Characterization of the effect of the membrane on in vitro dissolution profiles for pulmonary drug delivery
Maral Simonides
Supervisors: Irès van der Zwaan & Göran Frenning
Degree Project in Pharmaceutical Technology and Quality Assurance, 30.0 hp, A9
Examiner: Göran Alderborn
Division of Pharmaceutics
Department of Pharmaceutical Biosciences Faculty of Pharmacy
Uppsala University
Characterization of the effect of the membrane on in vitro dissolution profiles for pulmonary drug delivery
Abstract
It has always been a challenge to imitate the lung environment, therefore there is a constant development of standardized in vitro dissolution methods for inhaled products. Dissolution in vitro has been considered as an important parameter, because low solubility determines the bioavailability of inhaled drugs. The in vitro dissolution data generated by the dissolution test experiment can be correlated with in vivo pharmacokinetic data through in vitro-in vivo correlation (IVIVC), because a completed predictive IVIVC model is very useful for drug formulation design and manufacturing changes after approval. The aim of this study was to investigate the effect of the membrane on the dissolution profile of orally inhaled drugs with different solubility, Budesonide (BUD) and Fluticasone propionate (FP) in the different pore sizes of the membrane 8.0 μm, 3.0 μm and 0.4 μm. The method in this study builds on previous dissolution methods, a Transwell® setup to dissolve the drugs with a small amount of dissolution medium, which mimics more the limited lung fluid capacity in vivo. In order to collect the dose from the drugs, Andersen Cascade Impact was used. The dissolution rate of BUD was first in the ranking in all of the pore sizes in the membrane.
Keywords: Oral inhalation, in vitro dissolution, Budesonide, Fluticasone propionate, Transwell
1. Introduction
Inhalation is the prioritized route for treating lung diseases. In order to develop inhalation products in an effective way, the relationship between particle parameters and in vivo performance must be determined. Dissolution in vitro has been considered as an important parameter, because low solubility determines the bioavailability of inhaled drugs (Radivojev et al, 2019).
Inhaled corticosteroids (ICS) combined with long-acting beta2-agonists (LABA) are the
suggested first-line treatment for persistent asthma. However, the efficacy and safety of ICS may vary due to differences in dosages, formulations and delivery systems. In particular, the chemical properties of the active ingredient, its lung bioavailability, and the size and physical properties of the particles themselves can significantly affect the efficacy of inhalation therapy (Kupczyk et al, 2020).
The ICS particles have high lipophilicity therefore the dissolution is considered to be the step which is rate limiting for the drug to enter the cells or being absorbed in the systematic circulation (Rohrschneider et al, 2015).
Budesonide (BUD) and Fluticasone propionate (FP) are ICS, BUD is a glucocorticoid that is being used to treat chronic obstructive pulmonary disease (COPD), asthma, and non-infectious rhinitis. FP is also a glucocorticoid and is used to treat asthma and allergic rhinitis (Crim et al, 2001). FP has higher lipophilicity than BUD and has also better affinity, which means that it binds more strongly to the binding site for the glucocorticoid receptor (Yeo et al, 2017). BUD is
comes to the solubility and the dissolution rate in the surfactant sodium dodecyl sulfate (SDS) ( Rohrschneider, 2012).
In the pharmaceutical industry, for quality control purposes, drug dissolution testing is often used to provide critical in vitro drug release information, i.e., to evaluate the consistency of solid oral dosage forms (such as tablets) between batches and drug development, to also predict the in vivo drug release profile. The in vitro dissolution data generated by the dissolution test experiment can be correlated with in vivo pharmacokinetic data through in vitro-in vivo correlation (IVIVC).
Complete predictive IVIVC model is very useful for drug formulation design and manufacturing changes after approval (Bai et al, 2011).
It has always been a challenge to imitate the lung environment, therefore there is a constant development of standardized in vitro dissolution methods for inhaled products.
There are currently no pharmacopoeial methods for assessing orally inhaled product dissolution rates (Radivojev et al, 2019). Methods include modifications to existing resolution or
permeability methodologies, such as μDiss, USP apparatus 4, Franz cell, and Transwell (Franek et al, 2018).
The Transwell® (Fig. 1) permeable supports are commonly used equipment for cell culture problems. In 2010, the static Transwell® system was introduced as a new method for inhalation powder dissolution testing. Both the modified Transwell® and Franz Cell have an air liquid interface on the membrane, but with the modified Transwell® the amount of dissolution medium is small, which mimics more the limited lung fluid capacity in vivo (May, 2013).
Transwells® can be used with polyester (PE), polycarbonate (PC) or collagen coated polytetrafluoroethylene (PTFE) membranes with different pore sizes (May, 2013).
Andersen Cascade Impactor (ACI) (Fig. 2) is a dose collection method. The respiratory tract distribution of inhaled particles depends on the aerodynamic particle size. In order to classify powders for inhalation, one of the techniques to determine the in vitro distribution of
aerodynamic particles is ACI. It consists of eight different stages (0-7), and the nozzle size is reduced by increasing the number of stages. Between each stage, a collection board or impact board is placed. Depending on the reduced nozzle size, the air flow is accelerated. The particles with too large aerodynamic diameter cannot follow the deviation of the airflow by nearly 90° and hit the collecting plate. The airflow is from top to bottom. As the number of stages increases, the air velocity increases, size separation of the particles can effectively be obtained and smaller particles hit the collecting plate firstly (May, 2013).
Figure 1: Schematic drawing of Transwell dissolution apparatus (May, 2013).
Using ACI to collect the dose by settling the drug substance onto the filter, this study will use the modified ACI (mACI) with preseparator, stage 0, stage 1 and custom hollow stage to disperse and collect the drug on the filter at room temperature. The cut-off diameter of stage 1 at 60 L/min is 4.4μm. The hollow platform setup has the same height as the standard platform, each empty ring is 2.6 cm(Franek, et al 2018).
Figure 2: Illustrates dose collection by using mACI precipitation onto the filter and subsequent dissolution using the Transwell setup (Franek et al, 2018).
The present study builds on previous dissolution studies by Franek et al and Rohrschneider et al.
Rohrschneider et al. used a Transwell setup to dissolve the drug precipitated on the filter and a commercial Transwell insert with a membrane pore size of 0.4 μm. Franek et al. used 8.0 μm
membrane pore size (Franek et al, 2018). A commercially available Transwell 6-well plate was fixed on the top of the rotating table, dissolution media, phosphate-buffered saline (PBS) pH 6.8 and 0.5% SDS, were preloaded in the wells (Franek et al, 2018). The dissolution starts with moistening the filter with amount of dissolving medium and needs to be immediately placed in the pre-filled hole, and the shaking table rotates at 210 rpm (Franek et al, 2018).
The quantification of the drug substances concentrations was done by ultra performance liquid chromatography-UV (UPLC-UV). The diffusion through the filter and Transwell membrane was measured, but instead of the powder, budesonide solution was added to the filter (Franek et al, 2018).
Data analysis, Weibull function (Eq.1) was used, where Fd is the experimental cumulative fraction of dissolution at time t, t63 is the fitting time until 63% of the dose is dissolved, and b is the dissolution curve shape factor (Franek et al, 2018).
The Weibull fit was used to evaluate the dissolution data and expressed as t63, that is, the dissolution time was 63% of the initial dose. They ranked several drug substances after the t63, BUD was first in the ranking and had a t63 of 10 min and FP had a t63 of 38 min (Franek et al, 2018).
This method showed that the Transwell method has been used successfully to rank the dissolution rate of BUD and FP from commercial metered dose inhalers (MDI) (Franek et al, 2018).
Equation 1: Weibull function (Franek et al, 2018)
The dissolution test method to be used in this study is Transwell to be able to investigate how the dissolution behavior is affected by varying the pore size of the Transwell membrane, namely 8.0 μm, 3.0 μm and 0.4 μm.
2. Aim
The aim of this study is to investigate the effect of the membrane on the dissolution profile of orally inhaled drugs. The influence of membrane pore size on dissolution will be analyzed to see the influence on the dissolution profile. The drugs to be used are the well-known orally inhaled drugs Budesonide (BUD) and Fluticasone propionate (FP) in the different pore sizes of the membrane 8.0 μm, 3.0 μm and 0.4 μm.
3. Materials & Methods
3.1 Materials
Pulmicort Turbuhaler (Budesonide) and Flutide diskus (Fluticasone propionate) were bought from Distansapoteket in Stockholm. The pure FP and BUD for the stock solutions were
Pharmaceutical secondary standards (Sigma-Aldrich, Germany). Organic solvents that were used were at least HPLC grade (VWR, France). The water that was used was Ultrapure water
(PURELAB flex). Whatman glass microfiber filters in 21 mm were used. PBS tablet was obtained from EC Diagnostics (Uppsala, Sweden), and SDS was bought from Sigma-Aldrich (Germany). The filter which was used to filter the buffer was a 2.0 μm filter from Filtropur S 0.2 (Sarstedt, Germany). The Transwells were Corning Transwell system with 24 mm inserts,
polycarbonate membranes with pore sizes of 8.0 μm, 3.0 μm and 0.4 μm (Sigma-Aldrich, Germany)
3.2 Stock solution preparation of BUD and FP, calibration curve and buffer
The buffer was prepared by dissolving 1 tablet PBS in 1000 ml water and 5 g SDS. The stock solution of BUD with a final concentration of 50 μg/ml was prepared by adding 40 μl from 5 mg/ml stock solution and 3960 μl buffer. From the stock solution the calibration curve was prepared, samples with a concentration of 0.05 μg/ml, 0.1 μg/ml, 0.5 μg/ml and 5.0 μg/ml were prepared and analysed with the UPLC-UV.
The stock solution 50 μg/ml FP was prepared in the same way and the calibration curve samples had concentrations of 0.05 μg/ml, 0.2 μg/ml, 0.5 μg/ml and 5,0 μg/ml.
3.3 Dose collection of BUD and FP via sedimentation onto filters using mACI
The mACI with preseparator, stage 0, stage 1 and custom hollow stages was used to deposit drugs from the clinical device onto the filter paper with a flow rate of 60 l/min, suction time of 0.3 seconds and a sedimentation time of 20 min. The cut-off diameter of stage 1 at 60 L/min is 4.4 μm. The suction of the several drugs was done every 20 min, 3 times for the Diskus for FP and 5 times for the Turbuhaler for BUD. 3 filters were placed on the collection plate in the mACI and were collected after all the suctions were done and moved to the Transwell inserts to
dissolve.
3.4 Dissolution with Transwell setup after dose collection of BUD and FP
The dissolution experiments for BUD using the Transwell setup started by using a 6 well plate, the 3 filters from the mACI were moved onto 3 upper well inserts in order to do triplicates of the experiment. The space in between the bottom of the plate and the well inserts was filled for all the 3 inserts with 2.3 ml buffer. Thereafter 3 inserts with the filters were moved in to the 6 well plate. 700 μl buffer was added on the top of the filters and the plate was immediately moved to the shaking table (Heidolp Unimax 1010) with a stirring speed of 150 rpm. Thereafter samples of 200 μl were taken using a multichannel pipette at 2 min, 5 min, 10 min, 15 min, 30 min, 60 min, 120 min from the bottom of the plate and piped into vials. After every sample was taken, 200 μl from the buffer was added to the solution in the bottom of the plate. After the 120 min timepoint, 3 ml of methanol was added to extract any remaining drug substance and to ensure mass balance, therefore samples were obtained after at least 30 minutes and pipetted into vials. The vials were then analyzed using UPLC-UV for quantification. The same experiment was done for the three different pore sizes, 8.0 μm, 3.0 μm and 0.4 μm, of the Transwell membrane.
The FP experiments were done in the same way as for BUD using the Transwell setup, but for FP an extra timepoint at 180 min was needed for all the pore sizes and therefore 3 ml of methanol was added after 180 min.
3.5 Diffusion of the solution of BUD and FP across filter and Transwell membrane
The diffusion was measured as described above for the dissolution but no deposition of particles with the mACI was done. Instead, 200 μl from the stock solution of BUD and FP was added to the filters.
3.6 Quantitative analysis of BUD and FP by UPLC-UV
BUD and FP samples were analyzed by a Waters Acquity UPLC-UV I-Class system with a BEH C18 column with 1.7 μm particle size. The UPLC-UV method for BUD had a run time of 1.8 min, the mobile phase A was water with 0.03% TFA, mobile phase B was acetonitrile with 0.03
% TFA. The wash solution was MeOH:water (50:50), the column temperature was 40 °C, the sample temperature was 18 °C, the wavelength was 249-259 nm and the retention time of BUD was at 0.78 min.
The UPLC-UV method for FP had a run time of 1.1 min, the mobile phase A was water with 0.03% TFA and the mobile phase B was acetonitrile with 0.03% TFA. The wash solution was MeOH:water (50:50), the column temperature was 40 °C, the samples temperature was 18 °C, the wavelength 232-242 nm and retention time of FP was at 0.92 min. Quantification was based on peak area measurement at wavelength with maximum absorption (λmax).
3.7 Data analysis
The data was compared according to the t63, the fitting time until 63% of the dose is dissolved.
Excel (Microsoft Corporation, Washington, USA) was used to compare t63 for BUD and FP.
Excel was also used to calculate the standard deviation (SD) from triplicate experiments.
4. Results and discussion
4.1 Diffusion across membrane
Figure 3. The diffusion rate profiles of BUD in 8.0, 3.0, 0.4 μm pore sizes.
The rate profiles the diffusion of BUD in Fig. 3 show that the diffusion is faster for the pore sizes 8 and 3 μm and much slower for the pore size 0.4 μm. However, there is a small difference at the beginning between the pore size 8.0 and 3.0 μm of the membrane where 3 μm is a bit faster, but the average fraction transferred reaches 1 at the same time.
0 0,2 0,4 0,6 0,8 1 1,2
0 20 40 60 80 100 120 140
Average fraction transferred
Time (min)
Figure 4. The diffusion rate profiles of FP in 8.0, 3.0, 0.4 μm.
The rate profiles of the diffusion of FP in Fig. 4 show that the diffusion is faster for the pore sizes 8.0 and 3.0 μm and much slower for the pore size 0.4 μm.
4.2 Drug dissolution
Figure 5. The dissolution rate profiles of BUD in 8.0, 3.0, 0.4 μm.
0 0,2 0,4 0,6 0,8 1 1,2
0 50 100 150 200
Average fraction transferred
Time (min)
-0,2 0 0,2 0,4 0,6 0,8 1 1,2
0 50 100 150 200
Average fraction dissolved
Time (min)
The rate profiles of the dissolution of BUD in Fig. 5 show that the dissolution is faster for the pore sizes 8.0 and 3.0 μm and much slower for the pore size 0.4 μm. Even here the pore size 3.0 μm is a bit faster than pore size 8.0 μm at the beginning but the average fraction dissolved for both reaches 1 at the same time.
Figure 6. The dissolution rate profiles of FP in 8.0, 3.0, 0.4 μm.
The rate profiles of the dissolution of FP in Fig. 6 show that the dissolution is faster for the pore sizes 8.0 and 3.0 μm and much slower for the pore size 0.4 μm.
4.3 Rate constants
Table 1. The average t63 and the SD values for the diffusion and the dissolution for BUD.
Pore size μm
Diffusion, average t63 (min)
Dissolution, average t63 (min)
8.0 6.1 (2.6) 10.1 (3.5)
3.0 3.4 (1.0) 6.4 (0.3)
0.4 20.9 (8.4) 43.2 (4.4)
-0,2 0 0,2 0,4 0,6 0,8 1 1,2
0 50 100 150 200
Average fraction dissolved
Time (min)
Table 2. The average t63 and the SD values for the diffusion and the dissolution for FP.
Pore size
μm Diffusion, average t63
(min) Dissolution,
average t63(min)
8.0 7.6 (1.6) 10.9 (4.3)
3.0 6.6 (0.7) 13.3 (2.6)
0.4 26.5 (0.0) 71.4 (1.9)
In Table 1 and 2, the t63 for the diffusion rate for BUD was a bit faster than the FP in all pore sizes and the dissolution rate was also faster for the BUD than FP.
The rate profiles in figure 1, 2, 3 and 4 show that the dissolution is faster for the pore sizes 8.0 and 3.0 μm and much slower for the pore size 0.4 μm. For the pore sizes 8.0 and 3.0 μm of the membrane, the dissolution rates are quite similar to each other, however, the pore size 3.0 μm is slightly faster than 8.0 μm but the average fraction dissolved reaches 1 at the same time.
The study by Franek et al. ranked several drug substances after the t63, BUD was first in the ranking and had a t63 of 10 min and FP had a t63 of 38 min, the pore size of the membrane was 8 μm. In this study in Table 1 and 2 BUD was also first in the ranking and had a t63 of 10.1 min and FP had a t63 of 10.9 min for the dissolution in the pore size 8.0 μm of the membrane and it vary from the results from the study by Franek et al. Thereafter a bigger difference in this study was noticed in Table 1 and 2 using the pore size 0.4 μm, the t63 of BUD is 43.2 min and for FP is 71.4 min. The dissolution of the drugs is faster through the membrane with bigger pore size. The difference between BUD and FP is the lipophilicity, it was found that FP dissolves slower than BUD in the several pore sizes of the membrane. The membrane with the pore size 0.4 μm has showed more significant difference between BUD and FP, Fig. 5 and 6 show that the dissolution
profile rate for BUD and FP is a lot slower than the dissolution rate in pore sizes 8.0 and 3.0 μm.
Table1 and 2 show the huge difference in the average t63 for the dissolution for BUD and FP.
The ICS particles have high lipophilicity in general and FP is more lipophilic than BUD
therefore the dissolution FP was much slower compared with BUD in every experiment that has been done in different pore sizes of the membrane.
The dissolution of less soluble drugs may significantly affect the lung absorption rate and drug concentration inside and outside the target site, this affects the safety, the efficacy even the treatment duration.
The method used was easy to perform, except when loading a new dose of BUD with mACI between each suction as the turbuhaler got stuck in the mouthpiece which made it hard to trust that the entire dose was sucked. Several experiments were performed to ensure that the correct dose was inhaled.
5. Conclusion
The method presented could successfully rank the dissolution rate of BUD and FP. A significant difference was found between the different sizes of the pores for the membrane, the larger size 8.0 μm allows the dissolution through the membrane faster than the smaller pore size 0.4 μm.
However, no major difference between pore size 8.0 and 3.0 μm. The different solubility of the substances also proved to impact the dissolution rate. FP dissolves slower than BUD, which made the dissolution rate faster for BUD and slower for FP in all the different pore sizes of the membrane that has been investigated. BUD was first in the ranking in all of the pore sizes for the dissolution in the membrane and had a t63 of 10.1 min and FP had a t63 of 10.9 min in the pore
size 8.0 μm. BUD, 6.4 min and FP 13.3 min in the pore size 3.0 μm. BUD 43.2 min and FP 71.4 min in the pore size 0.4 μm. The most significant difference between the dissolution rate for BUD and FP show in the pore size 0.4 μm which need further studies with different substances.
It would be interesting to do several experiments with different orally inhaled drugs with different properties such as different solubilities to compare the results and strength the study.
6. Acknowledgement
To the supervisors, Irès van der Zwaan and Göran Frenning.
7. References
Bai, G., Wang, Y., Armenante, P. 2011. Velocity profiles and shear strain rate variability in the USP Dissolution Testing Apparatus 2 at difawferent impeller agitation speeds. Int J
Pharm. 403(1-2), 1-14. DOI: 10.1016/j.ijpharm.2010.09.022
Crim, C., Pierre, L. N., & Daley-Yates, P.T. 2001. A review of the pharmacology and
pharmacokinetics of inhaled fluticasone propionate and mometasone furoate. Clin Ther. 23(9), 1339-1354. DOI: 10.1016/s0149-2918(01)80113-2
Franek, F., Fransson, R., Thörn, H., Bäckman, P., Andersson, PU., Tehler, U. 2018. Ranking in Vitro Dissolution of Inhaled Micronized Drug Powders including a Candidate Drug with Two Different Particle Sizes. Mol Pharm. 15, 5319–26. DOI: 10.1021/acs.molpharmaceut.8b00796
Kupczyk, M., Majak, P., Kuna, P., Asankowicz-Bargiel, B., Barańska, E., Dobek, R., Majorek- Olechowska, B., Olech-Cudzik, A., Poziomkowska-Gęsicka, I., Rulewicz-Warniełło, M.,
Świderska, A., Tarnowski, M., & Kopyto, P. 2020. A new formulation of fluticasone
propionate/salmeterol in a metered-dose inhaler (MDI HFA) allows for the reduction of a daily dose of corticosteroid and provides optimal asthma control-A randomized, multi-center, non- inferiority, phase IV clinical study. Respir Med. 176, 106274. DOI: 10.1016/j.rmed.2020.106274
May, S. (2013). "Dissolution testing of powders for inhalation." DOI: 10.22028/D291-22927
Radivojev S., Zellnitz S., Paudel A., Fröhlich E. 2019. Searching for physiologically relevant in vitro dissolution techniques for orally inhaled drugs. Int J Pharm. 556, 45–56.
https://doi.org/10.1016/j.ijpharm.2018.11.072
Rohrschneider M., Bhagwat S., Krampe R., Michler V., Breitkreutz J., Hochhaus G. 2015.
Evaluation of the Transwell System for Characterization of Dissolution Behavior of Inhalation Drugs: Effects of Membrane and Surfactant. Mol Pharm. 12(8), 2618–2624.
DOI: 10.1021/acs.molpharmaceut.5b00221
Rohrschneider, M. 2012. Correlation of ICS in vitro dissolution and pulmonary absorption.
Doctoral dissertation, Universitäts-und Landesbibliothek der Heinrich-Heine-Universität Düsseldorf, pp 2. Available at: [PDF] uni-duesseldorf.de
Yeo, S. H., Aggarwal, B., Shantakumar, S., Mulgirigama, A., & Daley-Yates, P. 2017. Efficacy and safety of inhaled corticosteroids relative to fluticasone propionate: a systematic review of randomized controlled trials in asthma. Expert Rev Respir Med. 11(10), 763-778.
DOI: 10.1080/17476348.2017.1361824
