(W1130-10-54) Streamlining Pharmaceutical Process Development: A User-Centric Approach to High-Throughput Analysis

Purpose: Throughout the stages of pharmaceutical process development, new reactions, work-ups, and crystallizations need to be invented and optimized. This work demands efficient analytical methodologies that not only deliver robust results but also minimize the time and expertise required from end-users. In response to this need, we present a tiered approach for supporting High-Throughput Experimentation (HTE), composed of direct-injection mass spectrometry (MS) without liquid chromatography (LC) separation, rapid LC-MS analysis with short analysis time, and a comprehensive LC-MS analysis with moderate analysis time, combining technical innovation with user empowerment to streamline high-throughput analysis in pharmaceutical settings. The adoption of HTE in process development has thus greatly increased the number of experiments a single scientist can perform in a given time period.

Methods: Direct Injection MS: A proof of concept direct injection MS method was evaluated. An internal standard was used to normalize the signal for quantitation, and a good agreement was obtained when comparing direct injection MS to the traditional LC-MS analysis. Flow rate, diluent composition, mobile phase composition and sample concentration were evaluated to achieve optimal signal intensity and reliable integration of the peak of interest.
Ultra fast LC-MS analysis: While the direct injection MS has the potential of benefiting analysis with a large number of samples, we explore the integration of LC-MS approaches for smaller sample sets, leveraging UHPLC columns with small column geometries, e.g., 30-50 mm in length, to achieve the fast runtimes. Impurity mixtures containing 6 characteristic compounds were used to evaluate the efficiency of the methods, to ensure the feasibility of the methods in a High-Throughput setup.
End-user oriented model: Recognizing the need for user-friendly methodologies, we discuss strategies for further enhancing accessibility and usability, including pre-configured methods and streamlined workflows. We also discuss how to make the option of batch processing/ exporting readily available to the end users.

Results: Direct Injection MS: A condition at flow rate of 0.75 mL/min with a solution composed of 30/70 water/acetonitrile with 0.05% TFA was selected to analyze the samples containing the internal standard with a diluent of methanol (MeOH). By eliminating the need for LC separation, our method significantly reduced sample analysis time and solvent consumption by 40% and 66%, respectively. 
Ultra fast LC-MS analysis: Herein, we build upon existing platform methods available in the department, enabled template sample sets with either 2-min or 4-min LC methods, making the options with either a focus on analysis speed or a balance between analysis speed and separation efficiency available to the end-users, based on their project needs.
End-user oriented model: Our discussion on how to make the option of batch processing/ exporting readily available to the end users, including bundling multiple methods to reduce the activation energy for end-users, was shown to potentially benefit over 50% of the users in their daily work from a recent survey. Our findings underscore the importance of integrating technical innovation with user-centric design principles in pharmaceutical process development.

Conclusion: By addressing both technical performance metrics and practical usability considerations, we strive to empower end-users with the tools and resources needed to drive innovation and efficiency in their workflows. Our research represents a pivotal step towards advancing high-throughput analysis in pharmaceutical settings. Moving forward, we are committed to optimizing and refining our methodologies to maximize user accessibility and impact, ultimately fostering a culture of innovation and collaboration within the industry.

Table 1: Resource consumption of direct injection MS and conventional LC-MS (96-sample plate)

Table 2: Metrics for LC methods with different focuses (20-sample plate)

Table 3: Areas of opportunity