The current paper shows how excipient properties impact the process parameters and the final tablet properties in a fully integrated continuous direct compression line. Blend properties of low-dose (1% w/w) and high-dose (40% w/w) paracetamol formulations were evaluated and linked to the blending and tableting performance via multivariate models (Partial Least Squares analysis, PLS). Feeding behavior was analyzed separately, as the amount of active pharmaceutical ingredient (API) that ended into tablets was driven by random fluctuations in the API feeding behavior. The developed PLS models elucidated that formulation behavior was mainly driven by the concentration of the active pharmaceutical ingredient (API), explained by the distinct API properties. Excipient properties also had a substantial impact on formulation behavior. Generally, formulations with microcrystalline cellulose as a filler showed better compactability, lower hold-up mass, lower flowability and higher cohesion than formulations with different lactose grades. The relative performance of a formulation with different fillers differed for 1% w/w and 40% w/w drug loading. Granular and spray dried lactose grades increased in compactability ranking compared to anhydrous lactose when evaluating higher drug loading, due to the difference in morphology. It was shown that besides understanding the impact of excipients on the formulation performance, processability of ingredients is crucial for formulation design.

• Department of Pharmaceutical Technology and Biopharmacy, University of Groningen, Antonius Deusinglaan 1, 9713, AV, Groningen, the Netherlands
• DFE Pharma, Klever Strasse 187, 47568 Goch, Germany
• Laboratory of Pharmaceutical Technology, Ghent University, Ottergemsesteenweg 460, 900 Gent, Belgium
• GEA Process engineering, Keerbaan 70, 2160 Wommelgem, Belgium

• Oral Solid Dose

1. Continuous manufacturing Continuous direct compression
2. Excipients
3. Lactose
4. Material characterization
5. Microcrystalline cellulose
6. Multivariate analysis
7. PLS regression
8. Tableting