The improved dissolution and prevention of ampoule breakage attained by the introduction of pretreatment into the production process of the lyophilized formulation of recombinant human Interleukin-11 (rhIL-11).

Lyophilized protein formulations sometimes pose problems such as the formation of a cloudy solution upon reconstitution. Ampoule or vial breakage can also occur during the production processes of lyophilized pharmaceutical products. Various efforts have been made to overcome those difficult problems. In this study, we introduce a particular temperature program into the production process of a recombinant human Interleukin-11 (rhIL-11) lyophilized formulation containing sodium phosphates (Na2HPO4/NaH2PO4, pH 7.0) and glycine in an attempt to improve its dissolution properties and to prevent ampoule breakage from occurring. The formulation was pretreated by nucleating ice and maintaining the solution overnight at a temperature of -6 degrees C. The solution was then completely frozen at a lower temperature. This pretreatment proved successful in not only producing a lyophilized cake which readily disintegrated and dissolved in the reconstitution media, but also prevented ampoule breakage from occurring during the production processes. In contrast, a lyophilized cake produced without the pretreatment created a cloudy solution particularly when reconstituted using water for injection contaminated with aluminum (Al3+), although the solution became transparent within 20-30 min. The pretreatment induced the crystallization of sodium dibasic phosphate (Na2HPO4) in the freeze-concentrate whereas direct freezing without the pretreatment did not crystallize the salt. Thermal analyses (DSC and TMA) showed that amorphous sodium dibasic phosphate in the freeze-concentrate became crystallized upon heating, accompanied by an increase in volume, which probably caused the ampoule breakage that occurred without the pretreatment. Although power X-ray diffraction (PXRD) experiments suggested that, with or without the pretreatment, glycine assumed the beta-form and sodium phosphate stayed amorphous in the final products, an electrostatic interaction between dibasic phosphate anions and rhIL-11, a highly cationic protein, would only exist in the lyophilized cake produced without the pretreatment. This interaction is highly likely because aluminum facilitates the formation of a cloudy solution upon reconstitution possibly by using the divalent anions which effectively reduce electrostatic repulsions between aluminum and the protein to form an aggregate structure that is not readily soluble. The pretreatment would circumvent the interaction by crystallizing the sodium salt before freezing creating a relatively soluble lyophilized cake that is much less sensitive to aluminum.

Potential use of cyclodextrins to enhance the solubility of YM466 in aqueous solution.

Various solubilizing agents for YM466, a new Factor Xa inhibitor, were investigated to begin designing the aqueous formulation for subcutaneous administration. The tentative target concentration was 5 mg/mL. First, three kinds of buffer solutions (glycine-HCl, citrate, and lactate) were examined for their solubilizing effects. The dissolution rate of YM466 in lactate buffer was the fastest, as determined by visual examination at room temperature. The dissolution rate of YM466 in lactate buffer was enhanced, without degradation, by heating at 40 degrees C, and YM466 solution at a concentration of 1 mg/mL became transparent 10 min after the start of heating. The solubility of YM466 increased along with lactate concentrations ranging from 50 mM to 200 mM and reached a high of 1.3 mg/mL after increasing lactate concentration to 200 mM at 5 degrees C. The addition of cyclodextrins beta-cyclodextrin (beta-CD), 2-hydroxypropyl-beta-cyclodextrin (HP-beta-CD), and gamma-cyclodextrin (gamma-CD), but not alpha-cyclodextrin (alpha-CD), had remarkable impact on its solubility, and 7-8 mg/mL of YM466 was dissolved by the addition of HP-beta-CD or gamma-CD. These results demonstrated that YM466 was included in cyclodextrins and that the inclusion formations required a cavity size larger than alpha-CD. Based on the calculation from the linear portion of the phase solubility diagrams, apparent stability constants of alpha-CD, beta-CD, HP-beta-CD, and gamma-CD at 5 degrees C were estimated to be 2M(-1), 206M(-1), 143M(-1), and 276M(-1), respectively. Therefore, we found that gamma-CD has the largest inclusion capacity.

Effect of colonic lactulose availability on the timing of drug release onset in vivo from a unique colon-specific drug delivery system (CODES).

PURPOSE: To test the hypothesis that the onset of drug release in vivo from a unique colon-specific drug delivery system (CODES) would depend on the colonic availability rate of lactulose. The site specificity of drug release in canine GI tract was also estimated. METHODS: CODES tablets were prepared by tableting the granulation or acetaminophen and lactulose, followed with film coating. The pharmacokinetic performance of different CODES formulations was evaluated in six beagle dogs under fasted conditions. The release of acetaminophen and lactulose was also characterized in vitro. RESULTS: The onset of acetaminophen release in beagle dogs was found to be dependent on the coating level of Eudragit E and lactulose loading in the core tablet. At Eudragit E coating levels of 4%, 8% and 12% (coating weight gain), the onset of in vivo drug release occurred 5.5 (+/- 1.9) h. 4.8 (+/-1.0) h. and 7.5 (+/-1.0) h, respectively, after dosing. A similar trend was observed when the loading of lactulose in the core tablet decreased from 78% to 58% and 38%. However, the rate and extent of acetaminophen absorption did not vary significantly in each situation based on the values of AUC and Cmax. CONCLUSION: The onset of drug release in vivo from CODES tablets is predominantly dependent on colonic availability rate of lactulose because drug release from this system is triggered by localized drop of colonic pH from the fermentation of lactulose.

Effect of food on gastrointestinal transit of liquids in cynomolgus monkeys.

We investigated the gastrointestinal transit of liquids, as well as various gastric pH profiles, in fed cynomolgus monkeys. Twelve grams of a biscuit-type solid food were provided 1 h before the test. The acetaminophen method was used to determine the gastric half-emptying time (t(50%)), which provided an estimate of the gastric emptying rate. The gastric emptying rate of liquids was significantly reduced after food intake in monkeys. The mean t(50%) value was 143.5 min and comparable to that of humans after eating. However, there was a large variability in the t(50%) between individual animals as shown by the coefficient of variance of approximately 80%. Next, the median oro-caecal transit time in fed monkeys was determined to be 1.8 h, using the sulfasalazine-sulfapyridine method. There was no significant difference in oro-caecal transit time between unfed and fed monkeys; thus, food intake has no significant effect on the oro-caecal transit time of liquids in either monkeys or humans. However, the oro-caecal transit time in humans is about 2 h longer than that in monkeys. Our experiments using several different foods suggested that the typical human gastric pH profile could not be simulated in fed monkeys.

Gastrointestinal transit of liquids in unfed cynomolgus monkeys.

In this study, the gastric emptying rate and oro-caecal transit time of two liquid volumes (20 and 60 ml) were compared in unfed cynomolgus monkeys. First, the acetaminophen method was used to determine the gastric half-emptying time (t(50%)). The mean t(50%) from seven monkeys was 21.2 min for the 20-ml volume and 27.8 min for the 60-ml volume. This mean t(50%) in monkeys is similar to that reported previously in fasting humans. Next, the sulfasalazine-sulfapyridine method was used to determine median oro-caecal transit times for the 2 liquid volumes; these times were 2.5 h for the 20-ml volume and 2.3 h for the 60-ml volume, which are about 1.5 h shorter than previously reported transit times in humans. An increase in volume administered did not significantly change either the t(50%) or oro-caecal transit time. The data also show that variability in both t(50%) and oro-caecal transit time within each monkey is not as great as the large variability between monkeys. Consequently, cynomolgus monkeys are good model animals to use for studies on the gastric emptying of drug-containing liquids after fasting; however, analysis of results from bioavailability studies must compensate for differences in the oro-caecal transit time between monkeys and humans.

Timed-release formulation to avoid drug-drug interaction between diltiazem and midazolam.

The purpose of this study was to investigate whether the use of a timed-release (TR) drug formulation can avoid unfavorable pharmacokinetic drug-drug interactions in vivo. First, the effects of the time interval between administration of midazolam and diltiazem on the known drug-drug interaction between these drugs were investigated in dogs. When dogs were given midazolam orally at the same time they were orally given an aqueous diltiazem solution, the area under the plasma concentration-time curves of midazolam increased significantly compared with that of midazolam given orally in the absence of diltiazem. However, there was no significant difference in pharmacokinetics of midazolam when the diltiazem solution was administered 1-2 h after midazolam. Tests on a TR formulation containing diltiazem demonstrated that the first appearance of diltiazem in plasma occurs at 2.6 +/- 0.5 h in dogs. Subsequent tests showed that the plasma concentration-time profile of midazolam after concurrent oral administration of the diltiazem TR formulation was almost the same as that of midazolam administered alone. These results demonstrate that a TR formulation of diltiazem can avoid the interaction between diltiazem and midazolam by creating a time interval between absorption of drugs in vivo, without the need for closely controlling the time of drug administration.

A new stressed test to predict the foreign matter formation of minodronic acid in solution.

A formulation containing 0.5 mg/ml minodronic acid, 40 mM citrate, pH 4.5, and sodium chloride, stored in regular flint glass ampoules, was stable without particulate increase under high temperature conditions, such as 40 degrees C for 6 months, or 50 or 60 degrees C for 3 months. However, when stored at 25 degrees C, there was an increase in >or=2 microm particles at the 5-month timepoint. This demonstrated that long-term stability cannot simply be predicted by the evaluation of samples just stored at higher temperatures. Therefore, a new stressed test was designed which is useful in the rapid selection of formulations that are stable and without particulate increase. Since the particulate matter is apparently a complex of minodronic acid and aluminum ions leaching from ampoules, samples were placed at 80 degrees C for up to 4 weeks to accelerate aluminum leaching. Although no particulate increase was observed directly after storage at 80 degrees C, 4 freeze-thaw cycles following the storage caused a drastic particulate increase. The evaluation of samples subjected to the freeze-thaw cycles indicated that the following formulation modifications have inhibitory effects on particulate generation: (1). addition of meglumine, diethanolamine, mannitol, or glycerol to the formulation; (2). increase of citric acid concentration; (3). decrease of minodronic acid concentration. These modifications also worked well for samples stored at 25 degrees C for 6 months, and particulate increase did not occur. This method is a powerful tool for predicting the stability of minodronic acid in solution.

Characteristics of the gastric pH profiles of unfed and fed cynomolgus monkeys as pharmaceutical product development subjects.

Gastric pH is an important factor which significantly affects the dissolution of drugs, and therefore their bioavailability. In this study, the gastric pHs were measured directly with a miniature pH electrode inserted through the nostril into the body of the stomach of cynomolgus monkeys. Results from three separate sets of measurements using the same male monkeys indicated that the median gastric pH profiles of unfed monkeys were low, fluctuating between pH 1 and pH 3. However, the median gastric pHs in fed monkeys given about 108 g of a biscuit-type solid food, which are commonly provided, shifted toward a more neutral range between pH 5 and pH 7, and remained in this range for about 9 h. This result contrasted with reported results for humans after eating a standard meal, which showed a neutral range between pH 5 and pH 7 for a brief period. Consequently, these results indicate that although the gastric pH of unfed cynomolgus monkeys is similar to that of fasting humans, there is a great difference in the gastric pH profiles between humans and monkeys after eating, which suggests that further studies are needed to establish optimal feeding conditions for bioavailability studies in monkeys.

Common solubilizers to estimate the Caco-2 transport of poorly water-soluble drugs.

Solubilizers are often used to enhance the bioavailability of drugs with poor aqueous solubility. This study focuses on the use of the Caco-2 system containing solubilizers to predict the absorption of poorly water-soluble drugs in humans. First, the effects of propylene glycol (PG), hydroxypropyl-beta-cyclodextrin (HP-beta-CD), polyethylene glycol 400 (PEG 400), and Tween 80 on the viability (transepithelial electrical resistance, TEER) of 3-day cultured Caco-2 monolayers were evaluated. These solubilizers, even at the low concentration, reduce the viability of Caco-2 monolayers; these results indicate the impossibility for 3-day cultured Caco-2 monolayers to be used for this test. Next, the effects of PG, Tween 80, PEG 400, HP-beta-CD, Pluronic F-68 (Pluronic), HCO-40, sodium lauryl sulfate (SLS), Gelucire 44/14, Transcutol P, and extract gall powder on the viability of 21-day cultured Caco-2 monolayers and the apparent permeability (P(app)) of propranolol (PPL), Nadolol (NDL), and FITC-dextran 4000 (FD-4) were investigated. Five different solubilizing methods (20% PG, 5% Tween 80, 5% PEG 400, 5% HP-beta-CD, and 5% Tween 80+5% PEG 400) did not affect the viability of 21-day cultured Caco-2 monolayers. Furthermore, the P(app) values of the three compounds containing these solubilizers did not differ from the values for control formulations (without solubilizers). These results clearly suggest that the use of PG, Tween 80, PEG 400, or HP-beta-CD as solubilizing excipients and the testing of these formulations on 21-day cultured Caco-2 monolayers can predict intestinal absorption of poorly water-soluble drugs in humans.

Failure of stability prediction for minodronic acid injectable by accelerated stability testing.

A liquid formulation containing 0.5 mg/ml minodronic acid, 40 mM, pH 4.5, citrate, and sodium chloride added to adjust the osmolarity of the final formulation was stored in flint glass ampoules at 25, 40, 50, and 60 degrees C. At specified times, the drug potency and pH, and the tendency to generate particulate matter, were measured. Test samples stored at 40 degrees C for 6 months or at 50 and 60 degrees C for 3 months were stable with no potency loss and no particulate increase. However, despite the satisfactory stability at high temperatures, the amount of particulate matter increased when the formulation was stored at 25 degrees C. Scanning electron microscopy-energy dispersive X-ray analysis of the particulate matter revealed that it contains aluminum and phosphorus, the latter thought to be derived from minodronic acid. In contrast, the number of the particulate matter did not increase, when the formulation was stored in either plastic containers or in SiO(2)-treated glass ampoules(.) The spike of minodronic acid solution with aluminum ions led to the particulate generation. These results demonstrate that the particulate matter is a complex of minodronic acid molecules and aluminum ions, which apparently leached from the glass of regular ampoules. Since the particulate generation could not be observed at higher temperatures, it was suggested that the complex formation was exothermic and accelerated testing did not predict the stability in terms of particulate generation.

Influence of water soluble fillers in hydroxypropylmethylcellulose matrices on in vitro and in vivo drug release.

The purpose of this study was to investigate the effect of fillers in gel-forming matrix on in vivo drug release after oral dosing. A further purpose was to predict the in vivo performance from in vitro dissolution test. Three controlled-release acetaminophen tablets containing hydroxypropylmethylcellulose (HPMC) with or without highly water soluble fillers, lactose or polyethylene glycol 6000 (PEG6000), were prepared. Water penetration into the matrix was enhanced by addition of fillers in the matrices, but the three tablets showed similar in vitro dissolution profiles, indicating that fillers in the HPMC matrices little affected the in vitro drug release. In contrast, the fillers in HPMC matrices did affect the in vivo performance in dogs. The absorption profile of HPMC matrix with PEG6000 was the fastest, followed by that with lactose and without water soluble filler, in that order. As the matrix with PEG6000 had a large amount of water and gelated a large portion of the matrix when in contact with water, the gel layer would be disintegrated by the gastrointestinal motility. It was found that dissolution of gel-forming HPMC matrices under mechanical stress by glass beads well correlated with the in vivo performance of the matrix, with little correlation by the conventional paddle method.