Synergistic Effect of Low‐Frequency Ultrasound and Sodium Lauryl Sulfate on Transdermal Transport

Synergistic Effect of Low-Frequency Ultrasound and Sodium Lauryl Sulfate on Transdermal Transport SAMIR MITRAGOTRI,1 DEBANJAN RAY,2 JOANNE FARRELL,2 HUA TANG,2 BETTY YU,2 JOSEPH KOST,2 DANIEL BLANKSCHTEIN,2 ROBERT LANGER2 1

Department of Chemical Engineering, University of California, Santa Barbara, California 93016

2

Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139

Received 17 May 1999; revised 15 December 1999; accepted 25 February 2000

Application of low-frequency ultrasound has been shown to enhance transdermal transport of drugs (low-frequency sonophoresis). In this paper, we show that the efficacy of low-frequency ultrasound in enhancing transdermal transport can be further increased by its combination with sodium lauryl sulfate (SLS), a well-known surfactant. The dependence of the ultrasound–SLS-mediated transport on ultrasound parameters, including intensity, net exposure time, and duty cycle, is discussed. The transdermal transport enhancement is proportional to ultrasound intensity as well as to exposure time, and is independent of duty cycle as long as the net exposure time is the same. The synergistic effect of SLS and ultrasound on transdermal transport increases linearly with SLS concentration. The enhancement is also proportional to the ultrasound energy density beyond a threshold value, which suggests that a certain minimum amount of energy density is required before noticeable changes in skin permeability occur. A similar dependence of the transdermal transport enhancement on energy density is observed in the case of the enhancement induced by ultrasound alone. Although the threshold energy density value in the presence of SLS is about 10 times lower than that in the case of ultrasound alone, the relationship between enhancement and energy density in the presence and in the absence of SLS is otherwise similar. Possible mechanisms for the synergistic effect of ultrasound and SLS are also discussed. © 2000 Wiley-

ABSTRACT:

Liss, Inc. and the American Pharmaceutical Association J Pharm Sci 89: 892–900, 2000

Keywords: sonophoresis; surfactant; transdermal; synergistic; ultrasound

INTRODUCTION Transdermal drug delivery offers an advantageous alternative to injections, although its applications are limited by the low skin permeability.1 Various methods have been attempted to enhance transdermal drug transport, including the use of (i) chemical enhancers,2 (ii) therapeutic,3,4 highfrequency,5,6 and low-frequency ultrasound (sonoCorrespondence to: S. Mitragotri (Telephone: 805-8937532; Fax: 805-893-4731; E-mail: [email protected]. edu) Journal of Pharmaceutical Sciences, Vol. 89, 892–900 (2000) © 2000 Wiley-Liss, Inc. and the American Pharmaceutical Association

892

phoresis),7,8 and (iii) electric field (iontophoresis9 and electroporation10). Although each of these methods has been shown to enhance transdermal drug transport by itself, some methods have been shown to work synergistically when applied simultaneously. For example, iontophoresis has been shown to operate in synergy with electroporation11 or with chemical enhancers.9 In addition, ultrasound under therapeutic conditions (frequency ∼1 MHz) has been shown to enhance transdermal drug transport synergistically with electroporation12 or with chemical enhancers such as linoleic acid.13 In this paper, we show that sonophoresis under low-frequency conditions (∼20 kHz) enhances

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 7, JULY 2000

SYNERGISTIC EFFECTS OF ULTRASOUND AND SURFACTANT ON TRANSDERMAL TRANSPORT

transdermal transport synergistically with sodium lauryl sulfate (SLS, C12H25SO4−Na+), a well-known surfactant. SLS was chosen as a model surfactant because its effects on skin permeability are well studied.14–16 In addition, we have recently shown that the combination of SLS and low-frequency ultrasound significantly enhances skin permeability in human subjects and allows the noninvasive transdermal extraction of glucose.17 In this paper, we discuss (i) the synergistic effect of ultrasound and SLS on transdermal transport, (ii) the dependence of the synergistic effect on ultrasound parameters and SLS concentration, and (iii) a mechanistic explanation for the synergistic effect.

MATERIALS AND METHODS In Vitro Experiments In vitro transport experiments were performed with full-thickness pig skin (Yorkshire). The skin was harvested with a scalpel immediately after sacrificing the animal, and the underlying fat was removed. Samples were cut into small pieces (2.5 × 2.5 cm2), and those with no visible imperfections (for example, scratches or abrasions) were stored in a −80 °C freezer and used within 12 weeks. Just prior to an experiment, skin was thawed at room temperature and then immediately mounted onto a Franz diffusion cell. To ensure intact skin barrier function, only skin having an initial resistivity of at least 30 k⍀cm2 was used (measurement described later). The Franz cell is a vertical diffusion cell that consists of two compartments, the donor and the receiver compartments. Two 4-mm Ag/AgCl disk electrodes (E242, Invivo Metrics) were introduced into the cell, one in the receiver compartment and the other in the donor compartment, to measure skin conductivity throughout the experiment (vide infra). The skin was subsequently mounted in the Franz diffusion cell with the stratum corneum (SC) side, the uppermost layer of the epidermis, facing the donor compartment. The donor and receiver compartments were then clamped. The receiver compartment was filled with phosphate buffered saline (PBS, phosphate concentration ⳱ 0.01 M, NaCl concentration ⳱ 0.137 M, Sigma Chemicals Company) prepared using partially deionized water (resistance ∼ 7 M⍀). The donor compartment was filled with a 10 ␮Ci/mL solution of tritium-labeled mannitol (ARC) in the same PBS solution or in a PBS solution of sodium lauryl sulfate (SLS), vary-

893

ing in concentration from 0 to 1% by weight. Mannitol, a monosaccharide (182 MW), was selected as the model drug. The diffusion cell was placed in a custom-made lucite mounting block to prevent horizontal vibration. The receiver compartment was stirred with a magnetic stir plate (Bellco). Ultrasound was applied by methods described later for either 10 or 90 min. The average skin permeability (measured over 20 h) was calculated based on the equation P ⳱ V (⌬Cr/⌬t)/ (ACd), where V is the volume of the receiver compartment (12 mL), A is the skin area (1.7 cm2), ⌬Cr/⌬t is the measured increase in the permeant concentration of the solution in the receiver compartment over time intervals (⌬t), and Cd is the permeant concentration of the solution in the donor compartment. Typically, two or three samples were taken over the 20-h period, with the last sample taken at the end of the 20 h. In some cases, the skin was first exposed to ultrasound from a 1% SLS solution, and subsequently, the skin was put in contact with a solution of mannitol in PBS for 20 h after ultrasound was turned OFF. The permeability enhancement of mannitol was calculated by determining the ratio of the measured mannitol permeability, under the various conditions just described, and the permeability measured in a control experiment, in which mannitol transport across the skin was measured from PBS alone in the absence of ultrasound or SLS. Ultrasound Application Ultrasound was applied with a sonicator (VCX 400, Sonics and Materials) operating at a frequency of 20 kHz. Before each experiment, the sonicators were “tuned” according to a procedure specified by the manufacturer. The sonicator horn was positioned 1 cm above the skin inside the donor compartment. The sonicators were operated in the duty-cycle; that is, pulsed mode (0.1 s ON and 0.9 s OFF, 1 s ON and 9 s OFF, or 5 s ON and 5 s OFF). This mode of pulsing parameters was adopted to minimize thermal effects and was otherwise arbitrary. The temperatures of the solutions in the donor and the receiver compartments were measured periodically with a thermocouple (Digithermo, VWR Scientific). When ultrasound was applied at a duty cycle of 10%, the temperature increased by ∼5 °C (starting from a temperature of ∼25 °C) over the first 30 min and remained constant thereafter. When ultrasound was applied at a duty cycle of 50%, the temperaJOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 7, JULY 2000

894

MITRAGOTRI ET AL.

ture typically increased by ∼13 °C in 2 min. At this time, ultrasound was turned OFF and the coupling medium was removed from the donor compartment. Fresh coupling medium was then placed in the donor compartment and sonication was continued for an additional 2 min. This process was continued for a total period of up to 10 min. Measurement of Ultrasound Intensity A commonly used calorimetric method was employed to calculate the power from the sonicator based on the change in the temperature of water exposed to the sonicator.18 For the ultrasound conditions used in this study, the generated ultrasound intensities were in the range from 1.6 to 14 W/cm2. These intensities correspond to spatially and peak-averaged values. Note that calorimetry measures the total energy delivered into the system. Assuming that this energy is emitted from the radiating face of the horn (1.3 cm2), we calculated the intensity of ultrasound at the face of the horn. Because some of this energy may be reflected, absorbed, or dissipated, the actual ultrasound intensity at the skin surface may differ from the intensity measured by calorimetry, but is related to the ultrasound intensity at the horn surface. Electrical Resistance Measurements At the outset of the experiment, two Ag/AgCl electrodes (E242, Invivo Metrics) were introduced in the donor and the receiver compartments of the diffusion cell to measure the electrical resistance of the skin. These measurements were taken approximately every 2 or 15 min during sonication, and periodically thereafter. To measure the electrical resistance of the skin, a 100-mV AC electric field (10 Hz) was applied across the skin for a short time (typically, 5 s) with a signal generator (model HP 4116 A). Measurements were made with an ammeter (Micronta, Tandy Corporation). The electrical resistance was then calculated from Ohm’s law. The resistance of the SLS solution with the model drug was measured separately using the same diffusion cell assembly but without mounting the skin. Because the measured skin resistance is the sum of the actual skin resistance and the solution resistance, the latter was subtracted from the measured skin resistance to obtain the actual skin resistance. The skin resistivity was then calculated by multiplying the skin JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 7, JULY 2000

electrical resistance by the skin area (1.7 cm2). The skin conductivity enhancement was calculated as the ratio of the conductivity of the skin exposed to ultrasound, or to a solution of SLS in PBS, and the skin conductivity of the control, in which the skin was only exposed to PBS. Aluminum Foil Measurements In an attempt to quantify the potential of the sonicator to induce transient cavitation in this experimental setup, aluminum foil (Reynolds) was mounted onto the diffusion cells in a manner identical to that of the skin. The receiver and the donor compartments were then filled with PBS. Sonication was performed for 20 s (0.1 s ON and 0.9 s OFF), and then the aluminum foil was removed from the cell. The number of pits on a foil was determined by visual inspection to assess the extent of cavitation bubble formation induced by ultrasound. The pits represent physical evidence of the effects of cavitation because the force of the microjets that results from the bubble implosions on the foil mechanically dents it and generates the pits. Note that the ultrasound exposure time for these experiments was typically much shorter than that for the transport experiments because a longer exposure results in a very high number of pits, thus making their counting difficult. These experiments were repeated by replacing PBS with a 1% solution of SLS in PBS. Note that the pitting data is used here only as a measure of transient cavitation in our system. A determination of jet formation on the skin and its mechanistic implications are beyond the scope of this paper.

RESULTS AND DISCUSSION Synergistic Effect of Low-Frequency Ultrasound and SLS Application of SLS alone, as well as of ultrasound alone, increases skin conductivity and permeability. Relative to the passive permeability (measured over 20 h), application of SLS alone (1% solution) for 90 min induced about a 3-fold increase in mannitol permeability, whereas application of ultrasound alone (10 W/cm2, 0.1 s ON and 0.9 s OFF) for 90 min induced an 8-fold enhancement. However, when combined, a 90-min application of ultrasound from a 1% SLS solution induced about a 200-fold increase in the skin permeability to mannitol. In each case, the enhancement in skin permeability was accompanied by an

SYNERGISTIC EFFECTS OF ULTRASOUND AND SURFACTANT ON TRANSDERMAL TRANSPORT

enhancement in skin conductivity. The average skin conductivity in the absence of ultrasound is 0.01 (k⍀-cm2)−1. The dependence of the skin conductivity enhancement on ultrasound parameters and SLS concentration is discussed later. Interestingly, skin conductivity did not change significantly after ultrasound was turned OFF (at 90 min) for the duration of the experiment (20 h). In other words, the skin conductivity 20 h after sonication was higher than that of the control skin. This finding suggests that the skin remains permeable well beyond the actual sonication time (90 min). Figure 1 shows the relationship between mannitol permeability (measured over 20 h) and skin conductivity (measured at the end of a 90min sonication period, or after skin exposure to a 1% SLS solution in PBS) under a variety of conditions including passive transport from PBS, SLS alone, ultrasound alone at three intensities (1.6, 6.5, and 10 W/cm2), and a combination of SLS and ultrasound at four intensities (1.6, 4.5, 6.5, and 10 W/cm2). Figure 1 shows that, after the indicated treatment, the skin permeability to mannitol is proportional to the skin conductivity under a wide variety of conditions. The proportionality factor is 1.05 in a linear fit of the data. Ultrasound may enhance transdermal transport by affecting the skin structure (through which enhanced diffusion may occur), by inducing convection, or by a combination of both effects. Because skin conductivity is an excellent indicator of the skin barrier properties, the data in Fig-

Figure 1. The relationship between mannitol permeability and skin conductivity exhibits a linear curve fit yielding an R2 value of 0.99. The x-axis indicates the skin conductivity measured at the end of the sonication period or after skin contact with a 1% SLS solution for 60 min. The y-axis displays the corresponding mannitol skin permeability measured over 20 h.

895

ure 1 suggest that ultrasound under the conditions examined here enhances skin permeability by altering its structure. We performed two additional experiments using pig skin in vitro to further test this hypothesis. In the first experiment, ultrasound (7 W/cm2, 5 s ON and 5 s OFF for 10 min) was applied simultaneously with a solution of radiolabeled mannitol containing 1% SLS as described in the Methods section. In the second experiment, ultrasound under the same conditions was applied for 10 min from a 1% SLS solution and, subsequently, a solution of radiolabeled mannitol in PBS was placed in the donor compartment after ultrasound was turned OFF. Because the skin conductivity in both experiments at the end of sonication was about the same, we assumed that the effect of ultrasound on the skin structure was about the same in both cases. We then hypothesized that if ultrasoundinduced convection across the skin is the primary cause of enhancement, we should see a higher enhancement of transdermal mannitol transport in the first case, in which mannitol and ultrasound were applied simultaneously. However, we did not observe any statistical difference in the mannitol transport enhancement in the two cases (200-fold in each case). This finding suggests that, under the conditions examined here, ultrasoundinduced convection across full-thickness pig skin does not contribute to the observed enhanced mannitol permeability as significantly as the ultrasound-induced changes in the structure of the skin. In view of this interesting finding and the relationship between the skin permeability to mannitol and its electrical conductivity shown in Figure 1, we decided to measure skin conductivity instead of skin permeability in all subsequent experiments to assess the effect of ultrasound on the skin. This approach is very convenient because the electrical conductivity of the skin can be measured nearly instantaneously and more accurately compared with the measurement of skin permeability. Dependence of Transdermal Transport Enhancement on Ultrasound Intensity, Duty Cycle, Net Exposure Time, Energy Density, and SLS Concentration Figure 2 shows the variation in skin conductivity enhancement with ultrasound intensity at a constant duty cycle of 10% (0.1 s ON and 0.9 s OFF) when applied from a solution of 1% SLS at a constant net exposure time of 3 min (net exposure JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 7, JULY 2000

896

MITRAGOTRI ET AL.

time is the total time for which ultrasound is ON). Figure 3 shows the measured dependence of the skin conductivity enhancement on the net exposure time at a constant intensity of 10 W/cm2 (with 1% SLS). The data in Figure 3 consists of experiments run at three different duty cycles [0.1 s ON and 0.9 s OFF (䊉), 1 s ON and 9 s OFF (䊏), and 5 s ON and 5 s OFF (䊊)]. In all cases, the net exposure time was calculated by multiplying the total application time by the duty cycle (the fraction of time when ultrasound was ON). The difference between the data collected at three duty cycles is not significant. Accordingly, the enhancement induced by ultrasound for a given net exposure time is the same regardless of the duty cycle. The dependence of enhancement on net exposure time at each duty cycle exhibits nonlinear behavior. This nonlinearity was not observed at intensities <10 W/cm2. The total energy density delivered from the horn is described by E ⳱ I␶, where I is the ultrasound intensity (W/cm2) and ␶ is the net exposure time (s). Figure 4 shows the dependence of the skin conductivity enhancement on the energy density, E. The filled circles correspond to the application of ultrasound from a 1% SLS solution, and the open circles correspond to the application of ultrasound from PBS. Figure 4 shows two important features: (i) there exists a threshold ultrasound energy density below which the effect of ultrasound on skin conductivity cannot be detected, and (ii) beyond that threshold value, the conductivity increases with the energy density. Figure 4 shows that the threshold energy density

Figure 2. Dependence of the skin conductivity enhancement on ultrasound intensity at a constant duty cycle of 10% (0.1 s ON and 0.9 s OFF) for 30 min (䊊) from 1% SLS. The R2 values for the linear curve fit of the data is 0.97. The error bars represent one standard deviation from the mean. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 7, JULY 2000

Figure 3. Dependence of the skin conductivity enhancement on ultrasound (with 1% SLS) net exposure time for an ultrasound intensity of 10 W/cm2 at three different pulse cycles: (䊉) 0.1 s ON and 0.9 s OFF; (䊊) 5 s ON and 5 s OFF; (䊏) 1 s ON and 9 s OFF.

for affecting skin permeability in the absence of SLS, Ethreshold, PBS, is about 141 J/cm2. This threshold energy density value probably reflects the ultrasound energy density required to induce minimum structural changes in the skin that are sufficient to yield a measurable change in skin conductivity. The threshold energy density value in the presence of 1% SLS, Ethreshold, SLS, is about 18 J/cm2, almost 8-fold lower than that in the absence of SLS. It is important to note that the val-

Figure 4. Dependence of the skin conductivity enhancement on ultrasound energy density, E, in the presence (䊉) and in the absence (䊊) of SLS for the broad spectrum of conditions covered by Figures 1–3. Ethreshold indicates the x-intercept of the best-fit line between enhancement and energy density. A linear fit yields R2 values of 0.94 and 0.55 for the data with and without SLS, respectively.

SYNERGISTIC EFFECTS OF ULTRASOUND AND SURFACTANT ON TRANSDERMAL TRANSPORT

ues of the threshold energy density and the relationship between enhancement and energy density are not unique. That is, they may vary with a number of variables, such as the ultrasound frequency, distance of the horn from the skin surface, and the skin area. Figure 4 shows the best fit equations between the enhancement and energy density in the presence as well as in the absence of SLS. The dependence of enhancement on energy density is close to linear for both cases. This result is interesting especially because the data in Figure 3 suggest the presence of a slight nonlinearity in the relationship between the enhancement and the net exposure time. One explanation for this difference is that the data utilized in Figure 4 is collected over a wide range of intensities (1.6–10 W/cm2) and net exposure times (0–120 min), whereas that in Figure 3 is collected at a single intensity (10 W/cm2) and net exposure times between 0 and 6 min. Furthermore, the nonlinear dependence of enhancement on net exposure time shown in Figure 3 is not observed at intensities <10 W/cm2 (data not shown). Therefore, the nonlinearity in Figure 3 is not visible in a large set of data collected over a variety of conditions. Although Figure 4 demonstrates that the primary relationship between the enhancement and the energy density is close to being linear, the nonlinearity observed in Figure 3 suggests the presence of secondary effects that require further investigation, especially under high intensity conditions. In addition, the relationship between the enhancement and the intensity measured by methods other than calorimetry (for example, using a hydrophone) should also be investigated. Hydrophone measurements may also help to better characterize the acoustic field in the diffusion cell. In all the experiments just described, the concentration of SLS in the donor compartment was maintained constant at 1%. The dependence of surfactant-induced enhancement on concentration has been well studied.19 Under typical circumstances, the enhancement is linear below the critical micelle concentration (cmc) of the surfactant and is either linear or nonlinear beyond that. We investigated the dependence of the skin conductivity enhancement on SLS concentration while keeping the ultrasound parameters at a constant value. Figure 5 shows the dependence of the skin conductivity enhancement on the concentration of SLS in the donor compartment (7 W/cm2, 5 s ON and 5 s OFF, 5 min net exposure time). As can be seen, the skin conductivity en-

897

Figure 5. Linear dependence of the skin conductivity enhancement on SLS concentration with ultrasound application (7 W/cm2, 50% duty cycle (5 s ON and 5 s OFF), 10 min exposure time). The R2 value for the curve fit is 0.99. The critical micelle concentration, 0.03%, of SLS in PBS is noted on the x-axis.

hancement varies linearly with SLS concentration. No noticeable deviation from the linear behavior is observed as one traverses the cmc of the SLS solution (cmc ⳱ 0.03% by weight in the presence of PBS). Possible Mechanisms for the Synergistic Effect of Ultrasound and SLS Cavitation has been shown to play an important role in sonophoretic transdermal transport. Cavitation involves the formation, growth, and collapse of gaseous bubbles, and may occur in either a stable (slow growth over hundreds of cycles) or transient (rapid growth and implosion) mode.20 The rapid implosion may lead to the generation of microjets (or shock waves). SLS, a well-known surfactant, may interact with cavitation bubbles as well as with the stratum corneum (SC), leading to a synergistic effect on transdermal transport. In view of these possible interactions, we next consider three possible mechanisms to explain the synergistic effect of SLS and low-frequency ultrasound on transdermal transport.

SLS Enhances Ultrasound-Induced Cavitation The presence of surfactants such as SLS is likely to affect the occurrence of cavitation due to their interaction with gaseous bubbles. To test the possibility that SLS enhances transient cavitation, we measured the occurrence of transient cavitation in our system in the presence, as well as in the absence, of SLS. The magnitude of cavitation JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 7, JULY 2000

898

MITRAGOTRI ET AL.

in a system can be detected by various techniques, including acoustic emission,18 pitting of metal surfaces,21 and sonochemistry.20 Stottlemyer et al.22 have studied the effect of surfactants on cavitation by sonochemistry and by measuring acoustic emission. Addition of surfactant (Triton X-100) was found to reduce the “strength” of cavitation. This reduction was attributed to a reduction in gas bubble size before collapse or to increased interfacial mobility. We investigated the effect of surfactants on cavitation by measuring pitting of aluminum foil. This method, utilized by others to estimate the magnitude of cavitation, was chosen because of its simplicity. The pits (indentations) generated on aluminum foil are mostly circular and of various diameters. The size distribution of the pits was typically Gaussian, with an average diameter of 200 ± 75 ␮m. Although the average number of pits is an indicator of the “number” of cavitation events, the average size of the pit is likely to be an indicator of the bubble size or “strength” of a cavitation event. A 5-s application of ultrasound (20 kHz, 7 W/cm2, continuous) resulted in 73 ± 20 pits from PBS. When PBS was replaced with a 1% SLS solution, the number of pits decreased to 6 ± 4. The pits were typically smaller than those in the PBS case. These data are consistent with the hypothesis that the presence of SLS decreases the strength, and potentially the magnitude, of bubble collapse. Based on our data and literature reports, it appears that the effect of SLS on cavitation is unlikely to explain the observed synergistic effect of SLS and ultrasound on transdermal transport.

Ultrasound Drives More SLS into the Skin SLS is a potent chemical enhancer. The penetration of SLS into the skin has been shown to enhance transdermal transport of various chemicals. Various mechanisms for this transdermal transport enhancing effect of SLS have been proposed, including fluidization of lipid bilayers, denaturation of keratin, dissolution of skin, and swelling of keratinocytes. Because ultrasound by itself increases transdermal transport of molecules, it can increase the uptake of SLS by the SC. This would increase SLS levels in the SC, which would likely result in increased transport (due to the aforementioned effects) relative to that induced by ultrasound alone. To test whether ultrasound enhances the penetration of SLS into the skin, we performed the following experiments. We sonicated the skin according to the JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 7, JULY 2000

protocol described in Methods (20 kHz, 7 W/cm2, 5 s ON and 5 s OFF, 10 min) from a 1% SLS solution containing trace amounts of radiolabeled SLS (0.5 ␮Ci/mL). At the end of sonication, the skin was rinsed in buffer to remove any residue of SLS attached to the skin surface, and was then dissolved in an aqueous solution containing 30% sodium hydroxide to determine the amount of SLS delivered into the skin during sonication using a scintillation counter. Control experiments were performed using a similar protocol except that no ultrasound was applied. The amount of SLS recovered from the skin exposed to ultrasound and SLS for 10 min was ∼7 times higher than that in the absence of ultrasound. Clearly, application of ultrasound increased the amount of SLS delivered into the skin. To test whether this is the only mechanism responsible for the synergistic effect, we soaked skin in a 1% SLS solution for ∼70 min in the absence of ultrasound. We hypothesized that this process would deliver ∼7 times more SLS into the skin compared with that delivered by soaking for 10 min. If the only mechanism responsible for the synergistic effect is the enhanced delivery of SLS into the skin, the enhancement in skin conductivity after about a 70-min contact with SLS alone should be the same as that induced by a 10-min simultaneous application of ultrasound and SLS. This similarity would exist because the total amount of SLS in the skin should be the same in both cases (which we indeed confirmed using radiolabeled SLS). However, we found that a 70-min contact of the skin with SLS enhanced its conductivity by only 20-fold compared with a 200-fold enhancement induced by a simultaneous application of ultrasound and SLS. This result shows that enhanced delivery of SLS into the skin cannot entirely explain the synergistic effect of ultrasound and SLS. A likely explanation for this synergism is the enhanced dispersion of SLS in the SC due to ultrasound. This possibility is discussed next.

Ultrasound May Enhance the Dispersion of SLS in the SC Chemical enhancers, such as oleic acid, have been shown to form segregated phases within the SC.23 In the absence of ultrasound, SLS may also tend to diffuse into the SC and become localized therein. Application of ultrasound may induce mixing and facilitate the dispersion of SLS in the SC, which, in turn, would result in a much larger fraction of SC exposed to SLS. If SLS does indeed

SYNERGISTIC EFFECTS OF ULTRASOUND AND SURFACTANT ON TRANSDERMAL TRANSPORT

segregate within the SC, we would expect that a prolonged (theoretically speaking, infinitely long) exposure of skin to SLS alone would eventually disperse SLS uniformly within the skin. At that stage, application of ultrasound may not lead to any additional enhancement in skin conductivity. We performed experiments to test this hypothesis. We soaked skin in a 1% SLS solution for various times ranging from 1 min to 24 h, and then exposed it to ultrasound from PBS. Figure 6 shows the skin conductivity at the end of soaking in 1% SLS alone (filled circles) for various times and that after an additional 10-min ultrasound exposure from PBS (open circles). Figure 6 clearly shows that the skin conductivity after soaking in SLS alone increases with increasing soaking time. The application of ultrasound further increased skin conductivity in each case. However, the difference between the conductivity at the end of soaking and after the ultrasound exposure is smaller for longer soaking times. If the two lines shown in Figure 6 are extrapolated to higher soaking times, they suggest that if the skin were soaked in SLS for ∼7300 min the additional enhancing effect of ultrasound on skin conductivity would be negligible. These data support the hypothesis that ultrasound increases dispersion of SLS within the skin. Thus, the data presented in this paper indicate that the synergistic effect between ultrasound and SLS is attributed to two

Figure 6. Conductivity of the skin soaked in a 1% SLS solution for various lengths of time (䊉), and skin conductivity at the end of 10 min of sonication (from PBS), following soaking of the skin in a 1% SLS solution for various lengths of time (䊊). Linear fits of both data sets yield R2 values of 0.99 and 0.97, respectively. Extrapolation of the data suggests that both treatments should yield a skin conductivity of 110 (k⍀cm2)−1 at 7300 min.

899

mechanisms: (i) enhanced delivery of SLS into the skin, and (ii) enhanced dispersion of SLS within the skin. Future studies in this area should include detailed investigations of the dispersion mechanism, with emphasis on the interactions of SLS with the SC lipids and keratinocytes, as well as of the effect of ultrasound on these interactions as a function of SLS concentration. Microscopic studies should also be performed to characterize structural changes induced in the skin by this treatment. The implications of the linear relationship between the enhancement and the energy density should also be further investigated.

CONCLUSIONS The data presented in this paper clearly show that SLS and low-frequency ultrasound synergistically enhance transdermal drug transport. Furthermore, the skin remains in a state of elevated permeability for a sustained period of time. The enhancement is proportional to the ultrasound energy density beyond a certain threshold value. A similar dependence of the enhancement on ultrasound energy density is observed in the case of ultrasound alone. The energy density threshold value in the case of PBS is about 141 J/cm2, whereas that in the presence of 1%SLS is about 18 J/cm2. Our mechanistic studies suggest that the synergistic effect between ultrasound and SLS is attributed to both enhanced penetration and dispersion of SLS in the skin due to ultrasound. This combined ultrasound–SLS treatment has several practical applications in diagnostics and drug delivery. This method has been used in human volunteers for the purpose of noninvasive transdermal glucose extraction.17 Specifically, a 2-min application of low-frequency ultrasound and SLS (under conditions described in this paper) was used to permeabilize the skin, and glucose was subsequently extracted through permeabilized skin for several hours. Application of ultrasound–SLS did not cause any visible side effects on the skin, and the permeability of the skin recovered to its baseline value in 20 h. Transdermally extracted fluxes correlated well with blood glucose values. Future studies should focus on advancing our mechanistic understanding of the synergistic effect between SLS and ultrasound using microscopy as well as on optimization of ultrasound parameters (frequency, intensity, etc.) to further inJOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 7, JULY 2000

900

MITRAGOTRI ET AL.

crease the enhancement. Studies should also focus on performing detailed analysis in human volunteers regarding the safety and efficacy of low-frequency sonophoresis for diagnostics and drug delivery applications.

ACKNOWLEDGMENTS This work was supported by NIH grant GM44884, a grant from the Juvenile Diabetic Foundation, a grant from the Centers for Disease Control, and a Merck Fellowship to Hua Tang. Joseph Kost thanks U.S.-Israel Binational Science Foundation grant 93-00244. We thank Peter Moore for very helpful discussions on surfactants.

REFERENCES 1. Bronaugh RL, Maibach HIE, editors. 1989. Percutaneous absorption: Mechanisms-MethodologyDrug Delivery. New York: Marcel Dekker, pp 1–12. 2. Walters KA 1990. Surfactants and percutaneous absorption. In Scott RC, Guy RH, Hadgraft J, editors. Predictions of percutaneous penetration. London: IBC Technical Services, pp 148–162. 3. Levy D, Kost J, Meshulam Y, Langer R. 1989. Effect of ultrasound on transdermal drug delivery to rats and Guinea pigs. J Clin Invest 83:2974–2978. 4. Mitragotri S, Edwards D, Blankschtein D, Langer RA. 1995. Mechanistic study of ultrasonically enhanced transdermal drug delvery. J Pharm Sci 84: 697–706. 5. Bommannan D, Menon GK, Okuyama H, Elias PM, Guy RH. 1992. Sonophoresis. II. Examination of the mechanism(s) of ultrasound-enhanced transdermal drug delivery. Pharm Res 9:1043–1047. 6. Bommannan D, Okuyama H, Stauffer P, Guy RH. 1992. Sonophoresis. I. The use of high-frequency ultrasound to enhance transdermal drug delivery. Pharm Res 9:559–564. 7. Mitragotri S, Blankschtein D, Langer R. 1995. Ultrasound-mediated transdermal protein delivery. Science 269:850–853. 8. Mitragotri S, Blankschtein D, Langer R. 1996. Transdermal drug delivery using low-frequency sonophoresis. Pharm Res 13:411–420. 9. Kalia YN, Guy RH. 1997. Interaction between penetration enhancers and iontophoresis: Effect on human skin impedance in vivo. J Controlled Release 44:33–42.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 7, JULY 2000

10. Prausnitz MR, Bose V, Langer R, Weaver JC. 1993. Electroporation of mammlian skin: A mechanism to enhance transdermal drug delivery. Proc Natl Acad Sci USA 90:10504–10508. 11. Bommannan D, Tamada J, Leung L, Potts R. 1994. Effects of electroporation on transdermal iontophoretic delivery of leutinizing hormone releasing hormone. Pharm Res 11:1809–1814. 12. Kost J, Pliquett U, Mitragotri S, Yamamoto A, Weaver J, Langer R. 1996. Enhanced transdermal delivery: Synergistic effect of ultrasound and electroporation. Pharm Res 13:633–638. 13. Johnson ME, Mitragotri S, Patel A, Blankschtein D, Langer R. 1996. Synergistic effect of ultrasound and chmeical enhancers on transdermal drug delivery. J Pharm Sci 85:670–679. 14. Faucher JA, Goddard, ED. 1978. Interaction of keratinous substrates with sodium lauryl sulfate: I. Sorption. J Soc Cosmet Chem 29:323–337. 15. Garcia MT, Leal JS, Ribosa I. 1992. Effect of various parameters on sodium dodecyl sulfate (SDS) flux through a collagen membrane. J Am Oil Chem Soc 69:20–24. 16. Rhein LD, Simion FA, Hill RL, Cagan RH, Mattai J, Maibach HI. 1990. Human cutaneous response to a mixed surfactant system: Role of solution phenomenon in controlling surfactant irritation. Dermatologica 180:18–23. 17. Kost J, Mitragotri S, Gabbay R, Pishko M, Langer R. 2000. Non-invasive extraction of blood analytes using ultrasound. Nature Med 6:347–350. 18. Liu J, Lewis TN, Prausnitz MR. 1998. Noninvasive assessment and control of ultrasoundinduced membrane permeabilization. Pharm Res 15:918–923. 19. Dugard PH, Scheuplein RJ. 1973. Effects of ionic surfactants on the permeability of human epidermis: An electrometric study. J Invest Dermatol 60(5):263–269. 20. Suslick KS. 1989. Ultrasound: Its chemical, physical, and biological effects. New York: VCH Publishers, pp 123–159. 21. Frederick JR. 1980. Ultrasound engineering. New York: John Wiley. 22. Stottlemyer TR, Apfel RE. 1997. The effects of surfactants additives on the acoustic and light emissions from a single stable sonoluminescing bubble. J Acoust Soc Am 103:1418–1423. 23. Ongpipattnakul B, Burnette RR, Potts RO, Francoeur ML. 1991. Evidence that oleic acid exists in a separate phase within stratum corneum lipids. Pharm Res 8:350–354.