design space mapping for quality by design compliant HPLC method development. However, these approaches were far from the robustness definition within ICH Q2(R1). The current work proposed and evaluated an alternative approach for design space representation as the probability to pass an improved in silico robustness test that complies with ICH Q2(R1) robustness definition. This proposed procedure test was performed as a full factorial design of simulated deliberate shifts in each method variable for each predicted point in knowledge space with propagation of peak location error into predicted retention times. This approach was applied on the optimization of the separation of a test mixture of six compounds as a function of three experimental variables: gradient time, pH and ternary solvent ratio and two calculated parameters; initial organic solvent composition and instrument dwell volume. Design space was calculated as zones exceeding a threshold probability to pass the simulated robustness testing. The proposed approach of robustness simulation had advantages upon comparison with other approaches of design space calculation with respect to ensuring robustness of calculated zones. The proposed approach had the value of providing design spaces that represent more confidence of predicted method robustness as compared with other approaches described in this field.