Hydroxypropyl Methyl Cellulose
A case study is given in this section as an example of the commercial manufacturing monitoring and continual improvement.53 Niacin Extended Release (ER) Tablets is a hydrophilic matrix polymer based ER tablet which contains about 16% Hypromellose (HPMC) and 81% of API. A control chart (Fig. 36.5)53 of niacin release at 20 hours (D20h) from more than 800 commercial lots indicated that the drug release was impacted by the HPMC lots, even though some variability was also apparent within the same HPMC lot, possibly due to additional variability in the manufacturing process and analytical method. Some of the tablet lots barely passed the lower specification limit of not less than 75% and a few lots failed. The apparent variability of the drug release at 20 hours of this product presented a significant challenge to maintain uninterrupted supply to patients of desired quality product.
Figure 36.5. Run chart of niacin release at 20 hours segmented by HPMC lots.
HPMC is a cellulose derivative by introducing methoxy (MeO) and hydroxypropyl (HP) substitutions simultaneously under alkaline conditions with inherent chemical heterogeneity due to the potential chemical variability of cellulose derivatives. Initial investigation performed on HPMC lots used in the commercial manufacturing indicated that the HP content is a primary impacting factor (p<0.0001) while the viscosity is of borderline significance (p=0.07). Neither MeO nor particle size was found significant. A regression model based on the HP content and viscosity (R2=0.79) predicts the D20h well in general. However, some HPMC lots deviated significantly from this model. As a part of the investigation, HPMC lots from two additional vendors (vendor B and vendor C) were also evaluated experimentally. The regression model did not predict dissolution for HPMC lots from vendor C (HP ~8.3–8.6%), although the predicted and tested values were consistent (~90% released at 20 hours) for HPMC from vendor B (HP ~9.5–9.7%).
An investigation using a two-dimensional NMR clearly differentiated two types of hydroxypropyl substituents: one as native where the hydroxyl is free (designated as HP1) and the other is inner substitution where the hydroxyl group is further substituted by either a MeO group or another HP group (designated as HP2). The HP1/HP2 ratios were found to vary significantly among the vendors: HP1:HP2 ~ 3.3:1 for the commercial HPMC vendor, while HP1:HP2 ~ 2:1 for both vendor B and vendor C. The HP1:HP2 ratio is perceived to be important because HP2 represents a longer and larger substitution and greatly weakens the polymer/polymer interaction, thus, favoring faster drug release. Indeed, even though the overall HP contents for vendor C HPMC lots were much lower (HP ~8.3–8.6%), they produced fastest release rates (>90%) while the regression model predictions were low (<80%). It was postulated that the apparent conformance of vendor B materials to the regression model was purely a coincidence because with this highly substituted HP make up (HP1:HP2 ~ 2:1), the HP content does not need to be as high as in HPMC from the commercial vendor to release fast.
Even though the vendor-to-vendor differences were explained by the difference in the HP makeups, the fact that not all HPMC lots from the commercial vendor conformed to the regression model still warrants further investigation. It was likely that additional chemical heterogeneity such as the nonuniform distribution of the substitution along the cellulose backbone may play a role. It has been found that more heterogeneously substituted HPMC have impacted its solution properties such as cloud point54 thus, may impact drug release from hydrophilic matrices.
Based on reviewing drug release mechanisms from hydrophilic matrices, the erosion rate of polymer was concluded to play a significant role in the drug release process. Therefore, a method was developed to measure the erosion rate of pure HPMC samples. The results confirmed that niacin release from the ER tablets correlated highly with the erosion rate of HPMC, regardless of the vendor (Fig. 36.6).
Figure 36.6. Correlation between niacin release at 20 hours and HPMC erosion rate.
Based on the investigation, a specification was then set on this property of HPMC. With the improved control on HPMC, no further dissolution issue has surfaced. In addition to that, predicted and observed drug release has been consistent. The Niacin ER tablets drug release at 20 hours data from 311 commercial batches manufactured post the implementation of the change are presented in Fig. 36.7, and a process capability analysis plot is given in Fig. 36.8. The 95% lower confidence bound of the performance index (LB for Ppk=1.43) indicates the additional control led to great improvement in meeting the quality requirement of drug release at 20 hours.
Figure 36.7. Run chart of niacin release at 20 hours, after the additional HPMC specification is implemented.
Figure 36.8. Process capability analysis of niacin release at 20 hours, after the additional HPMC specification was implemented.
In the above example, the lot-to-lot and vendor-to-vendor variability of HPMC was not apparent from the typical compendia type of properties conventionally measured. In the end, a specification was developed based on measurement of an unconventional property of HPMC to control its variability. Searching beyond the compendia properties and understanding the dosage form design are necessary to understand better and control excipient variability. The case study also demonstrated that control chart can be an effective tool to detect the presence of special cause variation in the manufacturing process and to ascertain if the process has reached a state of statistical control. The control charts also serve as indicators of impending problems and allow operating personnel or process engineers to take proactive actions to prevent product quality failure from happening.