Oral Drug Delivery of a Poorly Water-Soluble Drug to the Rat Model
3.3 Polymorphs and Amorphous Forms
3.3.4 Solubility and Bioavailability Enhancement
the least-soluble form (Singhal and Curatolo 2004 ) . However, in some cases the advantages of enhanced solubility, dissolution rates, and bioavailability outweigh potential disadvantages and a metastable form is developed.
In addition to relative stability, Gibbs free energy values can be utilized to estimate relative solubility between solid forms (Parks et al. 1928, 1934 ). This is accomplished by recognizing that free energy is related to the activity of a com-pound through the following defi nitions:
GI =RT·ln aI, (3.4) GII =RT·lnaII, (3.5) where aI and aII are activities of the respective forms. Activities are a refl ection of
“escaping tendencies” and are thus proportional to solubility, s (Gupta et al. 2004 ) . By substituting the solubility terms and subtracting ( 3.4 ) from ( 3.5 ), we can obtain:
·ln .
II II
I
I
G RT æs ö
D = çèσ ÷ø (3.6)
Hence, the solubility ratio between two forms is shown to be proportional to Gibbs free energy.
Experimental solubilities were found to be signifi cantly less than the predicted values, as shown in Table 3.6 . However, all higher-energy polymorphic forms pro-vided some degree of solubility enhancement. It is interesting to note that solubility of the amorphous forms were all noted to be much greater than those predicted or measured for the corresponding crystalline forms at all temperatures studied. The large discrepancy between predicted and experimental solubility values for the amorphous substances was attributed to a strong driving force for recrystallization in the dissolution media, as illustrated in Fig. 3.8 .
A comprehensive literature review on the actual solubility ratio between poly-morphs was conducted by Pudipeddi and Serajuddin ( 2005 ) . In total, the authors evaluated 55 compounds which resulted in 81 solubility ratios due to the existence of multiple polymorphic forms. Overall, the average solubility ratio for the poly-morphs evaluated was 1.7 (excluding the premafolxacin outlier). The solubility ratios from the literature were in agreement with and even included the data pre-sented by Hancock and Parks ( 2000 ) . These values are summarized in Fig. 3.9 . Additionally, the authors evaluated 23 anhydrate/hydrate solubility ratios and found that, in general, the values were less than about two. However, there were cases in which the solubility ratio was signifi cantly higher than this value. The anhydrate/
hydrate values are summarized in Fig. 3.10 .
Although relative improvements in solubility are modest between different poly-morphs or pseudopolypoly-morphs, for poorly water-soluble drugs that exhibit rate-limiting absorption, this difference may provide a signifi cant increase in therapeutic activity.
Table 3.5 Predicted solubility ratios for drug compounds
Compound Forms Solubility ratio a Comment
Indomethacin a -crystal/ g -crystal 1.1–1.2 45°C
amorphous/ g -crystal 38–301 5°C
25–104 25°C
16–41 45°C
Carbamazepine III-crystal/I-crystal 1.7–2.1 2°C
1.7–2.0 12°C
1.6–2.0 17°C
1.6–1.9 26°C
1.6–1.8 40°C
1.5–1.7 58°C
Chloramphenicol palmitate A-crystal/B-crystal 3.6 30°C
Iopanoic acid II-crystal/I-crystal 2.3–2.8 37°C
Mefnamic acid I-crystal/II-crystal 1.5 30°C
Glibenclamide amorphous/crystal 112–1652 23°C
Glucose amorphous/crystal 16–53 20°C
Griseofulvin amorphous/crystal 38–441 21°C
Hydrochlorothiazide amorphous/crystal 21–113 37°C
Iopanoic acid amorphous/I-crystal 12–19 37°C
Polythiazide Amorphous/crystal 48–455 37°C
Adapted from Hancock and Parks ( 2000 )
a The range of values refl ects the use of different D C p values for the calculations
In a study conducted by Kobayashi et al., the effect of crystalline carbamazepine polymorphs on solubility, dissolution rate, and oral bioavailability was investigated (Kobayashi et al. 2000 ) . Calculated aqueous solubility values of form I, form III, and the dihydrate were found to be 460.2 m g/mL, 501.9 m g/mL, and 311.1 m g/mL, respectively. In agreement with the ranges outlined by Pudipeddi and Serajuddin, the solubility ratio between the low and high-energy polymorphic states ranged from 1.5 to 1.6 for form I and form III, respectively (Pudipeddi and Serajuddin 2005 ) . Dissolution profi les of the three drug substances are illustrated in Fig. 3.11 . Forms I and III exhibited a transient dissolution rate improvement, ultimately converting to the more stable form in solution. While form I initially provided the greatest solubility, it converted to the more stable dihydrate form relatively rapidly.
However, form III exhibited sustained supersaturation which is a desirable charac-teristic for poorly water-soluble compounds. Each polymorphic form was evaluated for in vivo drug absorption using a crossover technique in male beagle dogs. The in vivo performance of the polymorphic forms was compared to a solubilized amor-phous formulation, representing 100% bioavailability. The study demonstrated that at high dose, form I provided the greatest C max and AUC 0 –12h values when compared
Table 3.6 Experimental solubility ratios for drug compounds
Compound Forms Solubility ratio a Comment Indomethacin a -crystal/ g -crystal 1.1–1.2 45°C, water
amorphous/ g -crystal 38–301 5°C, water 25–104 25°C, water 16–41 45°C, water Carbamazepine III-crystal/I-crystal 1.7–2.1 2°C, 2-propanol
1.7–2.0 12°C, 2-propanol 1.6–2.0 17°C, 2-propanol 1.6–1.9 26°C, 2-propanol 1.6–1.8 40°C, 2-propanol 1.5–1.7 58°C, 2-propanol Chloramphenicol
palmitate
A-crystal/B-crystal 3.6 30°C, 35% t -butanol (aq.) Iopanoic acid II-crystal/I-crystal 2.3–2.8 37°C, phosphate buffer
(aq.)
Mefnamic acid I-crystal/II-crystal 1.5 30°C, dodecyl alcohol Glibenclamide amorphous/crystal 112–1652 23°C, buffer (aq.)
Glucose amorphous/crystal 16–53 20°C, methanol
20°C, ethanol 20°C, isopropanol Griseofulvin amorphous/crystal 38–441 21°C, water
Hydrochlorothiazide amorphous/crystal 21–113 37°C, HCl & PVP (aq.) Iopanoic acid amorphous/I-crystal 19-Dec 37°C, phosphate buffer
(aq.)
Polythiazide Amorphous/crystal 48–455 37°C, HCl & PVP (aq.)
a The range of values refl ects the use of different D C p values for the calculations Adapted from Hancock and Parks ( 2000 )
Fig. 3.9 Solubility ratios for polymorphs ( n = 81). The data do not include the premafl oxacin (I/III) ratio which was found to be 23.1. From Pudipeddi and Serajuddin ( 2005 )
Fig. 3.8 Experimental aqueous solubility profi les for amorphous and crystalline indomethacin at ( a ) 5°C, ( b ) 25°C, and ( c ) 45°C. From Hancock and Parks ( 2000 )
to the other two forms. Bioavailability of form I, form III, and the dihydrate, relative to the amorphous solubilized formulation, was found to be 68.7%, 47.8%, and 33%, respectively. This result is consistent with the probable conversion of form III to the dihydrate form in situ. The plasma concentration versus time curve for this arm of the study is shown in (Fig. 3.12 ). This result indicated that while there was a very small difference in measured solubility between form I and form III, the ability of form I to remain in a supersaturated state for an extended period time allowed for improved oral bioavailability (Fig. 3.12 ).
Fig. 3.10 Anhydrate/hydrate solubility ratios ( n = 23). Compound 6 is expressed as hydrate/anhy-drate due to an anomaly. From Pudipeddi and Serajuddin ( 2005 )
Fig. 3.11 Dissolution patterns of carbamazepine polymorphs and dihydrate at 37°C in pH 1.2 media:
square -form I; circle -form III; and triangle -dihydrate.
From Kobayashi et al. ( 2000 )
In a recent study conducted by Kim et al., the oral bioavailability of amorphous atorvastatin hemi-calcium prepared by various techniques was evaluated (Kim et al.
2008 ) . Specifi cally, the researchers evaluated spray-drying and supercritical anti-solvent (SAS) processes against unprocessed crystalline material with low aqueous solubility (142.2 m g/mL). Material processed by spray-drying and SAS processes, which utilized acetone or tetrahydrofuran as the solvent, exhibited aqueous solubili-ties ranging from 467.1 to 483.2 m g/mL. Amorphous material prepared from the SAS and spray-drying processes were found to have particle size ranges from 68.7 to 95.7 nm and 3.62 to 7.31 m m, respectively. Powder dissolution analysis revealed that amorphous material provided signifi cant improvements in dissolution rate, as illustrated in Fig. 3.13 . Amorphous particles prepared by the SAS processing method were found to have a faster rate of dissolution than those prepared by spray drying, consistent with its small particle size. Unprocessed and amorphous materi-als were evaluated for in vivo drug absorption in male rats. The hypothesis tested with this study was that not only would amorphous materials provide enhanced absorption, but that material prepared by SAS, due to its small particle size and
Fig. 3.12 Plasma concentration–time curves of carbamazepine polymorphs and dihydrate after oral administration to dogs ( n = 4; mean±S.E.). Dose: 200 mg/body. Diamond : solution; square : form I; circle : form III; and triangle : dihydrate. From Kobayashi et al. ( 2000 )
large surface area, would provide the highest absorption. From this study, it was determined that amorphous materials provided a signifi cant increase in drug absorp-tion, with SAS prepared material providing the greatest improvement. The amor-phous form prepared by SAS with acetone showed a threefold improvement in the AUC 0–8h , a fourfold improvement in C max , and a twofold improvement in T max . Similarly, spray-dried amorphous material also provided a signifi cant improvement over crystalline drug. The plasma concentration versus time curve from this study is shown in Fig. 3.14 . It is clear from this study that the amorphous form can have a marked impact on bioavailability.
The use of metastable solids has been shown to be an effective method to enhance both solubility and bioavailability of drug substances. While most polymorphic forms generally provide relatively small improvements in solubility, the improve-ment by be signifi cant from a pharmacological standpoint. However, these systems are inherently unstable and extreme care should be taken to ensure that crystalliza-tion into an undesirable form does not occur during processing (e.g., granulacrystalliza-tion, drying, tableting, etc.) or storage. This is particularly diffi cult in the case of amor-phous solids, which are normally stabilized by excipients or formulated as solid dispersion systems prior to being incorporated into solid dosage forms (Leuner and Dressman 2000 ) .
Fig. 3.13 Powder dissolution profi les of unprocessed atorvastatin particles ( fi lled square ), SAS-processed amorphous atorvastatin calcium precipitation from acetone ( fi lled circle ), SAS-processed amorphous atorvastatin calcium precipitated from a tetrahydrofuran solution ( fi lled triangle ), spray-dried amorphous atorvastatin calcium from an acetone solution ( empty circle ), and spray-dried amorphous atorvastatin calcium from a tetrahydrofuran solution ( empty square ) ( n = 3), (mean±S.D.). From Kim et al. ( 2008 )