A QbD Method Development Approach for a Generic pMDI

csheezio/getty images

The uniformity of delivered dose, or dose content uniformity, was assessed to understand the performance of the Sirdupla generic product versus the Seretide Evohaler innovator product.

Analytical method development and validation are performed using a generic quality-by-design (QbD) framework for lifecycle management (1-3). This approach is analogous to the framework recommended for product development in the International Council on Harmonization (ICH) guidelines (4) and is a regular practice within the pharmaceutical industry (5-11). All methodology has a defined analytical target profile (ATP), which is a set of predefined objectives that defines performance requirements and is subject to risk management and continuous improvement processes.

The methodology for uniformity of delivered dose (UoDD)/dose content uniformity (DCU) is a critical quality attribute (CQA) of orally inhaled and nasal drug products (OINDP). In line with the QbD paradigm, this critical variable was assessed by designed experiments to understand the performance of the Sirdupla (Mylan) generic product versus the Seretide Evohaler (GlaxoSmithKline) innovator product (salmeterol/fluticasone propionate, 120 actuations).

Methodology
Two Sirdupla product strengths were evaluated: a high-strength (HS) product in respect to Seretide Evohaler 25/250 mcg/actuation salmeterol/fluticasone propionate, 120 actuations; and a medium-strength (MS) product in respect to Seretide Evohaler 25/125 mcg/actuation salmeterol/fluticasone propionate, 120 actuations. Samples were prepared at the start, middle, and end of the pressurized metered dose inhaler (pMDI) container life (SOL, MOL, EOL) using unit spray collection apparatus (USCA), in accordance with the European Pharmacopoeia (Ph.Eur)

and United States Pharmacopeia (USP) guidelines (12,13). Two actuations were collected per test sample. While priming and collection actuations were performed manually, an MDI FD-10 instrument (InnovaSystems, New Jersey, US) was used to automate the waste actuations between the stages of container life. Samples were then analyzed by high-performance liquid chromatography-ultraviolet detection (HPLC-UV) methodology validated in accordance with ICH guidelines (14), quantifying amounts of salmeterol xinafoate (SX) and fluticasone propionate (FP). Dose proportionality was demonstrated for these actuations for all studies. In this article, only FP data are shown. Critical method parameters are listed in Table I.

The lifecycle management tools supplement the traditional approach to method development/validation. Initially, the pre-prepared generic versions of the ATP and risk assessment (RA) for the methodology are reviewed, with any specific product considerations factored in. The generic assay ATP can be broadly described as accurate, precise, and robust quantification of the active substances over their specification range. More specifically for this Sirdupla UoDD method, the ATP requirements were that the average dose was within 85-115% of Seretide Evohaler label claim (15), with acceptable test variability (e.g., relative standard deviation not more than 10%) for both APIs. It is imperative for generic products that there is this ATP prerequisite for accuracy. Such pre-defined objectives are peer-reviewed and define the method development (MD) end point. The RA uses suitable risk management tools such as fishbone diagrams, cause-and-effect matrices, and failure mode effect analyses. These tools define appropriate controls and identify required MD and method validation (MV) experimentation for each method variable, and are again peer-reviewed. MD/MV experimentation includes appropriate robustness/ruggedness studies to understand and control critical variables and define design space, as outlined in detail in this publication.

Results and discussion

CLICK FIGURE TO ENLARGE: Figure 1: Fishbone diagram of sources of variability for uniformity of delivered dose (UoDD) data. CDS is clinical decision support.

There are multiple variables that can influence UoDD data as shown in Figure 1. Aside from the product-related factors, risk assessments of the other variables are required. While some parameters have no impact if they are suitably controlled (e.g., cleaning), other more critical parameters require experimentation to determine their effects (e.g., shake/fire parameters for suspension pMDIs). Further risk assessments were required to determine the critical shake/fire parameters the experimentation would focus on.

Firstly, screening experiments were performed for the automated waste actuations to define appropriate settings for shake speed and depression hold time. Other high-risk parameters were then assessed, particularly for their effects on container EOL data (see Figure 2). Actuation force and post actuation delay were not found to be statistically significant in their effect on UoDD at EOL. A number of main effects were statistically significant: temperature, number of shake cycles, and post-shake delay (data in red). The settings for these parameters were optimized to minimize potential for through container life trends (e.g., sedimentation will be slower at a lower temperature with a reduced post-shake delay). As well as these statistically significant main effects, there are a number of statistically significant interactions present involving all five parameters, highlighting the importance of conducting experiments in a balanced design. As there are significant interactions involving actuation force and post-actuation delay, which were not highlighted as significant main effects, these parameters are also appropriately controlled. It should be noted that no main effect or interaction was of large practical significance (all affect dose by less than 5%), demonstrating reasonable robustness over the entire experimental design space. In practice, a more constricted control space is applied for these parameters, which will therefore contribute significantly less than 5% to the overall test variability. All experiments on the waste actuations were performed with the MS Sirdupla product.  

Figure 2: Pareto chart indicating statistical significance for medium-strength Sirdupla fluticasone propionate uniformity of delivered dose data at end of life (EOL) following waste actuations using the MDI FD-10 instrument (any standardized effect >2.365 indicates a p value <0.05).

Next, a risk assessment was conducted to determine the shake/fire parameters to be assessed for the priming/collection actuations, performed manually by the analysts. Four parameters in total were chosen for experimentation. Screening experiments were performed using the HS product to define boundaries for optimization studies with both product strengths. The screening experiments, and the subsequent optimization tests, showed that shake speed and post actuation delay were not practically significant. Only depression hold time and post-shake delay were shown to be practically significant parameters, and are minimized to reduce sedimentation of the API and potential enrichment of the dose. As shown for the waste actuation studies, and for other products (16), post-shake delay is shown to be significant and requires tight control (see Figure 3). In practice, a zero second post-shake delay cannot be achieved. It is, therefore, important to consider the control that can be accomplished in a robust and rugged manner day after day by different analysts.  

Figure 3: Plot of individual and mean medium-strength and high-strength Sirdupla fluticasone propionate start of life (SOL) data (% label claim) versus post-shake delay.

Lastly, control space settings were chosen from the experimental design space evaluated for both the waste actuations and the priming /collection actuations. These optimized settings were verified/validated as shown in Table II.

Table II: Method verification of fluticasone propionate data for Sirdupla versus Seretide Evohaler at the start, middle, and end of container life. MS is medium strength. HS is high strength.

The data generated were shown to be comparable with data generated for the Seretide Evohaler product. Data will continue to be assessed as part of the method review process, completing the lifecycle loop.

Conclusion
Analytical method development and validation was performed using a generic QbD framework for lifecycle management. UoDD methodology was assessed using QbD tools such as risk assessments, ATP, and designed experiments to understand the performance of the Sirdupla generic product versus the Seretide Evohaler innovator product. Suitable control of the shake/fire parameters for priming, collection, and waste actuations is critical for the analytical method.

This control is delivered via use of automation for the waste actuations onlyautomation could be extended to the priming/collection actuations, and even the sample recovery process, to exercise greater control. The optimized methodology demonstrates that the Sirdupla generic product generates comparable UoDD data to the Seretide Evohaler innovator product. This methodology is considered appropriate for the product lifespan, although it will be subject to continuous improvement processes in line with the QbD paradigm.

Sirdupla is a trademark of Mylan Inc. Seretide and Evohaler are registered trademarks of GlaxoSmithKline Group.

References
1. C. Blatchford, “From Powder to Patientoptimization of particle sizing techniques,” presentation at the Drug Delivery to the Lungs 24 Conference (Edinburgh, Scotland, UK, 2013).
2. A. Cooper, C. Blatchford, M. Kelly, “Laser Diffraction Methodology for Particle Size Distribution Determination during pMDI Product DevelopmentA QbD approach,” presentation at the Respiratory Drug Delivery Europe Conference (Berlin, Germany, 2013).
3. A. Cooper, C. Blatchford, S. Stein, “A QbD Method Development Approach for the Ex-actuator Particle Size Distribution (PSD) determination of pMDIs by Laser Diffraction,” presentation at the Respiratory Drug Delivery Europe Conference (Fajardo, Puerto Rico, 2014).
4. ICH Harmonized Tripartite Guideline, Pharmaceutical Development, Q8(R2) (August 2009).
5. G. Martin et al., “Lifecycle Management of Analytical Procedures: Method Development, Procedure Performance Qualification, and Procedure Performance Verification,” www.usp.org/sites/default/files/usp_pdf/EN/USPNF/revisions/lifecycle_pdf.pdf.
6. M. Schweitzer et al., supplement to Pharm. Technol. 34, s52-59 (2010).
7. P. Borman et al., supplement to Pharm. Technol. 34, s72-86 (2010).
8. P. Borman et al., supplement to Pharm. Technol. 21, s142-152 (2007).
9. P. Nethercote et al., “QbD for better method validation and transfer,” Pharmaceutical Manufacturing, April 13, 2010.
10. E. Rozet et al., Anal. Chem., 84, s106-112 (2012).
11. A. Rignall et al., “QbD for analytical methods for use with OINDPs,” Pharm. Technol. Europe, October 1, 2008.
12. EDQM, Eur.Ph., General Text 2.9.1, 2.9.18 (EDQM, Strasbourg, France, 2013) 228-295, 2799-2811.
13. USP, <USP> General Chapter (601), “Aerosols, Nasal Sprays, Metered-Dose Inhalers, and Dry Podwer Inhalers.”
14. ICH Harmonized Tripartite Guideline, Pharmaceutical Development, Q2(R1), (November 2008).
15. FDA, Guidance for Industry-Metered Dose Inhaler (MDI) and Dry Powder Inhaler (DPI) Drug Products, October 1998.
16. J. Parker, L Hardaker, R Hatley, “In-Vitro Investigation into the Effect of a Delay between Shake and Fire on the Delivered Dose from Four HFA formulation pMDI’s,” Respiratory Drug Delivery Europe (2015).

Article Details
Pharmaceutical Technology
Vol. 40, No. 5
Pages: 30–36, 63

Citation:
When referring to this article, please cite it as A. Cooper, "A QbD Method Development Approach for a Generic pMDI," Pharmaceutical Technology 40 (5) 2016.