VISUALIZACIÓN DE RESULTADOS

The present study aimed to find out whether the presence of relatively high amounts of cosolvents or surfactants has a significant impact on the passive diffusion of ranitidine. It needs to be mentioned that the employed technique of dialysis diffusion has certain limitations and can only serve to explain a fraction of the processes taking place at the absorption site of ranitidine in vivo. While the employment of the hydrophilic dialysis membrane allowed a satisfying insight into the diffusion of ranitidine along its aqueous absorption pathways the present experiment may not reflect the movement of the drug through the tight junctions in every aspect. Tight junctions, for instance, have been reported to have a rather small pore size in the range of 0.5 to 5 nm (Macheras et al., 1995), which is higher in the small intestine than in the colon (Chadwick et al., 1977b). Larger molecules are, therefore, limited in their extent of absorption, as was shown for PEG 400, which has been described to be absorbed at around 50 % of the administered dose (Chadwick et al., 1977a; Ma et al., 1990). In the present study, however, it is suggested that the polymer diffused quantitatively as a result of the relatively high molecular weight cut-off of the dialysis membrane (12000 to 14000). The permeability of drugs has been shown to be dependent on the pore size of the dialysis membrane, although the impact appeared to increase in significance with increasing lipophilicity of

the compounds investigated (Yoon and Burgess, 1996). Dialysis membranes with a molecular weight cut-off as low as 1000 exist and it is likely that the overall diffusion rate of ranitidine is reduced by a lower membrane pore size. It is, however, assumed that diffusion studies employing such a membrane would not result in contradictory findings to the present study with respect to the influences of solubilizing agents since a molecular weight cut-off of 1000 is expected to also allow quantitative diffusion of the investigated cosolvents PEG 400 and propylene glycol. Nevertheless, it would have been preferably to quantify the diffusion of the administered excipients in addition to the qualitative assessment. Quantitative analysis via FAB mass spectrometry is possible when an internal standard is employed, e.g. ^^PEG400. The development of such a method, however, is not straightforward and very time consuming and as a result the input was thought to be out of proportion for the purposes of the present experiment. Alternative methods for quantitative determination of PEG 400, for instance, include refractive index diffraction.

Apart from the limitations of the dialysis technique to assess the in vivo effects of solubilizing agents on ranitidine, it is generally not hydrophilic drugs that are formulated with solubilizers, as they are sufficiently soluble in aqueous media. Poorly water-soluble drugs, e.g. class II drugs, on the other hand, require the addition of these excipients on a regular basis often at relatively high amounts in order to promote their solubility in the physiological fluids. It would have been of interest to also investigate the effects of the employed solubilizers on the passive diffusion of a hydrophobic compound such as amprenavir. The presence of high solubilizer concentrations in Agenerase® oral solution was found to have a significantly negative impact on the oral bioavailability of amprenavir, which might to a certain degree as well be attributable to an effect of the excipients on drug passive diffusion. However, the conduction of an experiment involving a class II drug dissolved in solubilized systems is faced with certain difficulties as a result of the sometimes extremely low aqueous solubility of the drug. In such cases, the solubilized system administered in the dialysis bag will contain comparatively high amounts of drug dissolved, which means drug levels close to saturation. With the start of the experiment water from the sink solution will rapidly penetrate into the dialysis bag concomitantly with efflux of the cosolvent out of the bag. This process results in a rapid decrease in solubilizer concentration, which in turn causes the system to supersaturate

and hence have the active ingredient precipitated. Such an effect was also observed by Levy and Benita (1990), who investigated the in vitro release of diazepam from a submicron o/w emulsion employing the dialysis bag technique. Here, the release rate of diazepam was drastically reduced as a result of drug precipitation from a 70 % alcohol solution, which was only obtained when water served as sink medium but not with 70 %

alcohol in the receiver vessel. It has been suggested to add non-aqueous solvents or solubilizers to the sink solution e.g. equal solubilizer-water mixtures in donor and receiver department, respectively. With this procedure, however, it is essential to investigate the drug diffusion rate as a function of the concentration of solubilizer in the sink (Washington, 1990). The results, however, are expected to be of little value if the sink additives predominate the effects on diffusion or, as in the present case, the solubilizer themselves are under investigation. It is possible, however, to reduce the concentration of the administered drug to such an extent that dilution of the solubilized system via water influx into the dialysis bag does not lead to supersaturation. This in turn is likely to result in drug concentration levels in the receiver solution that are below the level of detection at least with the set-up of automated UV detection used in the present experiment.

Apart from the mentioned difficulties in conducting the present diffusion study with a poorly water-soluble drug, it is doubtful that such results give valuable information on the absorption behaviour of the drug since hydrophobic molecules do not use aqueous absorption pathways but move across the lipophilic membrane of the enterocytes. It would, therefore, have been more interesting to investigate the effect of solubilizers on the transcellular passive diffusion employing a hydrophobic membrane instead. Membranes of polysiloxanes, for instance, serve as artificial hydrophobic barriers in in vitro diffusion cell studies to model intestinal as well as transdermal absorption. The influence of cosolvents and surfactants on passive diffusion of lipophilic drug molecules differs to their effects observed with hydrophilic molecules, which have been mentioned above. Besides their impact on drug solubility and precipitation the presence of solubilizing agents has also been shown to influence the octanol/water partition coefficient, logP value of a drug or more precisely the octanol/aqueous phase partition coefficient. LogP is a measure for the relative lipophilicity of a drug and its likeliness to diffuse through the lipid bilayer of the epithelium. Increasing concentrations of cosolvent

appear to have a decreasing effect on logP, which means that cosolvents reduce the affinity of a hydrophobic drug to leave the aqueous medium of the luminal contents and reach the lipophilic membrane. Significant reductions in logP may result in a low permeability of the therapeutic agent. A linear relationship has been described between the logarithm of the effective intestinal permeability, Peff, of carbamazepine and the octanol/water partition coefficient (Riad and Sawchuk, 1991). In this in situ perfusion technique on the duodeno-jejunum and ascending colon in the rabbit increasing percentages of PEG 400 decreased the logP value of the solution and concomitantly reduced the fraction of drug absorbed (Riad and Sawchuk, 1991). Hamza and Kata (1990) employed a Sartorius absorption simulator in an in vitro availability study on allopurinol, which uses artificial lipid barriers consisting of inert membrane filters, the pores of which are filled with a liquid lipid phase. The absorption simulator distinguishes between a stomach wall barrier and an intestinal wall barrier and uptake of drug from simulated gastric or intestinal juice to plasma is monitored. The authors investigated the solubilization effect of various cosolvents on allopurinol diffusion across the lipid barriers and observed a significant decrease in drug diffusion from intestinal juice to plasma with cosolvent solutions of PEG 400, propylene glycol, glycerol and sorbitol (Hamza and Kata, 1990). In contrast, an enhancement of allopurinol diffusion was observed with the addition of a surfactant (Tween® 80). Effects of the non-ionic surfactant Brij 97 on the effective permeability, Peff, of various model drugs administered as emulsions was investigated by Yoon and Burgess (1996) in an in vitro diffusion study using polysiloxane membranes. Permeation of the hydrophobic membrane by lipophilic drugs was found to increase with increasing surfactant concentrations up to 1.0 % (w/w) and then decrease with further increase in micellar concentration. At the same time increasing Brij 97 concentrations reduced the oil/surfactant-in-water partition coefficient of the drugs. Since only drug molecules present in the continuous aqueous phase of a micellar solution are available for absorption the presence of high amounts of micelles will result in entrapment of the drug molecules within the core of the micelles. In addition to entrapment inside the micelles, drug molecules traversing the continuous aqueous phase to reach the lipophilic absorptive membrane face reduced thermodynamic activity as a result of high micelle concentrations. The benefit of micellar solubilization regarding increased drug solubility at lower surfactant concentrations has, therefore, been found to have a negative impact on drug permeability at higher concentrations

(Yamada et al., 1966; Saski, 1968; Yoon and Burgess, 1996). In contrast, decreases in the drug’s thermodynamic activity as a result of complex formation between drug and solubilizer is not likely to affect lipophilic drugs to a significant extent since complexation primarily occurs via hydrogen bonding between hydrophilic moieties.