**Pharmacokinetic Profile and Anti-Adhesive E**ff**ect of Oxaliplatin-PLGA Microparticle-Loaded Hydrogels in Rats for Colorectal Cancer Treatment**

**Sharif Md Abuzar 1,2, Jun-Hyun Ahn 1,2, Kyung Su Park 3, Eun Jung Park 4,\*, Seung Hyuk Baik 4,\* and Sung-Joo Hwang 1,2,\***


Received: 14 June 2019; Accepted: 1 August 2019; Published: 5 August 2019

**Abstract:** Colorectal cancer (CRC) is one of the most malignant and fatal cancers worldwide. Although cytoreductive surgery combined with chemotherapy is considered a promising therapy, peritoneal adhesion causes further complications after surgery. In this study, oxaliplatin-loaded Poly-(d,l-lactide-co-glycolide) (PLGA) microparticles were prepared using a double emulsion method and loaded into hyaluronic acid (HA)- and carboxymethyl cellulose sodium (CMCNa)-based cross-linked (HC) hydrogels. From characterization and evaluation study PLGA microparticles showed smaller particle size with higher entrapment efficiency, approximately 1100.4 ± 257.7 nm and 77.9 ± 2.8%, respectively. In addition, microparticle-loaded hydrogels showed more sustained drug release compared to the unloaded microparticles. Moreover, in an in vivo pharmacokinetic study after intraperitoneal administration in rats, a significant improvement in the bioavailability and the mean residence time of the microparticle-loaded hydrogels was observed. In HC21 hydrogels, AUC0–48h, Cmax, and Tmax were 16012.12 ± 188.75 ng·h/mL, 528.75 ± 144.50 ng/mL, and 1.5 h, respectively. Furthermore, experimental observation revealed that the hydrogel samples effectively protected injured tissues from peritoneal adhesion. Therefore, the results of the current pharmacokinetic study together with our previous report of the in vivo anti-adhesion efficacy of HC hydrogels demonstrated that the PLGA microparticle-loaded hydrogels offer novel therapeutic strategy for CRC treatment.

**Keywords:** oxaliplatin; PLGA; hydrogel; intra-abdominal anti-adhesion barrier; colorectal cancer

#### **1. Introduction**

Colorectal cancer (CRC) is one of the most malignant cancers worldwide. Although the incidence of CRC is low, the number of new cases that are malignant and fatal is still the highest among both men and women. In addition, peritoneal carcinomatosis is one of the numerous manifestations of CRC identified at the first diagnosis in more than 10% of CRC patients, and it is extremely fatal, with median survival of approximately 6 months [1,2]. The treatment strategies for both CRC and peritoneal carcinomatosis patients depend on the type and stage, and optimal therapy comprises cytoreductive surgery combined with chemotherapy [3].

Peritoneal adhesion is one of the most common postoperative complications associated with cytoreductive surgery. Previous studies indicated their occurrence in more than 50% cases, and they have even higher recurrence rates (85–93%) [4,5]. As cytoreductive surgery and intraperitoneal chemotherapy are closely related to the completeness of cytoreduction, problems associated with cytoreductive surgery are still a big challenge in medical science. Therefore, hydrogel-based anti-adhesion barriers came into the spotlight to reduce adhesion by mechanical separation of injured tissue surfaces during peritoneal repair after surgery [6].

Oxaliplatin (Figure 1) is a third-generation, platinum-based, systemic chemotherapeutic agent for CRC [7], and it is expected to be similarly effective in peritoneal carcinomatosis. Therefore, it is currently being used as a part of the standard chemotherapy regimen, FOLFOX (oxaliplatin with 5-fluorouracil and leucovorin) [8], for the clinical treatment of metastatic CRC [9]. However, a recent study conducted in a murine model confirmed that intraperitoneally administered oxaliplatin enhanced peritoneal tissue concentration while reducing its systemic absorption, suggesting a possible decrease in toxicity associated with systemic chemotherapy [10]. Therefore, the new approach comprising cytoreductive surgery followed by intraperitoneal oxaliplatin delivery has resulted in significant improvement in the disease states of CRC and peritoneal carcinomatosis [11].

Biodegradable microparticles have been widely investigated for the controlled delivery of chemotherapeutic agents [12,13]. In addition, neither an initial burst release, nor the presence of a lag phase are desirable for the chemotherapeutic agents as it can be associated with adverse effects. Poly-(d,l-lactide-*co*-glycolide) (PLGA) is an unique copolymer approved by the regulatory agencies for the manufacturing of bioresorbable surgical sutures. Owing to its biodegradability and biocompatibility, PLGA become the first line materials for the production of injectable microparticle based controlled release system [14]. A recent report of metformin/irinotecan-loaded nanoparticles have shown an initial 2 h burst release, however, increase the drug retention time in tumor and increase drug circulation [15]. Moreover, alternative reports of paclitaxel-loaded PLGA microparticles have been evaluated for cancer therapy. These models revealed protection of the therapeutic payload from premature burst release, and enable the sustained release of paclitaxel [16,17].

**Figure 1.** Chemical structures of Poly-(d,l-lactide-*co*-glycolide) (PLGA) microparticles and hyaluronic acid (HA) and carboxymethyl cellulose sodium (CMCNa)-based cross-linked (HC) hydrogel components (**A**) HA; (**B**) CMCNa; (**C**) HC hydrogel; (**D**) Oxaliplatin; and (**E**) PLGA.

Recently we succeeded in synthesizing novel hyaluronic acid (HA) and carboxymethyl cellulose sodium (CMCNa)-based cross-linked (HC) hydrogels loaded with oxaliplatin [18]. These novel HC hydrogels significantly prevented intraperitoneal adhesion and offered the highest anti-adhesion barrier in an in vivo rat model. However, pharmacokinetic evaluation of oxaliplatin-loaded HC hydrogels was not carried out. Several reports of microparticles-loaded hydrogels in various drug delivery system showed controlled drug release for over a prolonged period [19,20]. Therefore, to provide an adequate anti-adhesion barrier effect after cytoreductive surgery and deliver intraperitoneal chemotherapy, oxaliplatin-loaded PLGA microparticles were prepared, characterized, and loaded into HC hydrogels. In addition, an in vitro oxaliplatin release study of PLGA microparticle-loaded HC hydrogels in comparison with that of commercially available Guardix-Sol® hydrogel was carried out. Furthermore, in vivo pharmacokinetic analysis was carried out to demonstrate the efficacy of intraperitoneal chemotherapy with oxaliplatin-PLGA microparticles loaded into HC hydrogels.

#### **2. Materials and Methods**

#### *2.1. Materials and Animals*

Oxaliplatin was a gift from Boryung Pharm (Ansan, Korea). Resomer® RG 502 H (Poly-(d,l-lactide*co*-glycolide)) PLGA, molecular weight (MW) approximately 7000–17,000, was purchased from Evonic Ind., (Darmstadt, Germany). Hyaluronic acid (HA, MW 1000 kDa) was a gift from Huons (Seongnam, Korea). Poly-(vinyl alcohol) (PVA, MW approximately 89,000–98,000), carboxymethyl cellulose sodium (CMCNa, MW approximately 700 kDa), and adipic acid dihydrazide (ADH) were obtained from Sigma Aldrich (St. Louis, MO, USA). 1-Ethyl-3-(3-(dimethylaminopropyl) carbodiimide (EDC) was purchased from Tokyo Chemical (Tokyo, Japan). All other chemicals were of reagent grade, and Milli-Q® water (Millipore1, Molsheim, France) was used throughout the study.

Male Sprague-Dawley (SD) rats were purchased from YoungBio (Seongnam, Republic of Korea). The animals were housed in a semi-specific pathogen free facility using standard cages at 19 ± 1 ◦C and 50 ± 5% relative humidity, with a 12-h light-dark cycle. The rats were fed a standard diet, provided with purified water, and allowed to move freely. All experiments were approved by the Institutional Animal Care and Use Committee (IACUC-201809-788-01) at Yonsei University, Seoul, Korea, and were performed according to IACUC guidelines.

#### *2.2. Preparation of Oxaliplatin-PLGA Microparticles*

Oxaliplatin-PLGA microparticles were prepared using a double emulsion method [21]. Briefly, oxaliplatin (20 mg) was dissolved in 5 mL of deionized water (first aqueous solution, W1) containing 0.5% (*w*/*v*) PVA. The resulting oxaliplatin containing W1 was added dropwise to 12.5 mL DCM, containing 1 g of PLGA 502 H, using a homogenizer ULTRA-TURRAX® (IKA-WERKE GMBH & Co. KG, Staufen, Germany) at 12,000 rpm for 2 min. This primary emulsion (W1/O) was slowly added to 100 mL of 1% (*w*/*v*) PVA solution (second aqueous solution, W2) and emulsified at 400 rpm using magnetic stirrer for 3 h while the DCM was allowed to evaporate completely under vacuum. The resultant microparticles were collected, washed using distilled water, centrifuged, and freeze-dried using an ilShinBioBase (Seoul, Korea) freeze-drier under vacuum (5 mTorr). Samples were pre-frozen for 2 h at −60 ± 1.0 ◦C prior to final drying at −88 ± 1.0 ◦C for 2 days. The collected freeze-dried microparticles were stored at 4 ◦C and further evaluated by scanning electronic microscopy (SEM) to characterize the morphological features.

#### *2.3. Synthesis of HA-CMCNa Cross-Linked Hydrogels*

Hydrogels were synthesized using variable ratios of HA and CMCNa (Figure 1) by chemical cross-linking, as described by our previous study [18]. Briefly, HA and CMCNa at ratios of 1:2 and 2:1 (*w*/*w*) were dissolved in deionized water with continuous stirring. After complete dissolution, 2 mM/L of ADH was added to the mixture, and the pH was adjusted to 4.75 ± 0.05 using 0.1 M HCl. After vortex mixing for 1 min, 2 mM/L EDC was added, and the resulting mixture was stirred thoroughly at room temperature (25 ± 1.0 ◦C) for at least 12 h to allow for complete cross-linking. The reaction was maintained at pH 4.75 ± 0.05 by the addition of 0.1 M HCl. Finally, the cross-linking reaction was terminated by elevating the pH of the mixture to 7.0 ± 0.05 via the slow addition of 0.1 M NaOH. The hydrogel samples were purified as previously described by Luo et al. [22]. Briefly,

the hydrogel samples were dispensed into a dialysis bag (Mw cut-off 18,000 Da) previously treated with a 70% ethanol (EtOH) solution and dried. The bag was submerged and dialyzed for 12 h in 0.1 M NaCl followed by alternating solutions of 25% EtOH for 6 h and DW for 6 h. The remaining hydrogels in the dialysis bag were rinsed using EtOH, and then centrifuged at 2000 rpm for 10 min to remove the remaining ADH. Finally, the precipitated hydrogels (HC12 and HC21) were collected and stored until further experiments.

#### *2.4. Preparation of Oxaliplatin-PLGA Microparticle-Loaded Hydrogel*

Guardix-Sol® is a commercial anti-adhesive hydrogel used for the prevention of post-operative adhesion. Above, we synthesized HC12 and HC21 hydrogel composites with variable weight ratios of HA and CMCNa. Next, oxaliplatin-PLGA microparticles were dispersed gently in the hydrogel (HC12, HC21, and Guardix-Sol®). Briefly, precisely weighed oxaliplatin-PLGA microparticles were mixed with 10 mL of hydrogels at room temperature (25 ± 1.0 ◦C) with continuous stirring for 30 min. The final concentration of oxaliplatin was 2 mg/ mL of hydrogel. After complete dissolution was visually observed, samples were evaluated for in vitro and in vivo release of oxaliplatin.

#### *2.5. Characterization of Oxaliplatin-PLGA Microparticles*

#### 2.5.1. Morphology

The morphology of oxaliplatin-PLGA microparticles was evaluated using SEM (JSM-6700F, JEOL, Tokyo, Japan). Briefly, a small amount of powder was sprinkled onto double-sided adhesive tape attached to an aluminum stub and was sputter-coated with gold under vacuum. Photographs were taken at 5× magnification with an accelerating voltage of 1–5 kV to reveal the surface characteristics of the particles.

#### 2.5.2. Particle Size Analysis

The mean particle size and distribution of oxaliplatin-PLGA microparticles were analyzed by dynamic light scattering (DLS) using an electrophoretic light scattering spectrophotometer (ELS-Z, Otsuka Electronics, Hirakata, Japan).

#### 2.5.3. Rheological Measurements

The rheological measurements for HC12, HC21, and Guardix-Sol® hydrogels were performed using a Brookfield rheometer (Brookfield Digital Rheometer Model DV-III, DV3T™ Rheometer, Middleboro, MA, USA) equipped with a Peltier system for temperature control. Precisely, about 0.5 g of sample was applied to the plate and allowed to equilibrate. Measurements were performed at 37 <sup>±</sup> 0.5 ◦C with shear rates ranging from 200–500 s−1. Before each measurement, the samples were allowed to rest for 5 min at 37 ± 0.5 ◦C. Results were analyzed with Brookfield software (Firmware version 1.2.2-9).

#### 2.5.4. Encapsulation Efficiency

Freeze-dried oxaliplatin-PLGA microparticles (eq. 2 mg of oxaliplatin) were dissolved in 5 mL of DCM. After 5 min of vortexing, particles were allowed to dissolve properly, and the tube was gently swung in incubator at 37 ± 0.5 ◦C for 1 h. Then, 5 mL of DW was added to the tube and vortexed vigorously for 1 h. The suspension was centrifuged at 12,000 rpm for 5 min to precipitate PLGA. The upper aqueous phase was collected, and the concentration of oxaliplatin was determined by HPLC. Encapsulation efficiency (EE %) was calculated using the following Equation (1):

$$\text{EE } \%= \text{M}\_{\text{actual}\,\text{coal}\,\text{aliplatin}} / \text{M}\_{\text{theoretical}\,\text{coal}\,\text{aliplatin}} \times 100\tag{1}$$

#### *2.6. In Vitro Oxaliplatin Release from PLGA Microparticles Loaded into Hydrogels*

To evaluate the in vitro oxaliplatin release rate from oxaliplatin-PLGA microparticles, an amount equivalent to 2 mg of oxaliplatin was weighed and suspended in 1 mL of HPLC grade water. The suspension was transferred into dialysis bags (MWCO 14,000) and then submerged in 20 mL of double distilled water in capped 50 mL Falcon® tube. Samples were shaken at 30 rpm at the predetermined time points (1, 2, 3, 4, 6, 8, and 12 h); 1 mL of sample was collected from the medium and 1 mL of pre-warmed medium was immediately added to the tubes. Experiments were performed in triplicate (n = 3), and the collected samples were analyzed with HPLC after necessary dilution. The release profile was expressed as the ratio of cumulative oxaliplatin release to initial oxaliplatin loading versus time, using the following Equation (2):

### Percent of oxaliplatin release = Mt/M0 × 100 (2)

For the in vitro oxaliplatin release from the PLGA microparticle-loaded hydrogels (from Section 2.4.), a similar method was used. Briefly, 1 mL PLGA microparticle-loaded hydrogel (HC12, HC21, or Guardix-Sol®) was transferred to one end-closed dialysis bags (MWCO 14,000). The other end of the bag was closed properly, and the samples were submerged in 20 mL of double distilled water in capped 50 mL Falcon® tube. Samples were shaken at 30 rpm at similar time points (1, 2, 3, 4, 6, 8, and 12 h). Subsequently, 1 mL of sample was collected from the medium and 1 mL of pre-warmed medium was immediately added to the tubes. Experiments were performed in triplicate (n = 3), and the collected samples were analyzed with HPLC after necessary dilution with the mobile phase. The release profile was expressed as cumulative oxaliplatin release calculated using Equation (2).

#### *2.7. In Vivo Oxaliplatin Release in SD Rat's Intraperitoneal Cavity*

In vivo oxaliplatin release from the prepared hydrogels was evaluated in the intraperitoneal cavity of SD rats, and oxaliplatin solution was also evaluated for comparison. Briefly, a total of fifteen male SD rats, aged 4–6 weeks, were randomly divided into three groups (n = 5, per group) and were anesthetized with isoflurane. Intra-abdominal adhesions were induced to mimic the postoperative surgical conditions, as described previously [23]. The peritoneum was exposed by a 5-cm ventral midline incision. The left abdominal sidewall was scraped with a 1 cm<sup>2</sup> piece of 100-grit sandpaper 200 times. In group 1, oxaliplatin solution was introduced, whereas, group 2 and 3 were exposed to oxaliplatin powder and oxaliplatin-PLGA microparticles, loaded into HC21 hydrogel, respectively. All the rats from each group were administered oxaliplatin in the form of a solution or loaded into hydrogel at a dose level of 5 mg/kg body weight [24]. The abdominal layers and skin incision were then completely closed, and each rat was kept separately in single cases. All surgeries were performed by the same individual. At predetermined time interval (0.5, 1, 1.5, 2, 3, 5, 8, 14, 24, and 48 h), approximately 1-mL blood samples were collected from the conjunctiva using a capillary tube and kept in an ice bath.

Blood samples collected from the rat's conjunctiva were centrifuged at 10,000 rpm (9425× *g*) for 10 min at 4 ◦C. The supernatant plasma was obtained and stored at −80 ◦C until analyzed. The frozen plasma samples were thawed, and approximately 200 ± 5 mg was weighed. The samples were oxidized into metals and organic materials by treating with 6 mL nitric acid using a microwave sample pre-treatment machine equipped with platinum (Pt) sensor (Microwave Reaction System, Multi-wave PRO, Anton Paar, Graz, Austria). Samples were heated at predefined temperature and pressure (200 ± 0.5 ◦C and 4.0 MPa) for 1 h and diluted up to 30 mL with deionized water, and the concentrations of Pt were analyzed using ICP-MS (NexION 300 D, PerkinElmer, Waltham, MA, USA). Pharmacokinetic parameters AUC0–48h (area under curve), Cmax (peak concentration), and Tmax (time to peak concentration) were calculated using noncompartmental analysis.

After the last blood sample collection at 48 h, rats were housed individually in standard case with a 12-h light-dark cycle. The rats were fed a standard diet, provided with purified water, and were allowed to move freely. After 10 days, they were sacrificed, and the peritoneum was opened. For all rats, adhesion type and extent were assessed, and photographs were taken.

#### *2.8. High-Performance Liquid Chromatography (HPLC) Analysis*

In vitro oxaliplatin release and % EE were determined by using HPLC Agilent 1200 Infinity Series HPLC system (Agilent Technologies, Waldbronn, Germany). Briefly, XTerra™ RPC18 column (particle size 5 μm, inside diameter 4.6 mm, and length 250 mm; Waters Corporation, Milford, MA, USA) at 210 nm using an HPLC-UV spectrometer (Agilent 1290 infinity). The mobile phase was a 20:80 mixture of ACN and deionized water and was set to a flow rate of 0.8 mL/min (Model 1260 Quat Pump VL). The samples were diluted as necessary, and 20 μL of each sample was injected using an autosampler (Model 1260 ALS).2.9. Statistical Analysis

In vitro percent (%) cumulative oxaliplatin release and in vivo pharmacokinetic study data are expressed as the means ± standard deviations. The *t*-test or two-sided RM ANOVA and Bonferroni test were applied to the analyses of the differences between the groups. A *p* value < 0.05 was considered as a statistically significant difference.

#### **3. Results and Discussion**

#### *3.1. Characterization of PLGA Microparticles*

PLGA-based microparticles were prepared using the double emulsion method. The microparticles were characterized for size, morphology, and encapsulation efficiency. As shown in Table 1, oxaliplatin-loaded PLGA microparticles had a particle size of more than 1 μm diameter, with a small standard deviation. Encapsulation efficiency was 77.9 ± 2.8%, which is excellent for a hydrophilic drug like oxaliplatin. Although the encapsulation of hydrophilic drugs in PLGA microparticles is challenging due to the partitioning of weakly associated drugs from the oil phases to the external water phase, the double emulsion technique is the most commonly used method for encapsulating hydrophilic drugs. In addition, the SEM image shown in Figure 2A reveals the morphology of the microparticles. Uniform size and spherical particles were observed, which is in agreement with the particle size results. To fulfill the aims of this study, PLGA microparticles were further loaded into HC hydrogels and the in vitro release and in vivo pharmacokinetic studies were conducted.


**Table 1.** Compositions and characterizations of Poly-(d,l-lactide-*co*-glycolide) (PLGA) microparticles and hyaluronic acid (HA) and carboxymethyl cellulose sodium (CMCNa)-based cross-linked (HC) hydrogels.

#### *3.2. Preparation and Characterization of HC Hydrogels*

Cross-linked HC hydrogels were prepared using variable weight ratios of HA and CMCNa. HC hydrogel synthesis was carried out based on our previous description [18]. In the presence of ADH (as a nucleophile), EDC (a water-soluble carbodiimide) linked both HA and CMCNa molecules with its amine group. EDC is not incorporated into the final product, but converts into a non-toxic water-soluble urea derivative, which is removed by dialysis. Moreover, CH2COO−anions provided by CMCNa react with H<sup>+</sup> ions, leading to carboxyl group formation and facilitation of cross-linking. Therefore, the reaction in the presence of EDC is pH-dependent and was performed at pH 4.75 ± 0.05. To evaluate the rheology of the HC hydrogel along with the commercial product Guardix-Sol®, a Brookfield Digital Rheometer was used. All measurements were performed at 37 ± 0.5 ◦C, with shear rates ranging from 200 to 500 s−1. All hydrogels followed non-Newtonian shear thinning (pseudo) plastic flow behavior (Figure 2B). Decreasing viscosity with increasing shear rates was observed, which demonstrated viscosity-dependent shear rates [25].

In addition, variable ratios of the composition (HA and CMCNa) alter the viscosity of the HC hydrogels. The order could be written as follows: HC21 > HC12, as markedly increased viscosity was observed with HA. The highest viscosity was observed with the HC21 formulation, whereas the lowest viscosity was observed with HC12. The viscosity of Guardix-Sol® was measured for comparison with those of the synthesized HC hydrogels, and it showed moderate viscosity. The rheology results correlated with the mechanical characteristics of the polymer, in which viscosity scaled linearly with molecular weight. The mechanical properties (entanglement phenomenon) of the polymers increased with increasing molecular weight. Finally, the hydrogels were loaded with oxaliplatin-PLGA microparticles and their in vitro and in vivo characteristics were evaluated.

**Figure 2.** (**A**) SEM image of oxaliplatin-PLGA microparticles (at X5.000 magnification); (B) rheology properties of hydrogel samples (-) HC21, (-) Guardix-Sol®, and () HC12 (measurements were performed at 37 ± 0.5 ◦C).

#### *3.3. In Vitro Oxaliplatin Release Study*

An in vitro oxaliplatin release study was conducted on oxaliplatin-PLGA microparticles generated using the double emulsion method. In addition, oxaliplatin-PLGA microparticles and oxaliplatin powder were dispersed into hydrogels (HC12, HC21, and Guardix-Sol®), and an in vitro oxaliplatin release study was conducted. Plots of cumulative oxaliplatin release (%) versus time (h) are shown in Figure 3A–C. In vitro release profile for oxaliplatin-PLGA microparticles (red closed circle; (-) showed immediate release. With an initial burst release, about 63.2% oxaliplatin was released at 4 h. Oxaliplatin release from PLGA microparticles is a combination of diffusion and bioerosion mechanism [26]. During the diffusion stage, oxaliplatin release occurs by diffusion through aqueous channels. The release could also be triggered by the external pores on the surface or through leaching of the oxaliplatin at or near the surface, leading to an initial burst release. Besides, the standard deviation values of the results of the in vitro release study for the hydrogels were low, which indicate excellent reproducibility of the release behavior. However, when oxaliplatin-PLGA microparticles were loaded into three different hydrogel systems, HC12, HC21, and Guardix-Sol®, slight initial burst release was observed. Although the rate of burst release was almost the same for the three systems, HC12, with the lowest viscosity, showed the maximum release. In contrast, the higher the viscosity of the hydrogel system, the lower the burst and sustained release observed. This may be due to low leaching of the drugs and high protection of the microparticles by the hydrogels.

Table 2 demonstrated the kinetic models adopted for evaluation. Where Mt was the cumulative drug released at time t and M was the initial drug present in the PLGA microparticles-loaded hydrogels. k1, kH, and kKP are the first order, Higuchi, and Korsmeyer-Peppas release constant, respectively. Mt/M<sup>∞</sup> was the fraction of drug released at time t, and n was the diffusional release exponent symbolic of the release mechanism. The rate of oxaliplatin release from the HC12, Guardix-Sol® and HC21-loaded PLGA microparticles were slower and constant compared with the oxaliplatin powder-loaded hydrogels, as indicated by the KH values of 18.9686, 17.5626, and 20.6559% h1/2, respectively. Release constants were determined using the slope of the appropriate plots, and the regression coefficient (R2) was obtained through linear regression analysis (Table 2). The coefficient of determination (R2) was used as an indicator of curve fit for each of the considered models [27]. The regression coefficients R2, from the Korsmeyer-Peppas plots were 0.9342, 0.9554, and 0.9366 for HC12, Guardix-Sol® and HC21, respectively, thus the log of cumulative drug release was proportional to the log of time. The best linear fits were observed for the PLGA microparticles-loaded hydrogels using both the Higuchi and Korsmeyer-Peppas models, suggesting that oxaliplatin release from the PLGA microparticles-loaded hydrogels were diffusive process. Oxaliplatin gradually dissolved into the fluid within the swelled hydrogels, then slowly diffused from polymeric networks of the hydrogels.

**Table 2.** Release rate constants and regression coefficient R<sup>2</sup> obtained from drug release profile based on kinetic equations.


<sup>a</sup> Higuchi model: (Mt = kH × t <sup>1</sup>/2); <sup>b</sup> First—Order model: (ln Mt = lnM + k1 × t); <sup>c</sup> Korsmeyer—Peppas model: (Mt/M∞ = kKP × t n).

**Figure 3.** *Cont.*

**Figure 3.** In vitro oxaliplatin release from PLGA microparticles (red, -). Oxaliplatin release profile from oxaliplatin powder and oxaliplatin-PLGA microparticles-loaded in (**A**) HC12; (**B**) Guardix-Sol®; and (**C**) HC21 hydrogel. Mean ± S.D. (n = 3).

Although hydrogels are swellable materials with high water content, which readily allow the release or leaching of hydrophilic drugs through the channels, high viscosity might hinder this process. Therefore, release may be completed through bioerosion. In this study, three different hydrogel systems—HC12, HC21, and Guardix-Sol®, were loaded with oxaliplatin powder and oxaliplatin-PLGA microparticles. The HC12 hydrogel system (low viscosity) was loaded with oxaliplatin powder and more than 90% oxaliplatin release occurred within 3 h, followed by saturated release behavior up to 12 h. However, oxaliplatin-PLGA microparticles loaded into the HC12 hydrogel showed a slower release pattern than that of the oxaliplatin powder. Approximately 50% oxaliplatin was released after 4 h, followed by 64.2% release up to the end of the experiment (*p* < 0.05). Due to the less hydrophilicity of the PLGA copolymers, they absorb less water, and subsequently degrade more slowly. Therefore, drug release from the PLGA microparticles was delayed. In the case of Guardix-Sol®, with moderate viscosity, a sustained release profile was observed for the oxaliplatin-PLGA microparticles. During the first 2 h of the release study, less than 30% of oxaliplatin was released; however, 59.25% of oxaliplatin was released after the same amount of time when loaded with oxaliplatin powder (*p* < 0.05).

In addition, the HC21 hydrogel system (high viscosity) showed a similar sustained release profile when loaded with oxaliplatin-PLGA microparticles. In all three systems, more rapid oxaliplatin release was observed when oxaliplatin was loaded as a powder compared to that when loaded as microparticles. The release rate was faster than those of both oxaliplatin-PLGA microparticles and oxaliplatin-PLGA microparticles-loaded hydrogels. Additionally, both Guardix-Sol® and HC21 showed similar sustained release patterns up to 12 h. Although PLGA microparticles were generated using the well-known double emulsion method, further studies are necessary to establish process parameters and extensive characterization of the particles.

#### *3.4. Pharmacokinetics in Rats*

The bioavailability of oxaliplatin-PLGA microparticles loaded into the HC21 hydrogel was evaluated in rats. Figure 4 shows the mean concentration–time profiles of oxaliplatin in rats after a single dose (5 mg/kg) of oxaliplatin solution, oxaliplatin-PLGA microparticles, and oxaliplatin powder loaded into the HC21 hydrogel. The samples were introduced through a midline incision to open the peritoneum immediately after introducing intra-abdominal adhesions by multiple scraping to mimic postoperative surgical conditions. Intraperitoneal absorption of oxaliplatin solution was obviously higher than that of the hydrogel loaded samples (oxaliplatin-PLGA microparticles and oxaliplatin powder). For oxaliplatin solution AUC0–48h, Cmax, and Tmax were 8181.51 ± 176.89 ng·h/mL, 2265.28 ± 192.51 ng/mL, and 1.0 h, respectively. In the case of oxaliplatin powder loaded into the HC21 hydrogel, AUC0–48h, Cmax, and Tmax were 16,571.37 ± 139.13 ng·h/mL, 690.63 ± 140.54 ng/mL, and 1.5 h, respectively. Intraperitoneal bioavailability of oxaliplatin was significantly increased by up to 2-fold in the HC21 hydrogel (Table 3). In particular, 1.9-fold higher bioavailability was

observed in the case of oxaliplatin-PLGA microparticles loaded into the HC21 hydrogel compared to that of the oxaliplatin solution. AUC0–48h, Cmax, and Tmax were 16,012.12 ± 188.75 ng·h/mL, 528.75 ± 144.50 ng/mL, and 1.5 h, respectively. Moreover, mean residence time (MRT) was increased by 2.7-fold for oxaliplatin-PLGA microparticles when loaded into the HC21 hydrogels (*p* < 0.05). When the oxaliplatin-PLGA microparticles were loaded into the HC21 hydrogel, oxaliplatin release rate decreased due to the multi-layered encapsulation, which was observed from the in vitro release study. Therefore, in the high, moderate, and low viscous hydrogels (HC21, Guardix-Sol®, and HC12, respectively), oxaliplatin release from the PLGA microparticles, followed by release from the hydrogels, was observed in a sustained manner (Figure 3). Furthermore, using sustained-release hydrogels allows a longer residence time of the component at the application site (intraperitoneal after cytoreductive surgery), which offers additional benefits through the prevention of intra-abdominal adhesion and improvement in the delivery of intraperitoneal oxaliplatin from the PLGA microparticles.

**Figure 4.** Plasma concentration-time profile of Pt in rats after intra-peritoneal administration of oxaliplatin powder and oxaliplatin-PLGA microparticles-loaded in HC21 hydrogel, and oxaliplatin solution. Data are expressed as the Mean ± standard deviation (n = 5). \*, *p* < 0.05 represents a significant difference, two-sided RM ANOVA and Bonferroni test.


Oxaliplatin-PLGA microparticle loaded HC21 528.75 ± 144.50 1.5 16012.12 ± 188.75 12.02 ± 0.41

**Table 3.** In vivo pharmacokinetic parameters after intra-peritoneal administration to rats (n = 5).

#### *3.5. Intraperitoneal Anti-Adhesion E*ff*ect*

An extensive investigation of the intraperitoneal anti-adhesion efficacy of HC hydrogels was conducted in our previous study, after introducing peritoneal injury to generate peritoneal adhesion [18]. HC hydrogel groups (treated) showed the highest anti-adhesion barrier compared to the control group (non-treated). In the present study, the pharmacokinetic parameters of the PLGA microparticles loaded into the HC21 hydrogel were evaluated and compared with those of oxaliplatin powder loaded into hydrogels. Artificial injury on the left abdominal sidewall was introduced to mimic surgical conditions and observe the anti-adhesion barrier effect of hydrogel samples followed by the intraperitoneal delivery of oxaliplatin. The method was adopted from experienced surgeons certainly feasible in human administration after cytoreductive surgery in colorectal cancer treatment. Intraperitoneal anti-adhesion effect was observed and recorded photographically. Figure 5 shows a photographical representation of adhesion characteristics of the study groups.

**Figure 5.** Photographical expression of in vivo anti-adhesion efficacy observed in rats after introducing intra-peritoneal injury. (**A**) oxaliplatin solution; (**B**) oxaliplatin powder-loaded HC21 hydrogel; and (**C**) oxaliplatin-PLGA microparticles-loaded HC21 hydrogel.

HC21 hydrogel samples exhibited an efficient anti-adhesion barrier effect towards the injury. Although hydrogels were not intended to cure the injury, long resident time hydrogels significantly prevented adhesion along with rapid recovery of the injury, possibly due to mechanical separation from the injury site [28]. As reported previously, the washing away or displacement of the oxaliplatin solution from the application site allows less barrier or protection to the injury site [29]. Therefore, dense adhesion on the abdominal wall was observed, and, although current study was not intended to be carried out extensively, adhesion scoring and extent (as disclosed in a previous report) and photographic representation demonstrated the superiority of the hydrogels in terms of the anti-adhesion barrier. In the hydrogel-loaded samples, no or minor adhesion was observed in the treated rat abdomen (Figure 5B,C). Although two different samples were evaluated after loaded in HC21 hydrogel both oxaliplatin powder and oxaliplatin-PLGA microparticles loaded hydrogels showed significant improvement as anti-adhesion barrier. Subsequently, both hydrogel systems showed rapid wound recovery effect may be due to mechanical separation- could be seen from the blood spot in Figure 5C. Instead, oxaliplatin solution presents no barrier to the injure site and dense abdominal adhesion was observed. The HC21 hydrogel-loaded sample showed enhanced anti-adhesion protection to the injury site, suggesting its potential therapeutic application in CRC patients who have undergone cytoreductive surgery. Although the current study evaluated the pharmacokinetic profile of PLGA microparticle-loaded HC hydrogels, a further pharmacodynamics study on a CRC model is required to evaluate the composition of the hydrogels.

#### **4. Conclusions**

Oxaliplatin-PLGA microparticles were loaded into hydrogels to improve intraperitoneal chemotherapy with an anti-adhesion barrier effect. In an in vitro release study, we observed sustained oxaliplatin release from PLGA microparticles loaded into hydrogels in comparison with the unloaded sample. Moreover, in the in vivo pharmacokinetics study on rats, PLGA microparticle-loaded hydrogels showed higher bioavailability compared to oxaliplatin solution. Furthermore, visual observations revealed that the HC hydrogels were effective as an intra-abdominal anti-adhesion barrier. Therefore, intraperitoneal delivery of PLGA microparticle-loaded hydrogels, which have an effective intra-abdominal anti-adhesion barrier, is expected to provide an optimized alternative therapy for CRC. Earlier reports of cytotoxicity study for PLGA (RG502H) microparticles exhibited non-cytotoxic at increasing concentration [30]. Additionally, both HA and CMCNa are considered as inert and non-cytotoxic. However, cytotoxicity profile for the novel oxaliplatin-PLGA microparticles-loaded HC hydrogels were yet to revealed. As the investigation undertaken in current study was preliminary, pharmacokinetic and anti-adhesion efficacy of this multi-complex drug delivery system, further studies

are required to address cytotoxicity, cell viability, toxicity in liver function, and efficacy in CRC model in vivo.

**Author Contributions:** Conceptualization, S.M.A., J.-H.A.; Analytical Support, S.M.A., K.S.P.; Methodology, J.-H.A., E.J.P.; Writing—Original Draft Preparation, S.M.A.; Writing—Review and Editing, J.-H.A., E.J.P., S.-J.H., S.H.B.; Supervision, S.H.B., S.-J.H.; Project Administrator, S.-J.H.

**Funding:** This study was supported by a grant from Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT & Future Planning, Republic of Korea (NRF-2017R1A2B2011520 and NRF-2019R1F1A1056350), and University-Centered Labs-2018R1A6A1A03023718).

**Acknowledgments:** The authors would like to acknowledge Yonsei Center for Research Facilities (YCRF, Seoul, Korea), Yonsei University for writing assistance.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Design and Evaluation of a Poly(Lactide-***co***-Glycolide)-Based In Situ Film-Forming System for Topical Delivery of Trolamine Salicylate**

#### **Yujin Kim 1, Moritz Beck-Broichsitter <sup>2</sup> and Ajay K. Banga 1,\***


Received: 5 June 2019; Accepted: 5 August 2019; Published: 12 August 2019

**Abstract:** Trolamine salicylate (TS) is a topical anti-inflammatory analgesic used to treat small joint pain. The topical route is preferred over the oral one owing to gastrointestinal side effects. In this study, a poly(lactide-*co*-glycolide) (PLGA)-based in situ bio-adhesive film-forming system for the transdermal delivery of TS was designed and evaluated. Therefore, varying amounts (0%, 5%, 10%, 20%, and 25% (*w*/*w*)) of PLGA (EXPANSORB® DLG 50-2A, 50-5A, 50-8A, and 75-5A), ethyl 2-cyanoacrylate, poly (ethylene glycol) 400, and 1% of TS were dissolved together in acetone to form the bio-adhesive polymeric solution. In vitro drug permeation studies were performed on a vertical Franz diffusion cell and dermatomed porcine ear skin to evaluate the distinct formulations. The bio-adhesive polymeric solutions were prepared successfully and formed a thin film upon application in situ. A significantly higher amount of TS was delivered from a formulation containing 20% PLGA (45 <sup>±</sup> 4 <sup>μ</sup>g/cm2) and compared to PLGA-free counterpart (0.6 <sup>±</sup> 0.2 <sup>μ</sup>g/cm2). Furthermore, the addition of PLGA to the polymer film facilitated an early onset of TS delivery across dermatomed porcine skin. The optimized formulation also enhanced the delivery of TS into and across the skin.

**Keywords:** NSAIDs; PLGA; polymeric film; topical drug delivery; trolamine salicylate

#### **1. Introduction**

Topical drug delivery has many advantages over systemic drug administration, such as minimizing side effects, bypassing first-pass metabolism, and ensuring better patient compliance [1]. However, dermal drug delivery is usually limited to certain drug molecules, owing to the well-known barrier function of the skin. The outermost layer of the skin, the stratum corneum, generally only allows permeation of molecules with a molecular weight <500 Da, a log*P* between 1 and 3, and a melting point of <250 ◦C [2].

Among the available topical formulations, non-steroidal anti-inflammatory drugs (NSAIDs) constitute approximately 18 world-wide marketed molecules (some examples include diclofenac, methyl salicylate, salicylic acid, and ketoprofen) [3]. As an example, topical salicylates are known to be absorbed by the dermal tissue and have been reported to be effective in relieving local pain [4]. Trolamine salicylate (TS), a derivative of salicylic acid offers many benefits over other topical analgesics, including a lack of distinct odor, low systemic absorption upon dermal or topical administration, and low skin irritation [5]. Although topical salicylates are widely used, compliance is often an issue, owing to the frequent dosing regimen (three to four times per day).

The development of topical formulations has made an important contribution to medical practice. To deliver active pharmaceutical ingredients into or through the skin, various types of formulations

are known, such as gels and emulsions. The type of vehicle chosen depends on the properties of the drug and the intended target area. In addition, the hydrophilicity and lipophilicity of the drug must be compatible with the vehicle. "Conventional" vehicles are inelegant and have the drawback of poor control of both the amount of drug applied and the area of skin exposed [6]. As a result, the use of topical formulations to deliver molecules to systemic circulation is less than ideal, thus resulting in substantial variability in the extent and duration of drug effects [7]. To overcome the shortcomings of skin drug delivery, the former Alza Corporation pioneered a transdermal patch in which system design and explicit control of the surface area led to improved passive drug delivery to the systemic circulation at a predetermined rate. Transdermal patches are recommended to be applied to a flat surface area to ensure adhesion to skin over a prolonged period of time. However, in the case of arthritis and other joint pain, the use of a patch is not suitable because of the uneven surface at the application sites. Pain associated with arthritis requires drug application at the pain region, which is usually difficult to cover with conventional patches [8]. In addition, clinical trials have shown that the topical application of NSAIDs delivers drugs at a higher local concentration to the tissue, resulting in better pain management [9].

Bioadhesive in situ film-forming systems provide many advantages such as higher dosing flexibility, extended release properties, higher patient compliance, improved cosmetic appearance, and less chance for loss of formulation by rubbing [10].

Poly(lactide-*co*-glycolide) (PLGA) is one of the most promising polymers used for the fabrication of drug delivery devices and tissue engineering applications [11]. PLGA is biocompatible and biodegradable, and its properties such as the erosion profile and mechanical strength can be controlled as needed. Furthermore, PLGA can be engineered to control the drug release behavior by changing the polymer molecular weight and the molar ratio of lactide to glycolide [12–14]. However, a detailed characterization of such systems is required to prevent potential dose dumping, and an inconsistent drug release profile [15].

In this study, a PLGA-based in situ bioadhesive film-forming system for the transdermal delivery of TS was designed and evaluated for drug permeation across dermatomed porcine ear skin. Different types of plasticizer and solvents were screened to optimize the film forming behavior. The effects of different ingredients were studied on the permeation profile of TS. The optimized formulations were developed by finding a proper ratio between polymer, adhesive, and plasticizer and showed enhancement of TS delivery into and across the skin.

#### **2. Materials and Methods**

#### *2.1. Materials*

All polymers (Table 1) were kindly donated by Merck KGaA (Darmstadt, Germany). Trolamine salicylate and ethyl-2-cyanoacrylate were purchased from Sigma (St. Louis, MO, USA). Glycerol and Poly(ethylene glycol) 400 was purchased from VWR (Richmond, BC, Canada) and Acros organics (Morris Plains, NJ, USA), respectively. Propylene glycol was obtained from EKI Industries (Joliet, IL, USA). Porcine ear skin was procured from a local slaughterhouse. Phosphate-buffered saline (PBS; 10 mM, pH 7.4), acetonitrile, ethyl acetate, and acetone were purchased from Fisher Scientific (Waltham, MA, USA).



#### *2.2. Development of Film-Forming Polymeric Solutions*

#### 2.2.1. Screening of Components

To choose a compatible solvent for PLGA, 1 g of PLGA and 0.25 g of cyanoacrylate (CA) were dissolved together in 1 g of four different solvents (ethanol, ethyl acetate, propylene glycols, and acetone) and left overnight mixing. To select a proper plasticizer, we added a few drops (0.25 g) of one of the three different plasticizers namely glycerol, poly(ethylene glycol) 400 (PEG 400), and propylene glycol to the polymeric mixture to increase the flexibility of the film. Each formulation (6.4 μL) was then applied on the dermatomed porcine ear skin, and the film forming behavior was investigated.

#### 2.2.2. Optimization of Formulations

Optimized film-forming solutions were prepared by the addition of the PLGA and CA to the solvent. After obtaining a clear solution, the plasticizer and the drug were added. The solutions were left overnight mixing in glass vials to allow the drug to dissolve completely at room temperature. A detailed composition of each formulation can be found in Table 2.


**Table 2.** Composition of the distinct formulations (% (*w*/*w*)).

2.2.3. Effect of Plasticizer PEG 400, Types of PLGA, and Various Concentration of PLGA for Topical Delivery of TS

PLGA, 50-2A, was used to test the effects of plasticizer on the permeation profiles of TS. The ratio of PLGA to CA was kept at four to one. Then, the proportion of the plasticizer was increased from 0 to 3. The highest amount PEG 400 was added to F7, at 11.25 g (11.25% (*w*/*w*)), then reduced to 8.75 g (F6; 8.75% (*w*/*w*)), 5 g (F'1; 5% (*w*/*w*)), and finally to no PEG 400 in F5 (Table 2).

For the comparison of different types of PLGA, the ratio of each component was kept at 4:1:1 (PLGA:cyanoacrylate:PEG 400). To investigate the effects of different amounts of PLGA in TS permeation profiles, we incorporated different amounts of PLGA into formulations. PLGA 50-2A was used in all formulations, and the ratio of PEG 400 and cyanoacrylate was kept at 1:1. To increase the amount of PLGA, the amount of cyanoacrylate was decreased from 25 to 0 in increments of 5.

#### *2.3. Evaluation of the Formulations*

#### 2.3.1. Crystallization Study and Solvent Evaporation Study

To observe the crystallization of the drug in the films, we placed 6.4 μL of formulations on a microscopic slide and evaluated the samples under a microscope after the evaporation of acetone. To study the extent and time of solvent evaporation, 6.4 μL of F1 formulation was placed on a microscopic slide using a positive displacement pipette. The slides were left in a convection oven (32 ◦C) for a pre-determined time and weighed at 2 min, 1 h and 24 h to calculate the percentage evaporation of the solvent over time.

#### 2.3.2. In Vitro TS Release Study

The release profiles of TS from the formulations F1, F2, F3, and F4 (different PLGA types) were determined to study the impact of the polymer properties on the drug release kinetics. Polymer films containing 64 μg of TS (equal to a surface area of 0.64 cm2) was added to the bottom of a glass vial. Then, 5 mL of PBS were added and left under constant shaking at 150 rpm and 30 ◦C. Samples (300 μL) were withdrawn from the vials at predetermined time points, and the same amount of fresh receptor solution was replaced. The samples were analyzed by high-performance liquid chromatography (HPLC) as outlined below.

#### 2.3.3. In Vitro Permeation Study

#### Skin Preparation

The outer region of full-thickness ear skin was removed carefully with a scalpel. The skin was then washed with PBS, dried, wrapped with Parafilm, and stored at −80 ◦C until use. The porcine ear skin was dermatomed with a Dermatome 75 mm (Nouvag AG, Goldach, Switzerland) prior to use. The average thickness of the dermatomed skin pieces was 0.50 ± 0.02 mm.

#### Evaluation of Skin Integrity

Before conducting a permeation study, we evaluated the integrity of the skin membrane by measuring the skin resistance value. The skin was mounted on a vertical Franz diffusion cell with the stratum corneum side facing up, and then 300 μL of PBS was added to the donor compartment. A silver electrode and silver chloride electrode were placed in the receptor and the donor chambers, respectively, without touching the skin membrane. Two electrodes were connected to a digital multimeter (34410A 6 <sup>1</sup> <sup>2</sup> digit multimeter; Agilent Technologies, Santa Clara, CA, USA) and waveform generator (Agilent 33220A, 20 MHz function/arbitrary waveform generator) [16]. The resistance of the employed skin (Rs) was calculated according to the following formula:

$$\mathbf{R\_s} = \mathbf{V\_s} \mathbf{R\_L} / (\mathbf{V\_0} - \mathbf{V\_s}) \tag{1}$$

where RL and V0 were 100 kΩ and 100 mV, respectively. Skin pieces with a resistance lower than 10 KΩ were discarded.

Next, permeation studies were performed on a jacketed Franz diffusion cell with a diffusion area of 0.64 cm2 (PermeGear, Bethlehem, PA, USA) on dermatomed porcine ear skin. PBS was used as the receptor solution (5 mL). The skin was clamped between the donor and receptor chambers of a vertical diffusion cell with the stratum corneum side in contact with the donor solution. The distinct formulations (64 μg of TS per 0.64 cm2) was added in the donor compartment. The temperature of the receptor medium was maintained at 37 ◦C, and the skin surface temperature was about 32 ◦C. The amount of drug diffused over the dermatomed porcine ear skin was determined by removal of aliquots of 300 μL at pre-determined time points over 72 h from the receptor compartment. The volume was immediately replaced with the same amount of fresh buffer. The samples were analyzed by HPLC as outlined below. A skin extraction study was carried out to determine the drug amount that had penetrated into the skin. After 72 h, the skin samples were removed from Franz diffusion cell and the residual polymer film was removed with D-squame tape. The skin surface was wiped off with two cotton buds dipped in receptor solution. Then, the skin was dried with two cotton buds. The utilized tape and the four cotton buds were pooled and extracted with 30 mL of receptor solution. After the cleaning procedure, the epidermis and dermis were separated. The tissue was minced manually and added to 1 mL of receptor solution. Samples were shaken for 4 h followed by filtration using a 0.45 μm membrane filter and analyzed by HPLC.

#### *2.4. Quantitative Analysis*

TS was quantified by HPLC analysis. A Waters Alliance HPLC system (2795 Separating Module; Waters Co., Milford, MA, USA) equipped with a Photodiode detector (Waters 2475) and a Kinetex EVO C18 column (5 μm, 100 Å, 150 × 4.6 mm; Phenomenex, Torrance, CA, USA) was used. The mobile phase consisted of a 70/30 (*v*/*v*) mixture of acetonitrile and 0.1% of trifluoroacetic acid in distilled water. The injection volume and the flow rate were set to 20 μL and 1.0 mL/min, respectively, and the detection wavelength of TS was 304 nm. The reversed-phase HPLC method provided a linear range of 0.1–100 μg/mL (*R*<sup>2</sup> = 0.999).

#### *2.5. Statistical Analyses*

All results are reported as the mean with the standard error of the mean (SE) from at least three replicates. Statistical calculations were performed with GraphPad Prism Version 8.0 (GraphPad Software, Inc., San Diego, CA, USA). One-Way ANOVA followed by Tukey HSD post hoc test was applied to compare the results of different groups. Statistically significant differences were denoted by *p* < 0.05.

#### **3. Results**

#### *3.1. In Situ Film-Forming Behavior*

PLGA and CA did not dissolve completely in propylene glycol and ethanol. Complete dissolution was achieved in ethyl acetate and acetone. However, ethyl acetate did not evaporate upon application and the obtained polymer films were difficult to handle. Therefore, acetone was chosen as the solvent for the system. A successful film formation behavior on the dermatomed porcine ear skin was observed with the plasticizer PEG 400. Upon administration, the solvent evaporated instantaneously and left a thin and transparent film layer behind.

#### *3.2. Crystallization Study and Solvent Evaporation Study*

Formulation (F1–F11) was observed under the light microscope after complete evaporation of the solvent. No drug crystals were observed in the polymer films during the slide crystallization study, indicating the solubility of TS in the film matrix after evaporation of the organic solvent (Figure 1). After 2 min in the convection oven, the result demonstrated 75.3 ± 4.1% (*n* = 3) solvent evaporation, and it did not further evaporate for 24 h.

**Figure 1.** Microscopic images of the 20% PLGA formulation (F1) on glass sides (**a**) without and (**b**) with added drug (10× magnification) did not show drug crystallization after complete evaporation of the solvent.

#### *3.3. In Vitro Release Study for Di*ff*erent Types of PLGA*

In the first hour, F3 and F4 released 34% and 38%, respectively. In contrast, a smaller amount was released after 1 h with F1 and F2 (20% and 13%, respectively). After 72 h, the drug release from F1, F2, and F3 was completed (>90%), whereas only 68% of TS was released from F4 (Figure 2).

**Figure 2.** In vitro drug release profile from different types of PLGA (*n* = 6).

#### *3.4. E*ff*ects of Each Ingredient and Permeation Study*

#### 3.4.1. Effects of Plasticizer

Four formulations with different amounts of PEG 400 were prepared, and a thin film was formed upon application on the skin. The amount of TS delivered in the receptor solution after 72 h was found to be 16 <sup>±</sup> <sup>3</sup> <sup>μ</sup>g/cm2 and 24 <sup>±</sup> <sup>3</sup> <sup>μ</sup>g/cm2 in F5 and F7, respectively. A significantly higher amount of TS was delivered to the receptor with F1 (47 <sup>±</sup> <sup>8</sup> <sup>μ</sup>g/cm2) and F6 (47 <sup>±</sup> <sup>5</sup> <sup>μ</sup>g/cm2). Moreover, formulation without any plasticizer (F5) delivered only 2.7 <sup>±</sup> 0.3 <sup>μ</sup>g/cm2, the least amount of drug into the skin (Figure 3). The highest amount of PEG 400 facilitated the largest amount of TS delivery into the skin (individual values in Table 3). Because there was no significant difference between the F1 and F6 groups in the amount of drug in the receptor, the F1 ratio was chosen to carry out permeation studies to investigate the effects of different types of PLGA.

**Figure 3.** Permeation profiles of TS through porcine ear skin to study the effects of plasticizer. The group is representative of the ratio between PLGA:CA:PEG 400. (**a**) Cumulative amount (**b**) Average amount in the epidermis, dermis, and total skin.


**Table 3.** Amounts of TS extracted after 72 h in the epidermis and dermis and the total amount of TS for all groups.

#### 3.4.2. Effects of Different Types of PLGA

All formulations were able to form a polymeric solution, and then a thin layer of film was formed upon application on the skin. After 72 h, all three formulations with the one to one ratio (G/L) delivered 17 <sup>±</sup> 3, 17 <sup>±</sup> 3, and 19 <sup>±</sup> 3 <sup>μ</sup>g/cm2, respectively, and no significant difference was observed in the cumulative amount of TS in the receptor. In contrast, F4 delivered 11 <sup>±</sup> <sup>1</sup> <sup>μ</sup>g/cm2, a significantly lower amount than was observed for the other three groups. There was no significant difference in the drug amount delivered into the skin (Figure 4).

#### 3.4.3. Effects of Concentration of PLGA

All formulations were able to form a polymeric solution and formed a thin layer of film after complete evaporation of the solvent. The highest amount of TS (45 <sup>±</sup> 4 <sup>μ</sup>g/cm2) was delivered with 20% PLGA (F1"), followed by the formulation with 25% of PLGA (F11) (21 <sup>±</sup> 4 <sup>μ</sup>g/cm2). The formulations with 0%, 5%, and 10% PLGA delivered a significantly lower amount of TS, at 0.6 ± 0.2, 1.6 <sup>±</sup> 0.3, and 5.6 <sup>±</sup> 0.7 <sup>μ</sup>g/cm2, respectively (Figure 5). The formulation with 10% (F10) and 20% (F1") PLGA delivered significantly higher amounts into the skin. However, the 0%, 5%, and 25% PLGA formulations delivered a lower amount of TS into the skin (values in Table 3).

**Figure 4.** Permeation profiles of TS through porcine ear skin for the different types of PLGA. The group is representative of different types of PLGA: (**a**) Cumulative amount (**b**) Average amount in the epidermis, dermis, and total skin.

**Figure 5.** Permeation profiles of TS through porcine ear skin for different concentrations of PLGA. The group is representative of different concentrations of PLGA: (**a**) Cumulative amount (**b**) Average amount in the epidermis, dermis, and total skin.

#### **4. Discussion**

#### *4.1. Parameters for the Development of Film-Forming Polymeric Solutions*

Limitations of conventional formulations for topical drug delivery include poor adherence to the skin, poor permeability, and a low compliance rate [17]. Bioadhesive film, because it is an intermediate between transdermal patches and semisolid dosage forms, has the advantages of both systems, such as transparency, no stickiness, convenience, less frequent dosing, sustained drug release, and resistance to wiping off. In addition, incorporating cosmetic or therapeutic agents is more convenient with this system than conventional topical drug delivery systems, owing to lower potential loss of the formulation by rubbing.

First, the most important parameters for polymeric solutions are the polymers. Suitable polymers for successful film forming systems may possess transparency, flexibility, and drug encapsulation ability at moderate temperature. To develop a suitable formulation to decrease the frequency of application of topical analgesics, common film-forming polymers, such as hydroxypropyl, methylcellulose, polyvinyl pyrrolidine, and acrylate copolymer, were searched in the literature. However, many of the polymers were hydrophilic and might not provide water resistance in daily life. PLGA was chosen to be incorporated in the polymeric solution because of its many benefits, such as its water resistance, nonirritating and nonallergic qualities, and capability for drug incorporation [15]. To provide an efficient local drug delivery system, Eskandari et al. have investigated the use of butyl-2-cyanoacrylate. To ensure good adhesion of the film to the skin, CA, which is compatible with PLGA, was chosen. CA gets stiffened by polymerization process in the presence of moisture [18–20]. When the polymeric film is applied on the skin, the film is formed upon polymerization of PLGA and CA due to the moisture of the skin. Polymerized CA is an excellent polymer candidate for drug delivery because it is a biodegradable, hemostatic, nonallergenic tissue adhesive with local antibacterial properties, and it is inexpensive and widely available [18,19,21]. In addition, it is commonly used and is acceptable to the public, owing to its use in liquid bandage formulations [22]. Currently, CA is used in surgical and clinical adhesives in various surgical procedures such as the treatment of arteriovenous malformation, retinal ruptures, and skin graft placements [23,24]. Eskandari et al. have shown that CA can successfully incorporate drugs for slow release of antibiotics to specific areas without causing an inflammatory response [25].

Solvents play an important role in film formation. The solvent used in the film forming system is responsible for drug-polymer solubility and also affects drug permeation. Some common solvents used in polymeric solutions are ethanol, ethyl acetate, propylene glycols, and acetone [26]. The combination of the PLGA and the adhesive was tested for dissolution in different solvents such as acetone, ethyl acetate, and ethanol. All the ingredients were dissolved readily in acetone and formed a film upon application to the skin. Acetone is approved in use of topical formulation per the inactive ingredient list of FDA [27]. Also, a toxicology report by US. Department of Health and Human Services reported that dermal exposure to acetone did not affect the human health negatively [28].

Apart from the polymer and solvent, other excipients such as plasticizers must be added into the formulation. Plasticizers are low molecular weight resins that may interact with polymer chains and affect polymer-polymer bonding [29]. These interactions may affect film flexibility as well as the permeability toward drug substances. The commonly used plasticizers in polymeric solutions include fatty acid esters, glycol derivatives, phthalate esters, and phosphate esters [30]. The plasticizer used should be compatible with the polymers and should have low skin permeability. Hence, choosing a proper plasticizer is important for the polymeric matrix. In addition, determining the right amount of plasticizer is critical for film-forming polymeric solutions. After the selection of two polymers to form the polymeric solution, proper plasticizers were screened for tolerance to mechanical stress, which may be exerted on the formed film by the movement of the skin. In selecting an appropriate plasticizer, compatibility with the polymers and plasticization efficiency were considered. After application on dermatomed porcine skin, the acetone evaporated, and a thin transparent film was formed in situ. Based on the results, PEG 400 was incorporated as a plasticizer to enhance the mechanical properties and flexibility of the film.

#### *4.2. Characterization Methods for Film-Forming Polymeric Solutions*

In a transdermal delivery system, drug solubility in the polymer matrix system is a critical factor that may affect the rate and extent of drug permeation [31]. If a drug is not soluble in a polymer matrix, it will become supersaturated and hence unstable. Moreover, when a drug is successfully incorporated in a polymeric matrix, many studies have reported that the stability of the drug improves [32,33]. PLGA formulations are widely used in transdermal and dermal delivery and have a safety profile showing a lack of skin irritation and uniform drug dosing [15,34].

The drug was successfully dissolved in the polymeric matrix, and the absence of crystallization, which might hinder drug permeation into the skin, was verified. In patch development, to improve the permeation and release profiles of drugs, the investigation of the supersaturated state is an important step. A possible strategy to study the solubility of a drug is to disperse it uniformly into a polymeric matrix to prevent drug crystallization [35]. Different techniques have been described to determine the solubility of drugs in polymer matrixes [31,36,37]. Such solubility testing can be time consuming and difficult. Among different techniques, Jain et al. have compared two different methods to predict drug solubility in the polymeric matrix: differential scanning calorimetry and slide crystallization study [31]. In that study, the author found that the experimental solubility potential and the theoretical values from a solubility calculator were similar. In addition, low drug solubility in the drug-polymer matrix was shown to negatively affect drug release. The author concluded that simple slide crystallization was able to predict the saturation solubility of a drug in the polymer. In the present study, after complete evaporation of solvents from our drug-polymeric matrix, crystallization of the drug was not observed under a light microscope, and we confirmed the drug solubility and stability in the system. The complete TS solubility in the polymer matrix system after evaporation of acetone was studied to ascertain that the stability and solubility of TS was not a factor affecting the rate and extent of drug permeation.

The amount of residual solvent was also measured and calculated. Instant solvent evaporation and complete dryness of the formulation was observed after 2 min. As a result, approximately 25% of the residual solvent was left on the application site. Per the International Conference on Harmonisation residual solvent guideline, solvents are classified in the three categories which set limits depending on the toxicity data for each solvent. Acetone is categorized as a Class III solvent, which has a low toxic potential and is safe to use in human [38]. Furthermore, Baino et al. reported that the drug concentration increases upon evaporation of the solvent, which may result in a greater driving force to deliver the drug into the system [39]. However, this variable was eliminated by keeping the solid content of all the formulations constant. Thus, residual solvent was not a factor that led to differences in the permeation profile of TS from the formulations tested.

#### *4.3. In Vitro Release Study for Di*ff*erent Types of PLGA*

In this study, the release profiles of TS from different PLGA polymers were determined to study the effects of the PLGA parameters on the drug release kinetics. In terms of developing a polymeric system, knowing the release mechanisms and the physicochemical properties is important. The two main release mechanisms related to drug release from a PLGA-based system are degradation and diffusion. Often, the release rate is initially determined by diffusion followed by degradation controlled at the final stage [40]. PLGA systems commonly release the drug in a bi-phasic or a tri-phasic pattern. In this study, for higher molecular weight PLGA polymers, we observed a bi-phasic profile of drug release, with a burst release followed by relatively slow diffusion. Lower molecular weight PLGA groups showed a monophasic release profile, which is preferable for drug delivery systems because it follows a zero-order release profile [41]. The formulation with higher molecular weight compounds 50-8A and 75-5A released almost 40% after 1 h, thus indicating dose dumping, whereas the release profiles for 50-2A and 50-5A showed a gradual release until 72 h. In this study, the film was submerged in the receptor solution, which was more hydrophilic than the polymeric solution. Because the drug was hydrophilic, it may be preferentially located in the receptor solution over the PLGA system. Thus, the higher molecular weight, and hence more lipophilic PLGA, resulted in a more rapid release of the drug than was observed in the groups with lower molecular weight. Based on the result of the release study, 50-2A was chosen for the final formulation to avoid dose dumping effect as well as provide a gradual release of TS until 72 h.

#### *4.4. In Vitro Permeation Study*

In vitro permeation studies, using the vertical Franz diffusion cell model, have been commonly used to evaluate drug delivery into and across the skin, and is well established as a reliable tool [42,43]. In addition, FDA recommends and endorses the use of an in vitro permeation test to evaluate topical products. FDA guidelines have included the merits of this approach and have established a strong correlation between the in vitro permeation test (IVPT) and in vivo bioavailability data with narrow inter- and intra-variability between data [44–46].

#### 4.4.1. Effects of Plasticizers

Plasticizers are important components of film forming systems because they improve the appearance of the film, prevent film cracking, increase film flexibility, and confer desired mechanical properties [47]. By selecting an appropriate plasticizer and optimizing its concentration in the formulation, the release rate of a therapeutic compound can be controlled. To observe the effects of the plasticizer, we kept the formulation ratio between the polymer and the adhesive at 4 to 1 (PLGA:CA), and the same amounts of other components were included, i.e., solvents and drug concentrations. The ratio of plasticizer was increased from 0 to 4 (0%, 5%, 8.75%, and 11.25% *w*/*w*). The polymeric films without any plasticizer and 11.25% *w*/*w* plasticizer delivered a significantly lower amount of TS after 72 h, than did those with 5 and 8.75% *w*/*w* PEG 400. The films with a plasticizer concentration of 8.75% *w*/*w* did not significantly improve the diffusion of the drug in the receptor (as compared with the results with 5% *w*/*w* plasticizer concentration). However, the amount of PEG 400 showed a linear relationship up to 8.75% *w*/*w* with regarding TS delivery into the skin. The formulation with 11.25% *w*/*w* did not show a significant increase over 8.75% *w*/*w* regarding TS delivery into the skin. Hence, 5% *w*/*w* was considered the optimum concentration for the plasticizer. These results may have been due to a polymer-plasticizer interaction that affected the release of the drug from the system. Barhate et al. have studied the carvedilol permeation profile with and without PEG 400 as a plasticizer and have found that the incorporation of PEG 400 increases the film flexibility and permeation rate [48]. The results suggested that too little or too much plasticizer in the polymeric solution can negatively affect the permeation profile of TS. In formulating a polymeric solution, minor variations might be acceptable; however, major changes in the composition should be carefully considered, because they

might unfavorably affect the properties of the film forming system and consequently the mechanical or cosmetic performance in drug delivery into the skin.

#### 4.4.2. Effects of Different Types of PLGA

PLGA is one of the most commonly used biodegradable polymers in the field of biomedical devices because of its favorable degradation characteristics and the possibility for sustained drug delivery. It can be engineered to control drug release by changing the polymer molecular weight and ratio of lactide to glycolide. Commercially available PLGA combinations are 50:50, 65:35, 75:25, and 85:15 (LA:GA) in different molecular weights [49]. These physical properties affect the solubility, glass transition temperature, and inherent viscosity, thus resulting in tensile strength and polymer chain flexibility. These parameters help delivery systems or medical devices achieve the desired dosage and release interval. However, evaluation of the system is required to minimize the potential toxicity from dose dumping, inconsistent release, and drug-polymer interactions [50]. In this study, four different types of PLGA were chosen and incorporated in the polymeric solution. Three formulations were 50:50 (LA:GA) with different molecular weights. One of the formulations was 75:25 (LA:GA). There was no significant difference among formulation groups with the same ratio of LA:GA, but the group with a higher proportion of lactide delivered less TS after 72 h. Moreover, the amount of drug delivered into the skin did not show significant differences across all four groups. PGA is a hydrophilic and highly crystalline polymer with a relatively fast degradation rate. Although PGA and PLA are similar structurally, they exhibit different physicochemical properties because of the presence of a methyl group on the alpha carbon. Delayed release of the drug from the higher ratio of LA delivered fewer drugs to the receptor at the end of the study. However, the same ratio with different molecular weights did not show a difference in drug delivery to the receptor and skin. Farahani has investigated the degradation mechanism of PLGA (50:50) and has reported that up to eight weeks are required for complete degradation [51]. In this study, drug diffusion was found to be important factor affecting in vitro release of the drug from the formulation; hence the rate of diffusion influenced drug delivery into and across the skin. Three formulations with 50:50 (LA:GA, different molecular weights) did not result in a difference in permeation profile of TS into and across skin; however, the formulation with 75:25 (LA:GA) delivered a significantly lower amount of TS due to its slower diffusion rate.

#### 4.4.3. Effects of Different Amounts of PLGA

To test the effects of the amounts of PLGA in TS permeation profiles, we varied the amount of PLGA without changing the ratio between CA and plasticizer. The amount of PLGA was increased from 0% to 25% *w*/*w*. Formulations with 20% and 25% PLGA delivered significantly higher amounts than the other groups. Formulations with lower amounts of PLGA required more CA. CA has been used as a synthetic adhesive for tissue adhesive application since the 1980s. CA polymerizes rapidly in the presence of weak basic conditions, such as in water. The interaction between skin and the polymer results in the impressive adhesive strength of CA. Moreover, CA is superior to rival polymers in terms of its strong wet adhesion, drug incorporation, and rapid curing [52–55]. The trend in permeation profile was observed because the drug's affinity toward CA was greater than that toward PLGA. Although CA has good adhesion to skin, the rapid polymerization of CA is associated with heat dissipation at the application site, and the brittleness can be problematic [56]. PLGA content was increased while the amount of CA was reduced proportionally to maintain a similar solid content of all formulation. The cumulative drug amount linearly increased with an increase in the amount of PLGA in the formulation. In the optimized formulation, 20% *w*/*w* PLGA was added, resulting in the highest delivery after 72 h without compromising its adhesive property. In this study, an in-situ film forming system was developed by minimizing the CA component and maximizing the amount of PLGA in solution, thus enhancing the delivery of TS into and across the skin.

#### **5. Conclusions**

Film forming solutions were successfully formulated with polymers from different chemical groups such as PLGA and polymerized CA. These formulations contained a combination of polymers, a volatile solvent, and other optional excipients such as plasticizers, and fixed concentrations were used for all excipients involved. The optimized formulation also enhanced the delivery of TS into and across the skin.

**Author Contributions:** Y.K. performed the experiments and drafted the manuscript. A.K.B. supervised the project. A.K.B. and M.B.-B. reviewed the entire contents of the manuscript.

**Funding:** This project was funded by Merck KGaA, Frankfurter Str. 250, 64293 Darmstadt, Germany, grant number [UR1710050]. The life science business of Merck KGaA, Darmstadt, Germany operates as MilliporeSigma in the U.S. and Canada. Product designations of Merck KGaA, Darmstadt, Germany, and third parties may appear in this material. For details on the ownership of mentioned trademarks, please refer to publicly available resources such as tmdn.org.

**Conflicts of Interest:** The founding sponsors had no role in the design of the study, writing of manuscript, or the decision to publish the results but provided scientific edits to the manuscript draft which were not related to the collection, analyses, or interpretation of data.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
