- Email: [email protected]
In vivo human skin permeability enhancement by oleic acid: a laser Doppler velocimetry study
Journal of Controlled Release 58 (1999) 97–104
In vivo human skin permeability enhancement by oleic acid: a laser Doppler velocimetry study Hanafi Tanojo a , *, Esther Boelsma b , Hans E. Junginger c , Maria Ponec b , Harry E. Bodde´ c 1 a
Department of Dermatology, University of California, Box 0989, 90 Medical Centerway, Surge 110, San Francisco, CA 94143 -0989, USA b Department of Dermatology, Leiden University Hospital, Leiden, Netherlands. c Division Pharmaceutical Technology, Leiden /Amsterdam Center for Drug Research, Leiden University, Leiden, Netherlands Received 24 April 1998; received in revised form 2 August 1998; accepted 10 August 1998
Abstract Topical application of a skin permeation enhancer such as oleic acid (OA) can result in: (i) higher skin permeability for many exogenous substances (ii) an irritation reaction. Laser Doppler velocimetry (LDV) is one of few techniques which can assess both effects using appropriate protocols. Direct LDV measurement, measuring cutaneous blood flow, has been preferred as a tool to evaluate skin irritation. By comparing the LDV value of an irritant-treated site with an untreated site, an irritation index for the irritant can be obtained. Occlusive application of 0.16 M OA in propylene glycol (PG) for either 3 or 24 h produced irritation in form of redness and slight swelling. Using LDV, we obtained an irritation index of 2 and 4, respectively. The vehicle, PG alone, produced an index of around 1, which corresponded well to the slight to almost no irritation observed visually. The duration of the high irritation index assessed by LDV after OA–PG application is comparable to the duration of the increase in transepidermal water loss following the same application. This indicates a correlation between skin irritation and barrier perturbation caused by OA. LDV can also be used to elucidate the effect of enhancers on skin using hexyl nicotinate (HN) as a model drug, since its vasodilative effect can be clearly assessed by LDV. Pre-treatment of PG for 3 h significantly reduced the t 0 and t max of HN. No further reduction could be observed when OA was added into PG, suggesting that OA–PG is not more effective than PG alone in enhancing the permeation of HN. 1999 Elsevier Science B.V. All rights reserved. Keywords: Skin permeability; Oleic acid; Topical application; Laser-Doppler velocimetry; Human, in vivo study
1. Introduction Skin gives protection to the body against penetration of most substances from outside, owing to its
*Corresponding author. Tel.: 11 415 4764997; fax: 11 415 7535304; e-mail: [email protected] 1 Deceased September 8, 1996.
barrier capacity. This barrier resides mainly in the uppermost layer, the stratum corneum [1]. Many exogenous substances have difficulty passing the stratum corneum to reach the blood circulation system. On the other hand, some compounds — so-called skin permeation enhancers — can improve the penetration of other substances by perturbing the barrier function of the stratum corneum. Many of them have been the subject of investigation for the
0168-3659 / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 98 )00144-8
98
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
development of transdermal drug delivery systems [2]. Several types of penetration enhancers especially fatty acids have gained much attention, because they are natural components of the skin lipids and therefore are not foreign to the body. Oleic acid (OA) is one of the fatty acids found abundantly in nature, including in the human skin. Its capacity as a skin penetration enhancer has been generally recognized in various in vitro studies [3]. In the last decade numerous in vivo studies with human subjects have been conducted in order to establish non-invasive techniques applicable for a large number of enhancers to measure the skin penetration enhancements in living subjects. In vivo techniques, such as transepidermal water loss measurements [4,5], skin impedance measurements [6], infrared spectroscopy (in combination with attenuated total reflectance) [7], skin blanching method for steroids [8], photopulse plethysmography [9], and laser-Doppler velocimetry (LDV), can be used to elucidate the capacity of enhancers. Some of these techniques have employed OA as a model enhancer [4,5,7]. In this paper, we focus on the use of LDV to assess the capacity of oleic acid as enhancer and its effect on the skin. LDV technique measures the skin blood flow. The working mechanism of this instrument is based on the measurement of a Doppler-shift in wavelength [10,11]. When applied to monitoring cutaneous blood flow, the coherent light impinging on the stationary tissues is reflected (scattered) back at the same frequency, but light reflected by moving structures (blood cells) undergoes a frequency shift. This phenomenon is known as Doppler effect. The end result is a spectral broadening of the almost monochromatic laser light applied. Part of the back scattered radiation / light is directed towards suitable photodetectors for subsequent electronic translation into electrical signals. The magnitude of the frequency shift can be related to the velocity of the moving particles, thus providing the basis for relative quantification, in which it is proportional to the number of blood cells multiplied by their velocity. The radiation penetrates to a depth of 1–1.5 mm, which reaches the capillary blood vessels in the dermis passing through the epidermis. In the beginning LDV was used along with visual scoring in the differentiation of irritation. A positive
relationship has been found between the dose of sodium lauryl sulfate (SLS) applied to the skin and blood flow values [11,12]. Lately, LDV has gained preference over visual scoring [13,14]. The use of LDV for the skin permeation study began when it was employed to measure the penetration of methyl nicotinate across the stratum corneum / epidermis to exert erythema as a result of vasodilation on cutaneous capillary blood vessels yielding increase in skin blood flow [9]. The vasodilation is triggered by direct action of the nicotinates on the blood vessel walls at the junction between the dermis and the epidermis [15]. The time of onset of erythema is found to be reproducible for the same experimental conditions [16]. Subsequently, the vasodilation capability of nicotinates was used to measure the penetration enhancement of pyrrolidone derivates [17] and various organic solvents, including alcohols, glycols and kerosine [18] as well as to characterize the drug release from a microemulsion [19].
2. Material and methods
2.1. Pre-experimental conditions Twelve subjects involved in this study were of both sexes, had normal, healthy skin and were fully informed of the nature of the study and the procedures involved. The age range was 20–40 years. This in vivo study was approved by the local ethical committee for human studies. Individuals had to rest for 20–30 min before LDV measurements, with the skin at the measuring site left uncovered, in an air-conditioned room, under constant temperature (20.5618C) and humidity (30–50%) and were kept relaxed throughout the procedure. Different treatments were applied randomly to the subjects. Only LDV values from the same anatomical area were expected to be comparable, so in all experiments only the median part of volar forearm was chosen as the site of applications and LDV measurements. Heat sources and air-draughts in the direction of the individuals were avoided.
2.2. Application of treatment solutions The occlusive application system consisted of a
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
99
In subsequent studies, the application of hexyl nicotinate (Sigma, St. Louis, MO, USA) solution followed by LDV measurements was performed 3 h after the removal of the application chambers (only for the 3-h applications).
2.3. LDV measurements
Fig. 1. Assembly of the occlusive application system.
Hill-Top chamber (Hill-Top, Cincinnati, OH, USA), Leukosilk adhesive tape (Beiersdorf, Hamburg, Germany) and plastic tape (3M, Borken, Germany), assembled as shown in Fig. 1. For each subject, four sites were marked on the volar forearm as shown in Fig. 2. One empty chamber was placed on site 1 and site 4 was left open without any treatment; both acted as blank controls. On site 2 a chamber containing 300 ml propylene glycol (J.T. Baker, Deventer, Netherlands) was attached. On site 3, the same procedure was repeated for 300 ml solution of 0.16 M oleic acid (Brocacef, Maarssen, Netherlands) in PG. Depending on the protocol, 3 or 24 h after the application, the chambers were removed. The application sites were wiped gently with a soft tissue paper but not stripped. LDV measurements were made immediately on the sites.
LDV was measured on the application sites on subject’s forearm using a Diodopp laser Doppler flowmeter (Applied Laser Technology, Maarheeze, Netherlands). The measuring probe contains a diode laser with a wavelength of 780 nm. The instrument was allowed to ‘warm-up’ for 15 min after being turned on, giving time for the laser light to stabilize and was never switched off between intermittent measurements during the day. Prior to the measurements the offset button was used to bring the value to a zero level and never used again during the same day of the experiments. Laser safety directions were followed. The probe was held gently on the skin to avoid vascular compression. Readings were performed after stabilization of the output signal. Only LDV values from the same anatomical area were expected to be comparable, so in all experiments only the median part of volar forearm was chosen for the sites of applications and LDV measurements. The LDV values of treated sites were directly compared to those obtained on the untreated area, i.e. area left completely open (application site 4), that were measured approximately at the same time on the same subject, to obtain an irritation index. Having determined the index for each individual subject, the overall average index and standard deviation were calculated. Statistical analysis of the data was performed using the Student’s t-test.
2.4. Application of hexyl nicotinate solution followed by LDV measurements
Fig. 2. Application sites on the volar forearm of the human volunteers.
Hexyl nicotinate (HN, 20 ml, 10 mM) in one of these solvents: (i) propylene glycol (PG); (ii) PG– isopropanol (60:40, v / v); (iii) 0.16 M oleic acid in PG (OA–PG), was applied for 60 s on the forearm using a filter paper disk. The technique has already been used [20] with a modification of the one applied in earlier studies [17,18]. The nicotinate-induced increase in skin blood flow was assessed continuous-
100
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
Fig. 3. Schematic curve of the vasodilative effect of hexyl nicotinate as recorded by LDV measurements.
ly using LDV by fixing the probe on the skin surface using a plastic probe holder. From the curves obtained, the onset time of action (t 0 ) and the time to reach maximum response (t max ) were determined (see Fig. 3). The division of the lag times of the untreated site by the lag times of the applied sites is presented as the enhancement ratio (ER) for the applied substances. Having determined the ER for each individual subject, the overall average ER and standard deviation were calculated. Statistical analysis of the data was performed using the Student’s t-test.
3. Results
3.1. Direct assessment of LDV on irritation effects Table 1 shows the irritation index of PG and OA–PG following a 3- or 24-h occlusive application. Following the occlusive application of OA–PG for a Table 1 LDV measurements after topical application of propylene glycol (PG) and oleic acid in PG (OA–PG) presented as irritation index in comparison to untreated site Application
3-h Application
24-h Application
PG 0.16 M OA–PG
1.160.4 (n56) 2.160.9 (n56)
1.260.3 (n510) 3.961.5 (n510)
Irritation index was calculated by dividing the LDV value of untreated site with the value of pretreated site (6standard deviation). Significant differences: PG to OA–PG, P,0.05; OA–PG 3-h to 24-h, P,0.05.
short period of time, it was noticed that visually the irritation did not always appear instantly after the application chambers were removed. However, LDV measurements recorded a consistent higher value of blood flow compared to the untreated site. This again proves the accuracy of this technique in comparison to visual scoring. A 3-h application of OA–PG caused a 2-fold increase in skin blood flow compared to the slight increase by PG alone. The degree of irritation was expressed as irritation index, calculated by dividing the LDV value at the treated site with the value at the non-treated site (control). Application for 24 h produced an irritation index of 4, while the application of pure PG yielded only slight erythema and an irritation index close to 1. The 24-h OA–PG applications produced a clearly visible irritation, which lasted longer than after 3-h applications before it completely disappeared. We could observe the diminishing of the erythema and the high blood flow on the OA–PG treated site up to 54 h after the removal of a 24-h application and up to 24 h for a 3-h application.
3.2. Assessment of LDV after the application of hexyl nicotinate In a preliminary study the influence of various vehicles on the flux of HN was compared (Table 2). The onset of action time, t 0 , and the maximum response time, t max , of the vasodilation assessed by LDV were used as parameters. The values of erythema caused by vasodilative effects of HN were much higher than those due to irritation caused by the vehicles. Hence, these two conditions are clearly distinguishable. Either dissolved in PG or in PG–isopropanol (60:40, v / v), the vasodilation effect of HN began to emerge and maximize at about the same time. The Table 2 Influence of solvents on the lag-times of hexyl nicotinate (10 mM) from the time of application to the initial (t 0 ) and maximal (t max ) responses (6standard deviation) Hexyl nicotinate 10 mM in
t 0 (min)
t max (min)
Propylene glycol (PG) PG–isopropanol (6:4, v / v) 0.16 M Oleic acid in PG
12.563.8 (n54) 12.463.2 (n54) 14.861.6 (n53)
20.163.6 (n54) 19.164.9 (n54) 21.266.9 (n53)
No significant difference among the lag-times: P.0.05.
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104 Table 3 Influence of 3-h occlusive pretreatments on the lag-time between application of hexyl nicotinate [10 mM in propylene glycol– isopropanol (60:40, v / v)] and initial or maximal responses (t 0 or t max ) Pre-treatment
ER t 0
ER t max
Propylene glycol (PG) 0.16 M Oleic acid in PG
1.860.3 (n55) 2.060.4 (n55)
1.660.4 (n55) 1.760.3 (n55)
ER, enhancement ratio, calculated by dividing the lag-time of untreated site with the lag-time of pretreated site (6standard deviation). No significant difference between PG and oleic acid in PG: t 0 and t max , P.0.05.
use of OA–PG as vehicle slightly increased the t 0 (not significant) compared to the other vehicles, but did not change the t max . For comparison reasons with other papers [17,20], we used the combination of PG and isopropanol as the vehicle for further studies. The enhancement effects of PG and OA–PG on the permeation of HN were studied by occlusively pre-treating the skin with PG and OA–PG for 3 h. The application chambers were subsequently removed. HN (dissolved in PG–isopropanol) was applied 3 h after the removal of the treatments. The data obtained as shown in Table 3 are presented as ER in which the lag times after HN application on pretreated sites are divided by the lag times after application on untreated (control) site. A pre-treatment of PG alone for 3 h prior to the application of HN in PG–isopropanol yielded a clear reduction of t 0 and t max , as expressed in an ER value of ¯2. Thus, PG exerted a significant enhancement effect to the flux of HN across the skin barrier. Pre-treatment with OA–PG for 3 h reduced both t 0 and t max two-fold compared to the condition when no treatment was carried out. This enhancement is only slightly higher than PG alone. The results show how fast HN reached the blood vessels in the dermal layer and exerted the vasodilation effects, but do not imply that an increased amount of HN is penetrating.
4. Discussion
4.1. Irritation effect of oleic acid LDV has been successfully applied to measure the
101
irritation potential along with visual scoring [11,13,14]. In the objective assessment of patch test responses, LDV was able to differentiate between negative, doubtful, and positive reactions [21]. In this study, LDV is used to quantify the degree of irritation after the application of OA. Some reports on the skin irritation potential of OA on human skin were based on visual scoring and gave merely qualitative results. Pure OA caused minor or hardly detectable irritation [4,22]. OA did not cause irritation when applied to the skin in n-propanol [23]. Irritation was reported when OA was dissolved in ethanol [24] and PG [22]. The LDV measurements in this study showed a 2-fold higher irritation index after a 3-h occlusive treatment with OA–PG compared to an irritation index of 1.1 after the treatment with PG alone. Application for 24 h of OA–PG increased the irritation index even more, i.e. three times the effect of PG alone. Therefore, these results give a good evaluation of the irritation side-effect induced by topical application of OA. The duration of irritation, which is almost the same as the duration of the increase in transepidermal water loss as reported in [25], indicates a close connection between barrier perturbation and irritation reaction. Boelsma et al. observed that the toxicity of OA to the viable epidermal cells was controlled by the thickness of the stratum corneum [26]. The toxic effect can only be observed in the absence of stratum corneum (after stripping or in early growth period of cultured human skin). An in vitro diffusion study showed that OA was not capable of penetrating across human stratum corneum in a significant amount within 20 h [27], but there can still be a possibility that a very minute quantity of OA may diffuse into the viable epidermis in vivo and result in local perturbation. Although the perturbation does not always produce visible morphological damage, it may trigger a reaction of the viable skin layers, causing a higher blood flow, which can be assessed by LDV. This makes LDV a very useful tool in detecting minor / moderate irritation. On the other hand, OA caused a detectable reaction in the skin while it was remaining in the stratum corneum and exerting an enhancement effect, e.g. to transepidermal water flux. The fact that application of 100% PG for 3 or 24 h did not cause consistent observable reactions is in
102
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
agreement with an earlier finding [28]. Skin reactions after application of PG as a standard in patch tests should therefore be interpreted carefully.
4.2. Enhancement effect on the flux of hexyl nicotinate For the use of LDV in the study of percutaneous absorption, one important step to be taken at the beginning of the studies is the choice of parameters to be used for data comparison. It was noticed that the pharmacodynamic effect of nicotinates can reach saturation [18,29]. After comparing many parameters the lag-time, t 0 , was considered consistent enough to be employed further [18]. In our study we noticed that t max to a lesser extent is also suitable to support the t 0 data for comparison. However, the use of lag-times as parameters made methyl nicotinate a less ideal model penetrant, because it has a very short t 0 (in the range of several minutes [9]). The enhancement of the flux of nicotinates would make the t 0 even shorter and more difficult to be measured. Another nicotinic acid ester, hexyl nicotinate (HN), has therefore become more attractive, since its lagtimes are much longer [17,18,20]. In a preliminary study, we have used nicotinic acid and nicotinyl alcohol as alternatives. The increase of blood flow by nicotinic acid was minute and could be related to its low skin penetration [30,31]. Even with the enhancement of OA the amount of the compound reaching the blood vessels may not be sufficient to exert high activity and the clearance of blood flow, being faster than the diffusion of the substance, may cause the fast decay of the effect. Nicotinyl alcohol showed a short t 0 similar to methyl nicotinate and was therefore not suitable as a model drug for a skin permeation study. Another advantage of the longer t 0 of the model penetrant HN is the possibility of prolonging the application time which allows better standardization of the experimental protocols. Instead of 30 s, we used a 60 s application time of HN solution, which resulted in less error values. The influence of solvents / vehicles on the permeation kinetics of HN was investigated with solvents containing isopropanol or OA. HN showed the same onset time of erythema, when it was dissolved in PG alone or in PG–isopropanol. Addition of 5% volume
(0.16 M) of OA in PG slightly delayed the onset of action, t 0 , but did not change the maximum response time, t max . Kohli et al. [18] found that HN shows a shorter t 0 compared to methyl and ethyl nicotinates when dissolved in a polar solvent. They gave the explanation that in a polar solvent (monopropylene glycol) the lipophilicity of HN increases, its solubility decreases and its thermodynamic activity is thereby increased at any given fixed concentration, whereas in an apolar solvent the opposite is true. Addition of OA could result in a more apolar character of the solvent and thereby slightly decrease the release of HN. The contrary explanation for the increase of HN release can also be applied for the addition of isopropanol, which is more polar. However, these effects are found to be very small in magnitude and do not produce a significant influence. The t 0 and t max of HN were clearly reduced upon application on the skin site pretreated with PG alone. Pre-treatment of the skin with OA–PG for 3 h significantly reduced the t 0 and t max of HN, but the extent of reduction was similar to that of PG. The application of PG alone had little influence on transepidermal water loss (TEWL) [25], but it clearly enhanced the flux of HN. By contrast, OA increased the TEWL significantly to a greater extent than PG, but to a lesser extent in the case of HN. As the skin barrier function is believed to be primarily located in the intercellular lipid domains [32], the lipid phase possesses the main role to act as a barrier against water loss. It is logical to correlate the increase of TEWL with the perturbation of the intercellular lipids of the skin. Hence, it can be stated that the water flux follows the intercellular route. As a nicotinate ester with high lipophilicity [31], HN is expected to follow the intercellular route as well and not the transcellular route [33]. From the results of this study it is obvious that the application of PG alone has a notable influence on the flux of HN, but not on the flux of water. Apparently under this condition the penetration of HN follows different pathways compared to that of water. The fact that PG alone does not highly enhance TEWL shows that PG has only a little effect on the intercellular lipids. On the other hand, the fact that PG enhances the permeation of HN suggests that, under the influence of PG and under the conditions of the experimental
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
protocol, HN follows an additional route of penetration than through the intercellular lipids. The possibilities are the transcellular pathway (by modifying the cellular domain [34]) or a pathway along the cornified cell envelope. The possibility of transcellular pathway contradicts the results showing the absence of tracers in the cell interior after the treatment with various solvents (including PG), which can damage the cell envelope [35,36]. Instead, those results might indicate a diffusion along the cell envelope, in which a unique lipid–protein interaction (suggested in [37]) occurs and could be perturbed by PG to form a new pathway for HN. A study with freeze fracture electron microscopy showed that PG mainly accumulates in the border between the corneocytes and intercellular lipid domain [38], while a portion of it might well penetrate into the cells or into the intercellular lipid bilayers [39]. The presence of OA perturbs the intercellular lipids resulting in another additional enhancement to the flux of HN. The less dramatic enhancing capability of OA for HN suggests that OA is a less effective enhancer for lipophilic compounds.
5. Conclusions The skin barrier perturbing property of OA may induce two effects: (i) higher skin permeability for certain exogenous substances; (ii) irritation reaction. Both effects could be demonstrated using LDV. The duration of the perturbation effect of OA has been shown to be closely related to the duration of its irritation effects. The in vivo enhancement effect of PG could be clearly determined using HN as a model penetrant and vasodilation activity of HN as the end-point, which could be measured non-invasively by LDV. The addition of OA did not induce further enhancement in the permeation of HN as the effect of OA–PG is comparable to PG alone.
Acknowledgements We are indebted to Dr. Angela Anigbogu for reviewing this manuscript.
103
References [1] R.J. Scheuplein, I.H. Blank, Permeability of the skin, Physiol. Rev. 51 (1971) 702–747. [2] J.E. Shaw, J. Urquhart, Transdermal drug administration — a nuisance becomes an opportunity, Br. Med. J. 283 (1981) 875–876. [3] B.J. Aungst, Fatty acids as skin permeation enhancers, in: E.W. Smith, H.I. Maibach (Eds.), Percutaneous Penetration Enhancers, CRC Press, Boca Raton, FL, 1995, pp. 277–287. [4] P.G. Green, R.H. Guy, J. Hadgraft, In vitro and in vivo enhancement of skin permeation with oleic and lauric acids, Int. J. Pharm. 48 (1988) 103–111. ´ [5] H. Tanojo, J.A. Bouwstra, H.E. Junginger, H.E. Bodde, Effects of oleic acid on human transepidermal water loss using ethanol or propylene glycol as vehicles, in: K.R. Brain, V.J. James, K.A. Walters (Eds.), Prediction of Percutaneous Penetration, Vol. 3b, STS, Cardiff, 1993, pp. 319–324. [6] R. Kohli, W.I. Archer, J.M.C. Roberts, A.J. Cochran, A. Li Wan Po, Impedance measurements for the non-invasive monitoring of skin hydration: a reassessment, Int. J. Pharm. 26 (1985) 275–287. [7] V.H.W. Mak, R.O. Potts, R.H. Guy, Oleic acid concentration and effect in stratum corneum determined in vivo by infrared spectroscopy, J. Invest. Dermatol. 90 (1988) 584. [8] A.W. McKenzie, R.M.M. Atkinson, Topical activities of betamethasone esters in man, Arch. Dermatol. 89 (1964) 741–746. [9] R.H. Guy, R.C. Wester, E. Tur, H.I. Maibach, Noninvasive assessments of the percutaneous absorption of methyl nicotinate in humans, J. Pharm. Sci. 72 (1983) 1077–1079. [10] M.D. Stern, In vivo evaluation of microcirculation by coherent light scattering, Nature 254 (1975) 56–58. [11] G.E. Nilsson, U. Otto, J.E. Wahlberg, Assessment of skin irritancy in man by laser-Doppler flowmetry, Contact Dermatitis 8 (1982) 401–406. [12] T. Agner, Noninvasive measuring methods for the investigation of irritant patch test reactions. A study of patients with hand eczema, atopic dermatitis and controls, Acta Derm. Venereol. (Stockh.), Suppl. 178 (1992) 3–26. [13] R. Blanken, P.G.M. Van der Valk, J.P. Nater, Laser-Doppler flowmetry in the investigation of irritant compounds on human skin, Dermatosen 34 (1986) 5–9. [14] E. Berardesca, H.I. Maibach, Bioengineering and the patch test, Contact Dermatitis 18 (1988) 3–9. [15] G.P. Fulton, E.M. Farber, A.P. Moreci, The mechanism of action of rubefacients, J. Invest. Dermatol. 33 (1959) 317– 325. ´ d’esters de [16] V. Henschel, F. Jaminet, Absorption percutanee l’acide nicotinique Influence des excipients, J. Pharm. Belg. 27 (1972) 743–754. [17] K.S. Ryatt, J.M. Stevenson, H.I. Maibach, R.H. Guy, Pharmacodynamic measurement of percutaneous penetration enhancement in vivo, J. Pharm. Sci. 75 (1986) 374–377. [18] R. Kohli, W.I. Archer, A. Li Wan Po, Laser velocimetry for the non-invasive assessment of the percutaneous absorption of nicotinates, Int. J. Pharm. 36 (1987) 91–98.
104
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
[19] M. Meloni, M.C. Poelman, M. Lavazza, LDV assessment of methyl nicotinate biological response in aqueous solution against that of a w / o microemulsion system, Int. J. Cosm. Sci. 16 (1994) 257–264. [20] E. Oestmann, A.P.M. Lavrijsen, J. Hermans, M. Ponec, Skin barrier function in healthy volunteers as assessed by transepidermal water loss and vascular response to hexyl nicotinate: intra- and inter-individual variability, Br. J. Dermatol. 128 (1993) 130–136. [21] B. Staberg, P. Klem, J. Serup, Patch test responses evaluated by cutaneous blood flow measurements, Arch. Dermatol. 120 (1984) 741–743. [22] T. Loftsson, N. Gildersleeve, N. Bodor, The effect of vehicle additives on the transdermal delivery of nitroglycerin, Pharm. Res. 4 (1987) 436–437. [23] M.A. Stillman, H.I. Maibach, A.R. Shalita, Relative irritancy of free fatty acids of different chain length, Contact Dermatitis 1 (1975) 65–69. [24] P.J. Frosch, A.M. Kligman, The chamber-scarification test for irritancy, Contact Dermatitis 2 (1976) 314–324. ´ In vivo human skin [25] H. Tanojo, H.E. Junginger, H.E. Bodde, permeability enhancement by oleic acid: Transepidermal water loss and Fourier-transform infrared spectroscopy studies, J. Controlled Rel. 47 (1997) 31–39. ´ M. Ponec, Assessment [26] E. Boelsma, H. Tanojo, H.E. Bodde, of the potential irritancy of oleic acid on human skin: Evaluation in vitro and in vivo, Toxic. in Vitro 10 (1996) 729–742. [27] H. Tanojo, E. Boelsma, H.E. Junginger, M. Ponec, H.E. ´ In vitro human skin permeability enhancement by Bodde, oleic acid and propylene glycol: Permeation of p-aminobenzoic acid and oleic acid across stratum corneum, submitted for publication. [28] J.O. Funk, H.I. Maibach, Propylene glycol dermatitis: reevaluation of an old problem, Contact Dermatitis 31 (1994) 236–241. [29] R.H. Guy, E. Tur, B. Bugatto, C. Gaebel, L.B. Sheiner, H.I. Maibach, Pharmacodynamic measurements of methyl nicotinate percutaneous absorption, Pharm. Res. 1 (1984) 76–81.
[30] H. Osamura, Penetration of topical corticosteroids through human epidermis, J. Dermatol. 9 (1982) 45–58. ¨ [31] R.H. Guy, E.M. Carlstrom, D.A. Bucks, R.S. Hinz, H.I. Maibach, Percutaneous penetration of nicotinates: in vivo and in vitro measurements, J. Pharm. Sci. 75 (1986) 968– 972. [32] P.M. Elias, Lipids and the epidermal permeability barrier, Arch. Dermatol. Res. 120 (1981) 95–117. [33] W.J. Albery, J. Hadgraft, Percutaneous absorption: in vivo experiments, J. Pharm. Pharmacol. 31 (1979) 140–147. [34] J.L. Zatz, U.G. Dalvi, Evaluation of solvent–skin interaction in percutaneous absorption, J. Soc. Cosmet. Chem. 34 (1983) 327–334. [35] M.L. Williams, P.M. Elias, The extracellular matrix of stratum corneum: role of lipids in normal and pathological function, Crit. Rev. Ther. Drug Carrier Syst. 3 (1987) 95– 122. [36] O. Simonetti, A.J. Hoogstraate, W. Bialik, J.A. Kempenaar, ´ M. Ponec, Visualization of A.H.G.J. Schrijvers, H.E. Bodde, diffusion pathways across the stratum corneum of native and in-vitro-reconstructed epidermis by confocal laser scanning microscopy, Arch. Dermatol. Res. 287 (1995) 465–473. [37] R.O. Potts, Physical characterization of the stratum corneum: the relationship of mechanical and barrier properties to lipid and protein structure, in: J. Hadgraft, R.H. Guy (Eds.), Transdermal Drug Delivery: Developmental Issues and Research Initiatives, Marcel Dekker, New York, 1989, pp. 23–57. [38] A.J. Hoogstraate, J. Verhoef, J. Brusee, A.P. IJzerman, F. ´ Kinetics, ultrastructural aspects and Spies, H.E. Bodde, molecular modelling of transdermal peptide flux enhancement by N-alkylazacycloheptanones, Int. J. Pharm. 76 (1991) 37–47. [39] J.A. Bouwstra, M.A. Salomons-de Vries, B.A.I. van den Bergh, G.S. Gooris, Changes in lipid organisation of the skin barrier by N-alkyl-azocycloheptanones: a visualisation and X-ray diffraction study, Int. J. Pharm. 144 (1996) 81–89.
In vivo human skin permeability enhancement by oleic acid: a laser Doppler velocimetry study Hanafi Tanojo a , *, Esther Boelsma b , Hans E. Junginger c , Maria Ponec b , Harry E. Bodde´ c 1 a
Department of Dermatology, University of California, Box 0989, 90 Medical Centerway, Surge 110, San Francisco, CA 94143 -0989, USA b Department of Dermatology, Leiden University Hospital, Leiden, Netherlands. c Division Pharmaceutical Technology, Leiden /Amsterdam Center for Drug Research, Leiden University, Leiden, Netherlands Received 24 April 1998; received in revised form 2 August 1998; accepted 10 August 1998
Abstract Topical application of a skin permeation enhancer such as oleic acid (OA) can result in: (i) higher skin permeability for many exogenous substances (ii) an irritation reaction. Laser Doppler velocimetry (LDV) is one of few techniques which can assess both effects using appropriate protocols. Direct LDV measurement, measuring cutaneous blood flow, has been preferred as a tool to evaluate skin irritation. By comparing the LDV value of an irritant-treated site with an untreated site, an irritation index for the irritant can be obtained. Occlusive application of 0.16 M OA in propylene glycol (PG) for either 3 or 24 h produced irritation in form of redness and slight swelling. Using LDV, we obtained an irritation index of 2 and 4, respectively. The vehicle, PG alone, produced an index of around 1, which corresponded well to the slight to almost no irritation observed visually. The duration of the high irritation index assessed by LDV after OA–PG application is comparable to the duration of the increase in transepidermal water loss following the same application. This indicates a correlation between skin irritation and barrier perturbation caused by OA. LDV can also be used to elucidate the effect of enhancers on skin using hexyl nicotinate (HN) as a model drug, since its vasodilative effect can be clearly assessed by LDV. Pre-treatment of PG for 3 h significantly reduced the t 0 and t max of HN. No further reduction could be observed when OA was added into PG, suggesting that OA–PG is not more effective than PG alone in enhancing the permeation of HN. 1999 Elsevier Science B.V. All rights reserved. Keywords: Skin permeability; Oleic acid; Topical application; Laser-Doppler velocimetry; Human, in vivo study
1. Introduction Skin gives protection to the body against penetration of most substances from outside, owing to its
*Corresponding author. Tel.: 11 415 4764997; fax: 11 415 7535304; e-mail: [email protected] 1 Deceased September 8, 1996.
barrier capacity. This barrier resides mainly in the uppermost layer, the stratum corneum [1]. Many exogenous substances have difficulty passing the stratum corneum to reach the blood circulation system. On the other hand, some compounds — so-called skin permeation enhancers — can improve the penetration of other substances by perturbing the barrier function of the stratum corneum. Many of them have been the subject of investigation for the
0168-3659 / 99 / $ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 98 )00144-8
98
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
development of transdermal drug delivery systems [2]. Several types of penetration enhancers especially fatty acids have gained much attention, because they are natural components of the skin lipids and therefore are not foreign to the body. Oleic acid (OA) is one of the fatty acids found abundantly in nature, including in the human skin. Its capacity as a skin penetration enhancer has been generally recognized in various in vitro studies [3]. In the last decade numerous in vivo studies with human subjects have been conducted in order to establish non-invasive techniques applicable for a large number of enhancers to measure the skin penetration enhancements in living subjects. In vivo techniques, such as transepidermal water loss measurements [4,5], skin impedance measurements [6], infrared spectroscopy (in combination with attenuated total reflectance) [7], skin blanching method for steroids [8], photopulse plethysmography [9], and laser-Doppler velocimetry (LDV), can be used to elucidate the capacity of enhancers. Some of these techniques have employed OA as a model enhancer [4,5,7]. In this paper, we focus on the use of LDV to assess the capacity of oleic acid as enhancer and its effect on the skin. LDV technique measures the skin blood flow. The working mechanism of this instrument is based on the measurement of a Doppler-shift in wavelength [10,11]. When applied to monitoring cutaneous blood flow, the coherent light impinging on the stationary tissues is reflected (scattered) back at the same frequency, but light reflected by moving structures (blood cells) undergoes a frequency shift. This phenomenon is known as Doppler effect. The end result is a spectral broadening of the almost monochromatic laser light applied. Part of the back scattered radiation / light is directed towards suitable photodetectors for subsequent electronic translation into electrical signals. The magnitude of the frequency shift can be related to the velocity of the moving particles, thus providing the basis for relative quantification, in which it is proportional to the number of blood cells multiplied by their velocity. The radiation penetrates to a depth of 1–1.5 mm, which reaches the capillary blood vessels in the dermis passing through the epidermis. In the beginning LDV was used along with visual scoring in the differentiation of irritation. A positive
relationship has been found between the dose of sodium lauryl sulfate (SLS) applied to the skin and blood flow values [11,12]. Lately, LDV has gained preference over visual scoring [13,14]. The use of LDV for the skin permeation study began when it was employed to measure the penetration of methyl nicotinate across the stratum corneum / epidermis to exert erythema as a result of vasodilation on cutaneous capillary blood vessels yielding increase in skin blood flow [9]. The vasodilation is triggered by direct action of the nicotinates on the blood vessel walls at the junction between the dermis and the epidermis [15]. The time of onset of erythema is found to be reproducible for the same experimental conditions [16]. Subsequently, the vasodilation capability of nicotinates was used to measure the penetration enhancement of pyrrolidone derivates [17] and various organic solvents, including alcohols, glycols and kerosine [18] as well as to characterize the drug release from a microemulsion [19].
2. Material and methods
2.1. Pre-experimental conditions Twelve subjects involved in this study were of both sexes, had normal, healthy skin and were fully informed of the nature of the study and the procedures involved. The age range was 20–40 years. This in vivo study was approved by the local ethical committee for human studies. Individuals had to rest for 20–30 min before LDV measurements, with the skin at the measuring site left uncovered, in an air-conditioned room, under constant temperature (20.5618C) and humidity (30–50%) and were kept relaxed throughout the procedure. Different treatments were applied randomly to the subjects. Only LDV values from the same anatomical area were expected to be comparable, so in all experiments only the median part of volar forearm was chosen as the site of applications and LDV measurements. Heat sources and air-draughts in the direction of the individuals were avoided.
2.2. Application of treatment solutions The occlusive application system consisted of a
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
99
In subsequent studies, the application of hexyl nicotinate (Sigma, St. Louis, MO, USA) solution followed by LDV measurements was performed 3 h after the removal of the application chambers (only for the 3-h applications).
2.3. LDV measurements
Fig. 1. Assembly of the occlusive application system.
Hill-Top chamber (Hill-Top, Cincinnati, OH, USA), Leukosilk adhesive tape (Beiersdorf, Hamburg, Germany) and plastic tape (3M, Borken, Germany), assembled as shown in Fig. 1. For each subject, four sites were marked on the volar forearm as shown in Fig. 2. One empty chamber was placed on site 1 and site 4 was left open without any treatment; both acted as blank controls. On site 2 a chamber containing 300 ml propylene glycol (J.T. Baker, Deventer, Netherlands) was attached. On site 3, the same procedure was repeated for 300 ml solution of 0.16 M oleic acid (Brocacef, Maarssen, Netherlands) in PG. Depending on the protocol, 3 or 24 h after the application, the chambers were removed. The application sites were wiped gently with a soft tissue paper but not stripped. LDV measurements were made immediately on the sites.
LDV was measured on the application sites on subject’s forearm using a Diodopp laser Doppler flowmeter (Applied Laser Technology, Maarheeze, Netherlands). The measuring probe contains a diode laser with a wavelength of 780 nm. The instrument was allowed to ‘warm-up’ for 15 min after being turned on, giving time for the laser light to stabilize and was never switched off between intermittent measurements during the day. Prior to the measurements the offset button was used to bring the value to a zero level and never used again during the same day of the experiments. Laser safety directions were followed. The probe was held gently on the skin to avoid vascular compression. Readings were performed after stabilization of the output signal. Only LDV values from the same anatomical area were expected to be comparable, so in all experiments only the median part of volar forearm was chosen for the sites of applications and LDV measurements. The LDV values of treated sites were directly compared to those obtained on the untreated area, i.e. area left completely open (application site 4), that were measured approximately at the same time on the same subject, to obtain an irritation index. Having determined the index for each individual subject, the overall average index and standard deviation were calculated. Statistical analysis of the data was performed using the Student’s t-test.
2.4. Application of hexyl nicotinate solution followed by LDV measurements
Fig. 2. Application sites on the volar forearm of the human volunteers.
Hexyl nicotinate (HN, 20 ml, 10 mM) in one of these solvents: (i) propylene glycol (PG); (ii) PG– isopropanol (60:40, v / v); (iii) 0.16 M oleic acid in PG (OA–PG), was applied for 60 s on the forearm using a filter paper disk. The technique has already been used [20] with a modification of the one applied in earlier studies [17,18]. The nicotinate-induced increase in skin blood flow was assessed continuous-
100
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
Fig. 3. Schematic curve of the vasodilative effect of hexyl nicotinate as recorded by LDV measurements.
ly using LDV by fixing the probe on the skin surface using a plastic probe holder. From the curves obtained, the onset time of action (t 0 ) and the time to reach maximum response (t max ) were determined (see Fig. 3). The division of the lag times of the untreated site by the lag times of the applied sites is presented as the enhancement ratio (ER) for the applied substances. Having determined the ER for each individual subject, the overall average ER and standard deviation were calculated. Statistical analysis of the data was performed using the Student’s t-test.
3. Results
3.1. Direct assessment of LDV on irritation effects Table 1 shows the irritation index of PG and OA–PG following a 3- or 24-h occlusive application. Following the occlusive application of OA–PG for a Table 1 LDV measurements after topical application of propylene glycol (PG) and oleic acid in PG (OA–PG) presented as irritation index in comparison to untreated site Application
3-h Application
24-h Application
PG 0.16 M OA–PG
1.160.4 (n56) 2.160.9 (n56)
1.260.3 (n510) 3.961.5 (n510)
Irritation index was calculated by dividing the LDV value of untreated site with the value of pretreated site (6standard deviation). Significant differences: PG to OA–PG, P,0.05; OA–PG 3-h to 24-h, P,0.05.
short period of time, it was noticed that visually the irritation did not always appear instantly after the application chambers were removed. However, LDV measurements recorded a consistent higher value of blood flow compared to the untreated site. This again proves the accuracy of this technique in comparison to visual scoring. A 3-h application of OA–PG caused a 2-fold increase in skin blood flow compared to the slight increase by PG alone. The degree of irritation was expressed as irritation index, calculated by dividing the LDV value at the treated site with the value at the non-treated site (control). Application for 24 h produced an irritation index of 4, while the application of pure PG yielded only slight erythema and an irritation index close to 1. The 24-h OA–PG applications produced a clearly visible irritation, which lasted longer than after 3-h applications before it completely disappeared. We could observe the diminishing of the erythema and the high blood flow on the OA–PG treated site up to 54 h after the removal of a 24-h application and up to 24 h for a 3-h application.
3.2. Assessment of LDV after the application of hexyl nicotinate In a preliminary study the influence of various vehicles on the flux of HN was compared (Table 2). The onset of action time, t 0 , and the maximum response time, t max , of the vasodilation assessed by LDV were used as parameters. The values of erythema caused by vasodilative effects of HN were much higher than those due to irritation caused by the vehicles. Hence, these two conditions are clearly distinguishable. Either dissolved in PG or in PG–isopropanol (60:40, v / v), the vasodilation effect of HN began to emerge and maximize at about the same time. The Table 2 Influence of solvents on the lag-times of hexyl nicotinate (10 mM) from the time of application to the initial (t 0 ) and maximal (t max ) responses (6standard deviation) Hexyl nicotinate 10 mM in
t 0 (min)
t max (min)
Propylene glycol (PG) PG–isopropanol (6:4, v / v) 0.16 M Oleic acid in PG
12.563.8 (n54) 12.463.2 (n54) 14.861.6 (n53)
20.163.6 (n54) 19.164.9 (n54) 21.266.9 (n53)
No significant difference among the lag-times: P.0.05.
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104 Table 3 Influence of 3-h occlusive pretreatments on the lag-time between application of hexyl nicotinate [10 mM in propylene glycol– isopropanol (60:40, v / v)] and initial or maximal responses (t 0 or t max ) Pre-treatment
ER t 0
ER t max
Propylene glycol (PG) 0.16 M Oleic acid in PG
1.860.3 (n55) 2.060.4 (n55)
1.660.4 (n55) 1.760.3 (n55)
ER, enhancement ratio, calculated by dividing the lag-time of untreated site with the lag-time of pretreated site (6standard deviation). No significant difference between PG and oleic acid in PG: t 0 and t max , P.0.05.
use of OA–PG as vehicle slightly increased the t 0 (not significant) compared to the other vehicles, but did not change the t max . For comparison reasons with other papers [17,20], we used the combination of PG and isopropanol as the vehicle for further studies. The enhancement effects of PG and OA–PG on the permeation of HN were studied by occlusively pre-treating the skin with PG and OA–PG for 3 h. The application chambers were subsequently removed. HN (dissolved in PG–isopropanol) was applied 3 h after the removal of the treatments. The data obtained as shown in Table 3 are presented as ER in which the lag times after HN application on pretreated sites are divided by the lag times after application on untreated (control) site. A pre-treatment of PG alone for 3 h prior to the application of HN in PG–isopropanol yielded a clear reduction of t 0 and t max , as expressed in an ER value of ¯2. Thus, PG exerted a significant enhancement effect to the flux of HN across the skin barrier. Pre-treatment with OA–PG for 3 h reduced both t 0 and t max two-fold compared to the condition when no treatment was carried out. This enhancement is only slightly higher than PG alone. The results show how fast HN reached the blood vessels in the dermal layer and exerted the vasodilation effects, but do not imply that an increased amount of HN is penetrating.
4. Discussion
4.1. Irritation effect of oleic acid LDV has been successfully applied to measure the
101
irritation potential along with visual scoring [11,13,14]. In the objective assessment of patch test responses, LDV was able to differentiate between negative, doubtful, and positive reactions [21]. In this study, LDV is used to quantify the degree of irritation after the application of OA. Some reports on the skin irritation potential of OA on human skin were based on visual scoring and gave merely qualitative results. Pure OA caused minor or hardly detectable irritation [4,22]. OA did not cause irritation when applied to the skin in n-propanol [23]. Irritation was reported when OA was dissolved in ethanol [24] and PG [22]. The LDV measurements in this study showed a 2-fold higher irritation index after a 3-h occlusive treatment with OA–PG compared to an irritation index of 1.1 after the treatment with PG alone. Application for 24 h of OA–PG increased the irritation index even more, i.e. three times the effect of PG alone. Therefore, these results give a good evaluation of the irritation side-effect induced by topical application of OA. The duration of irritation, which is almost the same as the duration of the increase in transepidermal water loss as reported in [25], indicates a close connection between barrier perturbation and irritation reaction. Boelsma et al. observed that the toxicity of OA to the viable epidermal cells was controlled by the thickness of the stratum corneum [26]. The toxic effect can only be observed in the absence of stratum corneum (after stripping or in early growth period of cultured human skin). An in vitro diffusion study showed that OA was not capable of penetrating across human stratum corneum in a significant amount within 20 h [27], but there can still be a possibility that a very minute quantity of OA may diffuse into the viable epidermis in vivo and result in local perturbation. Although the perturbation does not always produce visible morphological damage, it may trigger a reaction of the viable skin layers, causing a higher blood flow, which can be assessed by LDV. This makes LDV a very useful tool in detecting minor / moderate irritation. On the other hand, OA caused a detectable reaction in the skin while it was remaining in the stratum corneum and exerting an enhancement effect, e.g. to transepidermal water flux. The fact that application of 100% PG for 3 or 24 h did not cause consistent observable reactions is in
102
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
agreement with an earlier finding [28]. Skin reactions after application of PG as a standard in patch tests should therefore be interpreted carefully.
4.2. Enhancement effect on the flux of hexyl nicotinate For the use of LDV in the study of percutaneous absorption, one important step to be taken at the beginning of the studies is the choice of parameters to be used for data comparison. It was noticed that the pharmacodynamic effect of nicotinates can reach saturation [18,29]. After comparing many parameters the lag-time, t 0 , was considered consistent enough to be employed further [18]. In our study we noticed that t max to a lesser extent is also suitable to support the t 0 data for comparison. However, the use of lag-times as parameters made methyl nicotinate a less ideal model penetrant, because it has a very short t 0 (in the range of several minutes [9]). The enhancement of the flux of nicotinates would make the t 0 even shorter and more difficult to be measured. Another nicotinic acid ester, hexyl nicotinate (HN), has therefore become more attractive, since its lagtimes are much longer [17,18,20]. In a preliminary study, we have used nicotinic acid and nicotinyl alcohol as alternatives. The increase of blood flow by nicotinic acid was minute and could be related to its low skin penetration [30,31]. Even with the enhancement of OA the amount of the compound reaching the blood vessels may not be sufficient to exert high activity and the clearance of blood flow, being faster than the diffusion of the substance, may cause the fast decay of the effect. Nicotinyl alcohol showed a short t 0 similar to methyl nicotinate and was therefore not suitable as a model drug for a skin permeation study. Another advantage of the longer t 0 of the model penetrant HN is the possibility of prolonging the application time which allows better standardization of the experimental protocols. Instead of 30 s, we used a 60 s application time of HN solution, which resulted in less error values. The influence of solvents / vehicles on the permeation kinetics of HN was investigated with solvents containing isopropanol or OA. HN showed the same onset time of erythema, when it was dissolved in PG alone or in PG–isopropanol. Addition of 5% volume
(0.16 M) of OA in PG slightly delayed the onset of action, t 0 , but did not change the maximum response time, t max . Kohli et al. [18] found that HN shows a shorter t 0 compared to methyl and ethyl nicotinates when dissolved in a polar solvent. They gave the explanation that in a polar solvent (monopropylene glycol) the lipophilicity of HN increases, its solubility decreases and its thermodynamic activity is thereby increased at any given fixed concentration, whereas in an apolar solvent the opposite is true. Addition of OA could result in a more apolar character of the solvent and thereby slightly decrease the release of HN. The contrary explanation for the increase of HN release can also be applied for the addition of isopropanol, which is more polar. However, these effects are found to be very small in magnitude and do not produce a significant influence. The t 0 and t max of HN were clearly reduced upon application on the skin site pretreated with PG alone. Pre-treatment of the skin with OA–PG for 3 h significantly reduced the t 0 and t max of HN, but the extent of reduction was similar to that of PG. The application of PG alone had little influence on transepidermal water loss (TEWL) [25], but it clearly enhanced the flux of HN. By contrast, OA increased the TEWL significantly to a greater extent than PG, but to a lesser extent in the case of HN. As the skin barrier function is believed to be primarily located in the intercellular lipid domains [32], the lipid phase possesses the main role to act as a barrier against water loss. It is logical to correlate the increase of TEWL with the perturbation of the intercellular lipids of the skin. Hence, it can be stated that the water flux follows the intercellular route. As a nicotinate ester with high lipophilicity [31], HN is expected to follow the intercellular route as well and not the transcellular route [33]. From the results of this study it is obvious that the application of PG alone has a notable influence on the flux of HN, but not on the flux of water. Apparently under this condition the penetration of HN follows different pathways compared to that of water. The fact that PG alone does not highly enhance TEWL shows that PG has only a little effect on the intercellular lipids. On the other hand, the fact that PG enhances the permeation of HN suggests that, under the influence of PG and under the conditions of the experimental
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
protocol, HN follows an additional route of penetration than through the intercellular lipids. The possibilities are the transcellular pathway (by modifying the cellular domain [34]) or a pathway along the cornified cell envelope. The possibility of transcellular pathway contradicts the results showing the absence of tracers in the cell interior after the treatment with various solvents (including PG), which can damage the cell envelope [35,36]. Instead, those results might indicate a diffusion along the cell envelope, in which a unique lipid–protein interaction (suggested in [37]) occurs and could be perturbed by PG to form a new pathway for HN. A study with freeze fracture electron microscopy showed that PG mainly accumulates in the border between the corneocytes and intercellular lipid domain [38], while a portion of it might well penetrate into the cells or into the intercellular lipid bilayers [39]. The presence of OA perturbs the intercellular lipids resulting in another additional enhancement to the flux of HN. The less dramatic enhancing capability of OA for HN suggests that OA is a less effective enhancer for lipophilic compounds.
5. Conclusions The skin barrier perturbing property of OA may induce two effects: (i) higher skin permeability for certain exogenous substances; (ii) irritation reaction. Both effects could be demonstrated using LDV. The duration of the perturbation effect of OA has been shown to be closely related to the duration of its irritation effects. The in vivo enhancement effect of PG could be clearly determined using HN as a model penetrant and vasodilation activity of HN as the end-point, which could be measured non-invasively by LDV. The addition of OA did not induce further enhancement in the permeation of HN as the effect of OA–PG is comparable to PG alone.
Acknowledgements We are indebted to Dr. Angela Anigbogu for reviewing this manuscript.
103
References [1] R.J. Scheuplein, I.H. Blank, Permeability of the skin, Physiol. Rev. 51 (1971) 702–747. [2] J.E. Shaw, J. Urquhart, Transdermal drug administration — a nuisance becomes an opportunity, Br. Med. J. 283 (1981) 875–876. [3] B.J. Aungst, Fatty acids as skin permeation enhancers, in: E.W. Smith, H.I. Maibach (Eds.), Percutaneous Penetration Enhancers, CRC Press, Boca Raton, FL, 1995, pp. 277–287. [4] P.G. Green, R.H. Guy, J. Hadgraft, In vitro and in vivo enhancement of skin permeation with oleic and lauric acids, Int. J. Pharm. 48 (1988) 103–111. ´ [5] H. Tanojo, J.A. Bouwstra, H.E. Junginger, H.E. Bodde, Effects of oleic acid on human transepidermal water loss using ethanol or propylene glycol as vehicles, in: K.R. Brain, V.J. James, K.A. Walters (Eds.), Prediction of Percutaneous Penetration, Vol. 3b, STS, Cardiff, 1993, pp. 319–324. [6] R. Kohli, W.I. Archer, J.M.C. Roberts, A.J. Cochran, A. Li Wan Po, Impedance measurements for the non-invasive monitoring of skin hydration: a reassessment, Int. J. Pharm. 26 (1985) 275–287. [7] V.H.W. Mak, R.O. Potts, R.H. Guy, Oleic acid concentration and effect in stratum corneum determined in vivo by infrared spectroscopy, J. Invest. Dermatol. 90 (1988) 584. [8] A.W. McKenzie, R.M.M. Atkinson, Topical activities of betamethasone esters in man, Arch. Dermatol. 89 (1964) 741–746. [9] R.H. Guy, R.C. Wester, E. Tur, H.I. Maibach, Noninvasive assessments of the percutaneous absorption of methyl nicotinate in humans, J. Pharm. Sci. 72 (1983) 1077–1079. [10] M.D. Stern, In vivo evaluation of microcirculation by coherent light scattering, Nature 254 (1975) 56–58. [11] G.E. Nilsson, U. Otto, J.E. Wahlberg, Assessment of skin irritancy in man by laser-Doppler flowmetry, Contact Dermatitis 8 (1982) 401–406. [12] T. Agner, Noninvasive measuring methods for the investigation of irritant patch test reactions. A study of patients with hand eczema, atopic dermatitis and controls, Acta Derm. Venereol. (Stockh.), Suppl. 178 (1992) 3–26. [13] R. Blanken, P.G.M. Van der Valk, J.P. Nater, Laser-Doppler flowmetry in the investigation of irritant compounds on human skin, Dermatosen 34 (1986) 5–9. [14] E. Berardesca, H.I. Maibach, Bioengineering and the patch test, Contact Dermatitis 18 (1988) 3–9. [15] G.P. Fulton, E.M. Farber, A.P. Moreci, The mechanism of action of rubefacients, J. Invest. Dermatol. 33 (1959) 317– 325. ´ d’esters de [16] V. Henschel, F. Jaminet, Absorption percutanee l’acide nicotinique Influence des excipients, J. Pharm. Belg. 27 (1972) 743–754. [17] K.S. Ryatt, J.M. Stevenson, H.I. Maibach, R.H. Guy, Pharmacodynamic measurement of percutaneous penetration enhancement in vivo, J. Pharm. Sci. 75 (1986) 374–377. [18] R. Kohli, W.I. Archer, A. Li Wan Po, Laser velocimetry for the non-invasive assessment of the percutaneous absorption of nicotinates, Int. J. Pharm. 36 (1987) 91–98.
104
H. Tanojo et al. / Journal of Controlled Release 58 (1999) 97 – 104
[19] M. Meloni, M.C. Poelman, M. Lavazza, LDV assessment of methyl nicotinate biological response in aqueous solution against that of a w / o microemulsion system, Int. J. Cosm. Sci. 16 (1994) 257–264. [20] E. Oestmann, A.P.M. Lavrijsen, J. Hermans, M. Ponec, Skin barrier function in healthy volunteers as assessed by transepidermal water loss and vascular response to hexyl nicotinate: intra- and inter-individual variability, Br. J. Dermatol. 128 (1993) 130–136. [21] B. Staberg, P. Klem, J. Serup, Patch test responses evaluated by cutaneous blood flow measurements, Arch. Dermatol. 120 (1984) 741–743. [22] T. Loftsson, N. Gildersleeve, N. Bodor, The effect of vehicle additives on the transdermal delivery of nitroglycerin, Pharm. Res. 4 (1987) 436–437. [23] M.A. Stillman, H.I. Maibach, A.R. Shalita, Relative irritancy of free fatty acids of different chain length, Contact Dermatitis 1 (1975) 65–69. [24] P.J. Frosch, A.M. Kligman, The chamber-scarification test for irritancy, Contact Dermatitis 2 (1976) 314–324. ´ In vivo human skin [25] H. Tanojo, H.E. Junginger, H.E. Bodde, permeability enhancement by oleic acid: Transepidermal water loss and Fourier-transform infrared spectroscopy studies, J. Controlled Rel. 47 (1997) 31–39. ´ M. Ponec, Assessment [26] E. Boelsma, H. Tanojo, H.E. Bodde, of the potential irritancy of oleic acid on human skin: Evaluation in vitro and in vivo, Toxic. in Vitro 10 (1996) 729–742. [27] H. Tanojo, E. Boelsma, H.E. Junginger, M. Ponec, H.E. ´ In vitro human skin permeability enhancement by Bodde, oleic acid and propylene glycol: Permeation of p-aminobenzoic acid and oleic acid across stratum corneum, submitted for publication. [28] J.O. Funk, H.I. Maibach, Propylene glycol dermatitis: reevaluation of an old problem, Contact Dermatitis 31 (1994) 236–241. [29] R.H. Guy, E. Tur, B. Bugatto, C. Gaebel, L.B. Sheiner, H.I. Maibach, Pharmacodynamic measurements of methyl nicotinate percutaneous absorption, Pharm. Res. 1 (1984) 76–81.
[30] H. Osamura, Penetration of topical corticosteroids through human epidermis, J. Dermatol. 9 (1982) 45–58. ¨ [31] R.H. Guy, E.M. Carlstrom, D.A. Bucks, R.S. Hinz, H.I. Maibach, Percutaneous penetration of nicotinates: in vivo and in vitro measurements, J. Pharm. Sci. 75 (1986) 968– 972. [32] P.M. Elias, Lipids and the epidermal permeability barrier, Arch. Dermatol. Res. 120 (1981) 95–117. [33] W.J. Albery, J. Hadgraft, Percutaneous absorption: in vivo experiments, J. Pharm. Pharmacol. 31 (1979) 140–147. [34] J.L. Zatz, U.G. Dalvi, Evaluation of solvent–skin interaction in percutaneous absorption, J. Soc. Cosmet. Chem. 34 (1983) 327–334. [35] M.L. Williams, P.M. Elias, The extracellular matrix of stratum corneum: role of lipids in normal and pathological function, Crit. Rev. Ther. Drug Carrier Syst. 3 (1987) 95– 122. [36] O. Simonetti, A.J. Hoogstraate, W. Bialik, J.A. Kempenaar, ´ M. Ponec, Visualization of A.H.G.J. Schrijvers, H.E. Bodde, diffusion pathways across the stratum corneum of native and in-vitro-reconstructed epidermis by confocal laser scanning microscopy, Arch. Dermatol. Res. 287 (1995) 465–473. [37] R.O. Potts, Physical characterization of the stratum corneum: the relationship of mechanical and barrier properties to lipid and protein structure, in: J. Hadgraft, R.H. Guy (Eds.), Transdermal Drug Delivery: Developmental Issues and Research Initiatives, Marcel Dekker, New York, 1989, pp. 23–57. [38] A.J. Hoogstraate, J. Verhoef, J. Brusee, A.P. IJzerman, F. ´ Kinetics, ultrastructural aspects and Spies, H.E. Bodde, molecular modelling of transdermal peptide flux enhancement by N-alkylazacycloheptanones, Int. J. Pharm. 76 (1991) 37–47. [39] J.A. Bouwstra, M.A. Salomons-de Vries, B.A.I. van den Bergh, G.S. Gooris, Changes in lipid organisation of the skin barrier by N-alkyl-azocycloheptanones: a visualisation and X-ray diffraction study, Int. J. Pharm. 144 (1996) 81–89.