(M1230-08-53) An Investigation of Robust Sample Preparation on an Automated Tablet Processing Workstation, and Lessons Learned

Purpose: The intent of this research is to mitigate challenges, including variable API assay results, associated with automated sample preparation using the Tablet Processing Workstation (TPW) for low-volume and highly potent sterile formulations. This work ultimately enabled a complete, automated sample preparation capable of generating reliable and reproducible results while minimizing analyst exposure to highly-potent material.

Methods: The TPW automation platform was used during the development of highly-potent sterile solution-for-injection formulations. Wide replicate spreads were initially observed in several sequences, especially in longer sample sets (25-minute preparation time per sample). It was hypothesized that this resulted from the low formulation and analytical volumes and long run times: in a typical automated sample preparation, individual formulation samples (1 mL) were transferred into test tubes closed with metal caps and placed on a TPW to run overnight. The TPW process was monitored carefully, and changes to the automated setup, including the type of caps used on the test tubes, protection of the samples from heat, and additional mixing were applied.

Results: The observed high and variable assay values were found to result from solvent evaporation and concentration of the analyte in the test tubes. This was accelerated by heating the samples with the TPW electronics underneath the sample tray, leading to the evaporation of the sample solutions with subsequent condensation of water on the metal caps. To mitigate this, Styrofoam was added between the TPW electronics and the sample tray and resulting in decreased evaporation of the sample solutions. Additionally, metal caps were found to condense water and retain the resulting droplets to a larger extent than plastic caps. Plastic caps were therefore specified in the analytical method, and ten seconds of vortexing was added before sample preparation to dislodge and reincorporate any condensation that may occur. These changes enabled robust sample preparation even for long sample sequences.

Conclusion: The root cause of high and variable API assay values associated with automated sample preparation of solution formulation samples was identified as solvent evaporation exacerbated by adventitious sample heating and subsequent condensation on metal test tube caps. Mitigating these factors by minimizing heating and using plastic caps enabled a robust, reliable automated method for routine sample analysis. These highly effective strategies are expected to be generally applicable to the analysis of other solution formulations with low sample volumes and sensitivity to evaporation.

Representative data showing higher % LC for Research compound in longer sample preparation runs.

Representative photo of water condensation using diverse types of caps.

Eleven water samples were used to mimic longer run. Six Research compound sterile solutions were used to compare manual preparation VS. automation (with and without Styrofoam)