Poster Abstracts from Annual Meetings of the
American Association of Pharmaceutical Scientists (AAPS)
Process Analytical Technology Model-Application of NIR in the Analysis of Blend Homogeneity
Purpose: To develop a process analytical technology model for the analysis of pharmaceutical blend homogeneity using Near Infra Red Spectroscopy (NIR), which will be used to ultimately develop an online process analytical procedure to test blend homogeneity samples using NIR system fitted with a fiber optic sensor.
Method: A Foss NIR system Model 500 equipped with a reflectance detector and a software package of VISION 2.50® was used for this study. NIR spectra were collected, with 32 scans and the total acquisition time of approximately one minute. The blend samples consisted of EDTA as a model compound (in the range of 0 – 50 w/w %), and typical excipients; Lactose, Microcrystalline Cellulose, Sodium Starch Glycolate, and Magnesium Stearate. Each sample was shaken for 2 hours prior to analysis.
Results: NIR spectra were collected for EDTA and the individual excipients in the range from 1100 nm to 2500 nm. The analysis of those spectra showed the ideal range (1600 nm to 2100 nm) for quantitation of EDTA within blend samples. The calibration set consisted of blend samples with EDTA concentration from 0 % to 50%w/w. Partial Least Square (PLS) 2nd derivative calibration models for EDTA were constructed. The graph of computed concentration of EDTA vs. nominal concentration showed a linear calibration curve with R2 = 0.9910. Multiple Linear Regression (MLR) models were also tested, but PLS models provided the best fit. Accuracy and repeatability were tested using multiple blend samples. The results agreed with the predicted concentrations within 91.3% – 99.2% and %RSD values were between 3.3 % & 5.8 %.
Conclusion: A process analytical model was successfully developed using NIR spectroscopy for analysis of blend
homogeneity. The test method used was accurate and repeatable. The model will be extended to develop an on-line process analytical testing procedure using a NIR system equipped with a fiber optical sensor.
Authors: R. Bhatnagar 1, M. Hu 1, F. Fang 1, Y. Chen 2, A. Chan 2, B. Wu 2, N. Shah 1, G. Carr 1, A. Freedman 1
1 PDS, Toronto Region Operations, 2100 Syntex Court, Mississauga, Ontario L5N 7N9 Canada, 2 PDS,
York Mills Operations, 865 York Mills Road, Toronto, Ontario M3B 1Y5 Canada, Patheon Inc.
Blend Residence Time in a Tablet Press Feeder and its Effect on the Physical Characteristics and Dissolution of Folic Acid Tablets
Purpose: To investigate the over lubrication susceptibility of model folic acid formulations inside a tablet press force feeder and the effects on the physical characteristics and dissolution performance of the tablets
Methods: An instrumented Manesty Betapress equipped with a variable speed paddle force feeder was used to investigate the effects of the feeder residence time and paddle speed on the compression profile and dissolution of folic acid tablets. Two direct compression model formulas based predominantly on either microcrystalline cellulose (MCC) or dibasic calcium phosphate (DCP) were used to explore two types of excipients that have different mechanism of consolidation during compression (plastically deforming and fragmentation respectively). The blends consisted of 0.4 % folic acid, 2.0% croscarmellose sodium, 1.0% magnesium stearate and 96.6% of filler/carrier (MCC or DCP respectively). Folic acid was geometrically blended with either MCC or DCP and blended along with croscarmellose for 20 minutes using a V-blender; afterwards, the blend was lubricated with magnesium stearate for 3 minutes. Finally the blends were compressed into 100mg tablets.
Results: A pattern in the dissolution profiles of tablets from the folic acid formulation containing MCC suggested that long residence times in the tablet press feeder negatively affected the dissolution of the active, especially at the early dissolution time points (the effect of residence time was statistically significant at a probability less than 0.05); albeit the negative effect, all passed the USP dissolution standard. The least squares multiple regression analysis for the dissolution profiles of tablets from the folic acid DCP formulation revealed a statistically significant interaction between residence time and compression force at each dissolution time point. In addition, the DCP formulation tablets from most of the experimental runs failed the USP dissolution standard.
Conclusion: The residence time of a formulation inside a paddle force feeder of a rotary tablet press could be a contributing factor to over-lubrication issues that could affect the dissolution and physical characteristics of tablets when the formulation is susceptible to magnesium stearate. Dibasic calcium phosphate has a detrimental effect on the dissolution of folic acid tablets.
Authors: M. A. Hervas 1, A. Mehta 1, J. Mir 1, A. Singh 1, R. Balint 1, A. Freedman 1
1 Patheon Inc.
LIBS (Laser Induced Breakdown Spectroscopy) as an At-line Monitor of API
Coating Uniformity in a Film Coating Operation
Purpose: LIBS (Laser Induced Breakdown Spectroscopy) as an in-process monitor for uniformity during an API tablet coating process.
Method: PharmaLIBS™250 has shown to be an effective tool to rapidly probe the surface and the core of the tablet for various compounds and determine their distributions throughout the sample. The lack of uniformity in distribution can be clearly demonstrated in the surface plots obtained from a single tablet; both means of surface distribution and coating uniformity also can be plotted for single tablets. In this exercise LIBS was used at-line to monitor the in-process coating performance using a phenylephrine HCL active coating solution sprayed onto 500mg tablets using an AccelaCota 60 inch perforated pan (Thomas Engineering). In process measurements were taken on ten different coated tablets at ten different weight gain percentages up to coating endpoint. Measurements were evaluated to obtain information tablet coating thickness and uniformity.
Results were analyzed and plotted to examine inter and intra tablet thickness and uniformity as a function of coating time and weight gain %. Coating thickness and variation in coating uniformity within a tablet were able to be analyzed by monitoring chlorine and titanium elemental band intensity. A good correlation was obtained between elemental intensity and coating weight gain% (r2 = 0.99) Figure 1.
Conclusion: The use of PharmaLIBS™250 as an at line monitor of active coating processes can provide rapid and valuable information that can be used to optimize the process real time in order to minimize coating variation with tablets and across the tablet bed.
Authors: S. Closs 1, I. Akam 1, A. Freedman 1, M. Tourigny 2
1 Patheon Inc., 2 PharmaLaser
An Automated Dryer Monitoring Mechanism that Facilitates On-Line End Point Detection in a Fluidized Bed Process
Purpose: To evaluate the use of an enhanced automatic sampling device to facilitate the in-line extraction and measurement of thermal effusivity samples in order to better monitor and define fluid bed drying characteristics in a closed contained system.
Methods: Multiple ternary placebo granulations were dried in a Niro Aeromatic S2 Fluid Bed that had been equipped with an ISOLOK® sampler fitted with an integrated Mathis ESP effusivity sensor. The drying process was monitored at predetermined intervals by taking parallel measurements both on-line (effusivity) and near-line (LOD).
Results: The drying profiles confirmed that this automated and “closed” configuration repeatedly captured representative samples and, as a result, confirmed the linear relationship between effusivity and LOD that has been previously reported (e.nieves, f. estremera, AAPS, 2003). Further, the results indicate that the profile can be used to adapt the processing conditions real time to avoid particle attrition, over-drying, and the generation of fines at the final stage of drying.
Conclusion: In a development environment, extensive process knowledge generated allows a robust production protocol to be developed. This resulting process signature is completely transferable and responds to variations in incoming granule moisture and environmental conditions. The “J-Curve” that occurs due to creation of fines in some drying operations is avoidable by adjusting the dryer CFM and temperature. The on-line moisture monitoring significantly reduces test time, improves accuracy, and creates a safe, contained, real time in-process method for achieving drying endpoints.
Authors: J. Boodram 1, S. Closs 1, A. Freedman 1, C. Minchom 1, G. Chaplin 2, Y. Roy 2, N. Mathis 2, B. Werra 3
1 PDS, Patheon, 2 Applications Development, Mathis Instruments, 3 Process Sampling, Sentry Equipment Corporation
A Design Space Approach to Process Troubleshooting and Optimization for a Fluid Bed Granulation Process
Purpose: A process design space approach was used as a tool to guide process troubleshooting and optimization at production scale in a fluid bed granulation process.
Method: A fluid bed granulation was transferred from pilot scale to production scale using standard scale up methods, but with limited understanding regarding the effect of process variables on granule properties at either scale. In this exercise, the design space was evaluated within the filed parameter ranges at commercial scale in order to understand the effect of process variables on the granule properties. Since experimental work at this scale is very expensive, it was desirable to limit the number of experiments. Based on historical process evaluation, the following variables were fixed: air flow rate, inlet air dew point, atomization pressure, atomizer wand height, and batch size. The variables evaluated in the experimental work were inlet temperature and spray rate. The process design space was evaluated using a 22 full factorial design with one centre point, but in order to evaluate a physically meaningful range, the factors in the design were inlet temperature and exhaust humidity (actually a dependent variable).
Results: The results showed that the drier fluid bed environmental conditions (40 – 45% outlet RH and 65- 70ºC inlet temperatures) produced granules that had superior properties (bulk density and compressibility) as compared to the original condition. Furthermore, experiments in this area of the process space were more robust and led to granulation process improvements such as sustainable product fluidization and controlled granule growth.
Conclusion: Using this design space approach was useful to efficiently determine the optimum operating conditions within the existing filed process parameter ranges. It directed the scientists to an area of the process space that both produced granules with more desirable properties than the original condition and created more durable tablets for downstream coating operations.
Authors: S. Closs 1, J. Shockey 2, A. Freedman 1
1 Patheon Inc., 2 GreenRidge Consulting
Optimization of fast disintegrating direct compression tablets by statistical design
Purpose: To further optimize and assess the effect of refining magnesium stearate lubricant, sweetener, hardness and thickness (weight) levels on the design of a fast disintegrating formulation by direct compression.
Methods: A d-optimal statistical design was employed to study the effect of lubricant, and sweetener on disintegration time (DT), hardness(H) and friability (F) to further refine a fast disintegrating base. Formulation blends containing combinations of dicalcium phosphate, microcrystalline cellulose PH102, croscarmellose sodium, silicon dioxide, magnesium stearate, sucrose, mannitol and aspartame were prepared by screening, blending, and compressed on an instrumented rotary press (Bilayer Piccola) at 20 rpm with compression forces up to 9kN. Physical tests and disintegration by both USP and a modified disintegration test method. The modified method involved pipetting 1mL of deionized water in the well of a 1cm glass tube, placing a tablet within this well and the time for the tablet to completely disintegrate into fine particles was noted as the disintegration time.
Results: Hardness of the fast disintegrating tablets was optimized in the range from 10.3 (± 2.2) kp to 15.0 (± 1.0) kp. Friability ranged from 0.04 to 0.14%. Per USP<701>, disintegration averaged 10s for 100mg tablets, 12 s for 290mg tablets and 16s for 435mg tablets. Modified disintegration testing with limited media to simulate in vivo conditions gave disintegration times between 16.5 (± 0.7) s to 22.5 (±0.7) s.
Conclusion: Hardness of fast disintegrating tablets was improved (10 to 15 kp) for handling without compromising fast disintegration (both DT < 20s by USP and DT< 25s in fluid volume of an oral cavity). A statistical design of experiments was helpful in refining the design for a fast disintegrating tablet made on conventional equipment.
Authors: S. K. Lum 1, A. Heckman 2, L. Liu 1, J. Huang 1, M. McLaughlin 1, A. Freedman 1
1 Patheon Inc., 2 Consultant
Thermal effusivity blend lubrication data as a PAT platform to optimize tablet compression performance
Purpose: To investigate the use of thermal effusivity blend lubrication data as a PAT platform to optimize tablet compression performance.
Methods: Trial batches of a three component placebo (Lactose Monohydrate, Avicel PH102 and Magnesium Stearate) were manufactured at the 5 and 25kg scale in tote blenders. The trials focused on the measurement of the Δchange in thermal effusivity readings during lubrication blending using magnesium stearate and its resulting effect on tablet compression and in-process measurements. Blend lubrication trials were manufactured using different w/w percentages of magnesium stearate (0.20 – 2.0 %) at blend times ranging between 2 – 20 minutes. During blending, in-line thermal conductivity measurements were taken at regular in-process intervals using four Mathis ESP wireless sensors placed at different locations in the tumbling blender. The Δchange in average effusivity values were used to identify the optimum blend lubrication parameters required for tablet compression based on in-process test measurements.
Results: Effusivity measurements taken during the lubrication trials indicated a distinctive difference in the increase in average Δ effusivity at pre-determined blend measurement intervals. Blends containing high percentages of magnesium stearate (2 %) produced effusivity profiles with steeper slopes than blends containing low percentages (0.20 %) and took less time to reach “optimal” lubrication conditions. The resulting effusivity profiles could be used to identify optimal lubrication conditions that enhance tablet compression performance as measured by blend compressibility (compression force), tablet hardness, and tablet disintegration.
Conclusion: On the basis of the data collected from the trial batches, thermal effusivity could be an effective PAT instrument in the development of a blend lubrication process using magnesium stearate. This would allow optimization of blend lubrication and alleviate potential downstream tableting issues.
Authors: S. Closs 1, J. Boodram 1, A. Freedman 1, C. Minchom 1, D. Kirsch 2, D. Natoli 3
1 PDS, Patheon, 2 Technical Services, 3 Process Development, Natoli Engineering
Enhancement of bioavailability and stability of a low solubility drug using hot melting (lipid based) and wet granulation technologies
Purpose: Development of a solid dosage form for an unstable, low solubility drug (DRUG-PS) with poor wettability.
Methods: Selection of excipients for the DRUG-PS formulations were based on drug solubility in solubilizers and dissolution of drug excipient mixtures. Prototypes by hot melt granulation (HM -containing Gelucire 44/14 and Capryol 90) and by wet granulation (WG- containing micronized drug) were prepared and evaluated for dissolution, stability and bioavailability in rats. A lead process/formulation was selected and improved using a 2×2 factorial design of experiment (effect of binder and surfactant on dissolution).
Results: Both HM and WG formulations provided satisfactory dissolution profiles of 90 and 91% dissolved in 60 min, respectively. Approximately 25% drop in dissolution was observed in HM when stored @ 40°C/75RH for 1month and no major change observed in WG when stored @ 40°C/75RH for 6 month. The impurity content for HM was higher than WG after storage for 1 month. The AUC for WG was 55311 ng/ml/hr and better than with HM (30826 ng/ml/hr). Therefore WG was chosen for further development. The dissolution rates for the experimental design prototypes were similar (89 to 95% dissolved in 60min). However, a surfactant binder interaction was observed. After correcting for drug content in the formulations, the data showed that a low surfactant and low binder content could slightly improve the dissolution rate. Stability study on the confirmation batch indicated that the product is stable at 40°C/75%RH and 25°C/60%RH after storage for 6 and 12 months respectively. No significant changes of dissolution rate of were observed.
Conclusion: A wet granulated formulation was selected based on the bioavailability and physical and chemical stability for DRUG-PS. Hot melt prototypes with satisfactory processing characteristic were produced as template formulations for low solubility drugs.
Authors: S. Chowdari, S. Chowdari, S. Nimesh, A. Legore, A. Freedman, K. Chow
Small Scale Optimization of an Active Formulation Roller Compaction Process using a True Placebo DoE
Purpose: To evaluate the use of a true placebo DoE to create a platform process in which to define a viable low dose active IR tablet formulation process using roller compaction.
Methods: A true placebo formula was manufactured using a 2 (5-1) fractional factorial design (16 runs with 3 centre points) on a pilot scale roller compactor (Fitzpatrick L89). The input variables studied in the design included roll speed, roll pressure, vertical rpm, horizontal rpm and final mill screen size. Each 2 kg sublot run was tested for measurement responses including mean particle size, % material retained >500 micron, % material retained < 50 micron, Hausner Ratio, and Flow Orifice Index. Results of the experimental design were analyzed and optimal studied parameters were chosen for execution of a 5kg verification roller compaction run utilizing 6.25% w/w of API
(active pharmaceutical ingredient). Physical testing performed on the active granulation and on 400mg tablets compressed from the active granulation, as well as analytical testing of the tablets (weight averaged content uniformity, and dissolution) were used to evaluate the suitability of the DoE optimized process for the active formulation.
Results: The results showed that true placebo experimental design could be used to optimize critical roller compaction parameters to be utilized in a low dose active formulation. Physical data from the optimized process showed that API granulation incurred a slight shift in particle size from predicted (due to binding properties of the active) but still retained desired particle size and flow characteristics. Compression of these granules resulted in tablets with good content uniformity (94.8 – 99.4%, RSD 1.6%), weight variation (RSD 1.9%), and dissolution (> 85% in 12 minutes).
Conclusion: The use of placebo design of experiments (DOE) was effective in creating a platform process for active feasibility trials and scale up using minimal API. Although a slight shift in particle size and granulation density was experienced as a result of addition of API to the placebo platform process, the physical and chemical results of the active confirmation run met desired specifications.
Authors: S. Closs, S. McConnell, A. Freedman
Feed Frame Residence Time Effect on Tablet Physical Characteristics and Dissolution of Lubricant Sensitive Actives
Purpose: To investigate the effect of the residence time of a lubricant sensitive model drug in the paddle feed frame of a rotary tablet press on the tablet physical characteristics and dissolution performance.
Methods: An instrumented Piccola B-10 rotary tablet press equipped with a variable speed paddle feed frame was used to investigate the effects of the feed frame residence time and paddle speed on the compaction profile and dissolution of tablets manufactured from two model formulas: a formula based on microcrystalline cellulose (known to predominantly undergo plastic deformation during compaction); and a second formula based on dibasic calcium phosphate, the consolidation mechanism for which is primarily fragmentation. Both formulas contained magnesium stearate (1%), croscarmellose sodium (2%) and a model drug that has presented multiple dissolution problems. Over lubrication is also believed to play a role in the dissolution behaviour of the model drug.
Results: Preliminary results show that the hardness of tablets compressed at forces from 9 to 24 kN, using the microcrystalline cellulose based formulation was affected by longer feed frame residence times. Hardness decreased up to 35% compared to tablets subjected to short residence times. Some tablets presented capping at the long residence time, which is clear evidence of the deleterious effect of over-lubrication. In contrast, tablets manufactured with dibasic calcium phosphate based blends were not affected by feed frame settings. The dissolution profiles of tablets produced at long and short residence times will be presented. Granulate feed frame residence time is expected to have a significant effect on the dissolution of the model drug used in this investigation.
Conclusion: The residence time of a formulation in a variable speed paddle feed frame significantly affects the compression characteristics when the formulation is manufactured from an ingredient (eg. microcrystalline cellulose) known to be lubricant sensitive.
Authors: M. Hervas 1, A. Mehta 2, J. Ramsuran 2, A. Singh 2, S. Zhang 3, S. Yang 3, A. Freedman 1
1 Process Development, 2 Formulation Development, 3 Analytical Development, Patheon Inc.,
Pharmaceutical Development Services, Toronto Region Operations, Canada
For more information please contact Austin Freedman at email@example.com