Performance of supersaturating drug delivery systems in real-life conditions: in vivo characterisation of gastric drug behavior supporting the development of predictive tools.
Once marketed, drugs may suffer from suboptimal performance as they are often taken with meals and/or beverages not accounted for during clinical trials. Any ingested formulation will initially reside in the stomach where exposure to a variety of internal (e.g. pH and motility) and external factors (e.g. co-administration of beverages and foods) potentially influences intragastric formulation disintegration and drug dissolution. As this environment dictates how a drug is presented to the intestinal absorption compartment, these real-life dosing conditions can consequently affect systemic drug disposition. Moreover, a long intragastric residence time of a highly permeable drug creates a plausible scenario for gastric drug absorption. This thesis investigated (i) the potential influence of real-life dosing conditions on the gastrointestinal behavior of drugs and (ii) the potential of the gastric mucosa to absorb drugs.
In a first set of experiments in this research project, intraluminal ethanol concentrations (stomach and duodenum) were determined in fasted and fed healthy volunteers after the consumption of common alcoholic beverages: beer, wine, and whisky. Upon gastric arrival, a trend was seen where smaller ingested volumes resulted in higher degrees of ethanol dilution, though, interestingly, no clear differences were observed in the degree of gastric dilution between fasted and fed state volunteers. Overall, these studies revealed relatively low and rapidly declining intragastric ethanol concentrations after drinking two standard consumptions of beer, wine or whisky. Compared to the fasted state, postprandial gastric ethanol concentrations remained higher for a longer time following the Cmax. In all cases, lower ethanol concentrations were observed in the duodenum compared to the stomach. Current FDA guidelines for testing the alcohol resistance of formulations state that in vitro tests should investigate the impact of up to 40% ethanol for 2 h. The intragastric ethanol concentrations observed in the present study question the relevance of these strict guidelines.
Further studies in this research project investigated the effects of several real-life dosing conditions on the intraluminal behavior of drugs. The first study performed investigated the influence of (i) concomitant proton-pump inhibitor (PPI) intake and (ii) the administration of a liquid meal on the intraluminal dissolution, supersaturation and precipitation behavior of indinavir in healthy volunteers. A fasted state trial arm was included as a reference condition. When indinavir (Crixivan®) was administered to volunteers on a PPI regime, lower intragastric and intraduodenal indinavir concentrations were observed compared to the fasted state. An elevated intragastric pH hampered indinavir release from the capsule and resulted in a low solubilizing capacity for the drug in the stomach. Gastric supersaturation was observed in all volunteers on a PPI regime. The elevated intragastric pH resulted in lower duodenal indinavir concentrations through reduced dissolution and/or increased precipitation of indinavir in the stomach indicating that this real-life dosing condition can have a direct impact on the driving force for intestinal absorption.
When indinavir (Crixivan®) was administered to healthy volunteers with a liquid meal (Ensure® Plus) a notable delay in intragastric drug release was observed. The fraction of dissolved indinavir in the fed stomach was similar to the fasted state. Furthermore, reasonably high duodenal indinavir concentrations were observed compared to the fasted state. The bioaccessible fraction of indinavir was determined and appeared lower in a fed state human intestinal sample compared to a fasted state sample indicating micellar entrapment of the drug in the fed state. This difference in bioaccessible fraction of indinavir may cause reduced intestinal absorption irrespective of duodenal concentrations. In conclusion, both real-life dosing conditions tested notably affected intraluminal drug behavior and thus potentially systemic exposure, albeit through different mechanisms.
In a next study, the gastrointestinal behavior of the weakly acidic drug diclofenac was studied in healthy volunteers when given a solid meal. Ingestion of the FDA standard meal resulted in a gastrointestinal pH-profile similar to those observed in studies using a liquid meal. A delay in intragastric tablet disintegration was observed suggesting that the formation of a food-dependent precipitation layer on the tablet. Despite similar pH-profiles, the fraction of diclofenac dissolved in the gastrointestinal tract differed compared to observations following intake of the formulation with a liquid meal. Subsequent in vitro tests suggested that the FDA standard meal affects intragastric diclofenac dissolution from a Cataflam® tablet and local diclofenac solubility in a different way compared to the liquid meal. Furthermore, dissolved diclofenac molecules were observed to adsorb to several meal components present in the FDA standard meal. From this study, it becomes clear that meal composition and the consistency of meal (liquid or solid) given can markedly affect intraluminal drug disposition.
Literature research in the context of intraluminal ethanol and drug behavior revealed earlier experiments hinting at the ability of the stomach to absorb alcohol and other small molecules. To explore the potential gastric absorption of pharmaceutical compounds, an in situ gastric bolus administration rat model was used. In this model, gastrointestinal transfer was blocked by ligating the pylorus and drugs were injected into the stomach as solution (paracetamol and diclofenac) or suspension (posaconazole). In this set-up, both paracetamol (neutral) and diclofenac (weakly acidic) appeared in the systemic circulation of fasted and fed rats indicating absorption through the gastric mucosa. For paracetamol, the relative contribution of the gastric absorption was higher in the fed state compared to the fasted state. Very low systemic posaconazole concentrations were detected indicating negligible gastric absorption of a weakly basic compounds. Results from this study demonstrated the ability of the stomach to absorb alcohol and pharmaceutical compounds.
During drug development, clinical trials do not account for the large variety of beverages and meals taken with drugs, though these real-life dosing conditions can affect systemic drug exposure and thus therapy effectiveness. The potential influence of real-life dosing conditions is clear, yet further research is desired to elaborate this knowledge. Moreover, greater attention for the stomach as an absorptive organ is warranted.
The impact of real-life dosing conditions and gastric absorption on gastrointestinal and systemic drug disposition cannot be underrated; further research and discussion regarding potential implications in predictive in vitro and in silico tools are encouraged.