Research Article, J Pharm Drug Deliv Res Vol: 1 Issue: 1
Development of Extruded Starch Based Formulations Aimed for Local Drug Delivery to Oral Cavity
| Thomas Kipping and Hubert Rein* |
| Department of Pharmaceutical Technology, Institute of Pharmacy, University of Bonn, Germany |
| Corresponding author : Hubert Rein Institute of pharmaceutical Technology, university of Bonn, Gerhard-Domagk-Straße 3, 53121 Bonn, Germany Tel: +49-228-73 5243; Fax: +49-228-73 5268 E-mail: rein@uni-bonn.de |
| Received: June 14, 2012 Accepted: July 30, 2012 Published: July 31, 2012 |
| Citation: Kipping T, Rein H (2012) Development of Extruded Starch Based Formulations Aimed for Local Drug Delivery to Oral Cavity. J Pharm Drug Deliv Res 1:1. doi:10.4172/2325-9604.1000101 |
Abstract
Development of Extruded Starch Based Formulations Aimed for Local Drug Delivery to Oral Cavity
The suitability of different starches as a basic substance for a prolonged release oral dosage form was evaluated. Starches were extruded slightly above gelatinization temperature. The shape of the melt was predetermined by a slot nozzle. Circular, tablet like single dosage forms were produced by manual die-cutting of the still elastic strands. Five different starches were used as basic substances: corn starch, pea starch, potato starch, high-amylose corn starch Eurylon®5 and amylose-free corn starch Waxilys®200. Menthol and clove oil were used as active ingredients. Physical characteristics were analyzed to prove storage stability of the products.
Abbreviations |
|
| API: Active Pharmaceutical Ingredient; DSC: Differential Scanning Calorimetry; MSME: Mean Specific Mechanical Energy; TGA: Thermogravimetric Analyzer | |
Keywords |
|
| Hot-melt extrusion; Controlled release; Lozenges; Starch | |
Introduction |
|
| Starch has never been used as an excipient for the production of buccal dosage forms like lozenges. The current field of application for starch in tablet manufacturing is mainly focused on its function as a disintegrant. Water exposition leads to a swelling of starch granules resulting in an accelerated disintegration time, which should rather be avoided while developing a retard formulation. The most favoured way to produce lozenges is by applying direct compression techniques. Sorbit and isomalt are most commonly used as filling materials Due to an insufficient flowability or high segregation tendencies of the raw materials energy consuming granulation steps need to be included. Another manufacturing method is by melting the ingredients and pouring the liquefied pre-mixture into defined moulds. During the following cooling step the melt solidifies and adopts the shape of the form. A distinction is made between hard-candy type lozenges or soft chewable lozenges [1]. | |
| This work mainly focuses on the ability of the matrix systems to provide a continuous prolonged local drug delivery to oral cavity.www.scitechnol.com/scholarly/nanosensors-journals-articles-ppts-list.php | |
| Our developed dosage form shows also another great application potential. Usually if drugs are administered orally, they are affronted by several barriers, like first pass effects or high drug degradation while passing the gastrointestinal tract. Sublingual or buccal drug delivery systems bear the opportunity to bypass these effects. Passive diffusion across lipid membranes can be considered as the most prevalent way for transmucosal drug delivery; it might be limited by the fact that the transport mechanism is closely linked to the physicochemical drug properties [2]. During hot-melt extrusion of starch the API gets either solved in the surrounding matrix or forms a solid dispersion. These effects can significantly improve the permeability of an API. Repka et al. showed a significantly increased in vitro permeability of clotrimazol for a topical application of a hot-melt extruded film in relation to a gel [3]. | |
| During extrusion processing starches can form an amorphous, glassy solution which can be used for embedding of APIs. In food technology this effect has been studied for aroma retention. For extruded starches retention of aroma molecules could be linked to the amylose content [4]. A great advantage of starch in relation to already established polymers is the benefit of low toxicity and biodegradability. Only few studies have been carried out to investigate the physical stability during storage time. One of the main goals of our work is to provide a continuous production method for lozenges implicating a further industrial application of the technique. From an economic point of view hot-melt extrusion shows the great advantage of uniting several process steps. The continuous mode of operation is a great benefit in contrast to time wasting and expensive single batch processes. Hot-melt extrusion leads to a complete gelatinization of starch granules and provides the formation of an amorphous matrix. For our initial lab scale production the extruded, still elastic, biplanar strands where shaped manually into circular, tablet like matrices. An important step was to characterize the suitability of different starches as basic substances. Corn starch, Eurylon®5, pea starch, potato starch and Waxilys®200 were chosen to show the influence of varying amounts of amylose on flavour retention and physical stability of the products. | |
| Menthol and clove oil (high amount of eugenol) were chosen as active substances. Both show pharmacological effects but due to their natural origin they are not classified as APIs. This was an important aspect for further organoleptic studies. Menthol was selected because of its cooling effects and the associated topical analgesia [5]. Clove oil shows local anaesthetic effects and is therefore used in dental treatment. Usually clove oil is directly applied to the pulp, which involves the risk of local tissue damage [6]. At higher concentrations neurotoxic effects can be observed [7]. Systemic absorption of our tested substances should therefore rather be avoided. Local effects were favoured and should enable the connection between release characteristic behaviours of the produced matrices and organoleptic sensations. | |
| The aim of developing clove oil-lozenges was to prove the suitability of the lozenges as a sustained release dosage form. Lozenges with different amounts of clove oil were produced. To improve the taste, saccharin and sodium cyclamate were added. Incorporation of analgesic compounds leads to an alleviation of pain in mouth and throat areas, buccal administration of opioids might be used to avoid breakthrough pain. A great benefit of this administration form is the high patient acceptance [8]. For stability tests the samples were stored in a climatic chamber according to ICH-Q1A guideline. Physical characteristics were compared in relation to placebo-extrudates. | |
Materials and Methods |
|
| Materials | |
| Corn starch, Eurylon®5, pea starch, potato starch, and Waxilys®200 were a kind donation from Roquette (Lestrem, France). Menthol, clove oil and sodium cyclamate were purchased from Caelo (Hilden, Germany). Saccharin was purchased from Synopharm (Barsbüttel, Germany). Mint flavour was obtained from Firmenich (Genf, Switzerland). | |
| X-ray diffraction | |
| The studies were performed by using an x-ray diffractometer type PW 1830/40 from Philips (Kassel, Germany). The x-ray generator was running at 40 kV and 40 mA. Wavelength of the emitted radiation was about λ = 0.154 nm. The 2θ range was set from 5° to 35°. Evaluation was carried out by using X-Pert-High-Score software from Panalytical (Almelo, Netherlands). | |
| Hyper differential scanning calorimetry | |
| Hyper DSC© was used to determine the glass transition temperatures of the extrudates. The experiments were performed by using a Pyris 1 DSC from Perkin Elmer (Überlingen, Germany). About 5-7 mg of the grinded extrudates was sealed into a vented aluminium sample pan (50 µl). An empty aluminium pan was used as a reference. The sample pans were heated at a rate of 200°C/min from -60°C to 160°C. | |
| Thermogravimetric analysis | |
| The measurements were carried out by using a TGA 7 from Perkin Elmer (Überlingen, Germany). Sample atmosphere was purged with nitrogen at a flow rate of 10 ml/min. Sample mass was about 3-6 mg. Temperature was held isothermal at 60°C for 1 min., followed by a temperature scan (60°C-200°C) at a heating rate of 10°C/min. | |
| Density measurements | |
| A helium pycnometer type 1000T from Quantachrome (Odelzhausen, Germany) was used for measuring the densities of the samples. Before each measurement the system was purged by helium (20 times). Every sample was measured 10 times to calculate the mean value. | |
| Hardness and post-curing | |
| The hardness of the extrudates was determined by a Tablet Hardness Distribution Tester type TH-3 from Yamato scientific (Toyo, Japan). Clamped extrudates moved with a continuous speed of 1 mm/s towards a continuous rotating slot drill; developing dust was blown away by compressed air to avoid friction force. | |
| Scanning electron microscopy | |
| SEM studies were performed using a S2460 scanning electron microscope from Hitachi (Tokyo, Japan). During micrography an accelerating potential of 15 kV was used. The surfaces of the samples were sputter-coated with a thin gold layer in argon atmosphere to avoid charging. | |
| Extrusion | |
| All extrudates were produced by using a grooved barrel single screw extruder type 811201 from Brabender (Duisburg, Germany). The extruder is powered by a direct-current motor with a capacity of 3.7 kW. For shaping the extrudates a slot nozzle with an angular geometry of 8×2 mm was used. The screw rotation speed remained constant at 100 rpm. | |
| A laboratory blender type UMC 5 electronic from Stephan (Hameln, Germany) was used to homogenize the mixture. The mixing time was about 5 minutes at 500 rpm. Afterwards the mixture was sieved to equalize the particle sizes (mesh size: 1.4 mm). | |
| A uniform dosing of the mixture was assured by a steel screw hopper with integrated agitator. Both applied temperature and shear forces contribute to gelatinization of starch [9]. During the experiments the temperature profile remained static, where the feeding section, barrel and die were tempered to 66°C, 85°C and 96°C respectively. | |
| Mean specific mechanical energy | |
| The MSME characterizes the energy that is put into the extrudates either by the work of the motor, or by friction forces between material and screw or between material and barrel. It is a scale independent measurement. The calculation is based on detecting the torque by deformation of strain gauge sensors [10] | |
| MSME calculation [10]: | |
 |
|
| m: Extruded mass | |
| t: Time of extrusion | |
 : Average torque |
|
| Equation 1: Calculation of MSME. | |
| Organoleptic tests | |
| For our studies sensory impressions were linked to numeric scales. Short five point scales were chosen to avoid bias distortion and to allow the application of statistical analysis to the data [11]. Additionally a ranking test according to DIN 10963-A was performed [12]. The subjects had to put different test samples into descending order according to their overall impressions. The implementation of the tests was identically equal for both surveys. Each person had to evaluate five different lozenges by a questionnaire. To avoid systemic bias samples were encoded by a three digit numeric code and distributed in random order. | |
| The different menthol-lozenges were characterized concerning colour, smell, flavour, aftertaste, consistency, cooling effect, duration of taste and disintegration time. After evaluating the results of the organoleptic tests obtained from menthol-lozenges and the desired physical parameters of the extrudates, potato starch was chosen for producing clove oil-lozenges. The lack of an unpleasant aftertaste and a high dissolving time in mouth were the determinant criteria. | |
| No smoking was allowed until at least 10 minutes before the test, no sweets or chewing gums were allowed. Between each test there was a break about 5 minutes. In between the taste had to be neutralized with water. Statistical analysis was performed by using Winstat from Fitch Software (Bad Krozingen, Germany). | |
Results and Discussion |
|
| Physical properties of the extrudates | |
| The exact composition of the mixtures is shown in Table 1. Considering clove oil-lozenges, it was attempted to improve the sweetening by using a 10:1 ratio of cyclamate to saccharin which also attenuates the off-taste [13]. The production of the placebo-extrudates proved to be problematic. The absence of plasticizers results in a high viscosity of the elastic strands. Increased energy inputs are necessary to obtain homogenous strands. No white areas can be identified, which would be an indication for incomplete gelatinization. Potato- and pea starch show a high tendency towards melt fracture and lead to unshaped products. According to the high content of xanthophylls and carotenes, corn-starch-extrudates and Eurylon®5-extrudates show a yellow colour [14]. Extrusion of the menthol-mixture led to well shaped products. Menthol shows a plasticizing effect on the extruded melt, which becomes obvious by comparing the viscosities of the extruded strands (Table 2). Also the specific mechanical energy input is much lower. Remarkable is the white colouring of the extrudates which is probably caused by inclusion of finely dispersed air bubbles. | |
 |
Table 1: Composition of the mixtures. |
|
Table 2: Physical parameters of the extruded products. |
| Hardness measurements | |
| Hardness was determined by a tablet hardness distribution tester. All samples were stored in a climatic chamber at 25°C and a relative humidity of 60% according to ICH Q1A guideline, climatic zone 2 (subtropical and Mediterranean climates). All extrudates show a sharp increase of hardness during the first two weeks after extrusion. From that point on the material shows a nearly constant resistance (Figure 1a). A rapid solidification of the material is an important factor for providing constant and reproducible release rates. | |
|
Figure 1: Hardening of the placebo-extrudates in relation to storage time; a) placebo-extrudates, b) menthol-extrudates. |
| Considering the placebo-extrudates Eurylon®5, a high amylose starch, shows a greater hardness (9 N) compared to the other starches. Waxilys®200, a nearly amylose-free starch, shows lowest values (6 N). There seems to be a link between the ratio of amylose and amylopectin and the hardening process. A possible explanation is that retrogradation of starch seems to involve two steps. First there is the fast complexation of amylose followed by slower recrystallisation of amylopectin [15]. The lack of amylose content in Waxilys®200 modifies the retrogradation process and leads to a decrease of hardness. The menthol-extrudates show a similar tendency (Figure 1b). After four months pea starch, Eurylon®5 and potato starch show values between 9 N-9.5 N in contrast to Waxilys®200 and corn starch which show lowest values, of about 7-8 N. Above glass transition temperature amylopectin forms inter- and intramolecular double helices. Intermolecular crystallinity leads to a reinforcement of the network, whereas intramolecular crystallization leads to a limited mobility of amylopectin which favours the formation of cracks and increases the brittleness of the products. The addition of plasticizers can attenuate these effects and enables the improvement of mechanical stability [16]. At the beginning of the hardening process the reinforcement of the material is highly dependent on the amount of water remaining in material. In Figure 2 this correlation is shown for Eurylon®5-menthol extrudates. The addition of plasticizers is indispensable for the production of lozenges in order to reduce the fragility of the material. This relationship was confirmed while determining the hardness of the clove-oil-extrudates. At low concentrations an increase of clove oil content goes along with an increase of hardness. | |
|
Figure 2: Hardness in relation to water content (exemplary: Eurylon®5-menthol-extrudates). |
| Water content | |
| The water content of the extruded products was determined by thermogravimetry. | |
| Shortly after extrusion the water content rapidly increases and reaches highest levels (14%-15%) followed by a slowly decrease with time. After four months values of about 9% to 10% are reached. Both placebo- and menthol-extrudates show a similar trend (Figures 3a and 3b). | |
|
Figure 3: Time related changes of water content; a) placebo-extrudates, b) menthol-extrudates. |
| Water contents of native starches vary from about 11% to 14%. An interesting aspect is that the residual moisture of the extruded products reaches final values of less than 10%. | |
| Taking a close view at the structure of a starch granule might help to understand this phenomenon. Water distribution varies in between amorphous and crystalline regions. Though crystalline regions due to their helical structures show a high packing density, especially in B-type starches a high amount of incorporated water molecules can be observed [17]. The amorphization during extrusion processing causes a loss of crystalline regions. The lack of crystalline bond water might be reason for lowered residual moistures. The effect might also depend on the high heating rates used for the measurements. The extruded matrix might form a stable network inhibiting the release of water. | |
| Glass transition temperature | |
| Glass transition temperature was measured with differential scanning calorimetry. | |
| A link between glass transition temperature (Tg), residual moisture and hardness could be observed. Figure 4 demonstrates the correlation for the menthol extrudates. Due to its low Tg (-135 to -139°C) the remaining water shows a great plasticizing effect and decreases the Tg of the whole mixture [18]. A comparison between placebo-and menthol extrudates is shown in Figure 5. Comparing the entire mean values of the extruded products after the initial hardening time of about two weeks, the placebo-extrudates consequently show higher values than the menthol-extrudates. | |
|
Figure 4: 3D-scatter plot showing the correlations of the observed parameters Tg, water content and hardness (exemplary for the menthol extrudates). |
|
Figure 5: Glass transition temperatures of extruded products. A comparison between placebo- and menthol extrudates after the initial hardening process was completed (n=5). |
| For clove oil extrudates an increasing amount of clove oil leads to a decrease of Tg like shown in Figure 6. The resulting Tg in polysaccharide systems depends on the average molecule size [19]. It is decreased by the presence of small molecules like menthol or eugenol. This effect prevails in contrast to an estimated molecule degradation caused by high MSME inputs, shown for the placebo-extrudates (Table 2), which would lead to a further decrease of Tg. According to the high heating rates, reduced heat conduction and sample relaxation a delay of characteristic transitions might occur [20]. | |
|
Figure 6: Glass transition temperatures showing the plasticizing effect of clove oil addition; comparison in equilibrium state (n=3). |
| X-ray diffraction | |
| All x-ray diffractograms of native starches and mixtures show distinct peaks, which indicate crystallinity (Figure 7a). The strong peaks at 0°-2° 2Theta could be linked to the peaks of sodium cyclamate and saccharin. Extrusion leads to a complete amorphisation of the mixture. Considering the diffractograms of the menthol-lozenges, even after extrusion some signs of crystalline regions in the range of about 17°-20° 2Theta are still visible (Figure 7b). This fact indicates that the amount of energy applied to the material was not enough to achieve a complete amorphisation of the starch. Long term stability could be proved. For more than two years no further recrystallization could be observed. | |
|
Figure 7: a) x-ray diffractograms of the pre-mixtures; b) x-ray diffractograms of menthol-extrudates (after 2 years of storage time). From top to bottom: corn starch, Eurylon®5, potato starch, Waxilys®200, pea starch. |
| Scanning electron microscopy | |
| SEM was performed to get a detailed view of the surfaces in order to eventually detect crystalline regions. The investigations performed by x-ray diffractometry indicated the presence of crystalline regions in menthol-extrudates. The results from SEM are presented in Figure 8. According to SEM images apart from Waxilys®200-menthol-extrudates (Figure 8d), which clearly show contours of intact starch granules, no crystalline regions can be detected. Extrusion led to a complete gelatinization of the starches. | |
|
Figure 8: SEM images showing cross-sectional views of the menthol-lozenges; a): corn starch; b): Eurylon®5; c): potato starch; d): Waxilys®200; e): pea starch. |
| Organoleptic tests | |
| Since no specific instructions for determining dissolving times of lozenges are defined by the European Pharmacopoeia, it was necessary to develop an appropriate method providing most realistic results. This could be achieved by designing questionnaires related to organoleptic impressions obtained by test persons. According to the fact that the number of choices per question (n=5) are very limited, an evaluation including the mean values was favoured. This enables the detection of small deviations. The reliability is eventually restricted because the distances between possible answers are not exactly defined. Statistical analysis was performed by using the Friedman test. Low p-values (< 0.05) indicate a high significant difference within the group. | |
| Menthol-lozenges | |
| 36 test persons took part in this study. The mean age was 24 years. | |
| Candidates could choose between five options; from very weak flavour (1) up to very strong flavour (5). The taste at the beginning varied between optimal flavour (3.0) and strong flavour (4.0). Waxilys®200-lozenges showed the strongest intensity (3.33), whereas potato starch and Eurlyon®5 were rated lowest (2.38; 2.41). A high significance for a difference in between the group could be shown (p:1.5197×10-11). | |
| Considering the flavour intensity of taste after 5 minutes Waxilys®200 even more differed from the other starches. With a mean value of 3.85 the lozenges made form Waxilys®200 were considered as strong flavoured, whereas the other lozenges vary between 2.27 and 2.55. They are considered in between too weak (2.0) till optimal flavoured (3.0). The extremely low p-value (1.7547×10-12) indicates a high significance for a difference between the lozenges according to the intensity of taste after 5 minutes (Figure 9). | |
|
Figure 9: Evaluation of flavour intensity after 5 minutes (menthol-extrudates; percentage distribution). |
| The presence of a cooling effect was most frequently observed by Waxilys®200-lozenges (97.0%) and pea-starch-lozenges (94.3%). Considering Eurylon®5-lozenges only 81.8% recognized a cooling effect. Also the intensity of the cooling effect was most prevalent for Waxilys®200 (2.94). Eurylon®5 and potato starch showed lowest values (Figure 10). | |
|
Figure 10: Evaluation of cooling intensity (menthol extrudates; percentage distribution). |
| The evaluation of the aftertaste showed that most subjects recognized the existence of an aftertaste for corn-starch-lozenges 57.6%. Here the aftertaste was also mostly negative rated (78.9%). Potato starch showed the best results. Only 51% rated the aftertaste negatively. Lowest values considering the duration of taste could be shown for Eurylon®5 (3.27) whereas Waxilys®200 showed highest values (4.0). A significant difference between the lozenges was obtained (p: 0.0004). | |
| Most interesting was determining the disintegration time of the lozenges. Apart from the low values of Waxilys®200 (4.30), no major variations between the different lozenges could be shown (p: 0.5578). With values between 4.43 and 4.55 the other starches vary between the options 20-30 minutes (4.0) and more than 30 minutes (5.0). Lozenges made from pea starch show highest disintegration times (4.55). | |
| Amylose is able to form helical complexes with hydrophobic guest molecules like menthol since helication is not hindered by side chains. In contrast according to frequent occurrence of side chains, amylopectin shows a limited ability to form inclusion complexes [21]. It can be assumed, that the high amylose content encourages the formation of inclusion complexes, causing a delayed release. According to the fact that Waxilys®200 is a nearly amylose-free starch, the formation of inclusion complexes of flavour compounds with amylose is nearly impossible. Boutboul et al. showed in their work that flavour retention in extruded starches depends on the amylose content [4]. This might be an explanation for the high values of Waxilys®200 concerning cooling effect and intensity of taste. Short disintegration times seem to be a limiting factor, and eventually prevent further development. Corn starch also showed high values referring to intensity and lasting of taste. The negative rating of the aftertaste could prove problematic. Potato starch reached low values according to cooling effect and taste. Main advantages are the lack of an aftertaste and high dissolving times. Pea starch shows similar properties. | |
| Rank order test (DIN 10963-A) | |
| The samples were evaluated according to their overall impressions. Lozenges made from Waxilys®200 were rated higher than the other samples. Pea-starch-lozenges got a better evaluation than those made from corn starch, Eurylon®5 and potato starch. Corn starch, Eurlyon®5 and potato starch were equally rated. | |
| Clove-oil-lozenges | |
| 20 test persons took part in this study. The mean age was 23 years. | |
| By evaluating the flavour intensity at the beginning confusing results were obtained. The intensity of the lozenges containing 1% and 5% of clove oil were considered as the ones with weak flavour (2.47). The other lozenges were regarded as optimal flavoured (2.8-3.1, p: p:7.5350×10-5). After 5 minutes the impression of taste evolved. Enhancing of taste could be connected to increasing clove oil contents. A difference between the lozenges could be shown (p: 0.0060). With an increasing amount of clove oil also the proportion of subjects detecting an analgesic effect ascended (from 30% to 80%). The percentage distribution is shown in Figure 11. Duration of taste considerably raised at higher clove oil contents (from 2.75 at 1% to 3.95 at 5%; p:1.0731×10-5).Considering disintegration time, no significant variations can be detected (p: 0.9203). For nearly all lozenges sufficient disintegration times of more than 30 minutes are achieved. A correlation between the amount of clove oil added to the mixture and the intensity of flavour could be shown. At a clove oil content of 3% more than 80% of the subjects felt an analgesic effect. No decrease of disintegration time was noticed, but it can be assumed that due to the plasticizing effect further increase of clove oil content will cause faster relea. | |
|
Figure 11: Evaluation of the analgesic effect (clove oil-extrudates; percentage distribution). |
Conclusions |
|
| According to the available data the suitability of starch as an excipient for producing lozenges by applying hot-melt extrusion could successfully be demonstrated. Sufficient disintegration times were achieved. Added ingredients showed a plasticizing effect and reduced the fragility of the material. The variations in amylose contents can be used for providing a tailor-made release of an API. Our investigations proved that amylose-free starch Waxilys200 shows an immediately release, resulting in initially stronger flavour intensities. Low concentrations of amylose result in very fast drug-release rates, whereas high amounts enable the production of retard formulations. Results of our organoleptic studies are conforming to in vitro dissolution profiles of starch based pellets containing ibuprofen which were studied by Bialleck et al. [22]. Nevertheless the structure of the API is very important and highly affects the dissolution behaviour. The combination of different starches is an interesting option for further research. Soft factors, such as presence of an unpleasant aftertaste should not be neglected and might become an important aspect for selecting the right starch. For our tested products, the effect was strongly noticeable for corn starch. | |
| Long term stability of our products could be assured by thermoanalysis and x-ray diffraction. Glass transition temperatures proved to be at least more than 50°C above storage temperature which avoids molecular movements inside the solid solution [23]. | |
| An interesting further field of application is the incorporation of strong analgesics. A great advantage of our tested formulations is that the release of an API is strongly dependent on swelling processes of the matrix. Unexpected burst effects can be prevented. | |
| An important step for industrial production will be an automation of the cutting process. | |
| First attempts have already been realized by adapting a cutting machine derived from plastics industry into the production line [24,25]. | |
Acknowledgements |
|
| The authors would like to thank Roquette (Lestrem, France) for providing a great variety of starches. | |
References |
|
|
|
