Performance of Multiple-Batch Approaches to Pharmacokinetic Bioequivalence Testing for Orally Inhaled Drug Products with Batch-to-Batch Variability.

Batch-to-batch pharmacokinetic (PK) variability of orally inhaled drug products has been documented and can render single-batch PK bioequivalence (BE) studies unreliable; results from one batch may not be consistent with a repeated study using a different batch, yet the goal of PK BE is to deliver a product comparison that is interpretable beyond the specific batches used in the study. We characterized four multiple-batch PK BE approaches to improve outcome reliability without increasing the number of clinical study participants. Three approaches include multiple batches directly in the PK BE study with batch identity either excluded from the statistical model ("Superbatch") or included as a fixed or random effect ("Fixed Batch Effect," "Random Batch Effect"). A fourth approach uses a bio-predictive in vitro test to screen candidate batches, bringing the median batch of each product into the PK BE study ("Targeted Batch"). Three of these approaches (Fixed Batch Effect, Superbatch, Targeted Batch) continue the single-batch PK BE convention in which uncertainty in the Test/Reference ratio estimate due to batch sampling is omitted from the Test/Reference confidence interval. All three of these approaches provided higher power to correctly identify true bioequivalence than the standard single-batch approach with no increase in clinical burden. False equivalence (type I) error was inflated above the expected 5% level, but multiple batches controlled type I error better than a single batch. The Random Batch Effect approach restored 5% type I error, but had low power for small (e.g., <8) batch sample sizes using standard [0.8000, 1.2500] bioequivalence limits.

Evaluation of the Sensitivity and Robustness of Modified Chi-Square Ratio Statistic for Cascade Impactor Equivalence Testing Through Monte Carlo Simulations.

The objective of this work was to study the performance of the modified chi-square ratio statistic (mCSRS test) proposed for cascade impactor (CI) profile equivalence testing. The test (T) and reference (R) CI profile datasets were generated from different typical CI profile patterns either with or without inter-site correlation (ISC) through Monte Carlo simulations. The mCSRS test pass rate outcome employing previously published critical values was compared with that of critical values derived from different types of datasets. The influence of number of bootstrap iterations (B) on the consistency of the outcome was assessed within the range of 10-10,000 iterations. Power curves were constructed to study the effect of differences in T and R mean stage deposition, T/R variance ratios, differences between T and R profiles in high/low deposition sites, and sample size on the performance of the mCSRS test. The derived critical values exhibited trends based on R product variability: M1 rank-ordered without ISC (at low variability) and the previously published M8 critical values (at high variability) resulted in lowest pass rate outcomes. The precision of the outcome did not increase considerably beyond B = 2000 (default). The probability of showing equivalence between T and R CI profiles increased with (1) a decrease in mean deposition differences, (2) a decrease in T product variability, and (3) an increase in sample size. The mCSRS outcome is less sensitive to low deposition sites that are prone to analytical variability. In conclusion, the mCSRS test is a sensitive and robust method under most conditions.

Performance of the Population Bioequivalence (PBE) Statistical Test with Impactor Sized Mass Data.

This article extends previous work studying performance characteristics of the population bioequivalence (PBE) statistical test recommended by the US Food and Drug Administration (FDA) for orally inhaled and nasal drug products. Based on analysis of a metered dose inhaler database for impactor sized mass, a simulation study was designed to compare performance of the recommended PBE approach with several modified or alternative approaches. These included an extended PBE that separately modeled within-batch (can) and between-batch (batch) variability and average bioequivalence (ABE) tests that modeled with or without between-batch variability and with or without log-transformation. This work showed that separately modeling within- and between-batch variability while increasing the number of sampled batches addressed previously identified issues of the PBE approach when between-batch variability was present, namely, (a) increased risk for falsely concluding equivalence and (b) low probability of correctly concluding equivalence. The same modifications were also required of the ABE to achieve expected performance. However, these modifications did not successfully address the issue of equivalence conclusions that depended on the direction of product mean differences (asymmetric performance). This work highlights the importance of understanding decision-making error rates in developing regulatory recommendations to standardize bioequivalence outcomes across products.

Cascade Impactor Equivalence Testing: Comparison of the Performance of the Modified Chi-Square Ratio Statistic (mCSRS) with the Original CSRS and EMA's Average Bioequivalence Approach.

The performances of three statistical approaches for assessing in vitro equivalence was evaluated with a set of 55 scenarios of realistic test (T) and reference (R) cascade impactor (CI) profiles (originally employed by the Product Quality Research Institute to evaluate the chi-square ratio statistic: CSRS) by comparing the outcomes against experts' opinion (surrogate for the truth). The three methods were (A) a stepwise aerodynamic particle size distribution (APSD) equivalence test integrating population bioequivalence (PBE) testing of impactor-sized mass (ISM) with the CSRS (PBE-CSRS approach), previously suggested by the USFDA; (B) the combination of PBE testing of single actuation content and ISM with the newly suggested modified CSRS (PBE-mCSRS approach), a method employing reference variance scaling; and (C) EMA's average bioequivalence (ABE approach). Based on Monte-Carlo simulations, both PBE-CSRS and ABE approaches resulted in high misclassification rates, the former with highest false-pass rate and the latter with highest false-fail rate at both ≥ 50% and ≥ 80% classification threshold values (the % of simulations or experts necessary to judge a given scenario as equivalent). Based on DeLong's tests, the PBE-mCSRS approach showed significantly better overall agreement with experts' opinion compared to the other approaches. Comparison of CSRS with mCSRS (both without PBE) suggested that the more discriminatory characteristics of the mCSRS method is based on the integration of variance scaling into the mCSRS method. Contrary to the ABE approach, the application of PBE-mCSRS approach for assessing APSD profiles of three dry powder inhaler (DPI) formulations supported the pharmacokinetic bioequivalence assessment of these formulations.

Performance of the Population Bioequivalence (PBE) Statistical Test Using an IPAC-RS Database of Delivered Dose from Metered Dose Inhalers.

This article reports performance characteristics of the population bioequivalence (PBE) statistical test recommended by the US Food and Drug Administration (FDA) for orally inhaled products. A PBE Working Group of the International Pharmaceutical Aerosol Consortium on Regulation and Science (IPAC-RS) assembled and considered a database comprising delivered dose measurements from 856 individual batches across 20 metered dose inhaler products submitted by industry. A review of the industry dataset identified variability between batches and a systematic lifestage effect that was not included in the FDA-prescribed model for PBE. A simulation study was designed to understand PBE performance when factors identified in the industry database were present. Neglecting between-batch variability in the PBE model inflated errors in the equivalence conclusion: (i) The probability of incorrectly concluding equivalence (type I error) often exceeded 15% for non-zero between-batch variability, and (ii) the probability of incorrectly rejecting equivalence (type II error) for identical products approached 20% when product and between-batch variabilities were high. Neglecting a systematic through-life increase in the PBE model did not substantially impact PBE performance for the magnitude of lifestage effect considered. Extreme values were present in 80% of the industry products considered, with low-dose extremes having a larger impact on equivalence conclusions. The dataset did not support the need for log-transformation prior to analysis, as requested by FDA. Log-transformation resulted in equivalence conclusions that depended on the direction of product mean differences. These results highlight a need for further refinement of in vitro equivalence methodology.

Meeting report: Advancing accelerated regulatory review with Real-Time Oncology Review (RTOR), Project Orbis, and the Product Quality Assessment Aid.

The American Association of Pharmaceutical Scientists (AAPS) Chemistry, Manufacturing, and Controls (CMC) Community hosted two virtual panel discussions focusing on several novel regulatory review pathways for innovative oncology products: Real-Time Oncology Review (RTOR), Project Orbis, and the Product Quality Assessment Aid (PQAAid). The panel sessions were held on August 27, 2021, for the discussion of RTOR, and January 21, 2022, for the discussion of Project Orbis and the PQAAid. Both panel sessions included representatives from the US Food and Drug Administration (FDA) and subject matter experts from the pharmaceutical and biotechnology industries, with the aim of facilitating knowledge sharing on CMC-specific advantages, challenges, eligibility criteria for participation, and operational modifications instituted through the utilization of these acceleration initiatives. Key topics included managing cross-regional regulatory CMC requirements, adapting to expedited development timelines, coordinating interactions between health authorities and industry, and potential opportunities for future improvement and expansion of these programs. As RTOR, Project Orbis, and PQAAid are relatively new initiatives, the experiences shared by the panel experts are valuable for providing deeper insight into these new regulatory pathways and processes.

Challenges and opportunities in implementing the FDA default parametric tolerance interval two one-sided test for delivered dose uniformity of orally inhaled products.

The goal of this article is to discuss considerations regarding implementation of the parametric tolerance interval two one-sided test (PTI-TOST) for delivered dose uniformity (DDU) of orally inhaled products (OIPs). That test was proposed by FDA in 2005 as an alternative to the counting test described in the 1998 draft FDA guidance for metered dose inhalers and dry powder inhalers. The 2005 PTI-TOST, however, still has not found much use in practice despite the general desirability of parametric approaches in modern pharmaceutical quality control. A key reason for its slow uptake is that it rejects, with high probability, batches whose quality is considered acceptable by all other published regulatory and pharmacopeial standards as well as by the DDU specifications for many approved OIPs. Manufacturers therefore continue using nonparametric counting tests for control of DDU. A simulated case study presented here compares the consequences of the PTI-TOST compared to the counting test. The article discusses three possibilities that would help increase the uptake of the PTI-TOST approach, namely: product-specific quality standards, a different default standard suitable for the majority of OIPs, and integration of the PTI-TOST with a continuous verification control strategy rather than using it as an isolated-batch (transactional) end-product testing. In any of these efforts, if a parametric test is used, it is critical not to set the target quality close to, or at the boundary of the process/product capabilities, because PTI tests are designed to reject with high probability the identified target quality.

Relative precision of inhaler aerodynamic particle size distribution (APSD) metrics by full resolution and abbreviated andersen cascade impactors (ACIs): part 1.

The purpose of this study was to compare relative precision of two different abbreviated impactor measurement (AIM) systems and a traditional multi-stage cascade impactor (CI). The experimental design was chosen to provide separate estimates of variability for each impactor type. Full-resolution CIs are useful for characterizing the aerosol aerodynamic particle size distribution of orally inhaled products during development but are too cumbersome, time-consuming, and resource-intensive for other applications, such as routine quality control (QC). This article presents a proof-of-concept experiment, where two AIM systems configured to provide metrics pertinent to QC (QC-system) and human respiratory tract (HRT-system) were evaluated using a hydrofluoroalkane-albuterol pressurized metered dose inhaler. The Andersen eight-stage CI (ACI) served as the benchmark apparatus. The statistical design allowed estimation of precision with each CI configuration. Apart from one source of systematic error affecting extra-fine particle fraction from the HRT-system, no other bias was detected with either abbreviated system. The observed bias was shown to be caused by particle bounce following the displacement of surfactant by the shear force of the airflow diverging above the collection plate of the second impaction stage. A procedure was subsequently developed that eliminated this source of error, as described in the second article of this series (submitted to AAPS PharmSciTech). Measurements obtained with both abbreviated impactors were very similar in precision to the ACI for all measures of in vitro performance evaluated. Such abbreviated impactors can therefore be substituted for the ACI in certain situations, such as inhaler QC or add-on device testing.

Relative precision of inhaler aerodynamic particle size distribution (APSD) metrics by full resolution and abbreviated andersen cascade impactors (ACIs): part 2--investigation of bias in extra-fine mass fraction with AIM-HRT impactor.

The purpose of this study was to resolve an anomalously high measure of extra-fine particle fraction (EPF) determined by the abbreviated cascade impactor possibly relevant for human respiratory tract (AIM-HRT) in the experiment described in Part 1 of this two-part series, in which the relative precision of abbreviated impactors was evaluated in comparison with a full resolution Andersen eight-stage cascade impactor (ACI). Evidence that the surface coating used to mitigate particle bounce was laterally displaced by the flow emerging from the jets of the lower stage was apparent upon microscopic examination of the associated collection plate of the AIM-HRT impactor whose cut point size defines EPF. A filter soaked in surfactant was floated on top of this collection plate, and further measurements were made using the same pressurized metered-dose inhaler-based formulation and following the same procedure as in Part 1. Measures of EPF, fine particle, and coarse particle fractions were comparable with those obtained with the ACI, indicating that the cause of the bias had been identified and removed. When working with abbreviated impactors, this precaution is advised whenever there is evidence that surface coating displacement has occurred, a task that can be readily accomplished by microscopic inspection of all collection plates after allowing the impactor to sample ambient air for a few minutes.

Improved quality control metrics for cascade impaction measurements of orally inhaled drug products (OIPs).

This study of aerodynamic mass-weighted particle size distribution (APSD) data from orally inhaled products (OIPs) investigated whether a set of simpler (than currently used) metrics may be adequate to detect changes in APSD for quality control (QC) purposes. A range of OIPs was examined, and correlations between mass median aerodynamic diameter and the ratio of large particle mass (LPM) to small particle mass (SPM) were calculated. For an Andersen cascade impactor, the LPM combines the mass associated with particle sizes from impactor stage 1 to a product-specific boundary size; SPM combines the mass of particles from that boundary through to terminal filter. The LPM-SPM boundary should be chosen during development based on the full-resolution impactor results so as to maximize the sensitivity of the LPM/SPM ratio to meaningful changes in quality. The LPM/SPM ratio along with the impactor-sized mass (ISM) are by themselves sufficient to detect changes in central tendency and area under the APSD curve, which are key in vitro quality attributes for OIPs. Compared to stage groupings, this two-metric approach provides better intrinsic precision, in part due to having adequate mass and consequently better ability to detect changes in APSD and ISM, suggesting that this approach should be a preferred QC tool. Another advantage is the possibility to obtain these metrics from the abbreviated impactor measurements (AIM) rather than from full-resolution multistage impactors. Although the boundary is product specific, the testing could be accomplished with a basic AIM system which can meet the needs of most or all OIPs.

Comparison of two approaches for treating cascade impaction mass balance measurements.

The multistage cascade impactor (CI) is the most appropriate tool for measuring the aero-dynamic particle size distribution (APSD) of active pharmaceutical ingredient(s) (API) in the aerosol from an orally inhaled drug product. It is possible to determine the total emitted mass per actuation of the inhaler by summing the individual component results obtained when determining APSD. The determination of total mass per actuation recovered from the CI components (or "mass balance" [MB]) has inherently lower precision than that of a delivered dose (DD) determination. An FDA draft guidance for industry has proposed using CI-determined MB as part of the product specification, with acceptance criteria of +/-15% of the label claim (LC) dosage. We propose instead that MB be used to assess whether the CI measurement of APSD is reliable. Two multitiered test schemes for MB are evaluated that allow for retests to accommodate the variability of the MB measurement. We provide statistical evaluations of both test schemes by using operating characteristic (OC) curves. We find that a two-tiered procedure with broader acceptance criteria but limited opportunity for investigating and retesting MB failure results in a greater risk of rejection of good batches ("false positive" error) without the commensurate reduction in the risk of passing unacceptable batches ("false negative" error). In contrast, a three-tiered procedure with narrower acceptance criteria, but more opportunity to check for potential CI system malfunction/method misapplication and to rerun the CI test, provides a compromise that enables the MB measurement to be used without significantly increasing the probability of false positive errors.

Performance properties of the population bioequivalence approach for in vitro delivered dose for orally inhaled respiratory products.

Regulatory agencies, industry, and academia have acknowledged that in vitro assessments serve a role in establishing bioequivalence for second-entry drug product approvals as well as innovator post-approval drug product changes. For orally inhaled respiratory products (OIPs), the issues of correctly analyzing in vitro data and interpreting the results within the broader context of therapeutic equivalence have garnered significant attention. One of the recommended statistical tests for in vitro data is the population bioequivalence method (PBE). The current literature for assessing the PBE statistical approach for in vitro data assumes a log normal distribution. This paper focuses on an assessment of that assumption for in vitro delivered dose. Concepts in development of a statistical model are presented. The PBE criterion and hypotheses are written for the case when data follows a normal distribution, rather than log normal. Results of a simulation study are reported, characterizing the performance of the PBE approach when data are expected to be normally distributed, rather than log normal. In these cases, decisions using the PBE approach are not consistent for the same absolute mean difference that the test product is from the reference product. A conclusion of inequivalency will occur more often if the test product dose is lower than the reference product for the same deviation from target. These features suggest that more research is needed for statistical equivalency approaches for in vitro data.