(M1430-12-83) Common Deficiencies Using Alternative Bioequivalence Approach for Metered Dose Inhaler in Abbreviated New Drug Applications

Purpose: Recently, there has been an increasing interest in alternative bioequivalence (BE) approach in lieu of a comparative clinical endpoint (CCEP) study to support BE determination for generic pressurized metered dose inhaler (MDI) products. While an alternative BE approach may offer advantages of cost-effectiveness, efficiency and reproducibility, this type of approach may also present challenges due to limited understandings of correlation between in vitro performance and regional lung deposition, especially for suspension-based MDI products. Several alternative in vitro tests have been developed and submitted to the FDA for generic MDI products. They include aerodynamic particle size distribution (APSD) with idealized mouth-throat (MT) models, characterization of particle size distribution of active pharmaceutical ingredient (API) using morphologically directed Raman spectroscopy (MDRS), comparative dissolution of respirable API fractions, velocity profiles of the aerosol plume and evaporation rate, quantitative methods and modeling, and alternative pharmacokinetic BE study with charcoal block. These in vitro tests aim to assess the lung delivery equivalence of the test and reference MDI products at various stages: aerosolization, transit of API through airways, and deposition/dissolution/absorption at the local site of action. The objective for this project is to outline the common issues identified for these in vitro tests and to elaborate on the FDA’s expectations for each proposed test, as part of the weight-of-evidence approach to support BE demonstration of generic MDI products.

Methods: ANDAs submitted between 2020 to 2023, which contain an alternative BE approach (as mentioned above) in lieu of CCEP BE study were selected. The common deficiencies were categorized and analyzed according to the recommendations specified in the product specific guidance (PSG) on Beclomethasone Dipropionate MDI, which is the first PSG to contain alternative approach mentioned above.

Results: Deficiencies were identified across the tests of alternative BE approach submitted, including incomplete method development, inadequate method validation and insufficient pivotal study data. For instance, data supporting the selection of number of actuations, deposition fraction, and particle characteristic threshold were lacking. Justifications were insufficient for how the studies address the relevance of PK properties to equivalence in local delivery and the relationship of in vitro performance to regional lung deposition. Deficiencies were identified for APSD with MT study related to selection of MT models and flow profiles that represent the target patient population. The methods used to account for inter-subject variability were deemed insufficient in the chosen MT models. Moreover, the justification for the use of proposed statistical method (e.g., studentized t-test), which deviates from the traditional population BE for impactor-sized mass, was not adequately explained for BE determination. The MDRS studies lacked details such as sample preparation, selection of morphology filters, optic magnification, threshold setting, and particle counting method using Raman correlation technique. All reviewed ANDAs contain deficiencies in the method validation data for reproducibility, robustness, and sample size justification. Other deficiencies identified in the MDRS studies included improper selection of a single impactor stage and failure to provide data for drug particles < 1 μm using an orthogonal method. Additionally, the submission lacked comparative data demonstrating particles excluded by MDRS filters for the non-volatile excipients, such as Oleic Acid, that are used in the suspension-based MDIs. For dissolution studies, the validation and optimization of the sample fraction collection procedure using insert filters in cascade impactor were not adequately demonstrated to represent the respirable fraction entered into the lungs. The stability of the APIs under the dissolution media and the discriminatory ability of the dissolution method were also not proven, which are prerequisites for evaluating comparative dissolution data for profile comparison. For quantitative methods and modeling including PopPK, key deficiencies included inadequate model validation, due to the missing of established correlation between population pharmacokinetic lung compartments and regional lung absorption and/or the non-submission of compelling rationale for using partial AUC to establish BE.

Conclusion: The alternative BE approach may provide additional information to ensure equivalence at the local site of action (lung) between the generic and reference listed drug products. Additional supporting data, such as improved method validation, in vitro-in vivo correlations and comparative study addressing aerosol performance may help addressing the deficiencies identified for these tests, reducing the review cycle and increasing the likelihood for regulatory acceptance and ultimate approval.