In Vitro and In Vivo Studies of a Verapamil-Containing Gastroretentive Solid Foam Capsule

Gastroretentive systems may overcome problems associated with incomplete drug absorption by localized release of the API in the stomach. Low-density drug delivery systems can float in the gastric content and improve the bioavailability of small molecules. The current publication presents verapamil–HCl-containing solid foam prepared by continuous manufacturing. Production runs were validated, and the foam structure was characterized by micro-CT scans and SEM. Dissolution properties, texture changes during dissolution, and floating forces were analyzed. An optimized formulation was chosen and given orally to Beagle dogs to determine the pharmacokinetic parameters of the solid foam capsules. As a result, a 12.5 m/m% stearic acid content was found to be the most effective to reduce the apparent density of capsules. Drug release can be described by the first-order model, where 70% of verapamil dissolved after 10 h from the optimized formulation. The texture analysis proved that the structures of the solid foams are resistant. Additionally, the floating forces of the samples remained constant during their dissolution in acidic media. An in vivo study confirmed the prolonged release of the API, and gastroscopic images verified the retention of the capsule in the stomach.


Introduction
Nowadays, one of the most important challenges in pharmaceutical technology is the development of drug delivery systems (DDS). There are five generations of DDS. The first and second generations include conventional therapeutic drug carriers such as tablets, capsules, granules, or their enteric coating versions. The group of controlled DDS is the third generation, which is now a focus due to the reduced dosing frequency or enhanced drug efficacy compared with conventional formulations [1]. In addition, the number of newly synthesized active ingredients has decreased in recent decades, and the emphasis has been on converting existing active ingredients into renewed dosage forms. Drug delivery For manufacturing, the following parameters and method were used: SA were placed and heated in a melting container at 61 °C while being gent an IKA ® EURO-ST D stirrer (IKA®-Werke GmbH & Co. KG, Staufen, Germa Then, verapamil was dispersed in the molten mixture. The foaming step w cles. Firstly, the heated foaming cell was filled with the molten suspension 0.25 mL/s at 56 °C. A total of 2.0 mL of molten dispersion was pumped int cell, and the IKA ® ULTRA-TURRAX ® T-25 digital dispersing instrument dis the dispersion while pumping 2.0 mL of air slowly through the foaming c 0.25 mL/s. Lastly, the hot foam was dosed into size 00 hard gelatin capsules room temperature until solidification.

Determination of the Densities of the Samples
To characterize the densities of the solid formulations, the following The mass and volume of the empty capsules (00 size) were 118 ± 7 mg an spectively, in accordance with the manufacturer's specifications [28]. The d sample was calculated by the following formula after the capsules had be filled with foam. For manufacturing, the following parameters and method were used: PEG 4000 and SA were placed and heated in a melting container at 61 • C while being gently stirred with an IKA ® EURO-ST D stirrer (IKA®-Werke GmbH & Co. KG, Staufen, Germany) at 50 rpm. Then, verapamil was dispersed in the molten mixture. The foaming step was done in cycles. Firstly, the heated foaming cell was filled with the molten suspension at a speed of 0.25 mL/s at 56 • C. A total of 2.0 mL of molten dispersion was pumped into the foaming cell, and the IKA ® ULTRA-TURRAX ® T-25 digital dispersing instrument dispersed air into the dispersion while pumping 2.0 mL of air slowly through the foaming cell at a rate of 0.25 mL/s. Lastly, the hot foam was dosed into size 00 hard gelatin capsules and cooled to room temperature until solidification.

Determination of the Densities of the Samples
To characterize the densities of the solid formulations, the following test was used: The mass and volume of the empty capsules (00 size) were 118 ± 7 mg and 0.91 mL, respectively, in accordance with the manufacturer's specifications [28]. The density of each sample was calculated by the following formula after the capsules had been completely filled with foam.
where V capsule is the volume of the empty capsule body (0.91 mL); m sample is the mass of filled foam in the entire capsule (body and cap), as measured by an analytical balance; m capsule is the mass of the empty capsule (body and cap) (118 mg); and ρ foam is the density of the foam.

SEM Analysis
The solid foams were broken to characterize their structures with a Hitachi Tabletop microscope (TM3030 Plus, Hitachi High-Technologies Corporation, Tokyo, Japan). Samples were crushed into halves and mounted on a fixture with graphite-containing, double-sided adhesive tape. The broken surface of the samples was not coated with gold before the SEM examination. A vacuum and low accelerating voltage of 5 kV were used during the investigation. A chemical element analysis (oxygen or chlorine) was done on the broken surface using a Bruker EDX 70 detector.

Dissolution Test
Three random samples were taken for dissolution tests from every composition (V1-V3, preparation steps are described in Section 2.2). Nine hundred milliliters of hydrochloric acid media was used for the dissolution tests (pH = 1.2 without pepsin) using an Erweka DT 800 dissolution tester (rotating paddle method, 75 rpm and 37 • C). Samples (3 mL) were taken after 5, 15, and 30 min, and 1, 2, 3, 4, 5, 6, 7, 8, and 10 h and filtered through a 0.22 µm PES membrane syringe filter. The verapamil content of the samples was determined with a UV/VIS spectrophotometer (Shimadzu UV 1601, Shimadzu Corp. Kyoto, Japan) at 278 nm after dilution with a pH 1.2 buffer. Floatation was checked visually at regular intervals.

Mathematical Analysis of the Drug Release Profiles
The data from the dissolution tests was graphically fitted and evaluated by zero-order, first-order, and Korsmeyer-Peppas models in MS Excel (Microsoft Corporation, Redmond, WA, USA) ( Table 2). Table 2. Mathematical models of drug release profiles.

Model Equations [29,30] Graph
Zero-order The graph of the drug-dissolved fraction vs. time is linear.
Linear graph of the released amount of drug (expressed by decimal logarithm) vs. time.

Korsmeyer-Peppas model
Straight line graph of released drug vs. the square root of time.
where Q-amount of drug released at time t; Q 0 -the initial amount of drug at time 0; Q t -the amount of drug remaining at time t; Q t /Q ∞ -fraction of drug released at time t; k 0 , k 1 , and k kp -kinetic constants for zero-order, first-order, and Korsmeyer-Peppas models, respectively; n-release exponent, indicative of the drug release mechanism.

Validation of Production
During the validation process, 3 batches of the V2 composition were produced by the foaming cell at 3 different times, following the parameters of the standard protocol that were described in Section 2.2. The densities, API contents and dissolution profiles of the batches were compared.
The data obtained from the dissolution tests of the validation products were compared by similarity (f2) and difference (f1) factors as a model-independent approach for each sample.
where n is the sampling number and R j and T j are the percentages dissolved from the reference and test products at each time point j.
where w j is an optional weight factor. The dissolution efficacies (DE) were also calculated as follows.
where y is the drug dissolved at time t (%).

Microtomography and Size Distribution of Foam Cells
The structure of the solid foam was determined by a noninvasive method using a SkyScan 1272 (Bruker, Billerica, MA, USA) compact desktop micro-CT system. The following parameters were used to scan the capsules placed in the sample holder: image pixel size-5 microns, matrix size-1344 × 2016 (rows × columns), Source Voltage-50 kV, and Source Current-200 µA. Flat Field Correction and Geometrical Correction were used. The cross-sectional images were reconstructed from tomography projection images by the SkyScan NRecon package (Version: 2.0.4.2). Post-alignment, Beam-hardening correction, Ring artefact correction, and Smoothing were done. The output formats were DICOM and BPM.
CTAn software was used in the 2D/3D analyses. Thresholding, ROI shrink-wrap, Reload, and 2D and 3D Analysis plugins were applied based on the density analysis. Air bubbles had gray threshold values of 0-40, and the background was removed by ROI shrink-wrap before the analysis. The 3D visualization was done using CTVox software with color coding.

Texture Analysis
The texture analysis was used to characterize the mechanical properties and structures of the dry and wet foamy formulations. Unwetted control samples were not immersed into the dissolution media and tested at room temperature. At first, three random samples of verapamil-containing foam were placed in the Erweka dissolution tester, and the method described in 2.5 was used. Samples were taken after 1, 3, 5, 7, and 10 h, and excess water was absorbed from their surfaces. The Brookfield CT3 texture analyzer was used in the measurements. TA25/1000, an acrylic cylinder (d: 50.8 mm), compressed the samples at a speed of 0.50 mm/s until reaching the maximum load of 4500 g, and the test probe was fixed for a 5 s hold time at the target pressure. After this, the test probe returned to its starting position. Load values are depicted as a function of time (s) to demonstrate texture changes.

Water Uptake and Matrix Erosion Studies
Initial sample weights (W ini ) were measured before the study. Then, the specimens were put into the dissolution vessels, as described in Section 2.5. Samples were slowly removed from the dissolution media with a plastic net after 1, 3, 5, 7, and 10 h and the wet samples (W wet ) were weighed after removing excess water from the surfaces of samples. Then, the samples were dried at 45 • C (Memmert SFE 550, Memmert GmbH, Schwabach, Germany) until reaching a constant weight. Three samples were tested from the selected composition.
The water uptake % (WU%) was calculated as follows: The remaining % of foam was calculated by the next equation: where W wet is the the mass of the wet samples, W dry is the dried mass of the samples, and W ini is the initial mass of the samples before testing. Erosion was followed by microtomography by performing CT scans after 1, 3, 5, 7, and 10 h of dissolution.

Floating Strength Determination
An apparatus was built to measure the buoyancy force of samples based on Simons and Wagner [31]. Briefly, a modified tensiometer measured the weight changes of a net holding the sample (Sigma 700, Attension), and using this data, the buoyancy force was calculated. Measurements were made in 500 mL in a buffer of pH = 1.2 at 37 • C under continuous stirring (50 rpm) to ensure similar conditions to those used in in vitro release studies.

Dissolution Test after Long-Term and Accelerated Storage Conditions
To evaluate the drug release profiles after storage, in accordance with ICH guidelines [18], capsules were stored in airtight glass containers under two different conditions. Ten capsules were placed in a climate chamber (ICH110, Memmert GmbH + Co. KG, Schwabach, Germany) for 3 months under accelerated storage conditions (40 ± 2 • C, 75 ± 5% RH), and another ten capsules were stored in an airtight container and kept at room temperature for two years. After storage, samples were inspected for appearance and in vitro drug release. In vitro drug release was compared with the initial dissolution data from samples by similarity (f2) and difference (f1) factors.

In Vivo Pharmacokinetic Study
The pharmacokinetic study was approved by the University of Debrecen Committee of Animal Welfare and by the National Food Chain Safety Office (approval number: HB/06-ÉLB/1657-4/2019) in accordance with national (Act XXVIII of 1998 on the protection and sparing of animals) and European Union (Directive 2010/63/EU) regulations. Six female dogs (10 ± 0.5 kg) were involved in the experiment. Animals were kept at a temperature of between 15 and 21 • C at a relative humidity of 50 ± 10% and a 12 h-12 h light-dark cycle. For all dogs, free access to tap water and 300 g of pelleted food were provided daily. Acclimatization was allowed for 21 days prior to experiments. The experiments were carried out in two steps with a 7-day wash-out period between them. Twelve hours before each experiment, food was withdrawn, allowing only free access to water. Blood samples were collected by cannulation of the cephalic vein. After sample collection, sterile saline (3 mL) was injected to flush the cannula, and sodium heparin solution was used as an anticoagulant.
First, 50 mg of verapamil dissolved in 10 mL of purified water was administered orally to the animals (n = 6). Blood samples (2-3 mL) were collected 0.5, 1, 2, 4, 6, 8 and 24 h after administration. Blood samples were centrifuged immediately (3500 rpm for 10 min), and plasma samples were kept at −80 • C until further analysis. After the blood sample collection at 6 h, food was given to all dogs.
Following at 1 week washout period, 120 mg of verapamil-containing foamed capsules was given orally to all dogs (n = 6). The procedure (sampling time, sample amounts, plasma separation, and storage) was the same as described above. Food was also given 6 h after capsule administration.

Quantitative Determination of Verapamilin Plasma
A liquid-liquid extraction method was used for sample preparation. A phosphate buffer (pH 6.0) was prepared in accordance with the European Pharmacopoeia. Briefly, 6.8 g of sodium dihydrogen phosphate was dissolved in 1 L of purified water. The pH was adjusted with sodium-hydroxide. Dog plasma samples (250 µL) were mixed with 0.5 mL of phosphate buffer (pH 6.0). Then, they were extracted with 1.5 mL of a TBME-ethyl acetate (1:1, v:v) mixture. The organic phase was separated by centrifugation (4000× g for 8 min), collected, and evaporated completely with a nitrogen stream at 40 • C [32]. The residue was solubilized in a 500 µL LC mobile phase and measured by the Thermo Accela +LTQ XL LCMS instrument (Thermo Fisher Scientific, Waltham, MA, USA). Chromatographic separation was achieved by a Kinetex XBC18 column (100 × 2.1 mm, 2.6 µm). The mobile phase for elution comprised methanol + 0.1% formic acid (A) and water + 0.1% formic acid (B). The gradient conditions were 0.00 min 70% B, 2.00 min 0% B, 3.10 min 70% B, and 5.50 min 70% B. The column temperature was maintained at 40 • C with the flow rate set at 0.3 mL/min, and a 10.0 µL sample was injected. The optimal ESI ionization parameters were as follows: heater temperature-200 • C; sheath gas-N 2 ; flow rate-10 arbitrary units (arb); aux gas flow rate-5 arb; spray voltage-5 kV; capillary temperature-275 • C; and capillary voltage-23.50 V. Sample measurements were run in positive ion mode (MS). The verapamil [M + H] + ion mass (455 m/z) was detected in SIM mode. For calibration, verapamil was dissolved in the LC mobile phase. To measure recovery, spiked samples containing 9.6, 19.2, and 38 ng/mL verapamil-HCL, prepared as above for injection, were used.
The results were analyzed by Graphpad Prism (Version 6.1 software, GraphPad Software Inc., San Diego, CA, USA) and the relative bioavailability was calculated with the following Equation (10): where f r is the relative bioavailability (%); AUC is the area under the curve (drug concentration in plasma versus time); and D is the dose of drug B reference sample and A test sample.

Statistical Analysis
The statistical analysis was performed with GraphPad Prism ® software (Version 6.01, GraphPad Software Inc.). One-way ANOVA and Tukey's post hoc test were applied to compare the density of the compositions described in Section 2.3. One-way ANOVA and Dunnett's post hoc test were used when the densities of 3 different V2 compositions were compared, as described in Section 2.7. Differences were considered significant at p < 0.05.

Density of Compositions
Three compositions were produced by continuous production, and the final densities were calculated by the equation mentioned above in Section 2.3. The densities of all samples were below 1000 mg/cm 3 , which is necessary to achieve gastric retention. From V1 to V3, the compositions contained increasing amounts of stearic acid as a lipophilic agent. The lowest density was measured for V2. Interestingly, we obtained higher values for V3 despite increasing the lipophilic agent content in the formulation (see Figure 2). armaceutics 2022, 14, Dunnett's post hoc test were used when the densities of 3 different V2 compo compared, as described in Section 2.7. Differences were considered significan

Density of Compositions
Three compositions were produced by continuous production, and the fi were calculated by the equation mentioned above in Section 2.3. The densitie ples were below 1000 mg/cm 3 , which is necessary to achieve gastric retention V3, the compositions contained increasing amounts of stearic acid as a lipo The lowest density was measured for V2. Interestingly, we obtained higher v despite increasing the lipophilic agent content in the formulation (see Figure   Figure 2. Densities of compositions V1-V3. V2 had the significantly lowest density v cates statistically significant differences at p < 0.0001. The measured values, averag standard deviations are presented (n = 20).

SEM Analysis
The SEM analysis showed the different properties of the compositions (F was shown to have a totally different broken surface from V1 and V2. The A lated on the surface of the bubbles in general; however, in the case of V3, the pamil crystals were found on the inner surfaces of the cavities, while for V1 crystals were covered by the matrix on the outer surfaces of the bubbles. In a a high porosity was observed, and this was detected in the form of holes on surface. The chemical analysis of V3 proved that the API existed on the walls of p 4). The oxygen-rich component (red) of the matrix surrounded the verapamil in chlorine, blue).

SEM Analysis
The SEM analysis showed the different properties of the compositions (Figure 3). V3 was shown to have a totally different broken surface from V1 and V2. The API accumulated on the surface of the bubbles in general; however, in the case of V3, the white verapamil crystals were found on the inner surfaces of the cavities, while for V1 and V2, the crystals were covered by the matrix on the outer surfaces of the bubbles. In all 3 samples, a high porosity was observed, and this was detected in the form of holes on the fractured surface.
The chemical analysis of V3 proved that the API existed on the walls of pores ( Figure 4). The oxygen-rich component (red) of the matrix surrounded the verapamil crystals (rich in chlorine, blue).

Dissolution Test
The dissolution test showed prolonged drug release for up to ten hours in all cases ( Figure 5). Flotation was checked visually, and no drugs sunk before the end of the test. The V1 composition reached the highest dissolution rate of 85.3%, while V2 and V3 reached lower values of 79.4% and 77.4%. When the dissolution efficiencies were determined, it was found that V1 showed the fastest release with a value of 70.03%, while for V3, the DE value was only 68.43%. The kinetic profiles are summarized in Table 3. All compositions fitted the best to first-order kinetics pattern; however, V2 fitted best with the target zero-order kinetics (the correlation coefficient was 0.7572). As V2 showed the best properties, this composition was mainly examined in further experiments.

Validation of Production
Validation of the production was then performed with sample V2, and this composition was further analyzed as a final product. Three batches were made using the method described in Section 2.2, and the validation properties of the batches are summarized in

Dissolution Test
The dissolution test showed prolonged drug release for up to ten hours in all cases ( Figure 5). Flotation was checked visually, and no drugs sunk before the end of the test. The V1 composition reached the highest dissolution rate of 85.3%, while V2 and V3 reached lower values of 79.4% and 77.4%. When the dissolution efficiencies were determined, it was found that V1 showed the fastest release with a value of 70.03%, while for V3, the DE value was only 68.43%. The kinetic profiles are summarized in Table 3. All compositions fitted the best to first-order kinetics pattern; however, V2 fitted best with the target zero-order kinetics (the correlation coefficient was 0.7572). As V2 showed the best properties, this composition was mainly examined in further experiments.

Dissolution Test
The dissolution test showed prolonged drug release for up to ten hours in all cases ( Figure 5). Flotation was checked visually, and no drugs sunk before the end of the test. The V1 composition reached the highest dissolution rate of 85.3%, while V2 and V3 reached lower values of 79.4% and 77.4%. When the dissolution efficiencies were determined, it was found that V1 showed the fastest release with a value of 70.03%, while for V3, the DE value was only 68.43%. The kinetic profiles are summarized in Table 3. All compositions fitted the best to first-order kinetics pattern; however, V2 fitted best with the target zero-order kinetics (the correlation coefficient was 0.7572). As V2 showed the best properties, this composition was mainly examined in further experiments.

Validation of Production
Validation of the production was then performed with sample V2, and this composition was further analyzed as a final product. Three batches were made using the method described in Section 2.2, and the validation properties of the batches are summarized in

Validation of Production
Validation of the production was then performed with sample V2, and this composition was further analyzed as a final product. Three batches were made using the method  Table 4. The average product weight was 890.7 ± 26.7 mg, and no samples deviated more from the average weight than the amount allowed in the pharmacopoeia (Supplementary Materials Figure S1). The average verapamil content was 115.4 mg, and the API content of all samples was in the range of 85-115%. The dissolution profiles of batches were compared with the V2 profile, and none differed from that ( Figure S2).

Microtomography
The foaming process dispersed air into the molten suspension containing verapamil, and the creation of cavities was confirmed by micro-CT images ( Figure 6). The distribution of cavities in the matrix and the size distribution of the bubbles were homogenous. A total of 89.2% of them were in the 20-120 µm range for diameter, and the average diameter of the cavities was around 78 µm (Figure 6b).   Figure S1). The average verapamil content was 115.4 mg, and the API content of all samples was in the range of 85-115%. The dissolution profiles of batches were compared with the V2 profile, and none differed from that ( Figure S2).

Microtomography
The foaming process dispersed air into the molten suspension containing verapamil, and the creation of cavities was confirmed by micro-CT images ( Figure 6). The distribution of cavities in the matrix and the size distribution of the bubbles were homogenous. A total of 89.2% of them were in the 20-120 μm range for diameter, and the average diameter of the cavities was around 78 μm (Figure 6b). The reconstructed model of the foam structure shows a closed spheroid cell structure, but some of the bubbles (bubbles marked with red) can be observed at the edges open to the surface. Some of them form a cavity system in the matrix, although their number is small compared to the whole matrix (Figure 7).

Texture Analysis
The results of the texture analysis are presented in Figure 8. Despite the high porosity, a hard structure is presented. Using a compression load of 4500 g on the foam, no cracks, fractures, or other injuries were detected on the dry sample at 25 °C. During the drug release, the initial hardness of the samples decreased. After 60 and 180 min of dissolution, a soft layer was found around the hard core, which was easily removable, and the two parts easily separated by micro-CT scans, as shown in Figure S3. At 300 and 420 min of dissolution, some cracking was seen on the curves of samples. The hard core became fragile, and the core was broken from 3810 g (for the 300 min sample) and 3770 g (for the 420 min sample). Thorough wetting was detected after 600 min, and the breakable hard core disappeared.

Texture Analysis
The results of the texture analysis are presented in Figure 8. Despite the high porosity, a hard structure is presented. Using a compression load of 4500 g on the foam, no cracks, fractures, or other injuries were detected on the dry sample at 25 • C. During the drug release, the initial hardness of the samples decreased. After 60 and 180 min of dissolution, a soft layer was found around the hard core, which was easily removable, and the two parts easily separated by micro-CT scans, as shown in Figure S3. At 300 and 420 min of dissolution, some cracking was seen on the curves of samples. The hard core became fragile, and the core was broken from 3810 g (for the 300 min sample) and 3770 g (for the 420 min sample). Thorough wetting was detected after 600 min, and the breakable hard core disappeared.

Texture Analysis
The results of the texture analysis are presented in Figure 8. Despite the high porosity, a hard structure is presented. Using a compression load of 4500 g on the foam, no cracks, fractures, or other injuries were detected on the dry sample at 25 °C. During the drug release, the initial hardness of the samples decreased. After 60 and 180 min of dissolution, a soft layer was found around the hard core, which was easily removable, and the two parts easily separated by micro-CT scans, as shown in Figure S3. At 300 and 420 min of dissolution, some cracking was seen on the curves of samples. The hard core became fragile, and the core was broken from 3810 g (for the 300 min sample) and 3770 g (for the 420 min sample). Thorough wetting was detected after 600 min, and the breakable hard core disappeared.

Water Uptake and Matrix Erosion Studies
During dissolution, the mass of the matrix decreased continuously, as shown in Figure S3. A summary of the erosion and water uptake data is shown in Figure 9. The matrix did not show swelling, and the sum of remaining matrix and the water content was constant. Only 30% of the sample remained after the test.

Water Uptake and Matrix Erosion Studies
During dissolution, the mass of the matrix decreased continuously, as shown in Figure S3. A summary of the erosion and water uptake data is shown in Figure 9. The matrix did not show swelling, and the sum of remaining matrix and the water content was constant. Only 30% of the sample remained after the test.

Floating Strength Determination
At the zero-time point, the sample was able to float on top of the acidic media and generate a buoyancy force of 1.5 mN. During the swelling and erosion of the capsule shell, the buoyancy decreased to 1.2 mN. Then, it started to increase and entered a plateau phase at around 2 mN until the API was completely released (Figure 10).

Dissolution Test after Long-Term and Accelerated Storage Conditions
Three parallel dissolution studies were performed to prove appropriate drug liberation after 3 months and 2 years, respectively. The dissolution profiles of the samples were compared by similarity and difference factors ( Figure S4 and Table S1). The difference factor was less than 5.00 in all cases, and the similarity factor was greater than 50.00.

Floating Strength Determination
At the zero-time point, the sample was able to float on top of the acidic media and generate a buoyancy force of 1.5 mN. During the swelling and erosion of the capsule shell, the buoyancy decreased to 1.2 mN. Then, it started to increase and entered a plateau phase at around 2 mN until the API was completely released (Figure 10).

Water Uptake and Matrix Erosion Studies
During dissolution, the mass of the matrix decreased continuously, as shown in Figure S3. A summary of the erosion and water uptake data is shown in Figure 9. The matrix did not show swelling, and the sum of remaining matrix and the water content was constant. Only 30% of the sample remained after the test.

Floating Strength Determination
At the zero-time point, the sample was able to float on top of the acidic media and generate a buoyancy force of 1.5 mN. During the swelling and erosion of the capsule shell, the buoyancy decreased to 1.2 mN. Then, it started to increase and entered a plateau phase at around 2 mN until the API was completely released (Figure 10).

Dissolution Test after Long-Term and Accelerated Storage Conditions
Three parallel dissolution studies were performed to prove appropriate drug liberation after 3 months and 2 years, respectively. The dissolution profiles of the samples were compared by similarity and difference factors ( Figure S4 and Table S1). The difference factor was less than 5.00 in all cases, and the similarity factor was greater than 50.00.

Dissolution Test after Long-Term and Accelerated Storage Conditions
Three parallel dissolution studies were performed to prove appropriate drug liberation after 3 months and 2 years, respectively. The dissolution profiles of the samples were compared by similarity and difference factors ( Figure S4 and Table S1). The difference factor was less than 5.00 in all cases, and the similarity factor was greater than 50.00.

In Vivo Study
The solutions containing verapamil-HCl (50 mg) and the solid gastroretentive foam were orally administered to the animals in the bioavailability test. The verapamil plasma concentrations were compared and are shown in Figure 11.

In Vivo Study
The solutions containing verapamil-HCl (50 mg) and the solid gastroretentive foam were orally administered to the animals in the bioavailability test. The verapamil plasma concentrations were compared and are shown in Figure 11. Gastroscopy was performed, and gastric retention after two hours was proven (Figure 12a). After 4 h, the capsule was eliminated from the stomach (Figure 12b). The main pharmacokinetic properties are presented in Table 5. The capsule reached the maximum concentration in plasma 4 h after administration, while the solution reached this level earlier, after 0.5 h. In the case of the foam, the plasma concentration was above 50 ng/mL for 5 times longer than in the case of the solution, while the initial concentration was just 2.5 times higher. The relative bioavailability was 99.3%. Gastroscopy was performed, and gastric retention after two hours was proven (Figure 12a). After 4 h, the capsule was eliminated from the stomach (Figure 12b).

In Vivo Study
The solutions containing verapamil-HCl (50 mg) and the solid gastroretentive foam were orally administered to the animals in the bioavailability test. The verapamil plasma concentrations were compared and are shown in Figure 11. Gastroscopy was performed, and gastric retention after two hours was proven (Figure 12a). After 4 h, the capsule was eliminated from the stomach (Figure 12b). The main pharmacokinetic properties are presented in Table 5. The capsule reached the maximum concentration in plasma 4 h after administration, while the solution reached this level earlier, after 0.5 h. In the case of the foam, the plasma concentration was above 50 ng/mL for 5 times longer than in the case of the solution, while the initial concentration was just 2.5 times higher. The relative bioavailability was 99.3%. The main pharmacokinetic properties are presented in Table 5. The capsule reached the maximum concentration in plasma 4 h after administration, while the solution reached this level earlier, after 0.5 h. In the case of the foam, the plasma concentration was above 50 ng/mL for 5 times longer than in the case of the solution, while the initial concentration was just 2.5 times higher. The relative bioavailability was 99.3%.

Discussion
In the present study, our aim was to increase the bioavailability of verapamil-HCl by applying a new solid foam formulation based on melt foaming. Verapamil formulations must be administered frequently, but this can be reduced by the use of GR formulations [8,9,22]. From a technological point of view, the thermal stability of verapamil is also favorable for the formulation using technology developed by our research team previously [33]. During the production of verapamil solid foam, the foam cell temperature was higher by 2 • C than that previously described with the BaSO 4 composition due to the higher melting point of the verapamil mixture [3,24]. Three compositions were produced with 15% verapamil-HCl, containing 80-120 mg of verapamil per capsule, which corresponds to the active substance content available in the literature and to the marketed preparations [20,34,35]. An increase in the stearic acid content from 10% (V1) to 12.5% (V2) helped to form the foam structure, as described previously by Vasvári et al. [3], but it is important to mention that the excessive stearic acid content reduced foaming. For V3, the increased density could be the reason for the changes in the wetting of the API particles by the molten matrix [36][37][38]. The SEM images support the altered solid particle location. In the case of V3, verapamil crystals were found in the inner surface of the cavity, while in the cases of V1 and V2, they were dispersed in the matrix and localization in the cavities was not specific. The foaming efficacy does not depend on only the hydrophilicity of the matrix, and according to our experiences, the particle size of the solid phase also significantly affects the degree of foaming. Previously, we successfully produced a lower-density product using smaller drug particles (314 nm ± 115) of BaSO4 with the abovementioned technology [24], and the use of verapamil particles with an average particle size (13.4 µm ± 11.2) resulted in a higher density. The drug dissolution test showed prolonged drug release in all compositions. Compared with marketed preparations, the drug release was slower than from Isoptin SR but showed similar kinetics to Calaptin SR [22,39]. During validation, three parallel productions were performed and compared, and the validation results comply with FDA and GMP regulations [40]. The micro-CT scans showed a high porosity, a homogenous distribution of bubbles in the matrix, and a monodispersed cavity size distribution. In general, open pores were shown to accelerate the rate of dissolution. Our capsule has few open pores, so their presence does not contribute significantly to the erosion of the preparation [41]. The texture analysis during the dissolution showed that the samples remained hard until 300 min. At the end of the dissolution process, 30% of the initial weight was still present, and the matrix became plastic and was easily removed by the grinding or churning motions of the stomach [2,10]. The PEG and API dissolved from the matrix, while the SA remained undissolved during the dissolution test. The composition did not show the ability to swell. In a dissolution study performed after storage under accelerated conditions, the formulation was compared to the initial formulation with factors f1 and f2, and the results met expectations. A similar result was observed for capsules stored under extended 2-year standard conditions. The results of the dissolution studies showed that no major changes occurred in the verapamil-containing solid foam matrices which influence the drug dissolution. The in vivo pharmacokinetics study proved that the GR system is able to significantly change the pharmacokinetic parameters of verapamil. For the verapamil solution, the in vivo plasma concentration of verapamil was correlated with the published conventional verapamil-HCl pill, Staveran [9]. In this study, verapamil solution was used to reveal the absorption properties and pharmacokinetics of pure verapamil from the GI tract and to avoid the influence of excipients or the liberation of verapamil from a formulation. By comparing the results of the solid foam capsule with a retard pill, Isoptin SR, the t max shift was observed. Instead of a sharp peak, a plateau was observed, and plasma levels could be further maintained in the therapeutic range [22], reaching a relative bioavailability of 99.3%. The verapamil plasma levels correlated well with the residency of the capsule in the stomach, as shown by the gastroscopy images. Even if the stomach of the animal was empty and showed fast motility, the capsule could be detected 2 h after administration. The dissolution results of verapamil foams are reflected in the in vivo data. The in vitro-in vivo correlation is shown in Figure S5, and in vivo absorption was predicted by the in vitro tests. With the application of high porosity verapamil foam, the gastroretentive purposes of controlled release and stomach retention were achieved simultaneously [42]. In terms of our results, it can be said that our preparation is suitable for gastroretention and can reduce the frequency of administration of the preparation, thus achieving better adherence. Our results were confirmed by both in vitro and in vivo experiments.

Conclusions
In our current study we produced verapamil-loaded gastroretentive capsules by continuous production using a foaming device for the first time. After optimizing the apparent density of the product by considering the SA contents, production runs were validated. Micro-CT scans revealed a closed cell structure where the main fraction of the voids was smaller than 120 microns. The texture analysis results confirmed a hard structure, even after 5 h of dissolution. Despite continuous erosion of the PEG matrix, floating strengths of the samples remained stable during dissolution. Our studies confirm that the first-order drug release was preserved, even after 2 years of storage. The in vivo pharmacokinetic study verified the prolonged release of the API from the matrix. The maximum drug plasma concentration was reached after 4 h of administration. Compared to the verapamil oral solution, a relative bioavailability of 99.3% was reached. Summarizing our results, we can state that our foaming device could be used successfully to produce lowdensity molded capsules with in vivo gastroretentive properties by continuous operation.

Supplementary Materials:
The following supporting information can be downloaded from https: //www.mdpi.com/article/10.3390/pharmaceutics14020350/s1. Figure S1: Weight validation of the V2 product. The measured weight, average weight, and standard deviations are presented; Figure S2: The dissolution profiles of the batches and V2 preformulation; Figure S3: Micro-CT scans of matrix erosion during dissolution; Figure S4: Three parallel dissolution studies of fresh samples after 3 months in a climate chamber and after 2 years at room temperature. Figure S5: In vitro-in vivo correlation of formulation V2; Table S1: The dissolution profile of the stability test was compared by similarity and difference factors.