Prediction of the in vivo performance of enteric coated pellets in the fasted state under selected biorelevant dissolution conditions

Bicarbonate buffer dissolution test with gentle mechanistic stress for bioequivalence prediction of enteric-coated pellet formulations

This study aimed to develop a dissolution test that can predict the bioequivalence (BE) of enteric-coated pellet formulations. The original duloxetine hydrochloride capsule (reference formulation (RF); Cymbalta® 30 mg capsule) and four generic test formulations (two capsules (CP) and two orally disintegrating tablets (OD)) were used as model formulations. Clinical BE studies were conducted on 24-47 healthy male subjects under fasting conditions. Dissolution tests were performed using a compendial paddle method (PD) (paddle speed: 50 rpm) and a flow-through cell method (FTC) (flow rate: 4 mL/min). For a further test, cotton balls were added to the vessel to apply gentle mechanistic stress to the formulations, and paddle speed was reduced to 10 rpm (paddle with cotton ball method (PDCB)).All the dissolution tests were conducted with 0.01 M HCl (pH 2.0) for 0.5 h followed by 10 mM bicarbonate buffer solutions (pH 6.5) for 4 h. One each of the two CP and two OD showed BE with RF. PDCB was able to discriminate between BE and non-BE formulations, while this was not possible with PD and FTC. In PDCB, the cotton balls intermittently moved the pellets near the vessel bottom. PDCB is useful for predicting BE during formulation development.

Unpredictable Performance of pH-Dependent Coatings Accentuates the Need for Improved Predictive in Vitro Test Systems

First introduced in the second half of the 19th century, enteric coatings are commonly used to protect acid-labile drugs, reduce the risk of gastric side effects due to irritating drugs, or for local drug delivery to the lower gastrointestinal (GI) tract. The currently available enteric-coatings are based on pH-sensitive weakly acidic polymers. Despite the long history of their use, the causes behind their performance often being unpredictable have not been properly investigated with most of the attention being focused only on the gastric emptying. However, little attention has been given to the postgastric emptying disintegration and dissolution of these dosage forms. This lack of attention has contributed to the difficulty in predicting the in vivo behavior of these dosage forms and to cases of bioavailability problems with some enteric-coated products. Therefore, increased attention needs to be given to this issue.

Toward Biopredictive Dissolution for Enteric Coated Dosage Forms

The aim of this work was to develop a phosphate buffer based dissolution method for enteric-coated formulations with improved biopredictivity for fasted conditions. Two commercially available enteric-coated aspirin products were used as model formulations (Aspirin Protect 300 mg, and Walgreens Aspirin 325 mg). The disintegration performance of these products in a physiological 8 mM pH 6.5 bicarbonate buffer (representing the conditions in the proximal small intestine) was used as a standard to optimize the employed phosphate buffer molarity. To account for the fact that a pH and buffer molarity gradient exists along the small intestine, the introduction of such a gradient was proposed for products with prolonged lag times (when it leads to a release lower than 75% in the first hour post acid stage) in the proposed buffer. This would allow the method also to predict the performance of later-disintegrating products. Dissolution performance using the accordingly developed method was compared to that observed when using two well-established dissolution methods: the United States Pharmacopeia (USP) method and blank fasted state simulated intestinal fluid (FaSSIF). The resulting dissolution profiles were convoluted using GastroPlus software to obtain predicted pharmacokinetic profiles. A pharmacokinetic study on healthy human volunteers was performed to evaluate the predictions made by the different dissolution setups. The novel method provided the best prediction, by a relatively wide margin, for the difference between the lag times of the two tested formulations, indicating its being able to predict the post gastric emptying onset of drug release with reasonable accuracy. Both the new and the blank FaSSIF methods showed potential for establishing in vitro-in vivo correlation (IVIVC) concerning the prediction of Cmax and AUC0-24 (prediction errors not more than 20%). However, these predictions are strongly affected by the highly variable first pass metabolism necessitating the evaluation of an absorption rate metric that is more independent of the first-pass effect. The Cmax/AUC0-24 ratio was selected for this purpose. Regarding this metric's predictions, the new method provided very good prediction of the two products' performances relative to each other (only 1.05% prediction error in this regard), while its predictions for the individual products' values in absolute terms were borderline, narrowly missing the regulatory 20% prediction error limits (21.51% for Aspirin Protect and 22.58% for Walgreens Aspirin). The blank FaSSIF-based method provided good Cmax/AUC0-24 ratio prediction, in absolute terms, for Aspirin Protect (9.05% prediction error), but its prediction for Walgreens Aspirin (33.97% prediction error) was overwhelmingly poor. Thus it gave practically the same average but much higher maximum prediction errors compared to the new method, and it was strongly overdiscriminating as for predicting their performances relative to one another. The USP method, despite not being overdiscriminating, provided poor predictions of the individual products' Cmax/AUC0-24 ratios. This indicates that, overall, the new method is of improved biopredictivity compared to established methods.

In vitro-in vivo-in silico approach in biopharmaceutical characterization of ibuprofen IR and SR tablets

Within the last decades, physiologically based pharmacokinetic models have emerged into a biopharmaceutical toolkit that has been proven useful in understanding how physicochemical, formulation and physiological factors affect oral drug absorption. The purpose of this study was to develop a drug specific physiologically based pharmacokinetic model that will allow mechanistic interpretation of oral absorption from dosage forms exhibiting different in vitro and different in vivo performance (i.e. immediate release and sustained release tablets) and identification of bioperformance dissolution testing. Ibuprofen was chosen to be used for the "proof of concept" considering it is well characterised and the necessary physicochemical, biopharmaceutical and pharmacokinetic properties for model development could be found in the literature. Gastrointestinal simulation technology implemented in Simcyp® was successful in estimating ibuprofen oral absorption. The developed model exhibited good generalisation ability for the dosage forms studied. The obtained results indicate that the model was sensitive to input kinetics represented by the in vitro drug release profiles obtained under various dissolution conditions. According to the obtained results, reciprocating cylinder apparatus with biorepresentative change in media pH might be considered as bioperformance dissolution in the case of the two ibuprofen SR products studied. These results further justify the use of integrated in vitro-in vivo-in silico approach in estimating bioperformance of oral solid dosage forms.