(166d) The Development of Methods to Link Design Space Models to Product Stability

Zong, Z., The University of Iowa
Wildfong, P. L., Duquesne University
Suryanarayanan, R., University of Minnesota
Munson, E. J., University of Kansas
Kaushal, A. M., University of Minnesota
Barich, D. H., University of Kansas
Pingali, K., Rutgers University
Muzzio, F. J., Rutgers University
Kayrak-Talay, D., Purdue University
Desai, S. D., The University of Iowa
Buckner, I. S., Duquesne University

Members of the National Institute for Pharmaceutical Technology and Education (NIPTE) have been working on an FDA-sponsored project entitled: ?Development of Quality by Design (QbD) Guidance Elements on Design Specifications Across Scales with Stability Considerations?. The overall objectives of the project are to improve pharmaceutical product quality and maximize process innovation and continuous quality improvements by: a) developing QbD guidance elements on process design space, scale-up, and process validation for three unit operations: mixing, granulation, and drying; and b) developing a framework for optimizing design space specification across scales with consideration to stability.

The key to linking design space models to product instability is the identification of a linkage between physical change in API that result in a changes in the susceptibility of that API to chemical degradation. The presence of this linkage means that monitoring the physical state of the API may predict the product's susceptibility of chemical instability failure.

The model drug for this project is gabapentin which is known to degrade to gabapentin lactam. The degradation is sensitive to formulation and processing conditions. The topic of this presentation is an interim report on the development of methods and the assembly of correlations between physical methods and susceptibility of chemical instability based on the evaluation of materials generated in laboratory scale process simulations.

Our approach has been to 1) screen a series of potentially using solid-state characterization methods and to install and validation stability-indicating chromatographic methods and 2) to subject the various physical forms of the API and powder blends to typical processing and storage stresses to determine whether quantifiable changes can be observed. Specifically, we have evaluated stability-indicating HPLC, SS-NMR, differential scanning calorimetry, and powder X-ray diffraction as potential characterization and/or stability predictors. Samples of gabapentin were subjected to processing stresses associated with granulation, milling, drying, and blending.

Physical Characterization Methods:

A thermogravimetric analyzer (TGA, Model Q50, TA Instruments) and a differential scanning calorimeter (DSC, Model 2920, TA Instruments, New Castle, DE) were connected to a thermal analysis operating system (Thermal Analyst 2000, TA Instruments). Approximately 2 to 6 mg of the sample, in a crimped aluminum pan or an open aluminum pan, was heated in the DSC/TGA from room temperature to 250 C at varying rates under nitrogen purge. XRD analysis was conducted on samples in a glass holder and exposed to Cu Ká radiation (45 kV x 40 mA) in a wide angle X-ray diffractometer (model D5005, Bruker, Madison, WI) at ambient temperature. The instrument was operated in a step-scan mode, in 0.05°2q increments, and counts were accumulated for 1.0 second at each step over the angular range of 5 to 40°2q. Data analyses were performed with commercially available software (JADE, version 8.0, Materials Data Inc., Livermore, CA). Solid state NMR spectra were acquired on a CMX spectrometer operating at 300 MHz for proton and 75 MHz for carbon. Samples were packed into a Revolution NMR 7 mm (o.d.) zirconia rotor. The spectra were acquired with cross polarization (CP) using a linear ramp on the proton channel during the CP period. The optimal contact time varied between samples, but was typically around 0.9 ms. Magic angle spinning (MAS) was done at a rate of 4.0 kHz. Proton spin-lattice relaxation data were acquired via saturation recovery. Spinning sidebands were eliminated from non-relaxation spectra via sideband suppression techniques.

Chromatographic methods:

RP-HPLC analysis was carried using a high/low assay wherein trace amounts of degradation product (lactam) were measured using high concentrations of parent API (12 mg/mL) and sample API amounts were measured at about 2.4 mg/mL. These two results were then used to compute the lactam content as a percentage (w/w ) of API. Calibration standards for the lactam and gabapentin were prepared in the range of 0.5-5 µg/mL and 0.5 to 5 mg/mL, respectively. The HPLC system was a

Thermo Spectrum System P4000 pump, AS3000 auto injector, and UV 6000 LP photodiode array detection system. The column was ìBondapak Cyano column 3.9x300mm. Isocratic analysis was conducted using a mobile phase composed of 95 parts buffer (10 mM KH2PO4/10 mM K2HPO4 ) and 5 parts acetonitrile . Analysis was carried out using an analytical wavelength of 210 nm, flow rate of 1.0 mL/min and a 0.020 mL injection volume. Run times were 10 minutes. Retention volumes for gabapentin and lactam were 4 and 8.5 mL, respectively. Calibration plots were linear. To evaluate the susceptibility of gabapentin materials to thermally-induced degradation, an aliquot of API (approximately 15 mg ) was weighed and placed in type II glass vial, sealed with a Teflon-faced butyl rubber stopper and crimp-secured with an aluminum seal. Vials were wrapped with aluminum foil, and then stored at 50 °C for 24 hours. A 1.0 mL aliquot of mobile phase was injected to the vial by using 25 G needle and 1.0mL syringe. The vials were shaken until samples dissolved. Solutions were assayed for lactam (high concentration) and diluted 5-fold with mobile phase of analysis of gabapentin.

Preparation of process stress materials:

Samples were subjected to milling using a Planetary Micro Mill ?Pulviserette 7? mill from Fritsch GMBH (Idar-Oberstein, Germany) using two forty-five mL chromium hardened steel grinding bowls with four 15 mm (diameter) chromium hardened steel balls. Milling was conducted for 15 to 60 minutes. Spray-dried materials were prepared using a 5% w/w gabapentin solution in water:methanol (1:1 or 1:2 v/v) spray dried using a Buchi spray dryer. Inlet temperature was set between 70 °C, aspirator was set at 100% of capacity, and pump was used at 5% capacity. Freeze dried material was prepared using a 10 % (w/v) solution frozen at -45 °C. Primary drying was done at -30 °C, and secondary drying was done at +20 °C. Additionally a solution of gabapentin, gabapentin:PVPK90 (weight ratio of 6:1) and a suspension of gabapentin:HPC-L (weight ratio of 6:1) were also freeze dried. Granulation samples were prepared from single drop experiments with water or aqueous solutions containing 1 to 5% (w/v) HPC as the binder. Aliquots of these materials were subjected to marumerization using a flat disc marumerizer for 1 minute at 200 rpm. Granulation samples were also generated using a Diosna granulator. Blending samples were prepared using a shear rheometer at 5 rpm,20 rpm and 80 rpm. At each shear rate, the conditions of exposure time(shear strain) were 40, 80, 160, 320, 640 and 1280 revolutions. These shear strain conditions were used for every shear rate. At each stage of shear strain condition, a sample of 1g was collected.

Results: Lactam levels increased by greater than 50-fold for samples subjected to lyophilization and milling. Elevated lactam levels (4 to 20-fold) were also observed in samples after granulation and mixing stress. Susceptibility of processed samples to post-processing instability was evaluated by thermal-stress (50 °C for 24 hours). The initial rate of lactam formation ranged from 0.000060 mole %/hour to 0.04 mole %/hour depending on processing history of the sample. In general, the initial lactam formation rate was proportional to the sample surface area. The presence of expicients (hydropropyl cellulose or polyvinylpyrolidine) increased the initial lactam formation rate. Surprisingly, exposure of samples to high humidity in combination with thermal stress decreased the initial rate of lactam formation. Some processed samples (lyophilized and spray-dried) contained detectable levels of gabapentin hydrate (Form I) although no correlation between in-process lactam formation or post-processing instability could be ascertained. An increase in T1 relaxation time by SS-NMR was observed when samples were processed under rigorous conditions such as milling.

Conclusion: Correlations between in-process lactam formation or T1 relaxation time and post-processing instability were observed. Thus the detection of trace levels of lactam formation during processing (using chromatographic or spectroscopy methods) appears to be a predictive indicator of the susceptibility of gabapentin to post-processing instability.