- Email: [email protected]
In situ gelling hexagonal phases for sustained release of an anti-addiction drug
Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
In situ gelling hexagonal phases for sustained release of an anti-addiction drug Jessica Phelps a , M. Vitória L.B. Bentley b , Luciana B. Lopes a,∗ a b
Albany College of Pharmacy and Health Sciences, Albany, NY, USA Faculdade de Ciências Farmacêuticas de Ribeirão Preto, Universidade de São Paulo, Ribeirão Preto, SP, Brazil
a r t i c l e
i n f o
Article history: Received 30 March 2011 Received in revised form 16 May 2011 Accepted 25 May 2011 Available online 1 June 2011 Keywords: Hexagonal phase Naltrexone Subcutaneous delivery In situ gelling Swelling
a b s t r a c t In this study, fluid precursor formulations for subcutaneous injection and in situ formation of hexagonal phase gels upon water absorption were developed as a strategy to sustain the release of naltrexone, a drug used for treatment of drug addiction. Precursor formulations were obtained by combining BRIJ 97 with propylene glycol (PG, 5–70%, w/w). To study the phase behavior of these formulations, water was added at 10–90% (w/w), and the resulting systems were characterized by polarized light microscopy. Two precursor formulations containing BRIJ:PG at 95:5 (w/w, referred to as BRIJ-95) and at 80:20 (w/w, referred to as BRIJ-80) were chosen. Naltrexone was dissolved at 1% or suspended at 5% (w/w). Precursor formulations were transformed into hexagonal phases when water content exceeded 20%. Water uptake followed second-order kinetics, and after 2–4 h all precursor formulations were transformed into hexagonal phases. Drug release was prolonged by the precursor formulations (compared to a drug solution in PBS), and followed pseudo-first order kinetics regardless of naltrexone concentration. The release from BRIJ-80 was significantly higher than that from BRIJ-95 after 48 h. The relative safety of the precursor formulations was assessed in cultured fibroblasts. Even though BRIJ-95 was more cytotoxic than BRIJ-80, both precursor formulations were significantly less cytotoxic than sodium lauryl sulfate (considered moderate-to-severe irritant) at the same concentration (up to 50 g/mL). These results suggest the potential of BRIJ-based precursor formulations for sustained naltrexone release. © 2011 Elsevier B.V. All rights reserved.
1. Introduction The non-selective opioid antagonist naltrexone is used to prevent relapse in former opioid- and alcohol-dependent patients [1,2]. Daily administration of oral doses of naltrexone at 50–100 mg has been clinically used for several years [3]. However, oral administration of naltrexone is complicated by its extensive firstpass hepatic metabolism, which results in variable and low drug bioavailability (5–40%), and increased costs [3]. In addition, the requirement of daily drug administration for treatment of chronic disorders often results in reduced patient compliance and increased probability of relapse [4,5]. Delivery systems that can sustain naltrexone delivery for longer periods of time offer an alternative to overcome these problems and improve patient compliance [2]. Based on these facts, the present study was aimed at developing hexagonal liquid crystalline gels for naltrexone delivery. Because the hexagonal phase is viscous and difficult to inject [6], we developed fluid, free-flowing precursor formulations that could be
∗ Corresponding author at: Department of Pharmaceutical Sciences, Albany College of Pharmacy and Health Sciences, 106 New Scotland Ave., Albany, NY 12208, USA. Tel.: +1 518 694 7113. E-mail address: [email protected] (L.B. Lopes). 0927-7765/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2011.05.048
injected subcutaneously to form a hexagonal liquid crystalline gel in situ upon absorption of water from the tissue. The hexagonal phase consists of cylinders of micelles arranged as a hexagon [7,8]. Depending on the structure-forming compound, water and hydrophilic drugs can be enclosed within the cylinders of the hexagonal phase (reverse type) or outside of them (normal type) [9,10]. So far, most of the studies have employed hexagonal phases of the reverse type for sustained drug release [6,11]. Such phases are formed by lipids (like monoolein), can resist dilution with body fluids for long periods of time and provide very slow release of hydrophilic compounds [6,12]. In this study, we propose to use a hexagonal phase of the normal type formed by the hydrophilic surfactant polyoxyethylene-10-oleyl ether (BRIJ 97) [9]. The rationale for this choice is based on the assumption that even though the hexagonal phase gel of the normal type should persist at the site of injection for days prolonging drug release, BRIJ hydrophilicity would account for a faster (though still prolonged) dissolution of the gel in body fluids compared to other long-lasting systems, decreasing incidence and duration of adverse reactions at the site of injection [13–15]. Different amounts of BRIJ were combined with propylene glycol to obtain injectable fluid precursor formulations, which gave rise to hexagonal phase approximately 2–4 h after contact with aqueous environment. The swelling kinetics of precursor formulations, drug
392
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
release properties and cytotoxic potential of the formulations were evaluated. 2. Materials and methods 2.1. Material Propylene glycol (PG), polyoxyethylene-10-oleyl ether (BRIJ 97), naltrexone hydrochloride, hydroxypropyl cellulose and phosphatebuffered saline (PBS) were obtained from Sigma–Aldrich (St. Louis, MO). 2.2. Methods 2.2.1. Development of precursor formulations Precursor formulations were obtained by mixing BRIJ with propylene glycol (PG) at ratios varying from 95:5 to 30:70 (w/w). Propylene glycol was included to facilitate drug incorporation and to allow producing fluid formulations. The precursor formulations were visually inspected for transparency and fluidity, and visualized under a polarized light microscope (Carl Zeiss, Oberkochen, Germany) for assessment of isotropicity. 2.2.2. Phase behavior of precursor formulations upon contact with water To evaluate the formation of hexagonal phases by the precursor formulations upon contact with water, each BRIJ:PG ratio was mixed with various amounts of water (10–90%, w/w), and the formulations were allowed to equilibrate in closed vials for 3 days at room temperature. Systems were characterized by visual inspection and polarized light microscopy. Visual inspection was used to verify sample homogeneity and phase separation. Polarized light microscopy allowed for the characterization of the hexagonal structure, which displays a fan like and angular texture under the microscope [16]. Phase diagrams were constructed to show the relationship between the formulation composition and phase behavior. Two precursor formulations (combinations of BRIJ:PG at 95:5 and 80:20, w/w) were chosen for further studies based on the fact that they were isotropic and fluid at room temperature, but formed stiffer hexagonal phase gels at relatively low water content (20 or 30%, see Section 3). For the sake of simplicity, the BRIJ:PG mixture at 95:5 (w/w) will be referred to as BRIJ-95, whereas the BRIJ:PG mixture at 80:20 (w/w), will be referred to as BRIJ-80. The influence of temperature on the structure of BRIJ-95 and BRIJ-80 containing water at 10–90% (w/w) was evaluated using a polarized light microscope coupled with a hot stage (Linkam, Surrey, UK). Temperature was increased at 1 ◦ C/min from 25 to 40 ◦ C, and changes in the structure of the systems were monitored by polarized light microscopy. To evaluate the influence of drug incorporation on formation of hexagonal phase, naltrexone was added to the precursor formulations at 0.5, 0.75, 1% or 5% (w/w). Samples containing up to 1% of naltrexone were clear and fluid, whereas those containing 5% were more turbid, visually similar to suspensions. Water was added at 10–90% (w/w), samples were left to equilibrate for 3 days at room temperature in closed vials and subsequently characterized by visual inspection and polarized light microscopy. 2.2.3. Swelling studies Swelling studies were performed to evaluate the rate and kinetics of water absorption by the two chosen precursor formulations (BRIJ-95 and BRIJ-80). Two experiments were performed in association with this study. The first set of experiments aimed to determine the length of time necessary for the fluid precursor formulation to be transformed into the hexagonal phase upon water uptake.
Twelve-well plates and cell culture inserts were used in this experiment. Each well of the 12-well cell culture plate was filled with 2 mL of 2% hydroxypropyl cellulose gel. This gel was chosen because it mimics better the subcutaneous tissue compared to aqueous solutions used in most swelling studies [17,18]. This is based on the fact that the intra and extracellular space resembles more closely gels than liquid solutions since water in tissues does not act like bulk water (i.e., is not freely available), but is mostly associated with surfaces, proteins and other cell components forming gels [19,20]. BRIJ-95 and BRIJ-80 fluid precursor formulations (200 mg) were added to cell culture inserts, and each insert was placed in contact with the hydroxypropyl cellulose gel. Samples of the formulations in the cell culture inserts were collected at 2, 4, 6, 8, 16, 24, 72 and 96 h and the structure of the system was identified under a polarized light microscope. In a parallel set of experiments, aimed to determine the rate of water absorption by the precursor formulation, cell culture inserts containing the precursor formulations (200 mg) were placed in contact with the hydroxypropyl cellulose gel, and weighed at 0.5 or 1 h intervals up to 8 h, and then at 24 and 36 h. A cellulose dialysis membrane (cut-off 1000 Da) was placed between the hydroxypropyl cellulose gel and cell insert to prevent adhesion of the gel on the inserts (which would prevent the weighing of the inserts). The amount of water absorbed was plotted as function of time and fitted using the first-order kinetic equation:
ln
W∞ W∞ − W
= kt,
(1)
or second-order order equation t 1 t + = , W k W ∞2 W ∞
(2)
where W∞ is the maximum water uptake, W is the water uptake at time t, and W∞ − W is the unrealized water uptake and k is the rate constant [21,22]. The rate of swelling was determined by plotting the difference of swelling at two consecutive time points as function of time. 2.2.4. In vitro drug release studies The in vitro release of naltrexone from BRIJ-95 and BRIJ-80 was evaluated using Franz diffusion cells, and phosphate-buffered saline (100 mM, pH 7.2) under constant stirring (350 rpm) at 37.0 ± 0.5◦C as receptor phase. A cellulose membrane (cutoff 1000 Da, Sigma, St Louis, MO) was placed between the receptor and donor compartments. The receptor phase was collected at 2, 4, 6, 8, 16, 24, 48 and 72 h post-application. As stated before, the precursor formulations are isotropic fluid systems that turn into hexagonal phases upon contact with water. Therefore, this transformation occurred during the in vitro release study. The formation of the hexagonal phase structure in the donor compartment was confirmed by polarized light microscopy. Naltrexone in the receptor phase was quantified by UVspectrophotometry at 215 nm. Standard curves of naltrexone in PBS in the range of 1–20 g/mL were used for calibration curves. The method was linear up to 50 g/mL. Interference of formulation components in the analytical method was assessed by adding unloaded (not containing drug) formulations in the donor compartment of diffusion cells and assaying the receptor phase after 24 h. No formulation component interfered with the analytical method. 2.2.5. Evaluation of cellular viability To evaluate the relative safety of the precursor formulations, we compared their cytotoxic effects to those of PBS (considered safe) and of sodium lauryl sulfate (considered a moderate-to-severe irritant) [23,24]. The classification of sodium lauryl sulfate as a moderate-to-severe irritant is based on its effects on the skin and
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
eyes of rabbits as determined in Draize tests [25]. In acute ocular tests, sodium lauryl sulfate at 5.1–10% caused moderate irritation to the rabbits’ eyes if not rinsed, while at 21%, it caused severe irritation. Similarly to its effects on the eyes, between 10 and 30% sodium lauryl sulfate caused skin corrosion and severe irritation [26,27]. Because of the ethical implications of the Draize test, measurement of viability and release of cytokines in cells and tissues in culture has been used as alternative in vitro assays to determine the irritation potential of formulations [28]. An in vitro cell culture assay was used in this study. Murine swiss 3T3 mouse fibroblasts were obtained from American Type Culture Collection (ATCC, Manassas, VA), and grown at 37 ◦ C and 5% CO2 atmosphere in Dulbecco’s modification of Eagle’s medium (DMEM, ATCC, Manassas, VA) containing 10% fetal bovine serum (FBS, Gibco, Carlsbad, CA) and additional penicillin and streptomycin (1%). For the cellular viability assay, cells were plated in 96-well plates (6000 cells/well), and treated for 12 h with either PBS, BRIJ-95, BRIJ-80 or sodium lauryl sulfate at concentrations ranging from 1 to 100 g/mL in cell culture medium. Cell survival was evaluated using a cell proliferation assay reagent (CellTiter 96 Aqueous One solution, Promega, Madison, WI) consisting of 3-(4,5-dimethylthiaziazol-2-yl)-5-(3carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium salt (MTS) and an electron coupling reagent. The MTS salt is reduced to a colored formazan product, and the amount of this product is directly proportional to the number of living cells. After treatment, cells were washed with PBS, and 100 L of cell culture medium plus 20 L of the cell proliferation assay reagent were added to each well. The plates were incubated for 2 h at 37 ◦ C in a humidified atmosphere of 5% CO2 in an incubator, and the absorbance was recorded at 490 nm using a plate reader (SpectraMax, Molecular Devices Corp., Sunnyvale, CA). These experiments were performed using 3–4 replicates; in other words, 3–4 wells containing cells (between passages 2 and 6) were treated with each concentration of test formulations or control solutions. 2.2.6. Statistical analysis The results were reported as means ± standard deviation. Data from swelling studies were statistically analyzed using one way ANOVA (followed by Tukey post-hoc test, GraphPad Prism 4 software), whereas analysis of data from the in vitro release studies was performed using repeated-measures ANOVA (followed by Fisher post-hoc test, Statistica Advanced 8.2 software). Values were considered significantly different when p < 0.05. 3. Results and discussion 3.1. Development of precursor formulations and evaluation of their phase behavior Because the hexagonal phase is too viscous to be injected, we developed fluid precursor formulations composed of BRIJ:PG mixtures at 95:5–30:70 (w/w). The precursor formulations were transparent and isotropic (under polarized light). The precursor formulations gave rise to several phases in the presence of water (Fig. 1). The hexagonal phase was observed predominantly when BRIJ was present at a concentration of 35% (w/w) and higher, PG was below 35% (w/w) and water was between 20 and 60% (w/w). This phase co-existed with excess water (hexagonal phase + water) when water content varied from 20 to 70% and PG varied from 5 to 40%. At higher (>80%) percentages of water, the hexagonal phases were transformed into liquid isotropic phases (probably micellar systems) [9]. This transformation occurred with hexagonal phase + water samples 8–10 days after preparation even if no additional water was added, and may be attributed to rear-
393
rangement of BRIJ and formation of micellar systems due to its hydrophilicity [9]. Phase transformation triggered by water uptake in BRIJ:PG mixtures at 95:5 or 90:10 (w/w) occurred in the following order: liquid isotropic system, lamellar phase, hexagonal phase, hexagonal phase + water and liquid isotropic systems (Fig. 1). The lamellar phase was observed only when PG content in the precursor formulation was equal or smaller than 10%. As PG content in the precursor formulation increased, the hexagonal phase existed at narrower aqueous contents, whereas the hexagonal phase + water was formed at smaller percentages of water (Fig. 1). In other words, the hexagonal phase could uptake less water. In precursor formulations containing 40% of PG (BRIJ:PG 60:40), the hexagonal phase + water was observed with water at only 30% (instead of 70% as in BRIJ:PG at 90:10). The hexagonal phase was no longer observed when PG content in the precursor formulation was higher than 50%. We chose precursor formulations with BRIJ:PG ratios of 80:20 (BRIJ-80) and 95:5 (BRIJ-95) for further investigation based on two facts: first, both allowed obtainment of the hexagonal phase at relatively low water content (20 or 30%), indicating that in situ absorption of low amounts of water would be necessary for formation of the gel; second, uptake of 20% of water by BRIJ-95 (but not by BRIJ-80) led to formation of lamellar phase (transformed into hexagonal phase with water content over 30%), which would give us the opportunity to evaluate whether drug release was affected by transient formation of the lamellar phase. No phase transition at temperatures ranging from 25 to 38 ◦ C was observed when either BRIJ-80 or BRIJ-95 was combined with various amounts of water (10–90%), suggesting that the hexagonal phase gel is stable at body temperature. We next evaluated the influence of drug load on phase behavior of the precursor formulations. Naltrexone could be dissolved in BRIJ-95 and BRIJ-80 at concentrations up to 1%. Compared to the unloaded (not containing drug) systems, naltrexone at 1% led to formation of the hexagonal phase + water at lower aqueous contents: 60–70% and 40–70% for BRIJ-95 and BRIJ-80, respectively (Fig. 1). Because we used naltrexone as hydrochloride, interactions between the dissolved drug and the polar headgroup of surfactant may be possible, affecting the interfacial curvature of the surfactant layer, and causing phase transition [29]. These types of interactions may cause a decrease in the amount of water the hexagonal phase can uptake. Previous studies reported the ability of other drugs (such as cyclosporine) to interact with the polar headgroup of structure-forming compounds, decreasing the amount of water that liquid crystalline phases can incorporate [11]. Addition of naltrexone at concentrations higher than 1% led to formation of suspensions. Nevertheless, we evaluated the phase behavior of BRIJ-80 containing 5% of the drug to assess the influence of drug crystals. Naltrexone was incorporated at 5% only in BRIJ-80 since this formulation had smaller cytotoxicity potential while still able to sustain drug release (see results for drug release studies and evaluation of cellular viability). As the formulation absorbed water, it became transparent due to naltrexone dissolution. There was no difference in the type of systems formed as function of water content when naltrexone was added at 1 or 5%, suggesting that presence of suspended drug within the studied range does not cause phase transition. Similar results were described in a previous study by Helledi et al. in which the authors incorporated acyclovir up to 45% in monoolein-based systems [30]. The use of materials that respond to the biological environment to trigger formation of delivery systems and/or to control drug release has been attracting increasing interest. Not only absorption of water, but also changes in temperature and pH (among other factors) can trigger the formation of modified-release systems [31–33]. Taking advantage of the fact that certain surfactants and polar lipids can uptake water, in situ-forming delivery systems
394
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
Fig. 1. Phase diagrams of BRIJ, PG and water without naltrexone (NTX) or containing the drug at final concentrations of 1 or 5% (w/w) at 25 ◦ C.
for intratumoral, oral, subcutaneous and periodontal administrations of drugs have been proposed [6,17,34,35]. To our knowledge, this is the first report employing in situ gelling hexagonal phases for sustained release of naltrexone. Other delivery systems have been investigated to prolong the release of naltrexone and decrease dosing frequency, including intramuscular microparticles, transdermal devices, buccal tablets and subcutaneous implants [36–40]. Compared to these systems, in situ gelling hexagonal phases offer several potential advantages: first, they provide sustained drug release for variable amounts of time depending on their composition; second, naltrexone potency in reducing ethanol self-administration seems to be increased when administered subcutaneously compared to other routes; and third, subcutaneous injection of fluid formulations does not require extensive training and surgical procedures like subcutaneous implants [6,15,31,41,42]. A previous study has investigated subcutaneous injectable thermosensitive poloxamer gels for naltrexone release, but sustained release for over 100 h was only obtained when the drug was complexed with cyclodextrins, which adds one more step to the preparation of precursor formulations, and increases cost [43]. The fact that BRIJ-95 and BRIJ-80 are very easy to prepare, can form hexagonal phase over a wide range of aqueous content, and naltrexone addition (within the concentration studied) does not preclude hexagonal phase formation suggests the potential of these formulations for use as in situ gelling delivery systems.
water uptake rate (data not shown). This is consistent with the phase behavior depicted in Fig. 1, which shows transition to hexagonal phase + excess water (after which additional water uptake is not expected) at aqueous content higher than 50% for BRIJ-80 and 60% for BRIJ-95. The maximum rate of water uptake was achieved after 1 h for both precursor formulations (Fig. 3B), but was significantly (p < 0.001) higher for BRIJ-95 (23 ± 5%/h for BRIJ-95 and 15 ± 3%/h for BRIJ-80). Fitting of the swelling data to Eqs. (1) and (2) suggests that swelling can be better described by second order kinetics (coefficients of determination of approximately 0.99 for BRIJ-95 and BRIJ-80, Fig. 3C and D). Similar kinetics of water uptake
3.2. Swelling studies The rate and kinetics of water absorption by BRIJ-95 and BRIJ-80 were determined through two swelling studies. Fig. 2 shows polarized light microscopy pictures of the precursor formulations at 0, 2, 4, and 48 h after contact with the aqueous phase. At 0 h, both fluid precursor formulations were clear, liquid isotropic phases. At 2 h, the BRIJ-95 was transformed into lamellar phase, and between 2 and 4 h, into the hexagonal phase. On the other hand, BRIJ-80 was transformed directly into the hexagonal phase after only 2 h of contact with the aqueous gel. Both systems sustained the hexagonal structure for approximately 96 h. After this period of time, the structure of the hexagonal phase started to collapse, and after 144 h, it was transformed into an isotropic fluid system. The rate of water uptake by the precursor formulations is shown in Fig. 3. Both precursor formulations quickly absorbed water up to 8 h, when they reached their respective equilibrium concentration. BRIJ-95 absorbed higher amounts of water compared to BRIJ-80: 62 ± 7% and 44 ± 7% of water was absorbed by BRIJ-95 and BRIJ-80, respectively, after 8 h. Drug incorporation did not interfere with
Fig. 2. Polarized light microscopy pictures of precursor formulations before and after contact with water for 2–48 h. Magnification = 200×, 25 ◦ C. At 0 h, both BRIJ-95 and BRIJ-80 were isotropic; at 2 h, BRIJ-95 was transformed into lamellar phase and BRIJ-80, into hexagonal phase. After 4 h, both formulations were transformed into hexagonal phase.
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
395
Fig. 3. Water uptake by precursor formulations. (A) Water uptake by BRIJ-95 and BRIJ-80 as function of time; (B) rates of water uptake by BRIJ-95 and BRIJ-80; (C) fitting of the swelling data using the first order equation; r2 = 0.79, p = 0.0004 for BRIJ-95 and r2 = 0.78, p = 0.0003 for BRIJ-80; (D) fitting of the swelling data using the second order equation, r2 = 0.99, p < 0.0001 for BRIJ-95 and BRIJ-80. Each point represents means ± standard deviation of 5–10 replicates. * p < 0.05 compared to BRIJ-80.
was described for monoolein and phytantriol, but not for the hydrophilic surfactant polyethoxylated 35 castor oil, which seems to uptake water according to pseudo-first order kinetics [6,17,22]. The difference in the swelling kinetics of polyethoxylated 35 castor oil and BRIJ may be related to the fact that water uptake by the former was measured during 2 h, and not until the system reached equilibrium. 3.3. In vitro drug release studies Lager amounts of drug were transported across the membrane when an aqueous solution of naltrexone at 1% (naltrexone in PBS, control) was used in the donor compartment compared to the investigated precursor formulations (containing 1% of drug) between 8 and 72 h (Fig. 4A), demonstrating the ability of the formed hexagonal phase to prolong drug release. Significantly higher amounts of naltrexone were released from BRIJ-80 compared to BRIJ-95 after 48 (p = 0.002) and 72 h (p = 0.016). At 72 h (the latest time point studied), 67 ± 7 and 59 ± 4% of drug was released from BRIJ-80 and BRIJ-95, respectively. Drug release seems to be influenced by the content of structure-forming surfactant in the liquid crystalline systems, with increased release being observed as initial content of surfactant decreases and content of water or other additives increase [22,44]. This effect may be related to small changes in system microstructure and/or decrease in viscosity, which may affect drug diffusivity and facilitate diffusion [44]. Independent of propylene glycol content, the amount of naltrexone released was found to be linear with the square root of time (at least during the first 72 h, Fig. 4), indicative of Higuchi model or diffusion-controlled release (coefficients of determination of 0.99). Similar findings were reported for drug release from reverse hexagonal phases [45], while there are inconsistencies among studies when it comes to drug release from normal type-systems. Previous studies from our group indicate that release of a lipophilic drug
over 16 h from normal-type hexagonal phases follows pseudo-first order, while a study by Farkas et al. described the release kinetics as zero-order [41,46]. A possible reason for this discrepancy may be the difference in the time window over which release was assessed: 72 h in this study and 6 h in the study by Farkas et al. [46]. Taking into consideration release only up to 8 h, linear relationships with coefficients of determination close to 0.98 for both pseudo-first order and zero-order release can be obtained. However, over 72 h, drug release kinetics could be best described by the Higuchi model. The ideal release kinetics for long-term controlled release is considered to be zero order, which means the incorporated drug is released at a constant rate to minimize the fluctuation of drug plasma levels [47]. In spite of this preference, formulations from which release is diffusion-controlled (following Higuchi model) have been demonstrated to provide sustained drug release in vivo with consistent blood levels of drug for not only hours, but days [45,48,49]. Drug release from the precursor suspension (containing naltrexone at 5%) also followed pseudo-first order kinetics (coefficient of determination = 0.985, Fig. 4B and C), suggesting that the mechanism of drug release does not depend on drug loading. Drug release from systems containing suspended drug is believed to follow three steps: dissolution of suspended drug in the matrix, followed by diffusion through the system and transfer from matrix to receptor medium [30]. Since kinetics of drug release from BRIJ-80 containing naltrexone at 5% can be described by the Higuchi model, it suggests that drug diffusion, and not the rate of drug dissolution into the matrix, controls drug release [30,50]. Further studies are necessary to address the in vivo release of the drug and whether effective plasma concentrations of naltrexone can be obtained. We expect that the precursor formulations described here will be able to uptake water, forming gels in vivo that can sustain naltrexone release. This is supported by previous studies showing a good relationship between in vitro and in vivo
396
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
Fig. 5. Concentration-dependent effect of PBS, sodium lauryl sulfate, BRIJ-80 and BRIJ-95 on the viability of fibroblasts. Each point represents means ± standard deviation of 3–4 replicates. * p < 0.05 compared to PBS, # p < 0.05 compared to BRIJ-95.
As to the obtainment of effective plasma concentrations, only speculation can be made based on the in vitro release data. It is known that plasma levels between 1 and 2 ng/mL are necessary to prevent relapses [38]. Considering that (i) pharmacokinetic parameters of naltrexone include t1/2 - 4 h and volume of distribution-1200 L [39], (ii) approximately 67% of drug is released from BRIJ-80 containing 1% of naltrexone within 72 h, and (iii) for constant blood levels, the rate of drug input and output (steady state × elimination constant) must be similar, we can predict that 100 mg of BRIJ-80 containing 1% of drug would provide steady state concentration of 0.05 ng/mL within 72 h. Approximately 3 mL of this formulation would have to be subcutaneously injected to provide steady state levels of 1.5 ng/mL, a volume at the upper limit for subcutaneous injection. On the other hand, considering the same factors, 1 mL of the precursor suspension containing 5% of naltrexone would provide steady state levels of 2 ng/mL. 3.4. Evaluation of cellular viability
Fig. 4. Cumulative in vitro release of naltrexone. (A) Release of naltrexone incorporated at 1% (w/w) in precursor formulations compared to a drug solution in PBS (control with 1% of drug); (B) release of naltrexone incorporated at 5% in BRIJ-80 compared to a drug solution in PBS (control with 5% of drug); (C) fitting of the cumulative release as function of square root of time (Higuchi model); r2 = 0.99 for BRIJ-95 and BRIJ 80 with 1% of naltrexone (p < 0.0001); r2 = 0.98 for BRIJ-80 with 5% naltrexone (p = 0.0001). Each point represents means ± standard deviation of 3–6 replicates. * p < 0.05 compared to BRIJ-95 and BRIJ-80, # p < 0.05 compared to BRIJ-95.
drug release from phytantriol-based formulations that swell and form cubic phase gels [6]. Even though the in vivo kinetics of gel formation was not studied, the formation of a system that slows down drug release was suggested by an increase in the tmax of glucose compared to a solution. It should be noted that formulation swelling may occur slower in vivo compared to in vitro due to the limited volume of fluids available, which might result in higher drug release at early time points; however, this remains to be evaluated.
Cell cultures have been widely used to evaluate irritation potential of formulations and their components [51,52], since there seems to be a good correlation between in vitro cytotoxicity assays and in vivo irritation. The cytotoxic potential of BRIJ-95 and BRIJ80 was assessed in comparison to that of PBS and sodium lauryl sulfate. PBS is considered safe, and as expected, the viability of fibroblasts (expressed as percent of controls, i.e., untreated cells) was not affected when increasing amounts of PBS (up to 100 g/mL) were added to the culture medium (Fig. 5). Compared to PBS, a significant decrease in cell viability was observed when sodium lauryl sulfate was used at a concentration as small as 1 g/mL, whereas no significant change on cell viability was detected using the same concentration of either precursor formulation. At a concentration of 10 g/mL, BRIJ-95 and BRIJ-80 significantly (p < 0.05) reduced cell viability (to 74.2 and 77.2%) compared to PBS, but not as strongly as did sodium lauryl sulfate (viability reduced to 52.3%). The cytotoxicity of BRIJ-95 and BRIJ-80 was similar at concentrations up to 10 g/mL. When a concentration of 50 g/mL was used, cell viability was significantly lower (p = 0.025) after treatment with BRIJ-95 compared to BRIJ-80, but still higher (p = 0.04) compared to sodium lauryl sulfate. Based on these cytotoxicity results, two conclusions can be drawn. First, both precursor formulations studied are safer com-
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
pared to sodium lauryl sulfate. Our observation that BRIJ-based formulations are less cytotoxic than sodium lauryl sulfate agrees with previous studies in which a reconstructured rabbit corneal epithelium model was used to compare the cytotoxicity of several surfactants [53]. The reported concentration of BRIJ to reduce cell viability to 50% was over 2 times higher than that of sodium lauryl sulfate. The second conclusion that can be drawn from the cytotoxicity results is that increasing BRIJ concentration in the precursor formulation (from 80 to 95%) increases cytotoxicity. Surfactants from the BRIJ series can affect membrane permeability, and thus, impact cell viability depending on their concentration [54]. Similar results were obtained as the concentration of polysorbate in topical microemulsions increased [24]. The fact that BRIJ-based formulations are milder to cells in vitro than sodium lauryl sulfate and BRIJ-80 is less cytotoxic than BRIJ-95 suggests that BRIJ-80 might be better tolerated in vivo. A good correlation exists between cell viability in cultured tissues and the Draize score following application of an irritant topically [55], suggesting that viability assays in cell and tissue cultures are good alternatives to predict in vivo irritation potential of formulations. However, we acknowledge that only future in vivo studies can fully address the safety of this formulation. To reduce the incidence and severity of local irritation, the search for milder surfactants for drug delivery systems is an on-going process. In recent years sugar-based surfactants (alkylpolyglucosides) have emerged as a novel class of surfactants with good biodegradability and tolerance (classified as slightly-tonon-irritating through measurement of viability of erythrocytes and bioengineered skin tissues) [56–58]. Since these surfactants have the ability to form certain liquid crystalline phases, they may also be studied as components for in situ forming gels. In conclusion, we have demonstrated that BRIJ-based precursor fluid formulations containing propylene glycol can uptake water and form hexagonal phase gels within 2–4 h of contact with an aqueous environment, providing sustained naltrexone release for 72 h. Increases in propylene glycol content from 5 to 20% in the precursor formulation decreased the amount of water necessary to form the hexagonal phase (since it suppressed lamellar phase transition), the cytotoxicity of the formulation, and increased the amount of drug released after 48 h, while having no significant effect on the kinetics of swelling and drug release. These results suggest the potential of BRIJ-based precursor formulations containing 20% of propylene glycol for sustained naltrexone release in the treatment of drug addiction. Acknowledgements The authors would like to thank Dr. Alexandre A. Steiner (Albany College of Pharmacy, Albany, NY) for critical comments on the manuscript. This work was supported by a Scholarship of Discovery Grant from Albany College of Pharmacy and Health Sciences to L.B. Lopes. J. Phelps received a Summer Research Award from the Department of Pharmaceutical Sciences of Albany College of Pharmacy and Health Sciences. References [1] P. Lobmaier, H. Kornor, N. Kunoe, A. Bjorndal, Sustained-release naltrexone for opioid dependence, Cochrane Database Syst. Rev. (2008) CD006140. [2] H.G. Roozen, R. de Waart, W. van den Brink, Efficacy and tolerability of naltrexone in the treatment of alcohol dependence: oral versus injectable delivery, Eur. Addict. Res. 13 (2007) 201–206. [3] P. Lobmaier, M. Gossop, H. Waal, J. Bramness, The pharmacological treatment of opioid addiction – a clinical perspective, Eur. J. Clin. Pharmacol. 66 (2010) 537–545. [4] K.J. Whittlesey, L.D. Shea, Delivery systems for small molecule drugs, proteins, and DNA: the neuroscience/biomaterial interface, Exp. Neurol. 190 (2004) 1–16. [5] S.L Tao, T.A. Desai, Microfabricated drug delivery systems: from particles to pores, Adv. Drug Deliv. Rev. 55 (2003) 315–328.
397
[6] W.K. Fong, T. Hanley, B.J. Boyd, Stimuli responsive liquid crystals provide ‘ondemand’ drug delivery in vitro and in vivo, J. Control. Release 135 (2009) 218–226. [7] C.C. Muller-Goymann, Physicochemical characterization of colloidal drug delivery systems such as reverse micelles, vesicles, liquid crystals and nanoparticles for topical administration, Eur. J. Pharm. Biopharm. 58 (2004) 343–356. [8] C.J Drummond, C. Fong, Surfactant self-assembly objects as novel drug delivery vehicles, Curr. Opin. Colloid Interface Sci. 4 (2000) 449–456. [9] Z. Wang, Z. Diao, F. Liu, G. Li, G. Zhang, Microstructure and rheological properties of liquid crystallines formed in Brij 97/water/IPM system, J. Colloid Interface Sci. 297 (2006) 813–818. [10] A. Otto, J. du Plessis, J.W. Wiechers, Formulation effects of topical emulsions on transdermal and dermal delivery, Int. J. Cosmet. Sci. 31 (2009) 1–19. [11] L.B. Lopes, J.L. Lopes, D.C. Oliveira, J.A. Thomazini, M.T. Garcia, M.C. Fantini, J.H. Collett, M.V. Bentley, Liquid crystalline phases of monoolein and water for topical delivery of cyclosporin A: characterization and study of in vitro and in vivo delivery, Eur. J. Pharm. Biopharm. 63 (2006) 146–155. [12] B.J. Boyd, D.V. Whittaker, S.M. Khoo, G. Davey, Lyotropic liquid crystalline phases formed from glycerate surfactants as sustained release drug delivery systems, Int. J. Pharm. 309 (2006) 218–226. [13] G.K. Hulse, V. Stalenberg, D. McCallum, W. Smit, G. O’Neil, N. Morris, R.J. Tait, Histological changes over time around the site of sustained release naltrexonepoly(dl-lactide) implants in humans, J. Control Release 108 (2005) 43–55. [14] J.E. Gomes-Filho, B.P. Gomes, A.A. Zaia, C.R. Ferraz, F.J. Souza-Filho, Evaluation of the biocompatibility of root canal sealers using subcutaneous implants, J.Appl.Oral Sci. 15 (2007) 186–194. [15] K. Al-Tahami, M. Oak, J. Singh, Controlled delivery of basal insulin from phase-sensitive polymeric systems after subcutaneous administration: in vitro release, stability, biocompatibility, in vivo absorption, and bioactivity of insulin, J.Pharm.Sci. 100 (2011) 2161–2171. [16] F Rosevear, The microscopy of the liquid crystalline neat and middle phases of soaps and synthetic detergents, J.Am.Oil Chem.Soc. 31 (1954) 628–639. [17] A. Feher, E. Urban, I. Eros, P. Szabo-Revesz, E. Csanyi, Lyotropic liquid crystal preconcentrates for the treatment of periodontal disease, Int.J.Pharm. 358 (2008) 23–26. [18] S.B. Rizwan, T. Hanley, B.J. Boyd, T. Rades, S. Hook, Liquid crystalline systems of phytantriol and glyceryl monooleate containing a hydrophilic protein: characterisation, swelling and release kinetics, J.Pharm.Sci. 98 (2009) 4191–4204. [19] G.H Pollack, Cells, Gels and the Engines of Life: A New, Unifying Approach to Cell Biology, Ebner and Sons, 2001. [20] M.A. Serban, Y. Liu, G.D. Prestwich, Effects of extracellular matrix analogues on primary human fibroblast behavior, Acta Biomater. 4 (2008) 67–75. [21] C.M. Chang, R. Bodmeier, Swelling of and drug release from monoglyceridebased drug delivery systems, J.Pharm.Sci. 86 (1997) 747–752. [22] M.G. Lara, M.V. Bentley, J.H. Collett, In vitro drug release mechanism and drug loading studies of cubic phase gels, Int. J. Pharm. 293 (2005) 241–250. [23] C. Verhulst, C. Coiffard, L.J. Coiffard, P. Rivalland, Y. De Roeck-Holtzhauer, In vitro correlation between two colorimetric assays and the pyruvic acid consumption by fibroblasts cultured to determine the sodium laurylsulfate cytotoxicity, J. Pharmacol. Toxicol. Methods 39 (1998) 143–146. [24] J. Hosmer, R. Reed, M.V. Bentley, A. Nornoo, L.B. Lopes, Microemulsions containing medium-chain glycerides as transdermal delivery systems for hydrophilic and hydrophobic drugs, AAPS PharmSciTech. 10 (2009) 589–596. [25] D.M. Bagley, J.R. Gardner, G. Holland, R.W. Lewis, H. Vrijhof, A.P. Walker, Eye irritation: updated reference chemicals data bank, Toxicol. In Vitro 13 (1999) 505–510. [26] Cosmetic Ingredient Review, Final report on the safety assessment of sodium lauryl sulfate and ammonium lauryl sulfate, Int. J. Toxicol. 2 (1983) 127–181. [27] V.C. Robinson, W.F. Bergfeld, D.V. Belsito, R.A. Hill, C.D. Klaassen, J.G. Marks Jr., R.C. Shank, T.J. Slaga, P.W. Snyder, F. Alan Andersen, Final report of the amended safety assessment of sodium laureth sulfate and related salts of sulfated ethoxylated alcohols, Int. J. Toxicol. 29 (2011) 151S–161S. [28] C. Faller, M. Bracher, N. Dami, R. Roguet, Predictive ability of reconstructed human epidermis equivalents for the assessment of skin irritation of cosmetics, Toxicol. In Vitro 16 (2002) 557–572. [29] F. Caboi, G.S. Amico, P. Pitzalis, M. Monduzzi, T. Nylander, K. Larsson, Addition of hydrophilic and lipophilic compounds of biological relevance to the monoolein/water system I. Phase behavior, Chem. Phys. Lipids 109 (2001) 47–62. [30] L.S. Helledi, L. Schubert, Release kinetics of acyclovir from a suspension of acyclovir incorporated in a cubic phase delivery system, Drug Dev. Ind. Pharm. 27 (2001) 1073–1081. [31] E. Ruel-Gariepy, J.C. Leroux, In situ-forming hydrogels—review of temperaturesensitive systems, Eur. J. Pharm. Biopharm. 58 (2004) 409–426. [32] J.O. You, D. Almeda, G.J. Ye, D.T. Auguste, Bioresponsive matrices in drug delivery, J. Biol. Eng. 4 (2010) 15. [33] L. Zhao, L. Zhu, F. Liu, C. Liu, D. Shan, Q. Wang, C. Zhang, J. Li, J. Liu, X. Qu, Z. Yang, pH triggered injectable amphiphilic hydrogel containing doxorubicin and paclitaxel, Int. J. Pharm. 410 (2011) 83–91. [34] S. Jauhari, A.K. Dash, A mucoadhesive in situ gel delivery system for paclitaxel, AAPS PharmSciTech. 7 (2006) E53. [35] A.S. Sallam, E. Khalil, H. Ibrahim, I. Freij, Formulation of an oral dosage form utilizing the properties of cubic liquid crystalline phases of glyceryl monooleate, Eur. J. Pharm. Biopharm. 53 (2002) 343–352. [36] B.A. Johnson, Naltrexone long-acting formulation in the treatment of alcohol dependence, Ther. Clin. Risk Manag. 3 (2007) 741–749.
398
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
[37] W. Yin, E.O. Akala, R.E. Taylor, Design of naltrexone-loaded hydrolyzable crosslinked nanoparticles, Int. J. Pharm. 244 (2002) 9–19. [38] H.T. Ngo, D.E. Arnold-Reed, R.C. Hansson, R.J. Tait, G.K. Hulse, Blood naltrexone levels over time following naltrexone implant, Prog. Neuropsychopharmacol. Biol. Psychiatry 32 (2008) 23–28. [39] L.I. Giannola, V. De Caro, G. Giandalia, M.G. Siragusa, C. Tripodo, A.M. Florena, G. Campisi, Release of naltrexone on buccal mucosa: permeation studies, histological aspects and matrix system design, Eur. J. Pharm. Biopharm. 67 (2007) 425–433. [40] M Milewski, A.L. Stinchcomb, Vehicle composition influence on the microneedle-enhanced transdermal flux of naltrexone hydrochloride, Pharm. Res. 28 (2011) 124–134. [41] J. Hosmer, S.H. Shin, A. Nornoo, H. Zheng, L.B. Lopes, Influence of internal structure and composition of liquid crystalline phases on topical delivery of paclitaxel, J.Pharm.Sci. (2010), doi:10.1002/jps.22370. [42] K.L. Williams, C.L. Broadbridge, Potency of naltrexone to reduce ethanol self-administration in rats is greater for subcutaneous versus intraperitoneal injection, Alcohol 43 (2009) 119–126. [43] K. Derakhshandeh, M. Fashi, S. Seifoleslami, Thermosensitive Pluronic hydrogel: prolonged injectable formulation for drug abuse, Drug Des.Dev. Ther. 4 (2010) 255–262. [44] M. Makai, E. Csanyi, Z. Nemeth, J. Palinkas, I. Eros, Structure and drug release of lamellar liquid crystals containing glycerol, Int. J. Pharm. 256 (2003) 95–107. [45] K.W. Lee, T.H. Nguyen, T. Hanley, B.J. Boyd, Nanostructure of liquid crystalline matrix determines in vitro sustained release and in vivo oral absorption kinetics for hydrophilic model drugs, Int. J. Pharm. 365 (2009) 190–199. [46] E. Farkas, R. Zelko, Z. Nemeth, J. Palinkas, S. Marton, I. Racz, The effect of liquid crystalline structure on chlorhexidine diacetate release, Int. J. Pharm. 193 (2000) 239–245. [47] Y. Tang, J. Singh, Thermosensitive drug delivery system of salmon calcitonin: in vitro release, in vivo absorption, bioactivity and therapeutic efficacies, Pharm. Res. 27 (2010) 272–284. [48] M.R Park, C. Chun, S.W. Ahn, M.H. Ki, C.S. Cho, S.C. Song, Sustained delivery of human growth hormone using a polyelectrolyte complex-loaded thermosensitive polyphosphazene hydrogel, J. Control Release 147 (2010) 359–367.
[49] Y. Ito, Y. Ochii, K. Fukushima, N. Sugioka, K. Takada, Three-layered microcapsules as a long-term sustained release injection preparation, Int. J. Pharm. 384 (2011) 53–59. [50] S.K. Chandrasekaran, D.R. Paul, Dissolution-controlled transport from dispersed matrixes, J. Pharm. Sci. 71 (1982) 1399–1402. [51] T. Benavides, V. Martinez, M. Mitjans, M.R. Infante, C. Moran, P. Clapes, R. Clothier, M.P. Vinardell, Assessment of the potential irritation and photoirritation of novel amino acid-based surfactants by in vitro methods as alternative to the animal tests, Toxicology 201 (2004) 87–93. [52] H.C. Korting, T. Herzinger, A. Hartinger, M. Kerscher, T. Angerpointner, H.I. Maibach, Discrimination of the irritancy potential of surfactants in vitro by two cytotoxicity assays using normal human keratinocytes, HaCaT cells and 3T3 mouse fibroblasts: correlation with in vivo data from a soap chamber assay, J. Dermatol. Sci. 7 (1994) 119–129. [53] S. Matsuda, M. Hisama, H. Shibayama, N. Itou, M. Iwaki, In vitro eye irritancy test of lauryl derivatives and polyoxyethylene alkyl derivatives with the reconstructed rabbit corneal epithelium model, J. Oleo Sci. 58 (2009) 437–442. [54] Y.L. Lo, Relationships between the hydrophilic–lipophilic balance values of pharmaceutical excipients and their multidrug resistance modulating effect in Caco-2 cells and rat intestines, J. Control Release 90 (2003) 37–48. [55] T. Watanabe, T. Hasegawa, H. Takahashi, T. Ishibashi, H. Itagaki, K. Sugibayashi, Utility of MTT assay in three-dimensional cultured human skin model as an alternative for draize skin irritation test: approach using diffusion law of irritant in skin and toxicokinetics–toxicodynamics correlation, Pharm. Res. 19 (2002) 669–675. [56] N. Pakpayat, F. Nielloud, R. Fortune, C. Tourne-Peteilh, A. Villareal, I. Grillo, B. Bataille, Formulation of ascorbic acid microemulsions with alkyl polyglycosides, Eur. J. Pharm. Biopharm. 72 (2009) 444–452. [57] I. Pezron, L. Galet, D. Clausse, Surface Interaction between a protein monolayer and surfactants and its correlation with skin irritation by surfactants, J. Colloid Interface Sci. 180 (1996) 285–289. [58] A. Mehling, M. Kleber, H. Hensen, Comparative studies on the ocular and dermal irritation potential of surfactants, Food Chem. Toxicol. 45 (2007) 747–758.
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
In situ gelling hexagonal phases for sustained release of an anti-addiction drug Jessica Phelps a , M. Vitória L.B. Bentley b , Luciana B. Lopes a,∗ a b
Albany College of Pharmacy and Health Sciences, Albany, NY, USA Faculdade de Ciências Farmacêuticas de Ribeirão Preto, Universidade de São Paulo, Ribeirão Preto, SP, Brazil
a r t i c l e
i n f o
Article history: Received 30 March 2011 Received in revised form 16 May 2011 Accepted 25 May 2011 Available online 1 June 2011 Keywords: Hexagonal phase Naltrexone Subcutaneous delivery In situ gelling Swelling
a b s t r a c t In this study, fluid precursor formulations for subcutaneous injection and in situ formation of hexagonal phase gels upon water absorption were developed as a strategy to sustain the release of naltrexone, a drug used for treatment of drug addiction. Precursor formulations were obtained by combining BRIJ 97 with propylene glycol (PG, 5–70%, w/w). To study the phase behavior of these formulations, water was added at 10–90% (w/w), and the resulting systems were characterized by polarized light microscopy. Two precursor formulations containing BRIJ:PG at 95:5 (w/w, referred to as BRIJ-95) and at 80:20 (w/w, referred to as BRIJ-80) were chosen. Naltrexone was dissolved at 1% or suspended at 5% (w/w). Precursor formulations were transformed into hexagonal phases when water content exceeded 20%. Water uptake followed second-order kinetics, and after 2–4 h all precursor formulations were transformed into hexagonal phases. Drug release was prolonged by the precursor formulations (compared to a drug solution in PBS), and followed pseudo-first order kinetics regardless of naltrexone concentration. The release from BRIJ-80 was significantly higher than that from BRIJ-95 after 48 h. The relative safety of the precursor formulations was assessed in cultured fibroblasts. Even though BRIJ-95 was more cytotoxic than BRIJ-80, both precursor formulations were significantly less cytotoxic than sodium lauryl sulfate (considered moderate-to-severe irritant) at the same concentration (up to 50 g/mL). These results suggest the potential of BRIJ-based precursor formulations for sustained naltrexone release. © 2011 Elsevier B.V. All rights reserved.
1. Introduction The non-selective opioid antagonist naltrexone is used to prevent relapse in former opioid- and alcohol-dependent patients [1,2]. Daily administration of oral doses of naltrexone at 50–100 mg has been clinically used for several years [3]. However, oral administration of naltrexone is complicated by its extensive firstpass hepatic metabolism, which results in variable and low drug bioavailability (5–40%), and increased costs [3]. In addition, the requirement of daily drug administration for treatment of chronic disorders often results in reduced patient compliance and increased probability of relapse [4,5]. Delivery systems that can sustain naltrexone delivery for longer periods of time offer an alternative to overcome these problems and improve patient compliance [2]. Based on these facts, the present study was aimed at developing hexagonal liquid crystalline gels for naltrexone delivery. Because the hexagonal phase is viscous and difficult to inject [6], we developed fluid, free-flowing precursor formulations that could be
∗ Corresponding author at: Department of Pharmaceutical Sciences, Albany College of Pharmacy and Health Sciences, 106 New Scotland Ave., Albany, NY 12208, USA. Tel.: +1 518 694 7113. E-mail address: [email protected] (L.B. Lopes). 0927-7765/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2011.05.048
injected subcutaneously to form a hexagonal liquid crystalline gel in situ upon absorption of water from the tissue. The hexagonal phase consists of cylinders of micelles arranged as a hexagon [7,8]. Depending on the structure-forming compound, water and hydrophilic drugs can be enclosed within the cylinders of the hexagonal phase (reverse type) or outside of them (normal type) [9,10]. So far, most of the studies have employed hexagonal phases of the reverse type for sustained drug release [6,11]. Such phases are formed by lipids (like monoolein), can resist dilution with body fluids for long periods of time and provide very slow release of hydrophilic compounds [6,12]. In this study, we propose to use a hexagonal phase of the normal type formed by the hydrophilic surfactant polyoxyethylene-10-oleyl ether (BRIJ 97) [9]. The rationale for this choice is based on the assumption that even though the hexagonal phase gel of the normal type should persist at the site of injection for days prolonging drug release, BRIJ hydrophilicity would account for a faster (though still prolonged) dissolution of the gel in body fluids compared to other long-lasting systems, decreasing incidence and duration of adverse reactions at the site of injection [13–15]. Different amounts of BRIJ were combined with propylene glycol to obtain injectable fluid precursor formulations, which gave rise to hexagonal phase approximately 2–4 h after contact with aqueous environment. The swelling kinetics of precursor formulations, drug
392
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
release properties and cytotoxic potential of the formulations were evaluated. 2. Materials and methods 2.1. Material Propylene glycol (PG), polyoxyethylene-10-oleyl ether (BRIJ 97), naltrexone hydrochloride, hydroxypropyl cellulose and phosphatebuffered saline (PBS) were obtained from Sigma–Aldrich (St. Louis, MO). 2.2. Methods 2.2.1. Development of precursor formulations Precursor formulations were obtained by mixing BRIJ with propylene glycol (PG) at ratios varying from 95:5 to 30:70 (w/w). Propylene glycol was included to facilitate drug incorporation and to allow producing fluid formulations. The precursor formulations were visually inspected for transparency and fluidity, and visualized under a polarized light microscope (Carl Zeiss, Oberkochen, Germany) for assessment of isotropicity. 2.2.2. Phase behavior of precursor formulations upon contact with water To evaluate the formation of hexagonal phases by the precursor formulations upon contact with water, each BRIJ:PG ratio was mixed with various amounts of water (10–90%, w/w), and the formulations were allowed to equilibrate in closed vials for 3 days at room temperature. Systems were characterized by visual inspection and polarized light microscopy. Visual inspection was used to verify sample homogeneity and phase separation. Polarized light microscopy allowed for the characterization of the hexagonal structure, which displays a fan like and angular texture under the microscope [16]. Phase diagrams were constructed to show the relationship between the formulation composition and phase behavior. Two precursor formulations (combinations of BRIJ:PG at 95:5 and 80:20, w/w) were chosen for further studies based on the fact that they were isotropic and fluid at room temperature, but formed stiffer hexagonal phase gels at relatively low water content (20 or 30%, see Section 3). For the sake of simplicity, the BRIJ:PG mixture at 95:5 (w/w) will be referred to as BRIJ-95, whereas the BRIJ:PG mixture at 80:20 (w/w), will be referred to as BRIJ-80. The influence of temperature on the structure of BRIJ-95 and BRIJ-80 containing water at 10–90% (w/w) was evaluated using a polarized light microscope coupled with a hot stage (Linkam, Surrey, UK). Temperature was increased at 1 ◦ C/min from 25 to 40 ◦ C, and changes in the structure of the systems were monitored by polarized light microscopy. To evaluate the influence of drug incorporation on formation of hexagonal phase, naltrexone was added to the precursor formulations at 0.5, 0.75, 1% or 5% (w/w). Samples containing up to 1% of naltrexone were clear and fluid, whereas those containing 5% were more turbid, visually similar to suspensions. Water was added at 10–90% (w/w), samples were left to equilibrate for 3 days at room temperature in closed vials and subsequently characterized by visual inspection and polarized light microscopy. 2.2.3. Swelling studies Swelling studies were performed to evaluate the rate and kinetics of water absorption by the two chosen precursor formulations (BRIJ-95 and BRIJ-80). Two experiments were performed in association with this study. The first set of experiments aimed to determine the length of time necessary for the fluid precursor formulation to be transformed into the hexagonal phase upon water uptake.
Twelve-well plates and cell culture inserts were used in this experiment. Each well of the 12-well cell culture plate was filled with 2 mL of 2% hydroxypropyl cellulose gel. This gel was chosen because it mimics better the subcutaneous tissue compared to aqueous solutions used in most swelling studies [17,18]. This is based on the fact that the intra and extracellular space resembles more closely gels than liquid solutions since water in tissues does not act like bulk water (i.e., is not freely available), but is mostly associated with surfaces, proteins and other cell components forming gels [19,20]. BRIJ-95 and BRIJ-80 fluid precursor formulations (200 mg) were added to cell culture inserts, and each insert was placed in contact with the hydroxypropyl cellulose gel. Samples of the formulations in the cell culture inserts were collected at 2, 4, 6, 8, 16, 24, 72 and 96 h and the structure of the system was identified under a polarized light microscope. In a parallel set of experiments, aimed to determine the rate of water absorption by the precursor formulation, cell culture inserts containing the precursor formulations (200 mg) were placed in contact with the hydroxypropyl cellulose gel, and weighed at 0.5 or 1 h intervals up to 8 h, and then at 24 and 36 h. A cellulose dialysis membrane (cut-off 1000 Da) was placed between the hydroxypropyl cellulose gel and cell insert to prevent adhesion of the gel on the inserts (which would prevent the weighing of the inserts). The amount of water absorbed was plotted as function of time and fitted using the first-order kinetic equation:
ln
W∞ W∞ − W
= kt,
(1)
or second-order order equation t 1 t + = , W k W ∞2 W ∞
(2)
where W∞ is the maximum water uptake, W is the water uptake at time t, and W∞ − W is the unrealized water uptake and k is the rate constant [21,22]. The rate of swelling was determined by plotting the difference of swelling at two consecutive time points as function of time. 2.2.4. In vitro drug release studies The in vitro release of naltrexone from BRIJ-95 and BRIJ-80 was evaluated using Franz diffusion cells, and phosphate-buffered saline (100 mM, pH 7.2) under constant stirring (350 rpm) at 37.0 ± 0.5◦C as receptor phase. A cellulose membrane (cutoff 1000 Da, Sigma, St Louis, MO) was placed between the receptor and donor compartments. The receptor phase was collected at 2, 4, 6, 8, 16, 24, 48 and 72 h post-application. As stated before, the precursor formulations are isotropic fluid systems that turn into hexagonal phases upon contact with water. Therefore, this transformation occurred during the in vitro release study. The formation of the hexagonal phase structure in the donor compartment was confirmed by polarized light microscopy. Naltrexone in the receptor phase was quantified by UVspectrophotometry at 215 nm. Standard curves of naltrexone in PBS in the range of 1–20 g/mL were used for calibration curves. The method was linear up to 50 g/mL. Interference of formulation components in the analytical method was assessed by adding unloaded (not containing drug) formulations in the donor compartment of diffusion cells and assaying the receptor phase after 24 h. No formulation component interfered with the analytical method. 2.2.5. Evaluation of cellular viability To evaluate the relative safety of the precursor formulations, we compared their cytotoxic effects to those of PBS (considered safe) and of sodium lauryl sulfate (considered a moderate-to-severe irritant) [23,24]. The classification of sodium lauryl sulfate as a moderate-to-severe irritant is based on its effects on the skin and
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
eyes of rabbits as determined in Draize tests [25]. In acute ocular tests, sodium lauryl sulfate at 5.1–10% caused moderate irritation to the rabbits’ eyes if not rinsed, while at 21%, it caused severe irritation. Similarly to its effects on the eyes, between 10 and 30% sodium lauryl sulfate caused skin corrosion and severe irritation [26,27]. Because of the ethical implications of the Draize test, measurement of viability and release of cytokines in cells and tissues in culture has been used as alternative in vitro assays to determine the irritation potential of formulations [28]. An in vitro cell culture assay was used in this study. Murine swiss 3T3 mouse fibroblasts were obtained from American Type Culture Collection (ATCC, Manassas, VA), and grown at 37 ◦ C and 5% CO2 atmosphere in Dulbecco’s modification of Eagle’s medium (DMEM, ATCC, Manassas, VA) containing 10% fetal bovine serum (FBS, Gibco, Carlsbad, CA) and additional penicillin and streptomycin (1%). For the cellular viability assay, cells were plated in 96-well plates (6000 cells/well), and treated for 12 h with either PBS, BRIJ-95, BRIJ-80 or sodium lauryl sulfate at concentrations ranging from 1 to 100 g/mL in cell culture medium. Cell survival was evaluated using a cell proliferation assay reagent (CellTiter 96 Aqueous One solution, Promega, Madison, WI) consisting of 3-(4,5-dimethylthiaziazol-2-yl)-5-(3carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium salt (MTS) and an electron coupling reagent. The MTS salt is reduced to a colored formazan product, and the amount of this product is directly proportional to the number of living cells. After treatment, cells were washed with PBS, and 100 L of cell culture medium plus 20 L of the cell proliferation assay reagent were added to each well. The plates were incubated for 2 h at 37 ◦ C in a humidified atmosphere of 5% CO2 in an incubator, and the absorbance was recorded at 490 nm using a plate reader (SpectraMax, Molecular Devices Corp., Sunnyvale, CA). These experiments were performed using 3–4 replicates; in other words, 3–4 wells containing cells (between passages 2 and 6) were treated with each concentration of test formulations or control solutions. 2.2.6. Statistical analysis The results were reported as means ± standard deviation. Data from swelling studies were statistically analyzed using one way ANOVA (followed by Tukey post-hoc test, GraphPad Prism 4 software), whereas analysis of data from the in vitro release studies was performed using repeated-measures ANOVA (followed by Fisher post-hoc test, Statistica Advanced 8.2 software). Values were considered significantly different when p < 0.05. 3. Results and discussion 3.1. Development of precursor formulations and evaluation of their phase behavior Because the hexagonal phase is too viscous to be injected, we developed fluid precursor formulations composed of BRIJ:PG mixtures at 95:5–30:70 (w/w). The precursor formulations were transparent and isotropic (under polarized light). The precursor formulations gave rise to several phases in the presence of water (Fig. 1). The hexagonal phase was observed predominantly when BRIJ was present at a concentration of 35% (w/w) and higher, PG was below 35% (w/w) and water was between 20 and 60% (w/w). This phase co-existed with excess water (hexagonal phase + water) when water content varied from 20 to 70% and PG varied from 5 to 40%. At higher (>80%) percentages of water, the hexagonal phases were transformed into liquid isotropic phases (probably micellar systems) [9]. This transformation occurred with hexagonal phase + water samples 8–10 days after preparation even if no additional water was added, and may be attributed to rear-
393
rangement of BRIJ and formation of micellar systems due to its hydrophilicity [9]. Phase transformation triggered by water uptake in BRIJ:PG mixtures at 95:5 or 90:10 (w/w) occurred in the following order: liquid isotropic system, lamellar phase, hexagonal phase, hexagonal phase + water and liquid isotropic systems (Fig. 1). The lamellar phase was observed only when PG content in the precursor formulation was equal or smaller than 10%. As PG content in the precursor formulation increased, the hexagonal phase existed at narrower aqueous contents, whereas the hexagonal phase + water was formed at smaller percentages of water (Fig. 1). In other words, the hexagonal phase could uptake less water. In precursor formulations containing 40% of PG (BRIJ:PG 60:40), the hexagonal phase + water was observed with water at only 30% (instead of 70% as in BRIJ:PG at 90:10). The hexagonal phase was no longer observed when PG content in the precursor formulation was higher than 50%. We chose precursor formulations with BRIJ:PG ratios of 80:20 (BRIJ-80) and 95:5 (BRIJ-95) for further investigation based on two facts: first, both allowed obtainment of the hexagonal phase at relatively low water content (20 or 30%), indicating that in situ absorption of low amounts of water would be necessary for formation of the gel; second, uptake of 20% of water by BRIJ-95 (but not by BRIJ-80) led to formation of lamellar phase (transformed into hexagonal phase with water content over 30%), which would give us the opportunity to evaluate whether drug release was affected by transient formation of the lamellar phase. No phase transition at temperatures ranging from 25 to 38 ◦ C was observed when either BRIJ-80 or BRIJ-95 was combined with various amounts of water (10–90%), suggesting that the hexagonal phase gel is stable at body temperature. We next evaluated the influence of drug load on phase behavior of the precursor formulations. Naltrexone could be dissolved in BRIJ-95 and BRIJ-80 at concentrations up to 1%. Compared to the unloaded (not containing drug) systems, naltrexone at 1% led to formation of the hexagonal phase + water at lower aqueous contents: 60–70% and 40–70% for BRIJ-95 and BRIJ-80, respectively (Fig. 1). Because we used naltrexone as hydrochloride, interactions between the dissolved drug and the polar headgroup of surfactant may be possible, affecting the interfacial curvature of the surfactant layer, and causing phase transition [29]. These types of interactions may cause a decrease in the amount of water the hexagonal phase can uptake. Previous studies reported the ability of other drugs (such as cyclosporine) to interact with the polar headgroup of structure-forming compounds, decreasing the amount of water that liquid crystalline phases can incorporate [11]. Addition of naltrexone at concentrations higher than 1% led to formation of suspensions. Nevertheless, we evaluated the phase behavior of BRIJ-80 containing 5% of the drug to assess the influence of drug crystals. Naltrexone was incorporated at 5% only in BRIJ-80 since this formulation had smaller cytotoxicity potential while still able to sustain drug release (see results for drug release studies and evaluation of cellular viability). As the formulation absorbed water, it became transparent due to naltrexone dissolution. There was no difference in the type of systems formed as function of water content when naltrexone was added at 1 or 5%, suggesting that presence of suspended drug within the studied range does not cause phase transition. Similar results were described in a previous study by Helledi et al. in which the authors incorporated acyclovir up to 45% in monoolein-based systems [30]. The use of materials that respond to the biological environment to trigger formation of delivery systems and/or to control drug release has been attracting increasing interest. Not only absorption of water, but also changes in temperature and pH (among other factors) can trigger the formation of modified-release systems [31–33]. Taking advantage of the fact that certain surfactants and polar lipids can uptake water, in situ-forming delivery systems
394
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
Fig. 1. Phase diagrams of BRIJ, PG and water without naltrexone (NTX) or containing the drug at final concentrations of 1 or 5% (w/w) at 25 ◦ C.
for intratumoral, oral, subcutaneous and periodontal administrations of drugs have been proposed [6,17,34,35]. To our knowledge, this is the first report employing in situ gelling hexagonal phases for sustained release of naltrexone. Other delivery systems have been investigated to prolong the release of naltrexone and decrease dosing frequency, including intramuscular microparticles, transdermal devices, buccal tablets and subcutaneous implants [36–40]. Compared to these systems, in situ gelling hexagonal phases offer several potential advantages: first, they provide sustained drug release for variable amounts of time depending on their composition; second, naltrexone potency in reducing ethanol self-administration seems to be increased when administered subcutaneously compared to other routes; and third, subcutaneous injection of fluid formulations does not require extensive training and surgical procedures like subcutaneous implants [6,15,31,41,42]. A previous study has investigated subcutaneous injectable thermosensitive poloxamer gels for naltrexone release, but sustained release for over 100 h was only obtained when the drug was complexed with cyclodextrins, which adds one more step to the preparation of precursor formulations, and increases cost [43]. The fact that BRIJ-95 and BRIJ-80 are very easy to prepare, can form hexagonal phase over a wide range of aqueous content, and naltrexone addition (within the concentration studied) does not preclude hexagonal phase formation suggests the potential of these formulations for use as in situ gelling delivery systems.
water uptake rate (data not shown). This is consistent with the phase behavior depicted in Fig. 1, which shows transition to hexagonal phase + excess water (after which additional water uptake is not expected) at aqueous content higher than 50% for BRIJ-80 and 60% for BRIJ-95. The maximum rate of water uptake was achieved after 1 h for both precursor formulations (Fig. 3B), but was significantly (p < 0.001) higher for BRIJ-95 (23 ± 5%/h for BRIJ-95 and 15 ± 3%/h for BRIJ-80). Fitting of the swelling data to Eqs. (1) and (2) suggests that swelling can be better described by second order kinetics (coefficients of determination of approximately 0.99 for BRIJ-95 and BRIJ-80, Fig. 3C and D). Similar kinetics of water uptake
3.2. Swelling studies The rate and kinetics of water absorption by BRIJ-95 and BRIJ-80 were determined through two swelling studies. Fig. 2 shows polarized light microscopy pictures of the precursor formulations at 0, 2, 4, and 48 h after contact with the aqueous phase. At 0 h, both fluid precursor formulations were clear, liquid isotropic phases. At 2 h, the BRIJ-95 was transformed into lamellar phase, and between 2 and 4 h, into the hexagonal phase. On the other hand, BRIJ-80 was transformed directly into the hexagonal phase after only 2 h of contact with the aqueous gel. Both systems sustained the hexagonal structure for approximately 96 h. After this period of time, the structure of the hexagonal phase started to collapse, and after 144 h, it was transformed into an isotropic fluid system. The rate of water uptake by the precursor formulations is shown in Fig. 3. Both precursor formulations quickly absorbed water up to 8 h, when they reached their respective equilibrium concentration. BRIJ-95 absorbed higher amounts of water compared to BRIJ-80: 62 ± 7% and 44 ± 7% of water was absorbed by BRIJ-95 and BRIJ-80, respectively, after 8 h. Drug incorporation did not interfere with
Fig. 2. Polarized light microscopy pictures of precursor formulations before and after contact with water for 2–48 h. Magnification = 200×, 25 ◦ C. At 0 h, both BRIJ-95 and BRIJ-80 were isotropic; at 2 h, BRIJ-95 was transformed into lamellar phase and BRIJ-80, into hexagonal phase. After 4 h, both formulations were transformed into hexagonal phase.
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
395
Fig. 3. Water uptake by precursor formulations. (A) Water uptake by BRIJ-95 and BRIJ-80 as function of time; (B) rates of water uptake by BRIJ-95 and BRIJ-80; (C) fitting of the swelling data using the first order equation; r2 = 0.79, p = 0.0004 for BRIJ-95 and r2 = 0.78, p = 0.0003 for BRIJ-80; (D) fitting of the swelling data using the second order equation, r2 = 0.99, p < 0.0001 for BRIJ-95 and BRIJ-80. Each point represents means ± standard deviation of 5–10 replicates. * p < 0.05 compared to BRIJ-80.
was described for monoolein and phytantriol, but not for the hydrophilic surfactant polyethoxylated 35 castor oil, which seems to uptake water according to pseudo-first order kinetics [6,17,22]. The difference in the swelling kinetics of polyethoxylated 35 castor oil and BRIJ may be related to the fact that water uptake by the former was measured during 2 h, and not until the system reached equilibrium. 3.3. In vitro drug release studies Lager amounts of drug were transported across the membrane when an aqueous solution of naltrexone at 1% (naltrexone in PBS, control) was used in the donor compartment compared to the investigated precursor formulations (containing 1% of drug) between 8 and 72 h (Fig. 4A), demonstrating the ability of the formed hexagonal phase to prolong drug release. Significantly higher amounts of naltrexone were released from BRIJ-80 compared to BRIJ-95 after 48 (p = 0.002) and 72 h (p = 0.016). At 72 h (the latest time point studied), 67 ± 7 and 59 ± 4% of drug was released from BRIJ-80 and BRIJ-95, respectively. Drug release seems to be influenced by the content of structure-forming surfactant in the liquid crystalline systems, with increased release being observed as initial content of surfactant decreases and content of water or other additives increase [22,44]. This effect may be related to small changes in system microstructure and/or decrease in viscosity, which may affect drug diffusivity and facilitate diffusion [44]. Independent of propylene glycol content, the amount of naltrexone released was found to be linear with the square root of time (at least during the first 72 h, Fig. 4), indicative of Higuchi model or diffusion-controlled release (coefficients of determination of 0.99). Similar findings were reported for drug release from reverse hexagonal phases [45], while there are inconsistencies among studies when it comes to drug release from normal type-systems. Previous studies from our group indicate that release of a lipophilic drug
over 16 h from normal-type hexagonal phases follows pseudo-first order, while a study by Farkas et al. described the release kinetics as zero-order [41,46]. A possible reason for this discrepancy may be the difference in the time window over which release was assessed: 72 h in this study and 6 h in the study by Farkas et al. [46]. Taking into consideration release only up to 8 h, linear relationships with coefficients of determination close to 0.98 for both pseudo-first order and zero-order release can be obtained. However, over 72 h, drug release kinetics could be best described by the Higuchi model. The ideal release kinetics for long-term controlled release is considered to be zero order, which means the incorporated drug is released at a constant rate to minimize the fluctuation of drug plasma levels [47]. In spite of this preference, formulations from which release is diffusion-controlled (following Higuchi model) have been demonstrated to provide sustained drug release in vivo with consistent blood levels of drug for not only hours, but days [45,48,49]. Drug release from the precursor suspension (containing naltrexone at 5%) also followed pseudo-first order kinetics (coefficient of determination = 0.985, Fig. 4B and C), suggesting that the mechanism of drug release does not depend on drug loading. Drug release from systems containing suspended drug is believed to follow three steps: dissolution of suspended drug in the matrix, followed by diffusion through the system and transfer from matrix to receptor medium [30]. Since kinetics of drug release from BRIJ-80 containing naltrexone at 5% can be described by the Higuchi model, it suggests that drug diffusion, and not the rate of drug dissolution into the matrix, controls drug release [30,50]. Further studies are necessary to address the in vivo release of the drug and whether effective plasma concentrations of naltrexone can be obtained. We expect that the precursor formulations described here will be able to uptake water, forming gels in vivo that can sustain naltrexone release. This is supported by previous studies showing a good relationship between in vitro and in vivo
396
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
Fig. 5. Concentration-dependent effect of PBS, sodium lauryl sulfate, BRIJ-80 and BRIJ-95 on the viability of fibroblasts. Each point represents means ± standard deviation of 3–4 replicates. * p < 0.05 compared to PBS, # p < 0.05 compared to BRIJ-95.
As to the obtainment of effective plasma concentrations, only speculation can be made based on the in vitro release data. It is known that plasma levels between 1 and 2 ng/mL are necessary to prevent relapses [38]. Considering that (i) pharmacokinetic parameters of naltrexone include t1/2 - 4 h and volume of distribution-1200 L [39], (ii) approximately 67% of drug is released from BRIJ-80 containing 1% of naltrexone within 72 h, and (iii) for constant blood levels, the rate of drug input and output (steady state × elimination constant) must be similar, we can predict that 100 mg of BRIJ-80 containing 1% of drug would provide steady state concentration of 0.05 ng/mL within 72 h. Approximately 3 mL of this formulation would have to be subcutaneously injected to provide steady state levels of 1.5 ng/mL, a volume at the upper limit for subcutaneous injection. On the other hand, considering the same factors, 1 mL of the precursor suspension containing 5% of naltrexone would provide steady state levels of 2 ng/mL. 3.4. Evaluation of cellular viability
Fig. 4. Cumulative in vitro release of naltrexone. (A) Release of naltrexone incorporated at 1% (w/w) in precursor formulations compared to a drug solution in PBS (control with 1% of drug); (B) release of naltrexone incorporated at 5% in BRIJ-80 compared to a drug solution in PBS (control with 5% of drug); (C) fitting of the cumulative release as function of square root of time (Higuchi model); r2 = 0.99 for BRIJ-95 and BRIJ 80 with 1% of naltrexone (p < 0.0001); r2 = 0.98 for BRIJ-80 with 5% naltrexone (p = 0.0001). Each point represents means ± standard deviation of 3–6 replicates. * p < 0.05 compared to BRIJ-95 and BRIJ-80, # p < 0.05 compared to BRIJ-95.
drug release from phytantriol-based formulations that swell and form cubic phase gels [6]. Even though the in vivo kinetics of gel formation was not studied, the formation of a system that slows down drug release was suggested by an increase in the tmax of glucose compared to a solution. It should be noted that formulation swelling may occur slower in vivo compared to in vitro due to the limited volume of fluids available, which might result in higher drug release at early time points; however, this remains to be evaluated.
Cell cultures have been widely used to evaluate irritation potential of formulations and their components [51,52], since there seems to be a good correlation between in vitro cytotoxicity assays and in vivo irritation. The cytotoxic potential of BRIJ-95 and BRIJ80 was assessed in comparison to that of PBS and sodium lauryl sulfate. PBS is considered safe, and as expected, the viability of fibroblasts (expressed as percent of controls, i.e., untreated cells) was not affected when increasing amounts of PBS (up to 100 g/mL) were added to the culture medium (Fig. 5). Compared to PBS, a significant decrease in cell viability was observed when sodium lauryl sulfate was used at a concentration as small as 1 g/mL, whereas no significant change on cell viability was detected using the same concentration of either precursor formulation. At a concentration of 10 g/mL, BRIJ-95 and BRIJ-80 significantly (p < 0.05) reduced cell viability (to 74.2 and 77.2%) compared to PBS, but not as strongly as did sodium lauryl sulfate (viability reduced to 52.3%). The cytotoxicity of BRIJ-95 and BRIJ-80 was similar at concentrations up to 10 g/mL. When a concentration of 50 g/mL was used, cell viability was significantly lower (p = 0.025) after treatment with BRIJ-95 compared to BRIJ-80, but still higher (p = 0.04) compared to sodium lauryl sulfate. Based on these cytotoxicity results, two conclusions can be drawn. First, both precursor formulations studied are safer com-
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
pared to sodium lauryl sulfate. Our observation that BRIJ-based formulations are less cytotoxic than sodium lauryl sulfate agrees with previous studies in which a reconstructured rabbit corneal epithelium model was used to compare the cytotoxicity of several surfactants [53]. The reported concentration of BRIJ to reduce cell viability to 50% was over 2 times higher than that of sodium lauryl sulfate. The second conclusion that can be drawn from the cytotoxicity results is that increasing BRIJ concentration in the precursor formulation (from 80 to 95%) increases cytotoxicity. Surfactants from the BRIJ series can affect membrane permeability, and thus, impact cell viability depending on their concentration [54]. Similar results were obtained as the concentration of polysorbate in topical microemulsions increased [24]. The fact that BRIJ-based formulations are milder to cells in vitro than sodium lauryl sulfate and BRIJ-80 is less cytotoxic than BRIJ-95 suggests that BRIJ-80 might be better tolerated in vivo. A good correlation exists between cell viability in cultured tissues and the Draize score following application of an irritant topically [55], suggesting that viability assays in cell and tissue cultures are good alternatives to predict in vivo irritation potential of formulations. However, we acknowledge that only future in vivo studies can fully address the safety of this formulation. To reduce the incidence and severity of local irritation, the search for milder surfactants for drug delivery systems is an on-going process. In recent years sugar-based surfactants (alkylpolyglucosides) have emerged as a novel class of surfactants with good biodegradability and tolerance (classified as slightly-tonon-irritating through measurement of viability of erythrocytes and bioengineered skin tissues) [56–58]. Since these surfactants have the ability to form certain liquid crystalline phases, they may also be studied as components for in situ forming gels. In conclusion, we have demonstrated that BRIJ-based precursor fluid formulations containing propylene glycol can uptake water and form hexagonal phase gels within 2–4 h of contact with an aqueous environment, providing sustained naltrexone release for 72 h. Increases in propylene glycol content from 5 to 20% in the precursor formulation decreased the amount of water necessary to form the hexagonal phase (since it suppressed lamellar phase transition), the cytotoxicity of the formulation, and increased the amount of drug released after 48 h, while having no significant effect on the kinetics of swelling and drug release. These results suggest the potential of BRIJ-based precursor formulations containing 20% of propylene glycol for sustained naltrexone release in the treatment of drug addiction. Acknowledgements The authors would like to thank Dr. Alexandre A. Steiner (Albany College of Pharmacy, Albany, NY) for critical comments on the manuscript. This work was supported by a Scholarship of Discovery Grant from Albany College of Pharmacy and Health Sciences to L.B. Lopes. J. Phelps received a Summer Research Award from the Department of Pharmaceutical Sciences of Albany College of Pharmacy and Health Sciences. References [1] P. Lobmaier, H. Kornor, N. Kunoe, A. Bjorndal, Sustained-release naltrexone for opioid dependence, Cochrane Database Syst. Rev. (2008) CD006140. [2] H.G. Roozen, R. de Waart, W. van den Brink, Efficacy and tolerability of naltrexone in the treatment of alcohol dependence: oral versus injectable delivery, Eur. Addict. Res. 13 (2007) 201–206. [3] P. Lobmaier, M. Gossop, H. Waal, J. Bramness, The pharmacological treatment of opioid addiction – a clinical perspective, Eur. J. Clin. Pharmacol. 66 (2010) 537–545. [4] K.J. Whittlesey, L.D. Shea, Delivery systems for small molecule drugs, proteins, and DNA: the neuroscience/biomaterial interface, Exp. Neurol. 190 (2004) 1–16. [5] S.L Tao, T.A. Desai, Microfabricated drug delivery systems: from particles to pores, Adv. Drug Deliv. Rev. 55 (2003) 315–328.
397
[6] W.K. Fong, T. Hanley, B.J. Boyd, Stimuli responsive liquid crystals provide ‘ondemand’ drug delivery in vitro and in vivo, J. Control. Release 135 (2009) 218–226. [7] C.C. Muller-Goymann, Physicochemical characterization of colloidal drug delivery systems such as reverse micelles, vesicles, liquid crystals and nanoparticles for topical administration, Eur. J. Pharm. Biopharm. 58 (2004) 343–356. [8] C.J Drummond, C. Fong, Surfactant self-assembly objects as novel drug delivery vehicles, Curr. Opin. Colloid Interface Sci. 4 (2000) 449–456. [9] Z. Wang, Z. Diao, F. Liu, G. Li, G. Zhang, Microstructure and rheological properties of liquid crystallines formed in Brij 97/water/IPM system, J. Colloid Interface Sci. 297 (2006) 813–818. [10] A. Otto, J. du Plessis, J.W. Wiechers, Formulation effects of topical emulsions on transdermal and dermal delivery, Int. J. Cosmet. Sci. 31 (2009) 1–19. [11] L.B. Lopes, J.L. Lopes, D.C. Oliveira, J.A. Thomazini, M.T. Garcia, M.C. Fantini, J.H. Collett, M.V. Bentley, Liquid crystalline phases of monoolein and water for topical delivery of cyclosporin A: characterization and study of in vitro and in vivo delivery, Eur. J. Pharm. Biopharm. 63 (2006) 146–155. [12] B.J. Boyd, D.V. Whittaker, S.M. Khoo, G. Davey, Lyotropic liquid crystalline phases formed from glycerate surfactants as sustained release drug delivery systems, Int. J. Pharm. 309 (2006) 218–226. [13] G.K. Hulse, V. Stalenberg, D. McCallum, W. Smit, G. O’Neil, N. Morris, R.J. Tait, Histological changes over time around the site of sustained release naltrexonepoly(dl-lactide) implants in humans, J. Control Release 108 (2005) 43–55. [14] J.E. Gomes-Filho, B.P. Gomes, A.A. Zaia, C.R. Ferraz, F.J. Souza-Filho, Evaluation of the biocompatibility of root canal sealers using subcutaneous implants, J.Appl.Oral Sci. 15 (2007) 186–194. [15] K. Al-Tahami, M. Oak, J. Singh, Controlled delivery of basal insulin from phase-sensitive polymeric systems after subcutaneous administration: in vitro release, stability, biocompatibility, in vivo absorption, and bioactivity of insulin, J.Pharm.Sci. 100 (2011) 2161–2171. [16] F Rosevear, The microscopy of the liquid crystalline neat and middle phases of soaps and synthetic detergents, J.Am.Oil Chem.Soc. 31 (1954) 628–639. [17] A. Feher, E. Urban, I. Eros, P. Szabo-Revesz, E. Csanyi, Lyotropic liquid crystal preconcentrates for the treatment of periodontal disease, Int.J.Pharm. 358 (2008) 23–26. [18] S.B. Rizwan, T. Hanley, B.J. Boyd, T. Rades, S. Hook, Liquid crystalline systems of phytantriol and glyceryl monooleate containing a hydrophilic protein: characterisation, swelling and release kinetics, J.Pharm.Sci. 98 (2009) 4191–4204. [19] G.H Pollack, Cells, Gels and the Engines of Life: A New, Unifying Approach to Cell Biology, Ebner and Sons, 2001. [20] M.A. Serban, Y. Liu, G.D. Prestwich, Effects of extracellular matrix analogues on primary human fibroblast behavior, Acta Biomater. 4 (2008) 67–75. [21] C.M. Chang, R. Bodmeier, Swelling of and drug release from monoglyceridebased drug delivery systems, J.Pharm.Sci. 86 (1997) 747–752. [22] M.G. Lara, M.V. Bentley, J.H. Collett, In vitro drug release mechanism and drug loading studies of cubic phase gels, Int. J. Pharm. 293 (2005) 241–250. [23] C. Verhulst, C. Coiffard, L.J. Coiffard, P. Rivalland, Y. De Roeck-Holtzhauer, In vitro correlation between two colorimetric assays and the pyruvic acid consumption by fibroblasts cultured to determine the sodium laurylsulfate cytotoxicity, J. Pharmacol. Toxicol. Methods 39 (1998) 143–146. [24] J. Hosmer, R. Reed, M.V. Bentley, A. Nornoo, L.B. Lopes, Microemulsions containing medium-chain glycerides as transdermal delivery systems for hydrophilic and hydrophobic drugs, AAPS PharmSciTech. 10 (2009) 589–596. [25] D.M. Bagley, J.R. Gardner, G. Holland, R.W. Lewis, H. Vrijhof, A.P. Walker, Eye irritation: updated reference chemicals data bank, Toxicol. In Vitro 13 (1999) 505–510. [26] Cosmetic Ingredient Review, Final report on the safety assessment of sodium lauryl sulfate and ammonium lauryl sulfate, Int. J. Toxicol. 2 (1983) 127–181. [27] V.C. Robinson, W.F. Bergfeld, D.V. Belsito, R.A. Hill, C.D. Klaassen, J.G. Marks Jr., R.C. Shank, T.J. Slaga, P.W. Snyder, F. Alan Andersen, Final report of the amended safety assessment of sodium laureth sulfate and related salts of sulfated ethoxylated alcohols, Int. J. Toxicol. 29 (2011) 151S–161S. [28] C. Faller, M. Bracher, N. Dami, R. Roguet, Predictive ability of reconstructed human epidermis equivalents for the assessment of skin irritation of cosmetics, Toxicol. In Vitro 16 (2002) 557–572. [29] F. Caboi, G.S. Amico, P. Pitzalis, M. Monduzzi, T. Nylander, K. Larsson, Addition of hydrophilic and lipophilic compounds of biological relevance to the monoolein/water system I. Phase behavior, Chem. Phys. Lipids 109 (2001) 47–62. [30] L.S. Helledi, L. Schubert, Release kinetics of acyclovir from a suspension of acyclovir incorporated in a cubic phase delivery system, Drug Dev. Ind. Pharm. 27 (2001) 1073–1081. [31] E. Ruel-Gariepy, J.C. Leroux, In situ-forming hydrogels—review of temperaturesensitive systems, Eur. J. Pharm. Biopharm. 58 (2004) 409–426. [32] J.O. You, D. Almeda, G.J. Ye, D.T. Auguste, Bioresponsive matrices in drug delivery, J. Biol. Eng. 4 (2010) 15. [33] L. Zhao, L. Zhu, F. Liu, C. Liu, D. Shan, Q. Wang, C. Zhang, J. Li, J. Liu, X. Qu, Z. Yang, pH triggered injectable amphiphilic hydrogel containing doxorubicin and paclitaxel, Int. J. Pharm. 410 (2011) 83–91. [34] S. Jauhari, A.K. Dash, A mucoadhesive in situ gel delivery system for paclitaxel, AAPS PharmSciTech. 7 (2006) E53. [35] A.S. Sallam, E. Khalil, H. Ibrahim, I. Freij, Formulation of an oral dosage form utilizing the properties of cubic liquid crystalline phases of glyceryl monooleate, Eur. J. Pharm. Biopharm. 53 (2002) 343–352. [36] B.A. Johnson, Naltrexone long-acting formulation in the treatment of alcohol dependence, Ther. Clin. Risk Manag. 3 (2007) 741–749.
398
J. Phelps et al. / Colloids and Surfaces B: Biointerfaces 87 (2011) 391–398
[37] W. Yin, E.O. Akala, R.E. Taylor, Design of naltrexone-loaded hydrolyzable crosslinked nanoparticles, Int. J. Pharm. 244 (2002) 9–19. [38] H.T. Ngo, D.E. Arnold-Reed, R.C. Hansson, R.J. Tait, G.K. Hulse, Blood naltrexone levels over time following naltrexone implant, Prog. Neuropsychopharmacol. Biol. Psychiatry 32 (2008) 23–28. [39] L.I. Giannola, V. De Caro, G. Giandalia, M.G. Siragusa, C. Tripodo, A.M. Florena, G. Campisi, Release of naltrexone on buccal mucosa: permeation studies, histological aspects and matrix system design, Eur. J. Pharm. Biopharm. 67 (2007) 425–433. [40] M Milewski, A.L. Stinchcomb, Vehicle composition influence on the microneedle-enhanced transdermal flux of naltrexone hydrochloride, Pharm. Res. 28 (2011) 124–134. [41] J. Hosmer, S.H. Shin, A. Nornoo, H. Zheng, L.B. Lopes, Influence of internal structure and composition of liquid crystalline phases on topical delivery of paclitaxel, J.Pharm.Sci. (2010), doi:10.1002/jps.22370. [42] K.L. Williams, C.L. Broadbridge, Potency of naltrexone to reduce ethanol self-administration in rats is greater for subcutaneous versus intraperitoneal injection, Alcohol 43 (2009) 119–126. [43] K. Derakhshandeh, M. Fashi, S. Seifoleslami, Thermosensitive Pluronic hydrogel: prolonged injectable formulation for drug abuse, Drug Des.Dev. Ther. 4 (2010) 255–262. [44] M. Makai, E. Csanyi, Z. Nemeth, J. Palinkas, I. Eros, Structure and drug release of lamellar liquid crystals containing glycerol, Int. J. Pharm. 256 (2003) 95–107. [45] K.W. Lee, T.H. Nguyen, T. Hanley, B.J. Boyd, Nanostructure of liquid crystalline matrix determines in vitro sustained release and in vivo oral absorption kinetics for hydrophilic model drugs, Int. J. Pharm. 365 (2009) 190–199. [46] E. Farkas, R. Zelko, Z. Nemeth, J. Palinkas, S. Marton, I. Racz, The effect of liquid crystalline structure on chlorhexidine diacetate release, Int. J. Pharm. 193 (2000) 239–245. [47] Y. Tang, J. Singh, Thermosensitive drug delivery system of salmon calcitonin: in vitro release, in vivo absorption, bioactivity and therapeutic efficacies, Pharm. Res. 27 (2010) 272–284. [48] M.R Park, C. Chun, S.W. Ahn, M.H. Ki, C.S. Cho, S.C. Song, Sustained delivery of human growth hormone using a polyelectrolyte complex-loaded thermosensitive polyphosphazene hydrogel, J. Control Release 147 (2010) 359–367.
[49] Y. Ito, Y. Ochii, K. Fukushima, N. Sugioka, K. Takada, Three-layered microcapsules as a long-term sustained release injection preparation, Int. J. Pharm. 384 (2011) 53–59. [50] S.K. Chandrasekaran, D.R. Paul, Dissolution-controlled transport from dispersed matrixes, J. Pharm. Sci. 71 (1982) 1399–1402. [51] T. Benavides, V. Martinez, M. Mitjans, M.R. Infante, C. Moran, P. Clapes, R. Clothier, M.P. Vinardell, Assessment of the potential irritation and photoirritation of novel amino acid-based surfactants by in vitro methods as alternative to the animal tests, Toxicology 201 (2004) 87–93. [52] H.C. Korting, T. Herzinger, A. Hartinger, M. Kerscher, T. Angerpointner, H.I. Maibach, Discrimination of the irritancy potential of surfactants in vitro by two cytotoxicity assays using normal human keratinocytes, HaCaT cells and 3T3 mouse fibroblasts: correlation with in vivo data from a soap chamber assay, J. Dermatol. Sci. 7 (1994) 119–129. [53] S. Matsuda, M. Hisama, H. Shibayama, N. Itou, M. Iwaki, In vitro eye irritancy test of lauryl derivatives and polyoxyethylene alkyl derivatives with the reconstructed rabbit corneal epithelium model, J. Oleo Sci. 58 (2009) 437–442. [54] Y.L. Lo, Relationships between the hydrophilic–lipophilic balance values of pharmaceutical excipients and their multidrug resistance modulating effect in Caco-2 cells and rat intestines, J. Control Release 90 (2003) 37–48. [55] T. Watanabe, T. Hasegawa, H. Takahashi, T. Ishibashi, H. Itagaki, K. Sugibayashi, Utility of MTT assay in three-dimensional cultured human skin model as an alternative for draize skin irritation test: approach using diffusion law of irritant in skin and toxicokinetics–toxicodynamics correlation, Pharm. Res. 19 (2002) 669–675. [56] N. Pakpayat, F. Nielloud, R. Fortune, C. Tourne-Peteilh, A. Villareal, I. Grillo, B. Bataille, Formulation of ascorbic acid microemulsions with alkyl polyglycosides, Eur. J. Pharm. Biopharm. 72 (2009) 444–452. [57] I. Pezron, L. Galet, D. Clausse, Surface Interaction between a protein monolayer and surfactants and its correlation with skin irritation by surfactants, J. Colloid Interface Sci. 180 (1996) 285–289. [58] A. Mehling, M. Kleber, H. Hensen, Comparative studies on the ocular and dermal irritation potential of surfactants, Food Chem. Toxicol. 45 (2007) 747–758.