- Email: [email protected]
Bioavailability of seocalcitol I: Relating solubility in biorelevant media with oral bioavailability in rats—effect of medium and long chain triglycerides
Bioavailability of Seocalcitol I: Relating Solubility in Biorelevant Media with Oral Bioavailability in Rats—Effect of Medium and Long Chain Triglycerides ¨ LLERTZ3 METTE GROVE,1,3 GITTE P. PEDERSEN,1 JEANET L. NIELSEN,2 ANETTE MU 1
Pharmaceutical Formulation, LEO Pharma A/S, Industriparken 55, DK-2750 Ballerup, Denmark
2
Department of Pharmacokinetics, LEO Pharma A/S, Industriparken 55, DK-2750 Ballerup, Denmark
3
Department of Pharmaceutics, The Danish University of Pharmaceutical Sciences, Universitetsparken 2, DK-2100 Copenhagen, Denmark
Received 8 November 2004; revised 8 March 2005; accepted 21 April 2005 Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.20403
ABSTRACT: Simulated intestinal media (SIM) containing bile salt (BS) and phospholipids (PL) with and without medium chain lipolytic products (MC-LP) or long chain lipolytic products (LC-LP) were developed to study the solubility of seocalcitol. Both MC-LP and LC-LP were studied in order to investigate the influence of fatty acid chain length on the in vitro solubility of seocalcitol. The same solubility of seocalcitol was found in media containing either MC-LP or LC-LP. The bioavailability after oral administration of seocalcitol dissolved in medium chain triglyceride (MCT), long chain triglyceride (LCT), and a reference formulation containing propylene glycol (PG) was studied in vivo in rats. The lipid formulations showed a twofold increase in bioavailability compared with the reference formulation, indicating positive effects of lipids on the bioavailability reflecting a better solubility in the intestine and protection against precipitation of seocalcitol in the gastro intestinal tract. There was no difference in the in vivo bioavailability of seocalcitol between the MCT and the LCT solutions, which correlates with the identical in vitro solubility of seocalcitol in SIM containing MC-LP or LC-LP. ß 2005 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 94:1830–1838, 2005
Keywords: oral bioavailability; lipid-based formulations; oral absorption; poorly soluble drug substances; solubility; simulated intestinal media; seocalcitol
INTRODUCTION Drug substances with low water solubility and high permeability are classified as Class 2 drug substances in the Biopharmaceutical Classification System (BCS).1 In the gastrointestinal tract (GIT) the limiting step in drug absorption for class 2 drug substances is the dissolution process. In order to overcome the dissolution process, poorly soluble drug substances can be dissolved in the
Correspondence to: A. Mu¨llertz (Telephone: þ45 35306440; Fax: þ45 35306030.; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 94, 1830–1838 (2005) ß 2005 Wiley-Liss, Inc. and the American Pharmacists Association
1830
formulation before entering the GIT, e.g., in propylene glycol (PG) solution or lipid vehicles. The advantage of applying a lipid vehicle in comparison to a simple water miscible system, is that the formed lipid degradation products may increase the drug absorption via enhanced solubilisation and dispersion of the drug substance in the intestine.2,3 During digestion of the lipid formulation by pancreas lipase, the drug substance is transferred from the lipid formulation into the intestinal environment, where it is incorporated into mixed micelles, containing bile salts (BS), phospholipids (PL), and lipolytic products (LP).4,5 Even small amounts of lipid are believed to induce the secretion of bile to facilitate these mechanisms.6 For poorly soluble drug substances the
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
SEOCALCITOL I: SOLUBILITY AND BIOAVAILABILITY
critical step in drug absorption is to avoid precipitation when transferred from the lipid formulation to the mixed micelles in the GIT. The capacity of the mixed micelles to keep the drug substance in solution depends on the characteristics of the drug substance, e.g. the lipophilicity, and on the composition of the mixed micelles. Solubility studies with simulated intestinal media (SIM) containing medium chain lipolytic products (MC-LP) or long chain lipolytic products (LC-LP) may give valuable information about what kind of lipids that most efficiently keep the drug substance in solution.7–9 It has been shown that drug substances with low log P (approximately 2) seem to have a better solubility in media containing MC-LP, whereas the solubility of drug substances with high log P (8–9) is favored by adding LC-LP to the media.8,9 For felodipine, having a log P of 4.8, the best solubilising capacity was found in SIM containing LC-LP8 whereas for danazol, having a log P of 4.5, the best solubility capacity was found in SIM containing MC-LP.9 Thus, as to drug substances with log P values around 4–5 no clear trend is seen regarding the type of lipid that will cause the highest increase in the solubility in biorelevant media why solubility studies in SIM with LP can be used as a tool to gain information about what kind of lipid to use. A strategy for a lipid based formulation of a given poorly soluble drug will include solubility studies in the excipients to ensure that the drug is in solution in the formulation. However, equally important is that the drug will be kept in solution during passage through the GIT, during digestion of the formulation and transfer between different formed colloid phases, until absorption from the mixed micelles. This makes solubility studies in SIM, simulating the composition of the GIT fluids during digestion of a given formulation, a possible and interesting formulation tool. Frequently used formulation vehicles in lipidbased formulations of poorly soluble drugs are medium chain triglycerides (MCT) and long chain triglycerides (LCT). Only a few studies have been performed comparing the bioavailability of drug substances dissolved in MCT or LCT, respectively.10–14 For halofantrine, cyclosporine A, probucol, and vitamin D3, the best bioavailability was obtained from LCT,10,12,13 whereas the best bioavailability for penclomedine and vitamin E was seen from MCT.11,14 Comparing these data, there seem to be no clear trend as to what characteristics
1831
of a compound that determines whether LCT or MCT will give the best bioavailability. The objectives of the present work were to: (a) compare the in vivo oral bioavailability of a model drug substance from pure oily solutions, MCT and LCT respectively, to a reference PG solution, (b) investigate the influence of SIM on the in vitro solubilisation of the model drug substance, including SIM containing either MC-LP or LC-LP in different concentrations, and (c) evaluate if in vitro solubility data can be related to in vivo behavior of the present model drug substance from formulations containing either MCT or LCT. Seocalcitol, a D-vitamin analogue and a class 2 drug, was chosen as a model drug in this study. Selected physicochemical properties of seocalcitol are shown in Table 1. The structure of seocalcitol is shown in Figure 1.
MATERIALS AND METHODS Materials Seocalcitol and tritium labeled seocalcitol (3Hseocalcitol, specific activity of 0.844 MBq/mg and radiochemical purity of 97%) were synthesized at LEO Pharma. Citric acid monohydrate and Sodium Chloride were purchased from Merck, Germany, Sodium Citrate from Jungbunzlauer GmbH, Germany, PG from Lyondell Chem. Comp., France, Viscoleo (MCT) from ICI Espana, Spain, and Sesame oil (LCT) was purchased from Henry Lamotte GmbH, Germany. The fatty acid composition of the triglycerides in MCT and LCT is as specified in Ph.Eur. Sodium taurocholate (97%) (NaTC), Trisma maleate (99%), Oleic acid (approx. 95%), Decanoic acid (99%–100%), Octanoic acid (99%), and Sodium Azide (99%) were all Table 1. Physicochemical Properties of Seocalcitol Physico-Chemical Parameter Log P Pappa (cm/s) Swaterb (ng/g) SMCTc (mg/g) SLCTd (mg/g)
4.8 6.35 106 20 5.3 1.7
a Apparent permeability coefficient (Papp), investigated in the CaCo-2 cell line at 378C. b Solubility in Milli-Q water (Swater) at 258C, n ¼ 3. c Solubility in Viscoleo (fractionated coconut oil) (SMCT) at 258C, n ¼ 2. d Solubility in sesame oil (SLCT) at 258C, n ¼ 2. Data from internal report, LEO Pharma A/S. Seocalcitol concentration determined by HPLC.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1832
GROVE ET AL.
Figure 1. Chemical structure of seocalcitol and 3 H-seocalcitol.
purchased from Sigma-Aldrich (St. Louis, MO). Lipoid S100 was purchased from Lipoid GMBH. RyloTM MG 19 Pharma (Glycerol Monooleate) and Emulsifier TS-PH 003 (Glycerol Monocaprate) were gifts from Danisco, Denmark. Hypnorm1 (fentanyl 0.2 mg/mL, fluanisone 10 mg/mL) was purchased from Janssen, Belgium, and Dormicum (Midazolam 5 mg/mL) from Roche, Switzerland. Heparin 10,000 IU/mL was produced at LEO Pharma A/S. Isotonic sterile NaCl was purchased at Dilab, Sweden. Water was obtained from a Milli-Q-water purification system (Millipore, MA). Pico-AquaTM and Pico Flour 40TM were purchased from Packard. All other chemicals were of analytical grade.
dissolved, the pH was adjusted to 6.5. The composition of the SIM used in this study is shown in Table 2. Oleic acid was chosen as a long chain fatty acid (LCFA) and glycerol monooleate as a long chain monoacyl glycerol (LCMG). Octanoic acid and decanoic acid in the ratio 60:40 were chosen as medium chain fatty acids (MCFA) and glycerol monocaprate as a medium chain monoacyl glycerol (MCMG). Solubility Studies of Seocalcitol in Simulated Intestinal Media The media were pre-heated to 378C and 10 mL were added to screw cap vials containing 30 mg (excess) of seocalcitol powder (n ¼ 3). The vials were gently rotated at 100 rpm in a heating chamber (Lab-Therm, Ku¨hner, Switzerland) at 378C. Two aliquots of 1 mL were removed after 24 and 48 h, respectively and centrifuged at 12000 rpm for 10 min (Hemle High Speed Centrifuge 2-200-M). 500 mL of the supernatant was mixed with 500 mL ethanol and 1 mL 50% methanol (v/v) was added. The solutions were mixed for 10 s and analyzed by HPLC.
Composition of Simulated Intestinal Media
HPLC Analysis of Seocalcitol in Simulated Intestinal Media
Trisma maleate was used as buffer in a concentration of 100 mM. NaN3 (3 mM) was added to prevent microbial growth. NaCl was used to give the same osmolality of 270 mosmol/L in the total media. For each media, all the excipients were accurately weighed into beakers, distilled water was added, and the solutions were heated to 508C while stirring. When all excipients were perfectly
The concentration of seocalcitol in simulated intestinal media was determined by reversed-phase HPLC consisting of a Waters Alliance Separation Module 2790 with cooling on the autosampler (88C), online vacuum degassing, column oven (408C), a Dual l Absorbance Detector 2487 set at 264 nm and data were processed by using Millennium Chromatography Manager version 32 (all
Table 2. The Composition of SIM. Horizontally: The Concentration of BS and PL. Vertically: The Concentration of FA and MAG
— 5 mM LCFA, 2.5 mM LCMG 10 mM LCFA, 5 mM LCMG 15 mM LCFA, 7.5 mM LCMG 20 mM LCFA, 10 mM LCMG 5 mM MCFA, 2.5 mM MCMG 10 mM MCFA, 5 mM MCMG 15 mM MCFA, 7.5 mM MCMG 20 mM MCFA, 10 mM MCMG
BS: 5.0 mM, PL: 1.25 mM
BS: 10 mM, PL: 2.5 mM
BS: 15 mM, PL: 3.75 mM
BS: 20 mM, PL: 5 mM
The marks in the table represent the media used in the solubility study. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
SEOCALCITOL I: SOLUBILITY AND BIOAVAILABILITY
Waters Associates, Milford, MA). The analytical column was a Symmetri C8, 50 2.1 mm i.d. 3.5 mm (Waters) and the injection volume was 150 mL. The mobile phase was a linear gradient from 70 to 99% methanol using two eluents: (A) methanol:water (20:80 v/v), (B) methanol:water (99:1 v/v), and the flow rate was 0.3 mL/ min. The retention time of seocalcitol was approximately 4 min and quantification was performed using a calibration curve of spiked blank media, the lower limit of quantification being 100 ng/mL. Preparation of Oral 3H-Seocalcitol Formulations to Bioavailability Study Two lipid based formulations, one MCT solution (viscoleo) and one LCT solution (sesame oil), and an oral reference solution (PG) were developed by dissolving the 3H-seocalcitol in the formulation, resulting in a concentration of seocalcitol of 59 mg/ mL and 50 MBq/mL.
1833
0.79 mm OD, Dilab, Sweden). The catheter was exteriorized at the back of the neck and connected to a swivel apparatus designed to allow free animal movement and computerized blood sampling (Dilab, Sweden). After the surgery, the animals received isotonic NaCl containing 25 U/ mL heparin through the catheter to stabilize the liquid balance of the animal. For the oral administrations, each rat received 800 mL formulation/ kg corresponding to 47 mg 3H-seocalcitol/kg (equal to 40 MBq/kg). For the intravenous administration, each rat received 1000 mL IV formulation/kg corresponding to 8 mg 3H-seocalcitol/kg (equal to 7 MBq/kg). Blood samples of 250 mL were collected at 0 (pre-dose), 0.5, 1, 1.5, 1, 2.5, 3, 5, 7, 9 h following oral administration and at 0 (pre-dose), 0.083, 0.25, 0.42, 0.583, 0.75, 1, 1.5, 2, 4, 6 h following intravenous administration. Serum was prepared and kept at 808C until analysis. Four rats received each formulation. HPLC Analysis of 3H-Seocalcitol in Serum
Preparation of IV 3H-Seocalcitol Formulation to Bioavailability Study The IV formulation was prepared by adding 3Hseocalcitol isopropanol stock solution to an aqueous solution containing 0.016 w/v% citric acid monohydrate, 0.68 w/v% sodium citrate, 8.0 w/v% ethanol, and 41.5 w/v% PG. Only a very small amount of isopropanol was added. The resulting IV solution had a concentration of 3H-seocalcitol of 8 mg/mL and a radioactivity of 7 MBq/mL. Bioavailability Study of 3H-Seocalcitol in Rats All surgical and experimental procedures were reviewed and approved by the local Animal Experimentation Ethics Committee. Male Sprague– Dawley rats weighing 280–320 g (Møllegaard Breeding Center, Lille Skensved, Denmark) maintained on a standard feeding and water ad libitum were included in the study. During the acclimatization, the housing conditions were two rats per cage maintained at 22 28C with a 50% relative humidity, an air change of 15 changes per hour and a 12-h light-dark cycle. During the experiment, the animals were anesthetized for the duration of the surgery by subcutaneous injection of 2.7 mL/kg of a solution consisting of Hypnorm1, Dormicum and water (1:1:2). The day prior to the dosing, the right carotid artery was cannulated with a tygon catheter (0.40 mm ID,
In general, the concentration of 3H-seocalcitol in serum was determined using the HPLC method previously described. However, the reversedphase HPLC-system was added a Packard Radiomatic Flow scintillation Analyzer with a TRLSCcell of 500 mL using Pico-AquaTM (Packard) as scintillator with a flow rate of 2.1 mL/min. The software used to integrate the radio-chromatograms was Flo-One for Windows version 3.61. Before injection, 100 mL serum was precipitated with 300 mL acetonitrile and centrifuged at 48C for 10 min at 4500 rpm (Hettich Rotanta 96R). The supernatant was then evaporated under a stream of nitrogen at 408C for 25 min (Zymark) and the residue reconstituted in 200 mL 50% methanol. The quantification of 3H-seocalcitol was performed using a calibration curve of spiked blank serum, where the lower limit of quantification was 0.5 ng/mL. Preparation of Samples for Liquid Scintillation Counting Duplicates of serum samples (10 mL) were mixed with 6 mL of Pico-FlourTM 40 (Packard) in glass scintillation vials and left overnight before being measured for 5 min by liquid scintillation counting (LSC) together with representative blank samples, using Packard Tri-Carb 2900 TR with automatic quench correction by an external method. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1834
GROVE ET AL.
Pharmacokinetic Analysis
followed by pairwise comparisons adjusted for multiplicity by the Tukey method. Regarding the response criterion bioavailability, a two-way analysis of variance was applied in which a correlation structure over time was incorporated for each rat. Again, the Tukey method was applied to adjust for multiplicity in the pairwise comparisons. In general, a 5% level of significance was used. The results are presented as means of their standard deviations (SD).
Pharmacokinetic parameters were determined using WinNonLin Professional version 3.3 (Pharsight Corp., Mountain View, CA). Maximum serum concentrations (Cmax) and time (tmax) at which Cmax occurred were found by inspection of the 3H-seocalcitol serum concentration–time data from individual animals. lz was calculated by logarithmic-linear regression of the serum concentration–time curve and serum half-life (t½) was calculated as ln 2/lz. The area under the curve (AUC) was calculated by non-compartmental analysis to the final measurable sample (AUC0 ! t) using the linear-log trapezoidal method. The linear trapezoidal rule was used up to Cmax, and the logarithmic trapezoidal rule was used after Cmax. The logarithmic trapezoidal rule was used as the serum concentration after Cmax fell exponentially. The oral bioavailability of seocalcitol was calculated as the percentage ratio of AUCp.o. and AUCi.v. with correction of the actual dose of 3H-seocalcitol (BioAUC). Using the total amount of radioactivity in serum (parent compound, metabolites, and degradation products), the oral absorption of seocalcitol was calculated as the percentage ratio of the AUCp.o. divided by AUCi.v. with correction of the actual dose of 3H-seocalcitol (AbsAUC).
RESULTS Bioavailability of 3H-Seocalcitol in Rats The pharmacokinetic parameters obtained after oral administration of 3H-seocalcitol from the three formulations as well as the AUC for the IV formulation are listed in Table 3. The serum concentration time profiles for the oral formulations are shown in Figure 2 and Figure 3 represents the IV profiles. Figure 2 clearly shows the difference in Cmax ( p < 0.05) and AUC ( p < 0.05) between the lipid formulations and the reference formulation containing PG. For all three formulations tmax and t½ were equal. The bioavailability of 3 H-seocalcitol from the two lipid formulations compared with the reference formulation exhibited a 2.2- to 2.4-fold increase in bioavailability ( p < 0.05) and no statistical difference in bioavailability, tmax, t½, and Cmax between the two lipid formulations was seen.
Statistical Analysis Two models were used to describe the four response criterions and hereby testing the influence of the regarded parameters. In the matter of Cmax, tmax, and AUC the data was logarithmically transformed in order to normalize variations. Hereby, a one-way analysis of variance could be used
Absorption of 3H-Seocalcitol in Rats Table 3 also presents the percentage of radio labeled substances absorbed, e.g., 3H-seocalcitol,
Table 3. Pharmacokinetic Parameters (Mean SD) Following Single Oral Administration of 47 mg/kg of 3 H-Seocalcitol (Equal to 40 MBq/kg) Formulated as a Reference Solution, MCT Solution, and LCT Solution to Rats (n ¼ 4) Pharmacokinetic Parameter tmax (h) Cmax (ng/mL) t½ (h) BioAUC,0 ! 9 (ng/mL h1) Bioavailability0 ! 9 (%) AbsAUC,0 ! 9 (ng-eqvi/mL h1) Absorption0 ! 9 (%)
IV
Reference Solution
MCT Solution
LCT Solution
— — 2.6 0.3 13 2 100 20 2 100
1.6 0.5 21 4.1 2.4 85 10 5 36 13 29 10
1.5 0.4 5 1a 2.6 0.5 18 3a 24 4a 57 9 47 7
1.6 0.3 4 1a 2.7 0.5 17 5a 22 6a 55 3 45 2
An IV formulation of 8 mg/kg (equal to 7 MBq/kg), (n ¼ 3), was dosed in order to calculate the absolute bioavailability and the absorption. a Statistical significantly different when compared to the reference formulation ( p < 0.05). JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
SEOCALCITOL I: SOLUBILITY AND BIOAVAILABILITY
Figure 2. Serum concentration-time profiles (mean values, n ¼ 4) following oral administration of 47 mg/kg of 3 H-seocalcitol (equal to 40 MBq/kg) to male rats. Three formulations were tested: Reference solution (), MCT solution (&), and LCT solution (&). Cmax and AUC0 ! 9 for the reference formulation is significantly different when compared to the lipid formulations.
metabolites, and degradation products calculated from the total amount of radioactivity in the serum samples. The amount of radio labeled substances absorbed from the two lipid formulations compared with the reference formulation was increased 1.5- to 1.6-fold. Half of the dosed amount is absorbed from the lipid formulations. Solubility of Seocalcitol in Simulated Intestinal Media
1835
the concentration of LP with constant concentration of BS and PL on the solubility of seocalcitol. In order to investigate the influence of the chain length of MAG and FA on the solubility of seocalcitol both MC-LP and LC-LP were used. pH, buffer concentration, and osmolality were constant in all media. The concentration of BS in the rat is between 8 and 25 mM,15 being highest during night when the animals are active.16 The concentration levels of BS used in this solubility study cover the level observed in the rat. Figure 4 shows the concentration of solubilized seocalcitol in the SIM after 24 h with increasing concentration of BS and PL. Figure 4 also shows the effect of adding 5 mM FA and 2.5 mM MAG to the BS and PL media. The solubility of seocalcitol increases with increasing concentration of BS and PL and for low concentrations of BS and PL addition of LP further increases the solubility of seocalcitol. For media containing 20 mM BS and 5 mM PL the solubility was not increased further by addition of LP. There is no difference in the solubility of seocalcitol between adding MC-LP or LC-LP to the media. The results obtained after 24 and 48 h were equal. Figure 5 shows that the solubility of seocalcitol in a medium with a concentration of 20 mM BS and 5 mM PL is not increased by adding either MC-LP or LC-LP. The slopes of the graphs are relatively small and close to 0 indicating that the effect of adding LP to a medium containing high concentrations of BS and PL is limited.
The SIM used in the present study were developed to investigate the influence of increasing the concentration of BS and PL as well as increasing
Figure 3. Serum concentration-time profile (mean value, n ¼ 3) following IV administration of 8 mg/kg of 3 H-seocalcitol (equal to 7 MBq/kg) to male rats.
Figure 4. The concentration of seocalcitol as a function of BS concentration (n ¼ 3). The influence on solubility of seocalcitol in absence and presence of 5 mM FA and 2.5 mM MAG (MC-LP or LC-LP) to the media containing increasing concentration of BS (n ¼ 3). Key: () BS and PL; (&) BS, PL, MCFA, and MCMG; and (&) BS, PL, LCFA and LCMG. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1836
GROVE ET AL.
solubilising a larger amount of 3H-seocalcitol minimizing or eliminating precipitation of 3Hseocalcitol in the intestine resulting in an improved bioavailability compared to the simple BS micelles being present after dosing the reference solution.
Solubility of Seocalcitol in Simulated Intestinal Media
Figure 5. The concentration of seocalcitol as a function of concentration of FA (n ¼ 3). The graphs show the effect of adding increasing concentration of either MCLP or LC-LP to a medium containing 20 mM BS and 5 mM PL (n ¼ 3). Key: (&) BS, PL, MCFA, and MCMG; and (&) BS, PL, LCFA, and LCMG.
DISCUSSION Bioavailability of 3H-Seocalcitol in Rats In the present study, the drug substance is dissolved in all the tested oral formulations, which makes it possible to discuss the mechanisms being responsible for keeping the drug substance solubilized in the GIT after entering the intestine and until absorption. There is no difference in the bioavailability of seocalcitol when comparing the two different lipid formulations (LCT and MCT). The lipid formulations had a positive effect on the solubilisation process resulting in a twofold higher bioavailability of 3H-seocalcitol compared with the reference formulation containing PG (Table 3). The dispersion and hydrolysis of the lipid formulations and trafficking of 3H-seocalcitol from the formulations to the mixed micelles prior to absorption may account for the difference in bioavailability between the lipid formulations and the reference formulation. As for the reference formulation no mixed micelles will be present, since no lipid is present in this formulation, thus absorption from the reference solution solely depends on transfer into simple BS micelles to be presented to the epithelial membrane in solution. Simple BS micelles will be present in the intestine in rats at all times,16 since rats have a continuous secretion of BS and PL17 whereas the formation of mixed micelles will only take place in the presence of LP. The in vivo data illustrates that the mixed micelles are capable of JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
Addition of either MC-LP or LC-LP to the SIM increases the solubility in the media representing the fasted state (5 mM BS and 1.25 mM PL). The same increase in the solubility is not observed in the media representing the fed state (20 mM and 5 mM PL), where the same solubility of seocalcitol is seen in absence and in presence of LP (Figure 4). By adding LP to the media representing the fasted state the observed difference in solubility between the fasted and the fed state is reduced, indicating that a formulation containing lipids will be beneficial and may reduce a possible food effect. The obtained solubilities with LC-LP or MC-LP did not differ, thus incorporation of MAG and FA with different chain length and hydrophobicity into the micelles does not affect the solubilisation of seocalcitol (Figure 5). This is in spite of a higher solubility of seocalcitol in the MCT compared with the LCT. The solubility data confirms the in vivo data by illustrating that LP increases the solubility capacity of seocalcitol in the fasted state. This observation makes the use of solubility data in formulation design interesting since it suggest the beneficial effect of developing a lipid based formulation for seocalcitol and also predicts that either lipid (MCT or LCT) can be used. As seen for other poorly soluble drugs,8,9,18,19 BS and PL increase the solubility of seocalcitol significantly, compared with the solubility in water. The solubility of seocalcitol in SIM (20 mM BS and 5 mM PL) increases more than 8000 times compared with the solubility in water (168 mg/mL vs. 20 ng/mL). In the range from 5 to 20 mM BS the increase is linear (R2 ¼ 0.9831). Sunesen et al. found a linear relationship between BS level and danazol solubility between 0 and 18.8 mM BS.19 Danazol has a log P value and a triglyceride solubility very similar to those of seocalcitol,20 indicating that the effect of BS and PL may depend on other factors than the log P value and the triglyceride solubility like e.g. molecular structure and surface area.
SEOCALCITOL I: SOLUBILITY AND BIOAVAILABILITY
Comparison of Solubility Data of Seocalcitol with Bioavailability For seocalcitol it seems to be possible to relate the in vitro solubility in the biorelevant media to the in vivo bioavailability study in rats. Lipid formulations that will generate mixed micelles in the GIT give a higher bioavailability of seocalcitol. This is related to the higher solubility of seocalcitol in media containing LP and simulating the fasted state (5 mM BS and 1.25 mM PL). Further the existence of a relationship between the solubility in SIM and the obtained bioavailablity from the corresponding lipid formulations is indicated, as there is no difference between the use of MCT and LCT in neither the solubility nor the bioavailability study. For halofantrine, a similar relation seems to be possible since the best solubility is seen in SIM with LC-LP9 as well as the best bioavailability is seen from LCT compared to MCT.21 The difference in lipophilicity between seocalcitol (log P 4.8) and halofantrine (log P 9) may account for the difference seen in solubility in SIM and bioavailability between MCT and LCT. The solubility of halofantrine is two times higher in MCT compared with LCT (89.0 vs. 47.3 mg/g), but when calculated as mmol/mol an equal solubility is seen (89.7 vs. 82.4 mmol/mol).20 The solubility of seocalcitol is three times higher in MCT than in LCT (Table 1) when measured as mg/g. When calculated in molar terms this factor decreases to 1.8, as the solubility of seocalcitol is 5.9 mmol/mol MCT versus 3.4 mmol/mol LCT. Thus the solubility in triglyceride cannot be used to predict in vivo performance. This emphasizes that the solubility that is important for prediction of in vivo performance may not be the solubility in the formulation per se, but the solubility in the GIT fluids, in this case reflected in the biorelevant media. A relation between solubility in SIM and bioavailability may not be surprising as the mixed micelles are present in both the SIM containing LP and in the intestine after hydrolysis of triglycerides. In this particular study, the solubility in SIM turned out to be very useful for the formulation scientist, as it indicated that lipid can be used to obtain an improved bioavailability in the fasted state, compared with the reference formulation. Furthermore, the solubility study and the bioavailability study showed that either lipid could be used since the solubility of seocalcitol in both MCLP and LC-LP containing media as well as the bioavailability from MCT and LCT solutions were
1837
similar. To elucidate whether it is a general trend in the field of lipid formulations, that the bioavailablity is better related to solubility in SIM containing LP than to the solubility in non-hydrolyzed lipids, needs to be investigated in studies using other drug substances.
ACKNOWLEDGMENTS The authors thank ATV (The Danish Academy of Technical Sciences) for financial support. The radio labeled compound was kindly synthesized by Gunnar Grue-Sørensen, LEO Pharma. Rie Andreassen, LEO Pharma, is thanked for skillful help in animal operation and dosing. Jonas Wiedemann, LEO Pharma, is thanked for help with the statistical analysis.
REFERENCES 1. Amidon GL, Lennernas H, Shah VP, Crison JR. 1995. A theoretical basis for a biopharmaceutic drug classification: The correlation of in vitro drug product dissolution and in vivo bioavailability. Pharm Res 12:413–420. 2. Gershanik T, Benita S. 2000. Self-dispersing lipid formulations for improving oral absorption of lipophilic drugs. Eur J Pharm Biopharm 50:179–188. 3. Humberstone AJ, Charman WN. 1997. Lipid-based vehicles for the oral delivery of poorly water soluble drugs. Adv Drug Deliv Rev 25:103–128. 4. Yamahira Y, Noguchi T, Takenaka H, Maeda T. 1979. Biopharmaceutical studies of lipid-containing oral dosage forms: Relationship between drug absorption rate and digestibility of vehicles. Int J Pharm 3:23–31. 5. Fouad FM, Farrell PG, Marshall WD, van de Voort FR. 1991. In vitro model for lipase-catalysed lipophile release from fats. J Agric Food Chem 39:150– 153. 6. Khoo SM, Shackleford DM, Porter CJ, Edwards GA, Charman WN. 2003. Intestinal lymphatic transport of halofantrine occurs after oral administration of a unit-dose lipid-based formulation to fasted dogs. Pharm Res 20:1460–1465. 7. MacGregor KJ, Embleton JK, Lacy JE, Perry EA, Solomon LJ, Seager H, Pouton CW. 1997. Influence of lipolysis on drug absorption from the gastrointestinal tract. Adv Drug Deliv Rev 25:33–46. 8. Nielsen PB, Mullertz A, Norling T, Kristensen HG. 2001. The effect of alpha-tocopherol on the in vitro solubilisation of lipophilic drugs. Int J Pharm 222: 217–224. 9. Kossena GA, Boyd BJ, Porter CJ, Charman WN. 2003. Separation and characterization of the colJOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1838
10.
11.
12.
13.
14.
15.
GROVE ET AL.
loidal phases produced on digestion of common formulation lipids and assessment of their impact on the apparent solubility of selected poorly watersoluble drugs. J Pharm Sci 92:634–648. Behrens D, Fricker R, Bodoky A, Drewe J, Harder F, Heberer M. 1996. Comparison of cyclosporin A absorption from LCT and MCT solutions following intrajejunal administration in conscious dogs. J Pharm Sci 85:666–668. Myers RA, Stella VJ. 1992. Systemic bioavailability of penclomedine (NSC-338720) from oil-in-water emulsions administered intraduodenally to rats. Int J Pharm 78:217–226. Palin KJ, Winson CG. 1984. The effect of different oils on the absorption of probucol in the rat. J Pharm Pharmacol 36:641–643. Holmberg I, Aksnes L, Berlin T, Lindback B, Zemgals J, Lindeke B. 1990. Absorption of a pharmacological dose of vitamin D3 from two different lipid vehicles in man: Comparison of peanut oil and a medium chain triglyceride. Biopharm Drug Dispos 11:807–815. Gallo-Torres H, Ludorf J, Brin M. 1978. The effect of medium-chain triglycerides on the bioavailability of vitamin E. Int J Vit Nutr Res 48:240– 249. Erhlinger S. 1987. Physiology of bile secretion and enterohepatic circulation. In: Johnson LR, editor.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
16. 17.
18.
19.
20.
21.
Physiology of the Gastrointestinal Tract. 2nd edn. New York: Raven Press. pp. 1557. Kararli TT. 1989. Gastrointestinal absorption of drugs. Crit Rev Ther Drug Carrier Syst 6:39–86. DeSesso JM, Jacobson CF. 2001. Anatomical and physiological parameters affecting gastrointestinal absorption in humans and rats. Food Chem Toxicol 39:209–228. Pedersen BL, Mullertz A, Brondsted H, Kristensen HG. 2000. A comparison of the solubility of danazol in human and simulated gastrointestinal fluids. Pharm Res 17:891–894. Sunesen VH, Pedersen BL, Kristensen HG, Mullertz A. 2005. In vivo in vitro correlations for a poorly soluble drug, danazol, using the flowthrough dissolution method with biorelevant dissolution media. Eur J Pharm Sci 24:305–313. Kaukonen AM, Boyd BJ, Porter CJ, Charman WN. 2004. Drug solubilization behavior during in vitro digestion of simple triglyceride lipid solution formulations. Pharm Res 21:245–253. Porter CJ, Kaukonen AM, Taillardat-Bertschinger A, Boyd BJ, O’Connor JM, Edwards GA, Charman WN. 2004. Use of in vitro lipid digestion data to explain the in vivo performance of triglyceridebased oral lipid formulations of poorly watersoluble drugs: Studies with halofantrine. J Pharm Sci 93:1110–1121.
Pharmaceutical Formulation, LEO Pharma A/S, Industriparken 55, DK-2750 Ballerup, Denmark
2
Department of Pharmacokinetics, LEO Pharma A/S, Industriparken 55, DK-2750 Ballerup, Denmark
3
Department of Pharmaceutics, The Danish University of Pharmaceutical Sciences, Universitetsparken 2, DK-2100 Copenhagen, Denmark
Received 8 November 2004; revised 8 March 2005; accepted 21 April 2005 Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.20403
ABSTRACT: Simulated intestinal media (SIM) containing bile salt (BS) and phospholipids (PL) with and without medium chain lipolytic products (MC-LP) or long chain lipolytic products (LC-LP) were developed to study the solubility of seocalcitol. Both MC-LP and LC-LP were studied in order to investigate the influence of fatty acid chain length on the in vitro solubility of seocalcitol. The same solubility of seocalcitol was found in media containing either MC-LP or LC-LP. The bioavailability after oral administration of seocalcitol dissolved in medium chain triglyceride (MCT), long chain triglyceride (LCT), and a reference formulation containing propylene glycol (PG) was studied in vivo in rats. The lipid formulations showed a twofold increase in bioavailability compared with the reference formulation, indicating positive effects of lipids on the bioavailability reflecting a better solubility in the intestine and protection against precipitation of seocalcitol in the gastro intestinal tract. There was no difference in the in vivo bioavailability of seocalcitol between the MCT and the LCT solutions, which correlates with the identical in vitro solubility of seocalcitol in SIM containing MC-LP or LC-LP. ß 2005 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 94:1830–1838, 2005
Keywords: oral bioavailability; lipid-based formulations; oral absorption; poorly soluble drug substances; solubility; simulated intestinal media; seocalcitol
INTRODUCTION Drug substances with low water solubility and high permeability are classified as Class 2 drug substances in the Biopharmaceutical Classification System (BCS).1 In the gastrointestinal tract (GIT) the limiting step in drug absorption for class 2 drug substances is the dissolution process. In order to overcome the dissolution process, poorly soluble drug substances can be dissolved in the
Correspondence to: A. Mu¨llertz (Telephone: þ45 35306440; Fax: þ45 35306030.; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 94, 1830–1838 (2005) ß 2005 Wiley-Liss, Inc. and the American Pharmacists Association
1830
formulation before entering the GIT, e.g., in propylene glycol (PG) solution or lipid vehicles. The advantage of applying a lipid vehicle in comparison to a simple water miscible system, is that the formed lipid degradation products may increase the drug absorption via enhanced solubilisation and dispersion of the drug substance in the intestine.2,3 During digestion of the lipid formulation by pancreas lipase, the drug substance is transferred from the lipid formulation into the intestinal environment, where it is incorporated into mixed micelles, containing bile salts (BS), phospholipids (PL), and lipolytic products (LP).4,5 Even small amounts of lipid are believed to induce the secretion of bile to facilitate these mechanisms.6 For poorly soluble drug substances the
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
SEOCALCITOL I: SOLUBILITY AND BIOAVAILABILITY
critical step in drug absorption is to avoid precipitation when transferred from the lipid formulation to the mixed micelles in the GIT. The capacity of the mixed micelles to keep the drug substance in solution depends on the characteristics of the drug substance, e.g. the lipophilicity, and on the composition of the mixed micelles. Solubility studies with simulated intestinal media (SIM) containing medium chain lipolytic products (MC-LP) or long chain lipolytic products (LC-LP) may give valuable information about what kind of lipids that most efficiently keep the drug substance in solution.7–9 It has been shown that drug substances with low log P (approximately 2) seem to have a better solubility in media containing MC-LP, whereas the solubility of drug substances with high log P (8–9) is favored by adding LC-LP to the media.8,9 For felodipine, having a log P of 4.8, the best solubilising capacity was found in SIM containing LC-LP8 whereas for danazol, having a log P of 4.5, the best solubility capacity was found in SIM containing MC-LP.9 Thus, as to drug substances with log P values around 4–5 no clear trend is seen regarding the type of lipid that will cause the highest increase in the solubility in biorelevant media why solubility studies in SIM with LP can be used as a tool to gain information about what kind of lipid to use. A strategy for a lipid based formulation of a given poorly soluble drug will include solubility studies in the excipients to ensure that the drug is in solution in the formulation. However, equally important is that the drug will be kept in solution during passage through the GIT, during digestion of the formulation and transfer between different formed colloid phases, until absorption from the mixed micelles. This makes solubility studies in SIM, simulating the composition of the GIT fluids during digestion of a given formulation, a possible and interesting formulation tool. Frequently used formulation vehicles in lipidbased formulations of poorly soluble drugs are medium chain triglycerides (MCT) and long chain triglycerides (LCT). Only a few studies have been performed comparing the bioavailability of drug substances dissolved in MCT or LCT, respectively.10–14 For halofantrine, cyclosporine A, probucol, and vitamin D3, the best bioavailability was obtained from LCT,10,12,13 whereas the best bioavailability for penclomedine and vitamin E was seen from MCT.11,14 Comparing these data, there seem to be no clear trend as to what characteristics
1831
of a compound that determines whether LCT or MCT will give the best bioavailability. The objectives of the present work were to: (a) compare the in vivo oral bioavailability of a model drug substance from pure oily solutions, MCT and LCT respectively, to a reference PG solution, (b) investigate the influence of SIM on the in vitro solubilisation of the model drug substance, including SIM containing either MC-LP or LC-LP in different concentrations, and (c) evaluate if in vitro solubility data can be related to in vivo behavior of the present model drug substance from formulations containing either MCT or LCT. Seocalcitol, a D-vitamin analogue and a class 2 drug, was chosen as a model drug in this study. Selected physicochemical properties of seocalcitol are shown in Table 1. The structure of seocalcitol is shown in Figure 1.
MATERIALS AND METHODS Materials Seocalcitol and tritium labeled seocalcitol (3Hseocalcitol, specific activity of 0.844 MBq/mg and radiochemical purity of 97%) were synthesized at LEO Pharma. Citric acid monohydrate and Sodium Chloride were purchased from Merck, Germany, Sodium Citrate from Jungbunzlauer GmbH, Germany, PG from Lyondell Chem. Comp., France, Viscoleo (MCT) from ICI Espana, Spain, and Sesame oil (LCT) was purchased from Henry Lamotte GmbH, Germany. The fatty acid composition of the triglycerides in MCT and LCT is as specified in Ph.Eur. Sodium taurocholate (97%) (NaTC), Trisma maleate (99%), Oleic acid (approx. 95%), Decanoic acid (99%–100%), Octanoic acid (99%), and Sodium Azide (99%) were all Table 1. Physicochemical Properties of Seocalcitol Physico-Chemical Parameter Log P Pappa (cm/s) Swaterb (ng/g) SMCTc (mg/g) SLCTd (mg/g)
4.8 6.35 106 20 5.3 1.7
a Apparent permeability coefficient (Papp), investigated in the CaCo-2 cell line at 378C. b Solubility in Milli-Q water (Swater) at 258C, n ¼ 3. c Solubility in Viscoleo (fractionated coconut oil) (SMCT) at 258C, n ¼ 2. d Solubility in sesame oil (SLCT) at 258C, n ¼ 2. Data from internal report, LEO Pharma A/S. Seocalcitol concentration determined by HPLC.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1832
GROVE ET AL.
Figure 1. Chemical structure of seocalcitol and 3 H-seocalcitol.
purchased from Sigma-Aldrich (St. Louis, MO). Lipoid S100 was purchased from Lipoid GMBH. RyloTM MG 19 Pharma (Glycerol Monooleate) and Emulsifier TS-PH 003 (Glycerol Monocaprate) were gifts from Danisco, Denmark. Hypnorm1 (fentanyl 0.2 mg/mL, fluanisone 10 mg/mL) was purchased from Janssen, Belgium, and Dormicum (Midazolam 5 mg/mL) from Roche, Switzerland. Heparin 10,000 IU/mL was produced at LEO Pharma A/S. Isotonic sterile NaCl was purchased at Dilab, Sweden. Water was obtained from a Milli-Q-water purification system (Millipore, MA). Pico-AquaTM and Pico Flour 40TM were purchased from Packard. All other chemicals were of analytical grade.
dissolved, the pH was adjusted to 6.5. The composition of the SIM used in this study is shown in Table 2. Oleic acid was chosen as a long chain fatty acid (LCFA) and glycerol monooleate as a long chain monoacyl glycerol (LCMG). Octanoic acid and decanoic acid in the ratio 60:40 were chosen as medium chain fatty acids (MCFA) and glycerol monocaprate as a medium chain monoacyl glycerol (MCMG). Solubility Studies of Seocalcitol in Simulated Intestinal Media The media were pre-heated to 378C and 10 mL were added to screw cap vials containing 30 mg (excess) of seocalcitol powder (n ¼ 3). The vials were gently rotated at 100 rpm in a heating chamber (Lab-Therm, Ku¨hner, Switzerland) at 378C. Two aliquots of 1 mL were removed after 24 and 48 h, respectively and centrifuged at 12000 rpm for 10 min (Hemle High Speed Centrifuge 2-200-M). 500 mL of the supernatant was mixed with 500 mL ethanol and 1 mL 50% methanol (v/v) was added. The solutions were mixed for 10 s and analyzed by HPLC.
Composition of Simulated Intestinal Media
HPLC Analysis of Seocalcitol in Simulated Intestinal Media
Trisma maleate was used as buffer in a concentration of 100 mM. NaN3 (3 mM) was added to prevent microbial growth. NaCl was used to give the same osmolality of 270 mosmol/L in the total media. For each media, all the excipients were accurately weighed into beakers, distilled water was added, and the solutions were heated to 508C while stirring. When all excipients were perfectly
The concentration of seocalcitol in simulated intestinal media was determined by reversed-phase HPLC consisting of a Waters Alliance Separation Module 2790 with cooling on the autosampler (88C), online vacuum degassing, column oven (408C), a Dual l Absorbance Detector 2487 set at 264 nm and data were processed by using Millennium Chromatography Manager version 32 (all
Table 2. The Composition of SIM. Horizontally: The Concentration of BS and PL. Vertically: The Concentration of FA and MAG
— 5 mM LCFA, 2.5 mM LCMG 10 mM LCFA, 5 mM LCMG 15 mM LCFA, 7.5 mM LCMG 20 mM LCFA, 10 mM LCMG 5 mM MCFA, 2.5 mM MCMG 10 mM MCFA, 5 mM MCMG 15 mM MCFA, 7.5 mM MCMG 20 mM MCFA, 10 mM MCMG
BS: 5.0 mM, PL: 1.25 mM
BS: 10 mM, PL: 2.5 mM
BS: 15 mM, PL: 3.75 mM
BS: 20 mM, PL: 5 mM
The marks in the table represent the media used in the solubility study. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
SEOCALCITOL I: SOLUBILITY AND BIOAVAILABILITY
Waters Associates, Milford, MA). The analytical column was a Symmetri C8, 50 2.1 mm i.d. 3.5 mm (Waters) and the injection volume was 150 mL. The mobile phase was a linear gradient from 70 to 99% methanol using two eluents: (A) methanol:water (20:80 v/v), (B) methanol:water (99:1 v/v), and the flow rate was 0.3 mL/ min. The retention time of seocalcitol was approximately 4 min and quantification was performed using a calibration curve of spiked blank media, the lower limit of quantification being 100 ng/mL. Preparation of Oral 3H-Seocalcitol Formulations to Bioavailability Study Two lipid based formulations, one MCT solution (viscoleo) and one LCT solution (sesame oil), and an oral reference solution (PG) were developed by dissolving the 3H-seocalcitol in the formulation, resulting in a concentration of seocalcitol of 59 mg/ mL and 50 MBq/mL.
1833
0.79 mm OD, Dilab, Sweden). The catheter was exteriorized at the back of the neck and connected to a swivel apparatus designed to allow free animal movement and computerized blood sampling (Dilab, Sweden). After the surgery, the animals received isotonic NaCl containing 25 U/ mL heparin through the catheter to stabilize the liquid balance of the animal. For the oral administrations, each rat received 800 mL formulation/ kg corresponding to 47 mg 3H-seocalcitol/kg (equal to 40 MBq/kg). For the intravenous administration, each rat received 1000 mL IV formulation/kg corresponding to 8 mg 3H-seocalcitol/kg (equal to 7 MBq/kg). Blood samples of 250 mL were collected at 0 (pre-dose), 0.5, 1, 1.5, 1, 2.5, 3, 5, 7, 9 h following oral administration and at 0 (pre-dose), 0.083, 0.25, 0.42, 0.583, 0.75, 1, 1.5, 2, 4, 6 h following intravenous administration. Serum was prepared and kept at 808C until analysis. Four rats received each formulation. HPLC Analysis of 3H-Seocalcitol in Serum
Preparation of IV 3H-Seocalcitol Formulation to Bioavailability Study The IV formulation was prepared by adding 3Hseocalcitol isopropanol stock solution to an aqueous solution containing 0.016 w/v% citric acid monohydrate, 0.68 w/v% sodium citrate, 8.0 w/v% ethanol, and 41.5 w/v% PG. Only a very small amount of isopropanol was added. The resulting IV solution had a concentration of 3H-seocalcitol of 8 mg/mL and a radioactivity of 7 MBq/mL. Bioavailability Study of 3H-Seocalcitol in Rats All surgical and experimental procedures were reviewed and approved by the local Animal Experimentation Ethics Committee. Male Sprague– Dawley rats weighing 280–320 g (Møllegaard Breeding Center, Lille Skensved, Denmark) maintained on a standard feeding and water ad libitum were included in the study. During the acclimatization, the housing conditions were two rats per cage maintained at 22 28C with a 50% relative humidity, an air change of 15 changes per hour and a 12-h light-dark cycle. During the experiment, the animals were anesthetized for the duration of the surgery by subcutaneous injection of 2.7 mL/kg of a solution consisting of Hypnorm1, Dormicum and water (1:1:2). The day prior to the dosing, the right carotid artery was cannulated with a tygon catheter (0.40 mm ID,
In general, the concentration of 3H-seocalcitol in serum was determined using the HPLC method previously described. However, the reversedphase HPLC-system was added a Packard Radiomatic Flow scintillation Analyzer with a TRLSCcell of 500 mL using Pico-AquaTM (Packard) as scintillator with a flow rate of 2.1 mL/min. The software used to integrate the radio-chromatograms was Flo-One for Windows version 3.61. Before injection, 100 mL serum was precipitated with 300 mL acetonitrile and centrifuged at 48C for 10 min at 4500 rpm (Hettich Rotanta 96R). The supernatant was then evaporated under a stream of nitrogen at 408C for 25 min (Zymark) and the residue reconstituted in 200 mL 50% methanol. The quantification of 3H-seocalcitol was performed using a calibration curve of spiked blank serum, where the lower limit of quantification was 0.5 ng/mL. Preparation of Samples for Liquid Scintillation Counting Duplicates of serum samples (10 mL) were mixed with 6 mL of Pico-FlourTM 40 (Packard) in glass scintillation vials and left overnight before being measured for 5 min by liquid scintillation counting (LSC) together with representative blank samples, using Packard Tri-Carb 2900 TR with automatic quench correction by an external method. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1834
GROVE ET AL.
Pharmacokinetic Analysis
followed by pairwise comparisons adjusted for multiplicity by the Tukey method. Regarding the response criterion bioavailability, a two-way analysis of variance was applied in which a correlation structure over time was incorporated for each rat. Again, the Tukey method was applied to adjust for multiplicity in the pairwise comparisons. In general, a 5% level of significance was used. The results are presented as means of their standard deviations (SD).
Pharmacokinetic parameters were determined using WinNonLin Professional version 3.3 (Pharsight Corp., Mountain View, CA). Maximum serum concentrations (Cmax) and time (tmax) at which Cmax occurred were found by inspection of the 3H-seocalcitol serum concentration–time data from individual animals. lz was calculated by logarithmic-linear regression of the serum concentration–time curve and serum half-life (t½) was calculated as ln 2/lz. The area under the curve (AUC) was calculated by non-compartmental analysis to the final measurable sample (AUC0 ! t) using the linear-log trapezoidal method. The linear trapezoidal rule was used up to Cmax, and the logarithmic trapezoidal rule was used after Cmax. The logarithmic trapezoidal rule was used as the serum concentration after Cmax fell exponentially. The oral bioavailability of seocalcitol was calculated as the percentage ratio of AUCp.o. and AUCi.v. with correction of the actual dose of 3H-seocalcitol (BioAUC). Using the total amount of radioactivity in serum (parent compound, metabolites, and degradation products), the oral absorption of seocalcitol was calculated as the percentage ratio of the AUCp.o. divided by AUCi.v. with correction of the actual dose of 3H-seocalcitol (AbsAUC).
RESULTS Bioavailability of 3H-Seocalcitol in Rats The pharmacokinetic parameters obtained after oral administration of 3H-seocalcitol from the three formulations as well as the AUC for the IV formulation are listed in Table 3. The serum concentration time profiles for the oral formulations are shown in Figure 2 and Figure 3 represents the IV profiles. Figure 2 clearly shows the difference in Cmax ( p < 0.05) and AUC ( p < 0.05) between the lipid formulations and the reference formulation containing PG. For all three formulations tmax and t½ were equal. The bioavailability of 3 H-seocalcitol from the two lipid formulations compared with the reference formulation exhibited a 2.2- to 2.4-fold increase in bioavailability ( p < 0.05) and no statistical difference in bioavailability, tmax, t½, and Cmax between the two lipid formulations was seen.
Statistical Analysis Two models were used to describe the four response criterions and hereby testing the influence of the regarded parameters. In the matter of Cmax, tmax, and AUC the data was logarithmically transformed in order to normalize variations. Hereby, a one-way analysis of variance could be used
Absorption of 3H-Seocalcitol in Rats Table 3 also presents the percentage of radio labeled substances absorbed, e.g., 3H-seocalcitol,
Table 3. Pharmacokinetic Parameters (Mean SD) Following Single Oral Administration of 47 mg/kg of 3 H-Seocalcitol (Equal to 40 MBq/kg) Formulated as a Reference Solution, MCT Solution, and LCT Solution to Rats (n ¼ 4) Pharmacokinetic Parameter tmax (h) Cmax (ng/mL) t½ (h) BioAUC,0 ! 9 (ng/mL h1) Bioavailability0 ! 9 (%) AbsAUC,0 ! 9 (ng-eqvi/mL h1) Absorption0 ! 9 (%)
IV
Reference Solution
MCT Solution
LCT Solution
— — 2.6 0.3 13 2 100 20 2 100
1.6 0.5 21 4.1 2.4 85 10 5 36 13 29 10
1.5 0.4 5 1a 2.6 0.5 18 3a 24 4a 57 9 47 7
1.6 0.3 4 1a 2.7 0.5 17 5a 22 6a 55 3 45 2
An IV formulation of 8 mg/kg (equal to 7 MBq/kg), (n ¼ 3), was dosed in order to calculate the absolute bioavailability and the absorption. a Statistical significantly different when compared to the reference formulation ( p < 0.05). JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
SEOCALCITOL I: SOLUBILITY AND BIOAVAILABILITY
Figure 2. Serum concentration-time profiles (mean values, n ¼ 4) following oral administration of 47 mg/kg of 3 H-seocalcitol (equal to 40 MBq/kg) to male rats. Three formulations were tested: Reference solution (), MCT solution (&), and LCT solution (&). Cmax and AUC0 ! 9 for the reference formulation is significantly different when compared to the lipid formulations.
metabolites, and degradation products calculated from the total amount of radioactivity in the serum samples. The amount of radio labeled substances absorbed from the two lipid formulations compared with the reference formulation was increased 1.5- to 1.6-fold. Half of the dosed amount is absorbed from the lipid formulations. Solubility of Seocalcitol in Simulated Intestinal Media
1835
the concentration of LP with constant concentration of BS and PL on the solubility of seocalcitol. In order to investigate the influence of the chain length of MAG and FA on the solubility of seocalcitol both MC-LP and LC-LP were used. pH, buffer concentration, and osmolality were constant in all media. The concentration of BS in the rat is between 8 and 25 mM,15 being highest during night when the animals are active.16 The concentration levels of BS used in this solubility study cover the level observed in the rat. Figure 4 shows the concentration of solubilized seocalcitol in the SIM after 24 h with increasing concentration of BS and PL. Figure 4 also shows the effect of adding 5 mM FA and 2.5 mM MAG to the BS and PL media. The solubility of seocalcitol increases with increasing concentration of BS and PL and for low concentrations of BS and PL addition of LP further increases the solubility of seocalcitol. For media containing 20 mM BS and 5 mM PL the solubility was not increased further by addition of LP. There is no difference in the solubility of seocalcitol between adding MC-LP or LC-LP to the media. The results obtained after 24 and 48 h were equal. Figure 5 shows that the solubility of seocalcitol in a medium with a concentration of 20 mM BS and 5 mM PL is not increased by adding either MC-LP or LC-LP. The slopes of the graphs are relatively small and close to 0 indicating that the effect of adding LP to a medium containing high concentrations of BS and PL is limited.
The SIM used in the present study were developed to investigate the influence of increasing the concentration of BS and PL as well as increasing
Figure 3. Serum concentration-time profile (mean value, n ¼ 3) following IV administration of 8 mg/kg of 3 H-seocalcitol (equal to 7 MBq/kg) to male rats.
Figure 4. The concentration of seocalcitol as a function of BS concentration (n ¼ 3). The influence on solubility of seocalcitol in absence and presence of 5 mM FA and 2.5 mM MAG (MC-LP or LC-LP) to the media containing increasing concentration of BS (n ¼ 3). Key: () BS and PL; (&) BS, PL, MCFA, and MCMG; and (&) BS, PL, LCFA and LCMG. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1836
GROVE ET AL.
solubilising a larger amount of 3H-seocalcitol minimizing or eliminating precipitation of 3Hseocalcitol in the intestine resulting in an improved bioavailability compared to the simple BS micelles being present after dosing the reference solution.
Solubility of Seocalcitol in Simulated Intestinal Media
Figure 5. The concentration of seocalcitol as a function of concentration of FA (n ¼ 3). The graphs show the effect of adding increasing concentration of either MCLP or LC-LP to a medium containing 20 mM BS and 5 mM PL (n ¼ 3). Key: (&) BS, PL, MCFA, and MCMG; and (&) BS, PL, LCFA, and LCMG.
DISCUSSION Bioavailability of 3H-Seocalcitol in Rats In the present study, the drug substance is dissolved in all the tested oral formulations, which makes it possible to discuss the mechanisms being responsible for keeping the drug substance solubilized in the GIT after entering the intestine and until absorption. There is no difference in the bioavailability of seocalcitol when comparing the two different lipid formulations (LCT and MCT). The lipid formulations had a positive effect on the solubilisation process resulting in a twofold higher bioavailability of 3H-seocalcitol compared with the reference formulation containing PG (Table 3). The dispersion and hydrolysis of the lipid formulations and trafficking of 3H-seocalcitol from the formulations to the mixed micelles prior to absorption may account for the difference in bioavailability between the lipid formulations and the reference formulation. As for the reference formulation no mixed micelles will be present, since no lipid is present in this formulation, thus absorption from the reference solution solely depends on transfer into simple BS micelles to be presented to the epithelial membrane in solution. Simple BS micelles will be present in the intestine in rats at all times,16 since rats have a continuous secretion of BS and PL17 whereas the formation of mixed micelles will only take place in the presence of LP. The in vivo data illustrates that the mixed micelles are capable of JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
Addition of either MC-LP or LC-LP to the SIM increases the solubility in the media representing the fasted state (5 mM BS and 1.25 mM PL). The same increase in the solubility is not observed in the media representing the fed state (20 mM and 5 mM PL), where the same solubility of seocalcitol is seen in absence and in presence of LP (Figure 4). By adding LP to the media representing the fasted state the observed difference in solubility between the fasted and the fed state is reduced, indicating that a formulation containing lipids will be beneficial and may reduce a possible food effect. The obtained solubilities with LC-LP or MC-LP did not differ, thus incorporation of MAG and FA with different chain length and hydrophobicity into the micelles does not affect the solubilisation of seocalcitol (Figure 5). This is in spite of a higher solubility of seocalcitol in the MCT compared with the LCT. The solubility data confirms the in vivo data by illustrating that LP increases the solubility capacity of seocalcitol in the fasted state. This observation makes the use of solubility data in formulation design interesting since it suggest the beneficial effect of developing a lipid based formulation for seocalcitol and also predicts that either lipid (MCT or LCT) can be used. As seen for other poorly soluble drugs,8,9,18,19 BS and PL increase the solubility of seocalcitol significantly, compared with the solubility in water. The solubility of seocalcitol in SIM (20 mM BS and 5 mM PL) increases more than 8000 times compared with the solubility in water (168 mg/mL vs. 20 ng/mL). In the range from 5 to 20 mM BS the increase is linear (R2 ¼ 0.9831). Sunesen et al. found a linear relationship between BS level and danazol solubility between 0 and 18.8 mM BS.19 Danazol has a log P value and a triglyceride solubility very similar to those of seocalcitol,20 indicating that the effect of BS and PL may depend on other factors than the log P value and the triglyceride solubility like e.g. molecular structure and surface area.
SEOCALCITOL I: SOLUBILITY AND BIOAVAILABILITY
Comparison of Solubility Data of Seocalcitol with Bioavailability For seocalcitol it seems to be possible to relate the in vitro solubility in the biorelevant media to the in vivo bioavailability study in rats. Lipid formulations that will generate mixed micelles in the GIT give a higher bioavailability of seocalcitol. This is related to the higher solubility of seocalcitol in media containing LP and simulating the fasted state (5 mM BS and 1.25 mM PL). Further the existence of a relationship between the solubility in SIM and the obtained bioavailablity from the corresponding lipid formulations is indicated, as there is no difference between the use of MCT and LCT in neither the solubility nor the bioavailability study. For halofantrine, a similar relation seems to be possible since the best solubility is seen in SIM with LC-LP9 as well as the best bioavailability is seen from LCT compared to MCT.21 The difference in lipophilicity between seocalcitol (log P 4.8) and halofantrine (log P 9) may account for the difference seen in solubility in SIM and bioavailability between MCT and LCT. The solubility of halofantrine is two times higher in MCT compared with LCT (89.0 vs. 47.3 mg/g), but when calculated as mmol/mol an equal solubility is seen (89.7 vs. 82.4 mmol/mol).20 The solubility of seocalcitol is three times higher in MCT than in LCT (Table 1) when measured as mg/g. When calculated in molar terms this factor decreases to 1.8, as the solubility of seocalcitol is 5.9 mmol/mol MCT versus 3.4 mmol/mol LCT. Thus the solubility in triglyceride cannot be used to predict in vivo performance. This emphasizes that the solubility that is important for prediction of in vivo performance may not be the solubility in the formulation per se, but the solubility in the GIT fluids, in this case reflected in the biorelevant media. A relation between solubility in SIM and bioavailability may not be surprising as the mixed micelles are present in both the SIM containing LP and in the intestine after hydrolysis of triglycerides. In this particular study, the solubility in SIM turned out to be very useful for the formulation scientist, as it indicated that lipid can be used to obtain an improved bioavailability in the fasted state, compared with the reference formulation. Furthermore, the solubility study and the bioavailability study showed that either lipid could be used since the solubility of seocalcitol in both MCLP and LC-LP containing media as well as the bioavailability from MCT and LCT solutions were
1837
similar. To elucidate whether it is a general trend in the field of lipid formulations, that the bioavailablity is better related to solubility in SIM containing LP than to the solubility in non-hydrolyzed lipids, needs to be investigated in studies using other drug substances.
ACKNOWLEDGMENTS The authors thank ATV (The Danish Academy of Technical Sciences) for financial support. The radio labeled compound was kindly synthesized by Gunnar Grue-Sørensen, LEO Pharma. Rie Andreassen, LEO Pharma, is thanked for skillful help in animal operation and dosing. Jonas Wiedemann, LEO Pharma, is thanked for help with the statistical analysis.
REFERENCES 1. Amidon GL, Lennernas H, Shah VP, Crison JR. 1995. A theoretical basis for a biopharmaceutic drug classification: The correlation of in vitro drug product dissolution and in vivo bioavailability. Pharm Res 12:413–420. 2. Gershanik T, Benita S. 2000. Self-dispersing lipid formulations for improving oral absorption of lipophilic drugs. Eur J Pharm Biopharm 50:179–188. 3. Humberstone AJ, Charman WN. 1997. Lipid-based vehicles for the oral delivery of poorly water soluble drugs. Adv Drug Deliv Rev 25:103–128. 4. Yamahira Y, Noguchi T, Takenaka H, Maeda T. 1979. Biopharmaceutical studies of lipid-containing oral dosage forms: Relationship between drug absorption rate and digestibility of vehicles. Int J Pharm 3:23–31. 5. Fouad FM, Farrell PG, Marshall WD, van de Voort FR. 1991. In vitro model for lipase-catalysed lipophile release from fats. J Agric Food Chem 39:150– 153. 6. Khoo SM, Shackleford DM, Porter CJ, Edwards GA, Charman WN. 2003. Intestinal lymphatic transport of halofantrine occurs after oral administration of a unit-dose lipid-based formulation to fasted dogs. Pharm Res 20:1460–1465. 7. MacGregor KJ, Embleton JK, Lacy JE, Perry EA, Solomon LJ, Seager H, Pouton CW. 1997. Influence of lipolysis on drug absorption from the gastrointestinal tract. Adv Drug Deliv Rev 25:33–46. 8. Nielsen PB, Mullertz A, Norling T, Kristensen HG. 2001. The effect of alpha-tocopherol on the in vitro solubilisation of lipophilic drugs. Int J Pharm 222: 217–224. 9. Kossena GA, Boyd BJ, Porter CJ, Charman WN. 2003. Separation and characterization of the colJOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1838
10.
11.
12.
13.
14.
15.
GROVE ET AL.
loidal phases produced on digestion of common formulation lipids and assessment of their impact on the apparent solubility of selected poorly watersoluble drugs. J Pharm Sci 92:634–648. Behrens D, Fricker R, Bodoky A, Drewe J, Harder F, Heberer M. 1996. Comparison of cyclosporin A absorption from LCT and MCT solutions following intrajejunal administration in conscious dogs. J Pharm Sci 85:666–668. Myers RA, Stella VJ. 1992. Systemic bioavailability of penclomedine (NSC-338720) from oil-in-water emulsions administered intraduodenally to rats. Int J Pharm 78:217–226. Palin KJ, Winson CG. 1984. The effect of different oils on the absorption of probucol in the rat. J Pharm Pharmacol 36:641–643. Holmberg I, Aksnes L, Berlin T, Lindback B, Zemgals J, Lindeke B. 1990. Absorption of a pharmacological dose of vitamin D3 from two different lipid vehicles in man: Comparison of peanut oil and a medium chain triglyceride. Biopharm Drug Dispos 11:807–815. Gallo-Torres H, Ludorf J, Brin M. 1978. The effect of medium-chain triglycerides on the bioavailability of vitamin E. Int J Vit Nutr Res 48:240– 249. Erhlinger S. 1987. Physiology of bile secretion and enterohepatic circulation. In: Johnson LR, editor.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
16. 17.
18.
19.
20.
21.
Physiology of the Gastrointestinal Tract. 2nd edn. New York: Raven Press. pp. 1557. Kararli TT. 1989. Gastrointestinal absorption of drugs. Crit Rev Ther Drug Carrier Syst 6:39–86. DeSesso JM, Jacobson CF. 2001. Anatomical and physiological parameters affecting gastrointestinal absorption in humans and rats. Food Chem Toxicol 39:209–228. Pedersen BL, Mullertz A, Brondsted H, Kristensen HG. 2000. A comparison of the solubility of danazol in human and simulated gastrointestinal fluids. Pharm Res 17:891–894. Sunesen VH, Pedersen BL, Kristensen HG, Mullertz A. 2005. In vivo in vitro correlations for a poorly soluble drug, danazol, using the flowthrough dissolution method with biorelevant dissolution media. Eur J Pharm Sci 24:305–313. Kaukonen AM, Boyd BJ, Porter CJ, Charman WN. 2004. Drug solubilization behavior during in vitro digestion of simple triglyceride lipid solution formulations. Pharm Res 21:245–253. Porter CJ, Kaukonen AM, Taillardat-Bertschinger A, Boyd BJ, O’Connor JM, Edwards GA, Charman WN. 2004. Use of in vitro lipid digestion data to explain the in vivo performance of triglyceridebased oral lipid formulations of poorly watersoluble drugs: Studies with halofantrine. J Pharm Sci 93:1110–1121.