Investigation of Mass Balance Achieved by Andersen Third Generation (A3G) Cascade Impactor During Aerodynamic Particle Size Analysis of Flovent HFA

Rajesh Maheshwari,Brajesh Kumar Thakur,GKesavan,Ritu Laddha, Ravindra K Kotak,Apurv J Patel, Sagar Chavan and Kaushal Patel

1 Lab Automate Technologies Inc, Millburn, NJ, USA

2 Zydus Cadila Healthcare Ltd, Ahmedabad, India

3 Lab Automate Technologies Inc, Vadodara, India

KEYWORDS: A3G, automated cascade impactor, high mass balance, APSD

 INTRODUCTION

The Andersen Cascade Impactor (ACI), and Next Generation Impactor (NGI) were designed for manual operation, which makes data generation slow and prone to operator-induced variability [1]. Since measurement of aerodynamic particle size distribution (APSD) is of critical importance to the development of inhaled aerosol products, and replicated testing is required during basic formulation, bioequivalence, and stability studies, these disadvantages can slow down product development and complicate the interpretation of results, which potentially leads to longer approval times. The Andersen third generation (A3G) cascade impactor was designed to take advantage of the validated geometry and particle sizing capabilities of the conventional ACI, but through the use of robotics, reduce the APSD sample preparation time to less than 25 minutes per run. In this abstract we investigate the ability of the A3G to consistently yield a high mass balance during ASPD experiments using Flovent® HFA pMDI. A high and reproducible mass balance is an important validation step in the aerodynamic sizing of inhaled products.

MATERIALS AND METHODS

 Test article

Flovent HFA (GlaxoSmithKline, fluticasone propionate inhalation aerosol, with a Label Claim of 220 mcg per spray) was used as purchased.

Fluticasone assay

A reverse phase High Performance Liquid Chromatography (HPLC) method for detection of fluticasone was developed based on the corresponding United States Pharmacopeial monograph (USP-39) [2]. Fluticasone propionate API and HPLC grade reagents (Zydus Cadila Healtcare Ltd., Gujarat, India) were used for assay development on a Shimadzu LC-2010C HPLC equipped with a XTerra RP 18 (5 µm) column maintained at 40°C. Injection volume was 50 µL, detection was at 239 nm and the total run time was 25 min. The mobile phase was comprised of acetonitrile and buffer solution (50:50 %v/v). The buffer solution contained 0.01 M of sodium dodecyl sulfate and 0.1 %v/v glacial acetic acid in 20 %v/v aqueous methanol. The column was isocratically eluted at a flow rate of 2.0 mL/min.

A3G design and operation

The A3G is able to robotically assemble, disassemble, and recover drug from the stages of an Andersen Cascade Impactor before  automatically  cleaning, drying  and  reassembling  column of stages containing impaction plates, and the terminal filter. Only the USP induction port, mouthpiece adaptor and pMDI actuator must be mounted and dismounted manually for drug recovery. The equipment is described in detail elsewhere [3, 4], but basically has two operational systems. (1) A Collapsed Stage Column in which aerosol particles are fractionated according to their aerodynamic size under controlled air flow, and (2) an Isolator Column, in which individual ACI stages and plates are separated and enclosed in sealed chambers to facilitate drug recovery by quantitative washing of all internal surfaces. Unlike the conventional ACI, the A3G accomplishes these operations repeatedly without operator intervention, and has the additional benefit of delivering the programmed volume of drug recovery solvent into each sealed chamber followed by programmed rocking a fixed number of times. The rocking algorithms used in the A3G are designed to force sample recovery solution through the stage jets to enhance drug dissolution and recovery. A subassembly of the A3G contains pumps, valves, and sample collection racks to accumulate samples for subsequent HPLC analysis. In these experiments the sample recovery solution was 60%v/v acetonitrile in water.

 The pMDI is automatically agitated and discharged into the USP induction port by the Shaker/ Actuator (Figure 1). Shake angle, acceleration, deceleration, velocity and number of times the pMDI is shaken are user selectable, as are the number of actuations and actuation force. The inhaler is robotically inserted in the mouthpiece adapter mounted on the induction port prior to each actuation.

Optimization actuation parameters

Actuation parameters were chosen during a series of preliminary experiments in which the Shaker/ Actuator of the A3G was used to discharge Flovent HFA into a Dosage Unit Sampling Apparatus (DUSA) tube (Lab Automate Technologies), using the range of actuation pressures shown in Table 1. Actuation parameters were varied in an attempt to (1) match the emitted dose to the Label Claim of Flovent HFA, and (2) ensure repeatable results (one actuation per test) during replicated actuations. The DUSA tube was operated at 28.3 L/min and drug was recovered in 25 mL of sample recovery solution for subsequent HPLC assay. Results are shown in Table 1. Each row is a single experiment.

Initial optimization of shaking and actuation parameters

We concluded that actuation pressures of 4 kg/cm2 yielded fluticasone propionate emitted doses closest to 100% Label Claim, and that automated shaking and actuation under these conditions yielded % Label Claim values closer to 100% than could be achieved by manual shaking and actuation. Following these initial experiments (as shown in Table 1), improvements were made to the actuation system hardware to minimize drug sedimentation in the inhaler between shaking and actuation. As shown in Table 2, the data obtained with this modified system operated at an inhaler actuation pressure of 4 kg/cm2 and by shaking the inhaler twelve times yielded fluticasone propionate recoveries close 100% of Label Claim with a high repeatability between duplicate experiments. 

Final optimization of shaking and actuation parameters.

Aerodynamic particle sizing and sample preparation 

Using an Actuation Pressure of 4 kg/cm2 with 12 shakes, a single canister of Flovent HFA, 220 mcg, was discharged four times during each of eight runs into the autonomously running A3G operated at

28.3 L/min. The mouth piece and all impaction plates were then automatically washed using 20 mL of sample recovery solution for each, while all stages and the inlet cone were washed using 30 mL of sample recovery solution for each.

RESULTS AND DISCUSSION

APSD results are shown in Table 3 as % of Label Claim recovered from each sample location. All shaking and actuation, particle fractionation by size within the Collapsed Stage Column, sample recovery in the Isolator Column, and collection of HPLC samples was fully automated, and except for interruptions caused by power outages, runs proceeded consecutively taking approximately 25 minutes each. The mean mass balance of fluticasone propionate was 100.5% of Label Claim of 220mcg with an RSD of 2.59%. On stages containing less the 0.3% Label Claim, RSD never exceeded 15%.

 Location-by-location fluticasone propionate recovery* following four sprays of Flovent HFA pMDI into the A3G Cascade Impactor using the automatic Shaker/Actuator (n=8).

*Fluticasone propionate recovery is expressed as % Label Claim (220mcg per spray). Reported values are derived from four sprays.

#Impaction Plate 1 is below Stage 0, Impaction Plate 2 is below Stage 1, etc. Reported fluticasone propionate recovery includes drug from each Stage and its corresponding Impaction Plate.

CONCLUSION

Eight aerodynamic particle size determinations performed using the fully automated A3G cascade impactor resulted in mass balances between 96.70% to 105.32% of Label Claim for Flovent HFA. These values corroborate unpublished previous results suggesting the A3G is able to generate highly accurate and reproducible data. The A3G represents a fast and cost-effective alternative to manual Anderson cascade impactor methods. Furthermore, adoption of the automated A3G system retains the flexibility of the conventional ACI. For example, air flow rates of 60 to 90 L/min can be used by simply swapping in and out the appropriate ACI stages, which takes less than 10 minutes.

ACKNOWLEDGEMENT

The authors of this work are grateful to the management and Analytical Research and Development Laboratory staff of Zydus Cadila Healthcare Limited for supporting this work.

REFERENCES

1. Bonam M, Christopher D, Cipolla D et al.: Minimizing variability of cascade impaction measurements in inhalers and AAPS PharmSciTech 2008, 9, 404–413.
2. United States Pharmacopeia, USP 39-NF34. Rockville, MD. United States Pharmacopeial Concention. Monograph for Fluticasone Propionate Inhalation Aerosol, 4001–4005.
3. Maheshwari RK, Patel K, Patel V, Maheshwari KK, Javia A, Sutaria V: A3G – Automated Andersen Cascade RDD Europe 2017. 2017, 2: 249–254.
4. Maheshwari RK, Sharma A, Maheshwari KK: Development of A3G – Automated Andersen Cascade RDD Asia 2014 2014, 1: 203–206.