Category: Formulation and Quality
Purpose: Formulation ruggedness studies are an integral part of justifying product quality attributes, confirming the design space and establishing a suitable control strategy during commercial product development. For biologics products, the scope of formulation ruggedness studies typically includes an assessment of the potential for variations in multiple formulation-related factors, such as solution pH, protein concentration, excipient concentration, etc., to impact product quality or stability. Ruggedness studies are often conducted using a statistical design of experiment (DOE) approach to enable efficient selection of conditions to cover the design space as broadly as possible. Even with these efficiencies, however, the number of formulations that must be prepared (frequently using tangential flow filtration or TFF) to support even a basic ruggedness study may be relatively large, and the experimental burden of sample preparation can quickly become a challenge. In addition, since surfactants such as polysorbate 80 (PS-80) are generally retained in the product solution during TFF, it is often preferred to obtain surfactant-free starting material in order to properly control relative surfactant levels in the final formulations for study. This presentation describes the development and implementation of a simple calculation-based approach to enable preparation of various histidine-based formulations, at different pH values, from a starting drug substance solution, without using TFF.
Methods: The initial drug substance used for these studies was formulated in 20 mM histidine buffer at pH 6.3, with 220 mM sucrose and 0.05 % (w/v) PS-80. The target pH values for the formulations to be evaluated in the ruggedness study ranged from pH 5.8 to 6.8. A series of mathematical equations was used to calculate the respective amounts of each formulation component (histidine, histidine hydrochloride, sucrose, PS-80, and pentetic acid that must be added to the starting drug substance solution to achieve the required final compositions of interest.
Results: Seventeen different formulations were then generated by aseptic addition of stock solutions of the individual buffer components and excipients to the drug substance, at the volumes specified by the calculation, using an automated liquid handler (Core Module 3, FreeSlate). The protein concentration and pH for each formulation prepared were measured and found to be in good agreement (+/- 6 % for protein concentration; +/- 1.9 % for pH) with the calculated values.
Conclusion: Using this approach, multiple compositions to support a wide formulation design space could be easily generated, with substantial savings in time and material compared with a conventional TFF-based buffer-exchange process. In addition, the new procedure allowed drug substance containing surfactant to be used as a starting material and may, therefore, provide needed flexibility when surfactant-free drug substance supply may not be readily available.