Tablet diversion strategy based on in-line NIR tablet press feed frame measurements

Alexander Ryckaert; Matthew Cannon; Thomas De Beer; Ashish Kumar

doi:10.1016/j.ijpharm.2025.125212

Tablet diversion strategy based on in-line NIR tablet press feed frame measurements

Int J Pharm. 2025 Jan 10:125212. doi: 10.1016/j.ijpharm.2025.125212. Online ahead of print.

Authors

Alexander Ryckaert¹, Matthew Cannon², Thomas De Beer³, Ashish Kumar⁴

Affiliations

¹ Laboratory of Pharmaceutical Process Analytical Technology, Department of Pharmaceutical Analysis, Ghent University, Ottergemsesteenweg 460, 9000 Ghent, Belgium; Pharmaceutical Engineering Research Group (PharmaEng), Department of Pharmaceutical Analysis, Ghent University, Ottergemsesteenweg 460, 9000 Ghent, Belgium.
² GSK Global Supply Chain, Priory Street, Ware SG12 0DJ, UK.
³ Laboratory of Pharmaceutical Process Analytical Technology, Department of Pharmaceutical Analysis, Ghent University, Ottergemsesteenweg 460, 9000 Ghent, Belgium.
⁴ Pharmaceutical Engineering Research Group (PharmaEng), Department of Pharmaceutical Analysis, Ghent University, Ottergemsesteenweg 460, 9000 Ghent, Belgium. Electronic address: Ashish.Kumar@UGent.be.

PMID: 39800000
DOI: 10.1016/j.ijpharm.2025.125212

Abstract

The tablet diversion strategy, based on in-line near-infrared (NIR) tablet press feed frame measurements, can be a key component of both batch and continuous oral solid dose manufacturing processes. It enables real-time, high-frequency monitoring and control, enhancing process understanding and compliance compared to conventional interval-based sampling methods. Central to this strategy are NIR spectrometers, which serve as PAT systems for in-line blend uniformity monitoring in the feed of the tablet press. These systems, when linked with the content uniformity of the corresponding tablets, are crucial for the tablet diversion strategy. In-line NIR blend uniformity measurements in the feed frame and off-line Raman tablet content uniformity measurements were performed during step change experiments for varying process settings. The Residence Time Distribution (RTD) from Tank-in-series (TIS) models for both spectroscopic measurement locations were used to derive the RTD from the NIR measurement location in the feed frame to the tablet diverter. This approach allowed prediction of API content in tablets based on in-line NIR measurements at the feed frame, avoiding yield loss by not diverting tablets that meet specifications. The impact of spectral loss during blend uniformity monitoring due to PAT sensor auto-correction was evaluated for all potential deviations during this 20-second period. Combinations of blends with an API content distinct from the target content, along with their durations observed by the NIR probe, that would exceed the upper limit were identified. However, materials that are not within specifications would not be diverted. It was concluded that diversions failing to trigger the diversion would only occur under very extreme and unlikely conditions.

Keywords: Near-infrared spectroscopy; Process Analytical Technology; Residence time distribution; Tablet diversion strategy; Tableting; Tank-in-series.