Characterisation of excipient-free nanoporous microparticles (NPMPs) of bendroflumethiazide.

The purpose of this study was to prepare excipient-free porous microparticles of bendroflumethiazide by spray drying and to characterise the physicochemical properties of the particles produced. Solutions of bendroflumethiazide in ethanol/water, ethanol/water/ammonium carbonate or methanol/water/ammonium carbonate were spray dried using a laboratory spray dryer. Spray dried products were characterised by scanning electron microscopy, X-ray powder diffraction, differential scanning calorimetry, FTIR, laser diffraction particle sizing and density measurement. Nanoporous microparticles (NPMPs) were prepared from the alcoholic solutions containing ammonium carbonate. NPMPs were amorphous in nature, had median particles sizes less than 3 microm and densities that were significantly reduced compared to non-porous spray dried bendroflumethiazide powder. The novel process may be used to produce excipient-free amorphous microparticles with desirable physical properties such as amorphous solid state, porosity and low bulk density. This new engineering technology has applications in the design of other therapeutic agents such as those used in pulmonary delivery.

Release kinetics of benzoic acid and its sodium salt from a series of poly(N-isopropylacrylamide) matrices with various percentage crosslinking.

Swelling and concomitant drug-release kinetics from a series of poly(N-isopropylacrylamide) (PNIPA) matrices were examined. Scanning electron microscopy indicated a decrease in polymer pore/mesh size above the lower critical solution temperature (LCST) with increasing percentage crosslinker. The release of sodium benzoate (NaB) or benzoic acid (BA) were investigated above and below the LCST of the gels and compared to the drug-loaded gel-swelling rates. The release rate of NaB increased with increasing percentage crosslinker above the LCST in contrast to a decrease in release rate with increasing crosslinker seen previously with non-thermoresponsive hydrogel systems. As the percentage crosslinker increased, there was therefore a decrease in the ability to thermally control the release of this small model drug. In contrast to the crosslinker-dependent pattern apparent with NaB, drug-PNIPA hydrophobic binding controlled the swelling rate of BA-loaded hydrogels. As a result, all the BA-loaded systems showed similar diffusion controlled swelling and release patterns, effectively independent of the inherent-swelling rates of the hydrogels. The crosslinking content of the hydrogel and the physicochemical nature of the loaded drug were therefore shown to be important in thermal control of drug release from PNIPA hydrogels.

Nanoparticles: pharmacological and toxicological significance.

Nanoparticles are tiny materials (<1000 nm in size) that have specific physicochemical properties different to bulk materials of the same composition and such properties make them very attractive for commercial and medical development. However, nanoparticles can act on living cells at the nanolevel resulting not only in biologically desirable, but also in undesirable effects. In contrast to many efforts aimed at exploiting desirable properties of nanoparticles for medicine, there are limited attempts to evaluate potentially undesirable effects of these particles when administered intentionally for medical purposes. Therefore, there is a pressing need for careful consideration of benefits and side effects of the use of nanoparticles in medicine. This review article aims at providing a balanced update of these exciting pharmacological and potentially toxicological developments. The classes of nanoparticles, the current status of nanoparticle use in pharmacology and therapeutics, the demonstrated and potential toxicity of nanoparticles will be discussed.

Combined effects of iontophoretic and chemical enhancement on drug delivery. II. Transport across human and murine skin.

This paper reports measurements of the release characteristics of the model drug salbutamol from a liquid crystalline vehicle across both human and hairless murine skin in vitro. The use of oleic acid and iontophoresis as penetration enhancement techniques, used separately and simultaneously, was also investigated. Over a period of 12h, salbutamol base did not diffuse from the vehicle across excised human skin while, in contrast, over a period of 2h, the drug passively transported across hairless murine skin. The diffusion co-efficient for the drug in this tissue was estimated to be 4.54+/-0.60x10(-9)cm(2)s(-1) with a permeability co-efficient of 7.03+/-0.83x10(-7)cms(-1). A current of density of 0.39mAcm(-2) facilitated a significant transport of salbutamol from the liquid crystalline vehicle across excised human skin but with a small (<0.1) transport number. The quantity of salbutamol transported across excised hairless murine skin under the same conditions was significantly greater with a transport number of 0.68. The alteration of the permeability of the tissue was less than that of the human skin and a full recovery of the pre-iontophoretic permeability of murine skin was consistently observed. The incorporation of either oleic or lauric acid into the monoglyceride component of the vehicle at a concentration of 0.1M had a marked effect on the transport of salbutamol across both human and murine skin. The initial passive permeation of the drug across the skin was not affected but the rate of drug delivery during iontophoresis was typically observed to increase by a factor greater than two. The post-iontophoretic transport of salbutamol across either tissue was also substantially enhanced in the presence of the fatty acid. The analogous use of stearic acid did not significantly influence the iontophoretic or the post-iontophoretic transport of salbutamol across excised human skin. The investigation also revealed a synergistic combination of the fatty acid and anodal iontophoresis to enhance the in vitro transport of other drug substances, including nicotine and diltiazem hydrochloride across murine skin. Oleic acid increased both the iontophoretic and post-iontophoretic transport of nicotine, so that the enhancement of drug delivery was greater than that caused by the current alone. The investigation also indicated that the barrier properties of the skin recover following the constant current iontophoresis in the presence of oleic or lauric acids.

Drug-polymer interactions and their effect on thermoresponsive poly(N-isopropylacrylamide) drug delivery systems.

Potential interactions between model drugs (benzoates, diltiazem, cyanocobalamin, dextrans) and a thermoresponsive poly(N-isopropylacrylamide) (PNIPA) hydrogel and corresponding linear polymer were investigated. The influence of the drugs on the equilibrium swelling level of the hydrogel was examined and drug-hydrogel binding isotherms were established where appropriate. Differential scanning calorimetry (DSC) was used to investigate the influence of the drugs on the lower critical solution temperature (LCST) of the linear polymer solution. Phase solubility studies were preformed to investigate binding. Drug-polymer co-precipitated blends were also prepared and analysed by X-ray diffraction (XRD), thermal analysis and Fourier transform infrared (FT-IR) spectroscopy. Hydrophobic binding was apparent between PNIPA and the aromatic ring/ester side chain of the unionised benzoate. The effect of this binding on hydrogel swelling was clarified in terms of the influence of the binding on the LCST of the system. The drug release rates of the benzoates from the hydrogel were shown to be dependent on drug binding properties. Ionisation of the benzoate prevented such hydrophobic binding, with a weaker salting out effect apparent with sodium benzoate. Significant interactions between diltiazem, cyanocobalamin (Vitamin B12) or the dextrans and PNIPA were not apparent. High concentrations of the hydrophilic drugs did, however, interfere with the magnitude of hydrogel equilibrium swelling. Hydrophobic binding, the salting out effect and the influence of the drugs on hydrogel swelling under non-sink conditions were therefore shown to be important effects which depended on the chemical nature of the drug present.

Hydrodynamic simulation (computational fluid dynamics) of asymmetrically positioned tablets in the paddle dissolution apparatus: impact on dissolution rate and variability.

The aim of this work was to investigate the dissolution rate from both the curved and planar surfaces of cylindrical compacts of benzoic acid, which were placed centrally and non-centrally at the base of the vessel of the paddle dissolution apparatus. The effect of fixing the compacts to a particular position on the variability of dissolution results was also examined. In addition, computational fluid dynamics (CFD) was used to simulate fluid flow around compacts in the different positions in the vessel, and the relationship between the local hydrodynamics in the region of the compacts and the dissolution rate determined. The dissolution rate was found to increase from the centre position to the off-centre positions for each surface examined. There was a corresponding increase in maximum fluid velocities calculated from the CFD fluid flow simulations at a fixed distance from the compact. There was less variability in dissolution from compacts fixed to any of the positions compared with those that were not fixed. Fluid flow around compacts in different positions could be successfully modelled, and hydrodynamic variability examined, using CFD. The effect of asymmetric fluid flow was evident visually from the change in shape of the eroded compacts.

Mechanistic aspects of the release of levamisole hydrochloride from biodegradable polymers.

The release of levamisole hydrochloride from poly-DL-lactide-co-glycolide compacts prepared at 5, 10 and 20% drug loading using two different particle size fractions of drug (90-125 and 125-250 microm) was investigated. Release profiles were significantly different from those previously reported for compacts prepared using the base form of the drug. Release was found to occur in a biphasic manner, with an initial fast release phase followed by a slower polymer degradation controlled release phase. The drug release profiles were successfully described by a model combining contributions from a first-order initial release phase and a polymer degradation controlled drug release phase. The fraction of drug released in the initial burst phase (F(B)) was attributed to the dissolution of drug domains situated at the surface of the polymer-drug compact and this fraction tended to increase with increasing drug particle size, as expected from the model. The increase in F(B) with increased loading was attributed to the clumping of dispersed drug particles which effectively increased the proportion of drug linked to the compact surface.

Effect of drug physicochemical properties on swelling/deswelling kinetics and pulsatile drug release from thermoresponsive poly(N-isopropylacrylamide) hydrogels.

The effect of drug physicochemical properties on swelling/deswelling kinetics and pulsatile drug release from a thermoresponsive hydrogel was examined. Hydrogels were loaded with drug and thermally triggered swelling/deswelling and release experiments were performed. Two series of drugs of contrasting hydrophilicity and varying physicochemical properties were examined. Benzoic acid (BA), its methyl and propyl esters, and diltiazem base were used as model hydrophobic drugs. Sodium benzoate (NaB), diltiazem HCl (DHCl), vitamin B12 (VB12) and various dextrans (MW 4300, 10,200, 42,000, 68,800) were used as model hydrophilic agents of increasing size. The hydrogel swelling rate was slowed by the presence of the hydrophobic drugs and this decreased rate was solubility dependant for the benzoates. The hydrophilic series increased the rate of swelling compared to the unloaded system. In all cases, the magnitude and rate of hydrogel contraction were proportional to the extent of swelling prior to temperature switch. Drug release was by diffusion below the lower critical solution temperature (LCST), while a solubility-dependent drug pulse release on temperature switch was observed for the hydrophobic series. Effectiveness of thermal control of hydrophobic drug release increased with increasing solubility. The hydrophilic series produced a molecular size-dependent drug pulse on temperature switch above the LCST. Pulsatile on-off drug release was shown with DHCl, VB12 and the various dextrans. Drug solubility, size and chemical nature were shown to be of particular importance in the control of hydrogel swelling and drug release from thermosensitive hydrogels.

Dissolution rate of cholesterol and palmitic acid mixtures in cholelitholytic cosolvent systems.

The dissolution rates and solubilities of cholesterol monohydrate, palmitic acid, and their mixtures in the cholelitholytic solvents monooctanoin (MO) and methyl tert-butyl ether (MTBE) and mixtures of these two solvents were determined. The dissolution rates obtained were consistent with the diffusion-controlled two-component noninteracting model. The addition of MTBE as cosolvent to MO resulted in an increase in the solubility of both cholesterol monohydrate and palmitic acid; in the case of the former, the solubility peaked at 80% MTBE. Neither solute exhibited a log-linear solubility relationship on addition of MTBE as cosolvent. Furthermore the increases in the dissolution rates of both components were much larger than could be explained by the solubility increases alone. Mass transfer coefficients increased dramatically with increasing MTBE content of the solvent, were consistently higher for palmitic acid, and reflected the decline in solvent viscosity. Incorporation of relationships among solubility, viscosity, and cosolvent composition into the two-component noninteracting model gave good correlation between predicted and observed rates over nearly 3 orders of magnitude.

Physicochemical characterization of diclofenac N-(2-hydroxyethyl)pyrrolidine: anhydrate and dihydrate crystalline forms.

This study on diclofenac N-(2-hydroxyethyl)pyrrolidine (DHEP) characterizes and compares the anhydrate (DHEPA) and dihydrate (DHEPH) solid state forms using powder X-ray diffraction, infrared spectroscopic, and thermal analyses. Heats of solution and intrinsic dissolution rates are determined. The thermodynamics of hydration are discussed and the entropic cost of dihydrate formation is calculated. Reported differences in the solution behavior of DHEP crystallized from different solvents are explained. The molecular structures of both solid forms were determined and are presented. Crystal data for DHEPA: triclinic, space group P-1 (No 2), a = 11.662(2) A, b = 11.874(2) A, c = 15.296(3) A, alpha = 76.183(14) degrees, beta = 84.575(12) degrees, gamma = 87.028(12) degrees V = 2046.8(6)A3, Z = 4. Crystal data for DHEPH: triclinic, space group P-1 (No 2), a = 9.356(3) A, b = 9.920(2) A, c = 13.5413(12) A, alpha = 69.915(12) degrees, beta = 82.05(2) degrees, gamma = 71.51(2) degrees, V = 1118.9(4) A3, Z = 2. The experimentally observed ease of dehydration under conditions of nitrogen purge is explained in terms of crystal packing within the dihydrate.

Factors influencing drug release from stearic acid based compacts.

Fatty acids are potentially suitable carriers for use in the design of drug delivery systems, being biocompatible, biodegradable inexpensive and of low toxicity. The release of the model compound benzoic acid from fatty acid compacts of stearic acid was evaluated using the USP Apparatus 2 dissolution assembly in phosphate buffer pH 7.4. Matrix controlled drug release was expected. Release profiles were approximated by square root of time kinetics. Release rate was independent of stirring speed in the rpm range 50-150, however, at 200 rpm a significant increase in release rate was observed particularly at later times, the amount released versus square root of time plots becoming non-linear. Release was independent of compression pressure in the range 1-7 tons. The particle size of the benzoic acid and stearic acid used had a significant influence on release. The use of particles in the range 250-500 microm gave release rate constants (k, g/cm(2) per min(0.5)) approximately 1.5 greater than those of smaller particle size (63-125 microm). The formation factor (F) tended to increase exponentially with drug loading, the increase being steeper for compacts prepared from the larger particle sizes. At 80% drug loading for large sized systems the matrix appeared to offer little resistance to drug release and F approached one.

Preparation and characterisation of a range of diclofenac salts.

Physicochemical properties of diclofenac salts prepared using eight different counterions and including five novel salts, obtained with the bases 2-amino-2-methyl-1,3-propanediol, 2-amino-2-methylpropanol, tert-butylamine, benzylamine and deanol, were compared. Four of the bases used to prepare these salts were related in their chemical structure, differing only in the number of hydroxy groups. Characterisation techniques included X-ray diffraction, differential scanning calorimetry, thermogravimetric analysis, thermomicroscopy, Karl Fischer titration, FT-IR spectroscopy and elemental analysis. In the case of salts prepared from 2-amino-2-methylpropanol and benzylamine, two polymorphic forms of each salt were identified. For the 2-amino-2-methyl-1,3-propanediol salt, a pseudopolymorphic form was identified. The aqueous solubilities of the salts studied ranged from 3.95 mM (tris(hydroxymethyl)aminomethane salt) to 446 mM (deanol salt), corresponding to a 113-fold difference in solubility. The solubility of diclofenac deanol was higher than the solubilities for diclofenac salts reported earlier. Correlation was found between the inverse of the salt melting point and the logarithm of salt solubility. A log-log relationship was observed between salt solubility and hydrogen ion concentration in the salt solution. Relationships between the properties of the salt-forming agents and those of the resulting diclofenac salts were explored. Reasonable correlation was found between the free base melting point and the salt melting point.

Comparison of the physicochemical properties of the N-(2-hydroxyethyl) pyrrolidine, diethylamine and sodium salt forms of diclofenac.

Non steroidal anti-inflammatory agents (NSAIDs) such as diclofenac have very low aqueous solubilities and consequently salt formation may be used to enhance solubility and dissolution rate. In this study, we examined the physicochemical properties of three diclofenac salts, diclofenac sodium (DNa), diclofenac N-(2-hydroxyethyl)pyrrolidine (DHEP) and diclofenac diethylamine (DDEA), and their different solid state forms to determine the influence of salt form on solubility, dissolution rate and membrane transport. The equilibrium solubility of DDEA at 25 degrees C was determined as 33 mM, lower than the solubilities of DHEP (273 mM) and DNa (66 mM) previously reported (Ledwidge and Corrigan, 1998). In addition to the dihydrate form of DHEP previously characterised, monohydrate forms of DHEP and DDEA were identified. Intrinsic dissolution rate studies were used to determine the solubility ratios of the hydrated and anhydrous forms. The monohydrate form of DHEP was found to be 1.8 times less soluble than the anhydrate, whereas DDEA anhydrate was approximately 1.7 times as soluble as the monohydrate form. On investigation of the pH-solubility profile (25 degrees C) of DDEA, appreciable supersaturation (76 mM) relative to the theoretical profile, was detected at the pH(max). This contrasts with values of >800 and 67 mM for DHEP and DNa, respectively. The transport of salt solutions through a porous membrane (Visking) was investigated. A linear relationship between concentration (mM) and rate of transport (mmol/h) was established for DNa and DHEP solutions. The mass transfer coefficient determined for DHEP was lower than that for the other two salts. Nevertheless, the maximum transport rate obtained for DHEP is almost six times higher than that obtained for DDEA.

Iontophoretic and chemical enhancement of drug delivery. Part I: across artificial membranes.

This paper reports on measurements of the release characteristics of the model drug salbutamol base from a liquid crystalline vehicle across a non-rate limiting synthetic membrane. The measured passive release rates were compared with analogous behaviour: (i) when a penetration enhancer such as oleic acid was incorporated into the vehicle; (ii) when the release was iontophoretically assisted; and (iii) when the penetration enhancer and iontophoretic assistance were used simultaneously. The effects of using isotonic phosphate buffer solution as the aqueous domain of the vehicle and in the receptor were also separately assessed. The passive release from the standard system was consistent with matrix diffusion control. The addition of oleic acid indicated association of the drug with the fatty acid so that its release into an aqueous medium was significantly retarded. With buffer ions present in the vehicle the release rate increased consistent with reduced association, and when phosphate buffer was used as a receptor medium the release rate exceeded that of the standard vehicle due to an ion exchange process. The delivery of salbutamol from the fatty acid containing systems was substantially enhanced by iontophoresis and the rates were shown to be approximately proportional to the assisting currents. The data clearly indicate the iontophoretic process to be significantly less efficient in the presence of buffer ions but with the iontophoretic delivery rates being enhanced by the presence of a fatty acid.

The problem of bias in training data in regression problems in medical decision support.

This paper describes a bias problem encountered in a machine learning approach to outcome prediction in anticoagulant drug therapy. The outcome to be predicted is a measure of the clotting time for the patient; this measure is continuous and so the prediction task is a regression problem. Artificial neural networks (ANNs) are a powerful mechanism for learning to predict such outcomes from training data. However, experiments have shown that an ANN is biased towards values more commonly occurring in the training data and is thus, less likely to be correct in predicting extreme values. This issue of bias in training data in regression problems is similar to the associated problem with minority classes in classification. However, this bias issue in classification is well documented and is an on-going area of research. In this paper, we consider stratified sampling and boosting as solutions to this bias problem and evaluate them on this outcome prediction problem and on two other datasets. Both approaches produce some improvements with boosting showing the most promise.

Mechanism of dissolution of cholesterol-calcium bilirubinate compressed discs in monooctanoin.

The dissolution of cholesterol monohydrate and calcium bilirubinate (neutral salt) mixtures in monooctanoin was investigated using the static disc method. The intrinsic dissolution rate of calcium bilirubinate was orders of magnitude (approximately 1000 fold) lower than that of cholesterol. Cholesterol release decreased as its weight fraction in the solid decreased. In model systems containing below 50% cholesterol dissolution became negligible. The release profiles deviated from the classical model for dissolution from two component mixtures. The observed dissolution profiles of both components were greater than predicted by theory. Anomalous positive curvatures in dissolution profiles suggested that calcium bilirubinate initially reduced the surface area available for cholesterol dissolution. A model, taking into account the change in surface area, was used to fit the cholesterol dissolution data. The results were consistent with the reported relationship between human gallbladder stone cholesterol content and average stone weight loss.

Physicochemical investigations of dipolar aprotic solvents as potential cholelitholytic agents.

A number of dipolar aprotic solvents have been examined as potential co-solvents for gallstone dissolution. The power of these agents to solubilize such gallstone components as cholesterol, calcium carbonate, calcium palmitate and palmitic acid was determined and compared with that of monooctanoin (Capmul), using the synthetic solubility method. The solubilities of cholesterol and palmitic acid were greater in the N-methyl, N-ethyl and N-butylpyrrolidone derivatives than in monooctanoin. In contrast, the solubilities of the calcium salts were very low (less than 0.25% w/w) in all solvents examined. The influence of N-methylpyrrolidone (NMP) on both the in-vitro dissolution of cholesterol from gallstones and the decrease in stone weight with time was determined. NMP proved to be a better solvent than monooctanoin for human stones. NMP, which is miscible with water and monooctanoin, may have potential as a co-solvent in the design of solvent systems for gallstones.

Amorphous spray-dried hydroflumethiazide-polyvinylpyrrolidone systems: physiochemical properties.

Hydroflumethiazide was spray-dried with polyvinylpyrrolidone (PVP) to produce products containing 0-30% PVP. These systems were amorphous and differed from previously prepared coprecipitates of similar composition. Differential scanning calorimetry (DSC) suggested that at low PVP weight fractions both amorphous drug and an amorphous drug-PVP complex can be present in spray-dried systems. The apparent solubility of hydroflumethiazide in spray-dried products increased with increasing PVP content reaching a plateau value approximately four times that of the pure crystalline drug. The estimated free energy and entropy of the spray-dried drug were greater than that of crystalline drug and also increased with increasing PVP content. Dissolution studies with compressed discs supported the apparent solubility data. The results suggest that amorphous phases having different orders of organization are formed in spray-dried systems with increasing PVP content.

The influence of polyvinylpyrrolidone on the dissolution properties of hydroflumethiazide.

The incorporation of hydroflumethiazide with polyvinylpyrrolidone (PVP) was found to retard and to enhance the dissolution of the drug from compressed discs, the magnitude of the effect being dependent on the proportion of PVP present and its method of incorporation. The most active system dissolved sixteen times faster than pure hydroflumethiazide. Low concentrations of PVP were also found to decrease the apparent solubility of hydroflumethiazide while at high concentrations solubility was enhanced. X-ray and infrared analysis of systems suggested the presence of an amorphous form of hydroflumethiazide in coprecipitate systems. The dissolution data were consistent with a physical model which takes account of the roles played by crystalline and amorphous hydroflumethiazide together with the complexing and crystal growth inhibiting effect of PVP on hydroflumethiazide.

Mechanism of drug dissolution rate enhancement from beta-cyclodextrin-drug systems.

The influence of beta-cyclodextrin on the physicochemical properties of bendrofluazide, chlorothiazide, hydrochlorothiazide and hydroflumethiazide was investigated using solubility, X-ray powder diffraction, differential scanning calorimetry (DSC) and intrinsic dissolution rate methods. The solubility of each drug was enhanced in the presence of beta-cyclodextrin, the effect being greatest with bendrofluazide. X-ray diffraction data on equimolar freeze dried systems indicated the formation of solid state inclusion complex only in the case of bendrofluazide. The relative increase in initial dissolution rate of drug from freeze dried systems varied from sixty fold for bendrofluazide to three fold for hydroflumethiazide. The observed dissolution rates were intermediate between those predicted by the classical two component soluble complex model and a carrier controlled model. It was concluded that the enhanced drug dissolution rates, which are above those predicted by the soluble complex model, are due to an extension of carrier phase dissolution control to higher drug weight fractions than predicted by the soluble complex model and that this was a consequence of the disparate solubilities of the carrier and drug.