270830 Validation of a Scale-up Model for the Lubrication Blending Process From Material Sparing to Commercial Scale

Tuesday, October 30, 2012: 2:10 PM
Allegheny II (Westin )
Joseph Kushner IV1, Holger Schlack2, Margit Karle2 and Deborah Booth-Schindler2, (1)Drug Product Design, Pfizer Worldwide Research and Development, Groton, CT, (2)Pfizer PPD, Freiburg, Germany

Turbula bottle blenders are often used in material sparing scale experiments during early-stage pharmaceutical product development.  However, applying knowledge gained with these blenders to larger scale diffusion mixers is limited by the lack of blending models that are validated with commercial scale equipment.  The present investigation attempts to incorporate Turbula bottle blenders and commercial scale blenders into a recently-developed lubrication blending process scale-up model: 

QA/QA0 = (1-β) + β*exp(-γ*L*FHeadspace*r)

where QA0 is initial quality attribute value, β is sensitivity of the QA to lubrication, γ is formulation-specific lubrication rate constant, L is characteristic mixing length scale, FHeadspaceis the blender headspace fraction, and r is the number of blending cycles.

2:1 blends of microcrystalline cellulose and spray-dried lactose or dibasic calcium phosphate were mixed with 1% magnesium stearate using Turbula bottle blenders and pilot/commercial scale blenders, varying blender volume, V, (30-1250 ml; 100-2000 Liters); Fheadspace(30, 70%); and r (24 to ~190,000 cycles).  Tablets were manufactured using a Huxley-Bertram compaction simulator to evaluate the change in tablet tensile strength as a function of the extent of lubrication mixing.  Other attributes of the blends/tablets were also evaluated including compressibility, bulk/tapped density, friability, and disintegration.

The validity of the proposed lubrication blending process scale-up model was shown for blenders from 30-ml to 2000-liter (an approximately five order of magnitude change in scale) for both placebo formulations.  This validated scale-up model can be used to replace commercial-scale experiments with bench-scale, material-sparing experiments to support QbD lubrication blending process understanding and design space verification.

Extended Abstract: File Not Uploaded