In Vitro Percutaneous Absorption of Thiamine Disulfide from a Mixture of Propylene Glycol and Fatty Acid YASUKO KOMATA~, MIHOINAOKA, AKIKOKANEKO, AND TADAOFUJIE Received April 16, 1991, from the Kyoritsu College of Pharmacy, 1-5-30, Shibakoen, Minato-ku, Tokyo 705, Japan. September 11, 1991. Ab8tract 0 The in vitro percutaneous transport of thiamine disulfide (TDS), an oxidized dimer of thiamine, from propylene glycol through
excised abdominal rat skin was studied. The application of saturated, long-chain fatty acids [stearicacid (18:0), myristic acid (14:0), and lauric acid (12:O)) as enhancers to the system was also studied. TDS permeated through rat skin from propylene glycol with a flux of 2.5 f 0.8 pg/cm2/min.The flux was enhanced 31 times by 12:O and 1.4 times by 14:O and was suppressed to 80% of its original value by 18:O. The absorption of TDS could not be explained by TDS permeating across a dialysis membrane, but the interaction between TDS and fatty acids may influence the system. The results show the possibility of developing a transdermal thiamine delivery system.
Thiamine is indicated for the prophylaxis and treatment of thiamine deficiency, which leads to beriberi or Wernicke's encephalopathy. It is administered orally (elixir or tablet) or parenterally (injection), but there are some pharmaceutical problems with these means of administration. Sugar coating, which masks the objectionable smell and taste of thiamine, causes bioinequivalence problems, because the dissolution of the drug is then affected by gastric acidity.1 Intramuscular injection results in some tenderness and induration. The instability of thiamine in multivitamin infusions administered with intravenous hyperalimentation restricts its period of storage.2 A transdermal system may resolve these problems. A thiamine transdermal system would be useful when oral administration is not acceptable (e.g., in malabsorption caused by use of alcohol). It would also be very useful for nutritional management during prolonged stress, heavy manual labor, and pregnancy because it is a convenient means of administration. In general, drugs face difficulty in permeating through the barrier presented by the skin. Because fat-soluble drugs have intrinsic percutaneous permeabilities, lipophilic thiamine derivatives would enhance the percutaneous absorption of thiamine, which is basically a water-soluble vitamin. Using excised rat skins, we studied the percutaneous absorption of thiamine disulfide (TDS) from propylene glycol (PG)in an in vitro model. TDS, an oxidized dimer of thiamine, is more lipophilic and more stable to both heat and alkalinity than thiamine and has thiamine activity.3 Kimura et al.4 reported that thiamine (water soluble) and thiamine tetrahydrohrfuryl disulfide (fat soluble) permeate rat skin from a simple ointment (Japanese Pharmacopoeia) containing 10% ethanol and that the concentration of thiamine in rat plasma is higher when thiamine tetrahydrofurfuryl disulfide, rather than thiamine, is applied. To optimize absorption, a study of appropriate vehicles is needed. Because fatty acids (FAs) are enhancers of the percutaneous absorption of many drugs,6.8 we tested the application of FAs as enhancers of percutaneous transdermal delivery of TDS. 744 / Journal of Pharmaceutical Sciences Vol. 81, No. 8, August 7992
Accepted for publication
Experimental Section Materials-TDS was purchased from Tokyo Kasei Industries Company, Ltd. (Tokyo, Japan). Lauric acid (12:O) was purchased from Sigma Chemical Company, Inc. (St. Louis, MO).Myristic acid (14:O) was purchased from Wako Pure Chemical Industries Company, Ltd. (Osaka, Japan). Stearic acid (18:O)was purchased from Koso Chemical Company, Ltd. (Tokyo, Japan). PG was purchased from Yamada Pharmaceutical Company, LM. (Ibaragi, Japan). All other chemicals and solvents were guaranteed reagent grade. Cellulose dialysis membranes were purchased from Sanko Junyaku Company, Ltd. (Tokyo, Japan). A n i m a l e M a l e rats (Wistar/ST strain) with an approximate weight of 230-270 g were supplied by Sankyo Laboratory Service. Preparation of Skin Membrane-The abdominal region of the rats was carefully shaved with electric and hand razors. A 2-cm2 section of skin with a thicknese of -0.60-0.75 mm was excised. Determination of Permeation through Skin-An excised section of skin was mounted between two half diffusion cells (horizontal; Nagoya Science Company, Ltd.), each with a volume of 5.0 mL and an effective diffusion area of 0.636 cm'. The dermis side of the skin was in contact with the receiver compartment, and the stratum comeum was in contact with the donor compartment. The receiver compartment of the cell was filled with 5 mL of phosphate-buffered saline (138 mM NaCl; 2.7 mM KCl; 8.1 mM Na2HP0,; 1.5mM K2HP0,, pH 7.3), and the donor compartment was filled with 5 mL of drug suspension in vehicle. The donor chamber was sealed from the atmosphere with plastic film (Parafilm). The diffusion cells were maintained at 37 "C in a water bath, and both the donor and receiver compartments were stirred throughout the experiment. At appropriate times, 200-pL samples were withdrawn from the receiver compartment and assayed for TDS. After sampling, 200 p.L of phosphate-buffered saline was added to the receiver compartment to keep the volume constant. The vehicles used were either PG or 10% (w/v) FA in PG. Mixtures of FAs and PG (FA-PG) were prepared at 50-60 "C to melt and dissolve the a4uvant. Clear solutions were obtained only for 12:O in PG (12% PG),whereas mixtures of 1 4 0 in PG (14:O-PG) and 18:O in PG (18:O-PG) were semisolid at the experimental temperature. TDS (2.5 g, a n amount greater than its solubility) was then added to 5 mL of each vehicle to maximize permeation through the skin. Determination of Permeation through the Dialysis MembraneThe same diffusion cell system was used to determine the rate of TDS transport across a dialysis membrane from each vehicle. Other conditions were the same as those used to determine drug permeation through skin. Analysis-The amount of TDS in the receiver phase was determined by high-performance liquid chromatography. Butyl p-hydmxybenzoate was used as an internal standard. The conditions were as follows: pump, HLDdO3D (Toyo Soda); column, 4.0 x 150 mm, Nucleosil 100 5C18 (Gasukuro Industries Company, Ltd., Japan); mobile phase, water:methanol (1:2, v/v); detector, UV-8 model II (Toyo Soda); wavelength, 254 nm. Retention times of TDS and butyl p-hydroxybemate at a flow rate of 0.8 m u m i n were 3.7 and 10.1 min, respectively. Peak areas were calculated by using a data treatment computer (Chromatopack C-R1A; Shimazu Seisakusho, Japan). Throughout the experiments, TDS was detected as a single peak.
Results and Discussion TDS permeates through ra t skin from PG, and the addition of FA to PG greatly affects the rate of permeation (Figure 1). 0022-3549/92//0800-0744$02.50/0 0 7992, American Pharmaceutical Assocation
20
-F
15
1
cn 0 I-
U
10
€ I*t C
a
E a
5
0 0
2
4
6
8
Time (hr) Flgum 1-Effect of FA on percutaneous absorption of TDS from PG through excised rat skin. Key to vehicle: (0)PG; (@) 12:O-PG; (a) 14:WG; (0)18:O-PG. Points and vertical bars represent means and standard deviations (n = 3), respectively.
TDS flux (pg/cm2/min) was calculated from the slope of the steady-state portion of a plot of the amount permeated versus time divided by the skin surface area. The enhancing effects of the various FAs were then evaluated (Table I). The greatest enhancing effect was exhibited by 12:0, which had an enhancement fador of 31; this was followed by 14:O. On the other hand, 18:O suppressed the permeation of TDS through skin. These results suggest that some FAs can be used to improve the percutaneous absorption of TDS from PG and that a decrease in the number of carbon atoms (C,) in the FA results in an increase in enhancement of TDS permeation. Other studies in which FAs were used as penetration enhancers6.6 have also reported the dependence of the enhancing effecton C,; there appears to be a parabolic curve relationship between the enhancing effect and C, (12:O is the most effective). The mechanism of permeation of these compounds could be explained by a nonpolar route in the stratum corneum. Therefore, TDS may also permeate mainly through a nonpolar route in the stratum corneum. FAs enhance permeation of compounds via a nonpolar route more than they enhance permeation via a polar route.'
The permeation lag time was determined from the intercept of the slope of the steady-state flux on the x-axis (Table I).The lag time for FA-PG was longer than for PG by -10 min, and there was little difference between 12:O-PG and 14:O-PG. A lag time for 18:O could not be determined. The retardation of lag time by FAs may be related to the penetration of FAs through the skin. Yamada and Uda6 suggested that it takes time for FAs to penetrate the skin and interact with skin lipids. Figure 2 shows the effectof FAs on the permeation of TDS from PG across a dialysis membrane. The flux of TDS from FA-PG was (determined from the amounts of TDS that permeated between 0 and 3 h; Table I) higher than that from PG alone. However, the enhancing effect did not depend on C,, and the factors were all between 1 and 2, less than the enhancement factors for permeation of TDS through skin. These results suggest that the permeation of TDS through the dialysis membrane does not predict its permeation through skin, perhaps because the permeation of TDS through rat skin is not explained simply by permeation through pores. It is, therefore, important to select a proper model membrane for evaluating the effect of enhancers on the permeability of a specific compound. From the data on permeation through a dialysis membrane (Table I), the increase in TDS flux caused by the addition of FA may be due to the increased solubility of TDS in the vehicle arising from an interaction between TDS and FA. The solubility of TDS in ethanol is increased by the addition of FA.&We have previously reported the formation of a complex between TDS and FA in methanol.9 The association constants for this complex do not depend on C, in FA. Golden et al.10 reported that 18:O (0.15 M) has no effect on the percutaneous absorption of salicylic acid, in contrast to the enhancing effect of cis-9-and cis-11-octadecenoic acids. Aungst" reported that 18:O (0.5 M) increases naloxone base flux, although the effect is less than that of unsaturated FAs of the same chain length. These results are explained as follows: FAs increase skin permeability by disrupting the packed structure of the stratum corneum; but 18:0, being like the hydrocarbon tails of normal stratum corneum lipids, does not disrupt the packed structure. Our results showing that
Tabk C E M of FA on TDS Permeatlon through Rat Skln and Dialyrls Membrane from Vehicles
Dialysis Membrane
Skin
Vehicle
2.5 2 0.8 12:WG 76.8 2 6.7 14:WG 3.6 2 1.7 18:O-PG 0.6 f 0.4
PG
a
Lag Time, FIUX, min W/cm2/min
FIUX,
pg/cm2/min 1
30.7 1.4 0.2
18.2 28.8 29.6
-'
25.4 2 0.2 28.5 f 0.1 41.8 2 0.8 39.7 2 0.7
Relative flux versus PG alone. Not determined.
1
1.1 1.7 1.6
0
2 Time (hr)
4
Figure 2-Effect of FA on permeation of TDS from PG through the dialysis membrane. Symbols are the same as for Figure 1. Points represent the means (n = 2). Journal of Pharmaceutical Sciences / 745 Vol. 81, No. 8, August 1992
18:O suppresses the flux of TDS through skin cannot be explained by this mechanism alone. TDS interacts with FA in l,2-dichloroethane1 whose dielectric constant (E = 10.65 a t 20 "C)is similar to that of 1-octanol( E = 10.34 at 20 "C),which is usually used to determine drug partition between skin and vehicle. The solubility of TDS in 1,2dichloroethane at 37 "C is 6.5 x lo-' M; the solubility increases to 1.8 x lo-' M (3.3 times) by adding 1.0 x lo-' M 180 and to 1.2 x lo-' M (2.2 M 14:O.lS TDS may be solubilized times) by adding 1.0 x in lower polarity solvents by the addition of FAs because of an interaction between TDS and FA. This interaction, therefore, may explain both the enhancing and the suppressing effects of FAs on the percutaneous absorption of TDS. Ogiso and Shintani12 proposed that the percutaneous absorption of propranolol is enhanced by the formation of a complex between the drug and FA, because the drug that partitioned into the lipid phase in the form of a complex does not bind to epidermal components. In vitro studies of percutaneous absorption with furry animal skin do not provide information in agreement with in vitro studies with human skin because of the difference in the thickness of the stratum corneum and/or the abundance of appendageal pathways. But the mechanism by which a certain enhancer changes the permeability of organic solutes may be similar for skim from different species, although the magnitude of the morphological changes brought about by damage or disruption of the skin may be different for skins of different species. The effect of FA as a percutaneous enhancer has been evaluated with rat skin for molsidominea and with human skin for naloxone base,B and results from both studies
746 I Journal of PharmaceuUcal Sciences Vol. 81, No. 8, August 1992
showed a similar enhancing effect depending on C,. Our results, therefore, are useful a s basic information showing the possibility of a thiamine transdermal system. Because 12:O and 18:O cause irritation,ll the development of a thiamine transdermal system must include a careful weighing of benefits and risks.
References and Notes 1. Aoyagi, N.; Ogata, H.; Kaniwa, N.; Koibuchi, M.; Shibazaki, T.;
EJima, A.; Mimbe, M.; Kohno, K.; Samejima, M. Chem. Pharm. Bull. 1986,34, 281-291. 2. Yoshida, S.; Ishikawa, S.;Su awara, K.; Kitamae, F.;Yokoyama, T. Byouin Yakugaku 1984,8,417-430. 3. Kawasaki, C. In Viturnins and Hormones; Harris,R. S.; Wool, I. G.; Loraine, J. A., Eda.; Academc: London, 1963; Vol. 21, pp
69-78. 4. Kimura, M.;Kueano, M.; Yokoi, K.; Itokawa, Y. Vitamins (Japan) 1989,63,621-625. 5. Yamada, M.;Uda, Y. Chem. Pharm. Bull. 1987,35,3390-3398. 6. Aunget, B. J.; Rogers, N. J.; Shefter, E.Znt. J . Phurm. 1986,33, 225-234. 7. Aunget, B: J.; Blake, J.A.; Ro ere, N. J.; Huseain, M.A. J . Pharm. Scr. 1990, 79, 1072-107f 8. Ueda F.; Hi ashi, T.; Ayukawa, Y.; Takada, A.; Fqjie, T.; Kaneko, A. dtamins (Japan) 1987,61,57-64. 9. Komata, Y.; F 'ie, T.; Kaneko, A.; Ueda, F.; Urano, S. Liplids 1988,23,525-#7 10. Golden, G. M.;McKie, J. E.;Potte, R. 0.J . Pharm.Sci. 1987, 76, 25-28. 11. Aungst, B. J. Pharm. Res. 1989, 6, 244-247. 12. Ogiao, T.; Shintani, M. J . Phurm.Sci. 1990, 79, 1065-1071. 13. Komata, Y., unpublished results.
excised abdominal rat skin was studied. The application of saturated, long-chain fatty acids [stearicacid (18:0), myristic acid (14:0), and lauric acid (12:O)) as enhancers to the system was also studied. TDS permeated through rat skin from propylene glycol with a flux of 2.5 f 0.8 pg/cm2/min.The flux was enhanced 31 times by 12:O and 1.4 times by 14:O and was suppressed to 80% of its original value by 18:O. The absorption of TDS could not be explained by TDS permeating across a dialysis membrane, but the interaction between TDS and fatty acids may influence the system. The results show the possibility of developing a transdermal thiamine delivery system.
Thiamine is indicated for the prophylaxis and treatment of thiamine deficiency, which leads to beriberi or Wernicke's encephalopathy. It is administered orally (elixir or tablet) or parenterally (injection), but there are some pharmaceutical problems with these means of administration. Sugar coating, which masks the objectionable smell and taste of thiamine, causes bioinequivalence problems, because the dissolution of the drug is then affected by gastric acidity.1 Intramuscular injection results in some tenderness and induration. The instability of thiamine in multivitamin infusions administered with intravenous hyperalimentation restricts its period of storage.2 A transdermal system may resolve these problems. A thiamine transdermal system would be useful when oral administration is not acceptable (e.g., in malabsorption caused by use of alcohol). It would also be very useful for nutritional management during prolonged stress, heavy manual labor, and pregnancy because it is a convenient means of administration. In general, drugs face difficulty in permeating through the barrier presented by the skin. Because fat-soluble drugs have intrinsic percutaneous permeabilities, lipophilic thiamine derivatives would enhance the percutaneous absorption of thiamine, which is basically a water-soluble vitamin. Using excised rat skins, we studied the percutaneous absorption of thiamine disulfide (TDS) from propylene glycol (PG)in an in vitro model. TDS, an oxidized dimer of thiamine, is more lipophilic and more stable to both heat and alkalinity than thiamine and has thiamine activity.3 Kimura et al.4 reported that thiamine (water soluble) and thiamine tetrahydrohrfuryl disulfide (fat soluble) permeate rat skin from a simple ointment (Japanese Pharmacopoeia) containing 10% ethanol and that the concentration of thiamine in rat plasma is higher when thiamine tetrahydrofurfuryl disulfide, rather than thiamine, is applied. To optimize absorption, a study of appropriate vehicles is needed. Because fatty acids (FAs) are enhancers of the percutaneous absorption of many drugs,6.8 we tested the application of FAs as enhancers of percutaneous transdermal delivery of TDS. 744 / Journal of Pharmaceutical Sciences Vol. 81, No. 8, August 7992
Accepted for publication
Experimental Section Materials-TDS was purchased from Tokyo Kasei Industries Company, Ltd. (Tokyo, Japan). Lauric acid (12:O) was purchased from Sigma Chemical Company, Inc. (St. Louis, MO).Myristic acid (14:O) was purchased from Wako Pure Chemical Industries Company, Ltd. (Osaka, Japan). Stearic acid (18:O)was purchased from Koso Chemical Company, Ltd. (Tokyo, Japan). PG was purchased from Yamada Pharmaceutical Company, LM. (Ibaragi, Japan). All other chemicals and solvents were guaranteed reagent grade. Cellulose dialysis membranes were purchased from Sanko Junyaku Company, Ltd. (Tokyo, Japan). A n i m a l e M a l e rats (Wistar/ST strain) with an approximate weight of 230-270 g were supplied by Sankyo Laboratory Service. Preparation of Skin Membrane-The abdominal region of the rats was carefully shaved with electric and hand razors. A 2-cm2 section of skin with a thicknese of -0.60-0.75 mm was excised. Determination of Permeation through Skin-An excised section of skin was mounted between two half diffusion cells (horizontal; Nagoya Science Company, Ltd.), each with a volume of 5.0 mL and an effective diffusion area of 0.636 cm'. The dermis side of the skin was in contact with the receiver compartment, and the stratum comeum was in contact with the donor compartment. The receiver compartment of the cell was filled with 5 mL of phosphate-buffered saline (138 mM NaCl; 2.7 mM KCl; 8.1 mM Na2HP0,; 1.5mM K2HP0,, pH 7.3), and the donor compartment was filled with 5 mL of drug suspension in vehicle. The donor chamber was sealed from the atmosphere with plastic film (Parafilm). The diffusion cells were maintained at 37 "C in a water bath, and both the donor and receiver compartments were stirred throughout the experiment. At appropriate times, 200-pL samples were withdrawn from the receiver compartment and assayed for TDS. After sampling, 200 p.L of phosphate-buffered saline was added to the receiver compartment to keep the volume constant. The vehicles used were either PG or 10% (w/v) FA in PG. Mixtures of FAs and PG (FA-PG) were prepared at 50-60 "C to melt and dissolve the a4uvant. Clear solutions were obtained only for 12:O in PG (12% PG),whereas mixtures of 1 4 0 in PG (14:O-PG) and 18:O in PG (18:O-PG) were semisolid at the experimental temperature. TDS (2.5 g, a n amount greater than its solubility) was then added to 5 mL of each vehicle to maximize permeation through the skin. Determination of Permeation through the Dialysis MembraneThe same diffusion cell system was used to determine the rate of TDS transport across a dialysis membrane from each vehicle. Other conditions were the same as those used to determine drug permeation through skin. Analysis-The amount of TDS in the receiver phase was determined by high-performance liquid chromatography. Butyl p-hydmxybenzoate was used as an internal standard. The conditions were as follows: pump, HLDdO3D (Toyo Soda); column, 4.0 x 150 mm, Nucleosil 100 5C18 (Gasukuro Industries Company, Ltd., Japan); mobile phase, water:methanol (1:2, v/v); detector, UV-8 model II (Toyo Soda); wavelength, 254 nm. Retention times of TDS and butyl p-hydroxybemate at a flow rate of 0.8 m u m i n were 3.7 and 10.1 min, respectively. Peak areas were calculated by using a data treatment computer (Chromatopack C-R1A; Shimazu Seisakusho, Japan). Throughout the experiments, TDS was detected as a single peak.
Results and Discussion TDS permeates through ra t skin from PG, and the addition of FA to PG greatly affects the rate of permeation (Figure 1). 0022-3549/92//0800-0744$02.50/0 0 7992, American Pharmaceutical Assocation
20
-F
15
1
cn 0 I-
U
10
€ I*t C
a
E a
5
0 0
2
4
6
8
Time (hr) Flgum 1-Effect of FA on percutaneous absorption of TDS from PG through excised rat skin. Key to vehicle: (0)PG; (@) 12:O-PG; (a) 14:WG; (0)18:O-PG. Points and vertical bars represent means and standard deviations (n = 3), respectively.
TDS flux (pg/cm2/min) was calculated from the slope of the steady-state portion of a plot of the amount permeated versus time divided by the skin surface area. The enhancing effects of the various FAs were then evaluated (Table I). The greatest enhancing effect was exhibited by 12:0, which had an enhancement fador of 31; this was followed by 14:O. On the other hand, 18:O suppressed the permeation of TDS through skin. These results suggest that some FAs can be used to improve the percutaneous absorption of TDS from PG and that a decrease in the number of carbon atoms (C,) in the FA results in an increase in enhancement of TDS permeation. Other studies in which FAs were used as penetration enhancers6.6 have also reported the dependence of the enhancing effecton C,; there appears to be a parabolic curve relationship between the enhancing effect and C, (12:O is the most effective). The mechanism of permeation of these compounds could be explained by a nonpolar route in the stratum corneum. Therefore, TDS may also permeate mainly through a nonpolar route in the stratum corneum. FAs enhance permeation of compounds via a nonpolar route more than they enhance permeation via a polar route.'
The permeation lag time was determined from the intercept of the slope of the steady-state flux on the x-axis (Table I).The lag time for FA-PG was longer than for PG by -10 min, and there was little difference between 12:O-PG and 14:O-PG. A lag time for 18:O could not be determined. The retardation of lag time by FAs may be related to the penetration of FAs through the skin. Yamada and Uda6 suggested that it takes time for FAs to penetrate the skin and interact with skin lipids. Figure 2 shows the effectof FAs on the permeation of TDS from PG across a dialysis membrane. The flux of TDS from FA-PG was (determined from the amounts of TDS that permeated between 0 and 3 h; Table I) higher than that from PG alone. However, the enhancing effect did not depend on C,, and the factors were all between 1 and 2, less than the enhancement factors for permeation of TDS through skin. These results suggest that the permeation of TDS through the dialysis membrane does not predict its permeation through skin, perhaps because the permeation of TDS through rat skin is not explained simply by permeation through pores. It is, therefore, important to select a proper model membrane for evaluating the effect of enhancers on the permeability of a specific compound. From the data on permeation through a dialysis membrane (Table I), the increase in TDS flux caused by the addition of FA may be due to the increased solubility of TDS in the vehicle arising from an interaction between TDS and FA. The solubility of TDS in ethanol is increased by the addition of FA.&We have previously reported the formation of a complex between TDS and FA in methanol.9 The association constants for this complex do not depend on C, in FA. Golden et al.10 reported that 18:O (0.15 M) has no effect on the percutaneous absorption of salicylic acid, in contrast to the enhancing effect of cis-9-and cis-11-octadecenoic acids. Aungst" reported that 18:O (0.5 M) increases naloxone base flux, although the effect is less than that of unsaturated FAs of the same chain length. These results are explained as follows: FAs increase skin permeability by disrupting the packed structure of the stratum corneum; but 18:0, being like the hydrocarbon tails of normal stratum corneum lipids, does not disrupt the packed structure. Our results showing that
Tabk C E M of FA on TDS Permeatlon through Rat Skln and Dialyrls Membrane from Vehicles
Dialysis Membrane
Skin
Vehicle
2.5 2 0.8 12:WG 76.8 2 6.7 14:WG 3.6 2 1.7 18:O-PG 0.6 f 0.4
PG
a
Lag Time, FIUX, min W/cm2/min
FIUX,
pg/cm2/min 1
30.7 1.4 0.2
18.2 28.8 29.6
-'
25.4 2 0.2 28.5 f 0.1 41.8 2 0.8 39.7 2 0.7
Relative flux versus PG alone. Not determined.
1
1.1 1.7 1.6
0
2 Time (hr)
4
Figure 2-Effect of FA on permeation of TDS from PG through the dialysis membrane. Symbols are the same as for Figure 1. Points represent the means (n = 2). Journal of Pharmaceutical Sciences / 745 Vol. 81, No. 8, August 1992
18:O suppresses the flux of TDS through skin cannot be explained by this mechanism alone. TDS interacts with FA in l,2-dichloroethane1 whose dielectric constant (E = 10.65 a t 20 "C)is similar to that of 1-octanol( E = 10.34 at 20 "C),which is usually used to determine drug partition between skin and vehicle. The solubility of TDS in 1,2dichloroethane at 37 "C is 6.5 x lo-' M; the solubility increases to 1.8 x lo-' M (3.3 times) by adding 1.0 x lo-' M 180 and to 1.2 x lo-' M (2.2 M 14:O.lS TDS may be solubilized times) by adding 1.0 x in lower polarity solvents by the addition of FAs because of an interaction between TDS and FA. This interaction, therefore, may explain both the enhancing and the suppressing effects of FAs on the percutaneous absorption of TDS. Ogiso and Shintani12 proposed that the percutaneous absorption of propranolol is enhanced by the formation of a complex between the drug and FA, because the drug that partitioned into the lipid phase in the form of a complex does not bind to epidermal components. In vitro studies of percutaneous absorption with furry animal skin do not provide information in agreement with in vitro studies with human skin because of the difference in the thickness of the stratum corneum and/or the abundance of appendageal pathways. But the mechanism by which a certain enhancer changes the permeability of organic solutes may be similar for skim from different species, although the magnitude of the morphological changes brought about by damage or disruption of the skin may be different for skins of different species. The effect of FA as a percutaneous enhancer has been evaluated with rat skin for molsidominea and with human skin for naloxone base,B and results from both studies
746 I Journal of PharmaceuUcal Sciences Vol. 81, No. 8, August 1992
showed a similar enhancing effect depending on C,. Our results, therefore, are useful a s basic information showing the possibility of a thiamine transdermal system. Because 12:O and 18:O cause irritation,ll the development of a thiamine transdermal system must include a careful weighing of benefits and risks.
References and Notes 1. Aoyagi, N.; Ogata, H.; Kaniwa, N.; Koibuchi, M.; Shibazaki, T.;
EJima, A.; Mimbe, M.; Kohno, K.; Samejima, M. Chem. Pharm. Bull. 1986,34, 281-291. 2. Yoshida, S.; Ishikawa, S.;Su awara, K.; Kitamae, F.;Yokoyama, T. Byouin Yakugaku 1984,8,417-430. 3. Kawasaki, C. In Viturnins and Hormones; Harris,R. S.; Wool, I. G.; Loraine, J. A., Eda.; Academc: London, 1963; Vol. 21, pp
69-78. 4. Kimura, M.;Kueano, M.; Yokoi, K.; Itokawa, Y. Vitamins (Japan) 1989,63,621-625. 5. Yamada, M.;Uda, Y. Chem. Pharm. Bull. 1987,35,3390-3398. 6. Aunget, B. J.; Rogers, N. J.; Shefter, E.Znt. J . Phurm. 1986,33, 225-234. 7. Aunget, B: J.; Blake, J.A.; Ro ere, N. J.; Huseain, M.A. J . Pharm. Scr. 1990, 79, 1072-107f 8. Ueda F.; Hi ashi, T.; Ayukawa, Y.; Takada, A.; Fqjie, T.; Kaneko, A. dtamins (Japan) 1987,61,57-64. 9. Komata, Y.; F 'ie, T.; Kaneko, A.; Ueda, F.; Urano, S. Liplids 1988,23,525-#7 10. Golden, G. M.;McKie, J. E.;Potte, R. 0.J . Pharm.Sci. 1987, 76, 25-28. 11. Aungst, B. J. Pharm. Res. 1989, 6, 244-247. 12. Ogiao, T.; Shintani, M. J . Phurm.Sci. 1990, 79, 1065-1071. 13. Komata, Y., unpublished results.
