Ketoprofen Tissue Permeation in Swine Following Cathodic Iontophoresis.

Multiple pharmaceutic methods exist for delivering medications transcutaneously. The most common and widely used method is injection. Alternative methods include the use of electromotive force (iontophoresis or ionization), mechanical force (phonophoresis or sonophoresis), or passive permeation. Parenteral transcutaneous delivery of pharmacologic agents has potential benefits. Avoidance of adverse effects on the gastrointestinal system when these same agents are introduced into the body orally is one example. Following oral administration and alimentary absorption of drugs, they are transported through the portal system to the liver where they may be metabolized. Parenteral drug delivery minimizes the metabolism of drugs by the liver, thus avoiding this first-pass effect.[1] Transcutaneous pharmacologic delivery also provides the potential for localized delivery to the tissues underlying the site of application.

Both electromotive and mechanical forces have been documented to transcutaneously deliver a wide variety of pharmacologic agents under either experimental or clinical conditions. Transcutaneous deliveries of polar and nonpolar agents of either small or large molecular weights have been documented.[2-10] Phonophoresis may have an additional advantage in the transcutaneous permeation of nonpolar agents due to the utilization of mechanical rather than electromotive force.[4,9,10] In contrast to iontophoresis and phonophoresis, passive permeation is limited in the variety of pharmacologic agents delivered.[2,5,9,11,12] Passive administration has not been shown to efficiently deliver proteins or hydrophilic anionic compounds transcutaneously. The potential for transcutaneous delivery of a wide variety of agents is clearly documented.

A recent literature review documented that both protein and nonprotein pharmacologic agents have been administered transcutaneously with the purpose of achieving therapeutic systemic pharmacologic levels while avoiding first-pass effects of the gastrointestinal system.[13] In contrast, transcutaneous administration of anti-inflammatory drugs has focused on local tissue permeation at the application site, regardless of whether the permeation was attempted by iontophoresis,[13,14] phonophoresis,[8,15] or passive administration.[16,17] The pharmacologic agents that have been transcutaneously administered for local tissue anti-inflammatory effects include steroidal and nonsteroidal drugs.[13] The advantages of transdermal delivery of anti-inflammatory drugs into the local tissue underlying the application site are reduced systemic drug levels and reduction of adverse gastrointestinal effects when these drugs are administered orally.[18]

Ketoprofen is an anionic nonsteroidal anti-inflammatory drug (NSAID) with approximately 160 times the anti-inflammatory potency of aspirin on a per weight basis.[18,19] Commercially, ketoprofen is formulated as a racemic mixture of "R" and "S" enantiomers, which are equivalent on a per weight basis. Enantiomers of a chemical are mirror images of the structure about a chiral center. Because of this difference, ketoprofen exhibits enantiomeric selectivity, with only the "S" enantiomer demonstrating pharmacodynamic activity.[20] Both oral and transcutaneously administered ketoprofen reduced carrageenan-induced acute inflammatory edema in the rat paw.[21] Passive transcutaneous administration of the ketoprofen gel at the inflammatory site was equivalent to oral administration of the drug in the degree to which the edema was reduced. The ketoprofen concentrations required for decreasing the inflammation by 50% for the topical gel and the oral formulation were 2.2 mg and 6.0 mg, respectively, per kilogram of animal weight. Thus, the topical formulation was nearly 3 times more effective on the basis of the drug weight than that of the oral formulation.

In humans, repeated daily passive topical administration of a ketoprofen gel for 3 days achieved detectable ketoprofen in the synovial fluid of the knee.[22] Ketoprofen has also been measured in venous blood flow from the application sites following 160-mA x min cathodic iontophoresis in humans.[23,24] Following these passive or iontophoretic applications, the detectable ketoprofen concentrations in the synovium and return venous blood were within the 0.4- to 6-[micro]g/mL potential therapeutic range of ketoprofen.[19,20]

Several clinical investigations[25-27] have also demonstrated positive clinical outcomes when ketoprofen was delivered transcutaneously. In these investigations, a twice-daily application of a ketoprofen gel for a minimum of 3 days resulted in reduced subjective assessment of pain, stiffness, movement limitations, or swelling associated with either acute soft tissue injuries or rheumatic dysfunction. Whether the positive clinical benefits observed with passive transcutaneous permeation of ketoprofen, and potentially with iontophoretic permeation of ketoprofen, were truly a local tissue effect or an effect following systemic distribution, however, remains open to discussion.[11,12,20,22,28,29]

Investigations have documented the potential for transcutaneous ketoprofen permeation following both passive administration[22] and iontophoretic administration.[23,24] The in vivo depth of local transcutaneous permeation for anionic anti-inflammatory drugs (eg, dexamethasone, ketoprofen) following a single iontophoretic or passive administration under clinically relevant conditions, however, has not been reported. Our investigation, therefore, was undertaken (1) to examine the depth of ketoprofen tissue permeation following a single iontophoretic or passive application and (2) to determine whether transcutaneous ketoprofen permeation following these applications was stereoselective to either the pharmacodynamically active "S" or the inactive "R" enantiomers. Ketoprofen was used as a model anionic anti-inflammatory drug due to previous evidence of transcutaneous permeation following passive administration[22] or cathodic iontophoresis,[23,24] edema reduction when applied passively,[21] reduction in the complaints of pain by patients when applied passively,[25-27] and detection of the drug in muscle tissue samples.[30]

Materials and Methods

Our experimental protocol using a swine model has been described previously,[23,24,30] and only additional relevant methods will be elaborated on in this report.

Materials

Ketoprofen, fenoprofen, S-(-)-[Alpha]-phenylethylamine, and anhydrous isopropanol were obtained from Sigma Chemical Company.(*) Acetonitrile (Optima), isa-octane, isopropanol, methylene chloride, methanol, glacial acetic acid, Na[HCO.sub.3], and HCl were obtained from Fisher Scientific.([dagger]) All borosilicate test tubes, Teflon([double dagger])-sealed 50- and 25-mL tubes and uncapped 20-mL tubes, and 30-mL polypropylene homogenization tubes were also obtained from Fisher Scientific.([dagger]) Polypropylene tubes (250 [micro]L) were from Cole Palmer International.([sections])

Solution Preparation

Standards of ketoprofen and fenoprofen (internal standard) were dissolved in methanol at 1 mg/mL and stored in Teflon-sealed brown glass containers (25 [degrees] C). The internal standard was a quality-control variable for ketoprofen quantification. The internal standard was added to all biologic standards and unknown samples to document the efficiency of ketoprofen extraction, dimerization, and chromatographic detection from the tissues in question. Ketoprofen used for in viva iontophoresis was prepared at 300 mg/mL in phosphate-buffered saline with 20% ethanol (volume per volume).

These reagents and all tissue extraction, derivatization, and high-performance liquid chromatography (HPLC) reagents were prepared as described previously.[23,24,30] Water used for all solutions was 18-M[Omega] eluant from a Barnstead Nanopure system.([parallel])

Animal Surgery and Tissue Sampling

Surgical anesthesia in the pigs was induced by atropine (0.05 mg/kg, administered intramuscularly), ketamine (25 mg/kg, administered intramuscularly), acepromazine (0.5 mg/kg, administered intramuscularly), and sodium pentobarbital (15 mg/kg, administered intravenously). Anesthesia in one animal was induced by succinylcholine (0.4 mg/kg, administered intravenously). Anesthesia was maintained by inhalation of 0.5% to 1.0% halothane in oxygen. The animals were ventilated at 12 to 16 respirations per minute at 300 to 400 mL per breath. At the termination of the experiment, the mean heart rate was 94 beats per minute, and the systolic and diastolic blood pressures were 70 and 38 mm Hg, respectively. These results confirm that the cardiovascular system was intact during the application period. The medial surface of each thigh was prepared by removal of hair with clippers, so as not to damage the stratum corneum. The application areas were precleaned with isopropyl alcohol swabs.

Iontophoresis was conducted using a medium EBIE applicator electrode and Dupel iontophoretic device.(#) A total of 2.5 mL of the ketoprofen (750 mg) solution was applied to the applicator. The first drug delivery applicator was placed on the medial surface of the animal's thigh and attached to the cathodic electrode of the Dupel device. The return anodic electrode was placed on the animal's abdomen 25 cm from the delivery applicator. The iontophoretic dosage was 160 mA x min (4 mA for 40 minutes) at a current density of 0.28 mA/[cm.sup.2]. The second drug delivery applicator was placed on the contralateral medial thigh of the animal, but without connection to the Dupel device, and served as the passive delivery system. Following the iontophoresis, the electrode was removed. A 2.54-cm-diameter keyhole bit with the center mandril removed was placed at the center of the iontophoretic application, and the circumference of the bit was scribed with a scalpel. The skin underlying the electrode was then excised from the underlying fascia and surrounding skin and stored for analysis. The fascia was then clipped from the underlying muscle and stored separately for later analysis. Finally, a core sample of muscle was taken. The keyhole drill was cleaned and placed on the surface of the muscle and then run in reverse to the lateral surface. The muscle biopsy specimen was stored separately for later analysis.

The procedure was repeated on the contralateral thigh receiving the passive ketoprofen delivery. Serum samples were obtained from the systemic circulation, either by cardiac puncture or from one of the great vessels in the central cavity, and then coagulated on ice, centrifuged (850 times per gram, 5 minutes, 25 [degrees] C) (Beckman TJ-6),(**) and stored as described below. All instruments were cleaned so as to be free of visual tissue residue between obtaining tissue samples following iontophoretic and passive ketoprofen deliveries. Finally, all samples were transported vertically on ice prior to storage (-80 [degrees] C).

Sample Preparation

Spiked tissue from skin, fascia, and muscle for standard curved determinations were processed as described previously.[30] Tissue standards were homogenized in 1 normal Na[HCO.sub.3], methylene chloride was added, and the homogenates were vortexed and centrifuged. The upper aqueous layer was acidified with 10 normal HCl, then combined with iso-octane:propanol. The samples were vortexed and centrifuged, and the organic layer was evaporated. Except for the following differences, tissue samples from the passive and iontophoretic application sites were processed in the same manner as in the spiked tissue standards. Upon thawing, the skin and fascia samples were weighed and sufficient nonexposed tissue was added to result in a final wet tissue weight of 0.5 g. Fenoprofen was added (40 [micro]L, 1 mg/mL), and the samples were subsequently processed as the standards. The muscle core sample was laid on a metal rule. Samples were obtained by slicing the core perpendicularly at 1-cm depths. Each 1-cm aliquot was weighed, and sufficient 1 normal Na[HCO.sub.3] was added to make the final sample weight:volume ratio 0.1 g/mL. Each muscle sample was homogenized as described earlier, and 2.5 mL of the homogenate was transferred to a 50-mL round-bottom flask. An additional 2.5 mL of 1 normal NaH[CO.sub.3] was added to make the final concentration 0.05 gm/mL, fenoprofen was added (40/[micro]L, 1 mg/mL), and the samples were subsequently processed as the standards. The total ketoprofen concentration in serum samples was determined[23,24] and compared with a standard curve generated by spiking 1 mL of serum with 10 [micro]L of ketoprofen (1 mg/mL) and 4 subsequent 1:1 (volume per volume) serial dilutions. The internal standard was 40/[micro]L of fenoprofen (1 mg/mL).

Sample Derivatization and High-Performance Liquid Chromatographic Determination

The extracted ketoprofen and fenoprofen were derivatized with ethylchloroformate/phenylethylamine in triethylamine. Sample derivatization and chromatographic separation and determination were conducted at 25 [degrees] C. Samples were transferred to 250-[micro]L polypropylene vials, injected via a 717 autosampler,([dagger][dagger]) and analyzed on an HPLC device, which consisted of an M45 solution delivery system,([dagger][dagger]) a Lambda Max Model 480 variable-wavelength ultraviolet monitor,([dagger][dagger]) a 746 data module,([dagger][dagger]) and a Z-module.([dagger] [dagger]) The 717 autosampler injected a 50-[micro]L sample for analysis, and the injector was rinsed with methanol:water (1:5, volume per volume) between samples. Analytes were eluted from a Waters Nova-Pak [C.sub.18] radial compression column([dagger][dagger]) (8 mm x 100 mm x 4 [micro]) with 10-[micro] Bondapak [C.sub.18] Guard-Paks([dagger][dagger]) and detected at 255 nm. The mobile phase was acetonitrile:water:glacial acetic acid:triethylamine (43:57:0.1:0.03, volume per volume) with a flow rate of 2.5 mL/min.

Calculations and Data Analysis

The peak areas under t-he curve for "R" and "S" diastereomers of ketoprofen or fenoprofen were added together for calculation of total ketoprofen or fenoprofen concentrations. Sample-to-sample variation was normalized by dividing the total ketoprofen peak area by the total fenoprofen peak area. Regression analyses were calculated by plotting ketoprofen concentration on the abscissa and ketoprofen-to-fenoprofen peak area ratios on the ordinate. Calculations for unknown ketoprofen concentrations were determined by utilizing a linear regression function for each tissue. The regression functions for the tissues were as follows:

skin (y=0.122x +0.010, [r.sup.2]=.992 and N=7), fascia (y=0.160x - 0.010, [r.sup.2]=.999 and N=7), muscle (y=0.086x - 0.061, [r.sup.2]=.982 and N=12), and serum (y=0.176x - 0.058, [r.sup.2]=.978 and N=7). Unless stated otherwise, all data are presented as means ([+ or -] standard deviation). Analyses between matched data from the same animals were determined by a 2-tailed paired t test (P [is less than or equal to] .05). A multivariate analysis of variance was used to detect differences among depths of drug delivery within a treatment protocol and differences between delivery protocols at a given depth. To determine specific differences among drug delivery depths within a protocol and differences between delivery protocols at a given depth, a Tukey post hoc groupwise comparison (P [is less than or equal to] .05) was conducted. All statistical analyses were conducted as previously described[31] using the SAS System for the PC([double dagger][double dagger]) on a Pentium Millennia Transport.([subsections])

Results

Both cathodic iontophoresis (160 mA x min; 4 mA for 40 minutes) and passive permeation for 40 minutes demonstrated measurable and equivalent total ketoprofen concentrations in the skin and fascia underlying the application sites (Fig. 1). Total ketoprofen concentrations in the fascia underlying the iontophoretic and passive delivery sites were 70% to 79% lower than the drug concentrations in the skin.

[Figure 1 ILLUSTRATION OMITTED]

Unlike the skin and fascia, the muscle tissues underlying the drug application sites were divided into sections 1 cm in depth and analyzed separately. The total ketoprofen concentrations in these different muscle samples were influenced by both the depth of the sample from the surface of the muscle and whether ketoprofen was applied passively or by iontophoresis (Tab. 1). An interaction between the depth of the muscle sample and ketoprofen delivery protocol was also detected in the statistical modeling. The ketoprofen concentrations at the various tissue depths following iontophoretic and passive administration are represented graphically in Figure 2.

Table 1. Statistical Modeling(a) of Total Ketoprofen Permeation Into Muscle Tissue

Variable               df   SS     MS      F       P>F

Model                   9   1914   213      7.45   .0001
  Depth                 4   1304   326     11.42   .0001
  Treatment             1    208   208      7.27   .01
  Treatment x depth     4    402   100      3.52   .015
Error                  40   1142    28.6
Corrected total        49   3056

(a) Statistical modeling examined the effects of ketoprofen permeation depth into muscle tissue (depth) by either iontophoretic or passive administration (treatment) and the interaction between drug administration protocol and depth of drug permeation (treatment x depth). Both main effect variables and their interaction were all found to demonstrate an influence on total ketoprofen concentration in the muscle tissue.

[Figure 2 ILLUSTRATION OMITTED]

At the iontophoretic site, the total ketoprofen concentration in the initial centimeter of muscle tissue was greater than the drug concentrations at depths of 2 to 5 cm. No differences in total ketoprofen concentrations were detected at depths of 1 to 5 cm following passive ketoprofen administration. Comparisons of the ketoprofen concentrations at each 1-cm muscle depth following iontophoretic or passive administration were also conducted (Fig. 2). Only the first centimeter of muscle depth demonstrated an increase in total ketoprofen concentrations following iontophoresis when compared with passive delivery. At depths of 2 to 5 cm, the total tissue ketoprofen concentrations following iontophoretic and passive administration were equivalent.

Subsequent power analyses were conducted to determine the number of additional experiments required to detect statistical differences in total ketoprofen concentrations between the first centimeter and deeper layers under passive administration and between iontophoresis and passive administration at the second centimeter and deeper layers. The common standard deviation of these analyses was derived from the mean square error shown in Table 1, which represented the pooled variation for all of the means under consideration. In brief, the analyses predicted that an additional 30 animals would be required to detect statistically significant differences in ketoprofen between the first and second layers under passive application. Additionally, approximately 6 times as many animals would be required to detect differences in ketoprofen between iontophoretic and passive administration at the second-centimeter muscle layer. These results suggest that further experiments to determine statistically significant differences in total ketoprofen concentrations for these variables were prohibitive both ethically and economically.

The ketoprofen levels measured in the skin, fascia, and muscle may represent ketoprofen that was absorbed into the systemic circulation and redistributed back to the tissue underlying the iontophoretic or passive application sites. To differentiate whether ketoprofen present within the tissue underlying the application sites represented local tissue drug delivery or was the result of systemic drug distribution, ketoprofen levels in the systemic blood from these animals were measured. Total ketoprofen concentrations from systemic serum were detectable in only 2 of the 5 animals. The detected ketoprofen concentrations in these animals were 0.44 and 0.5 [micro]g/mL. These results suggest that the ketoprofen tissue concentrations observed at the application sites represent delivery of local drug concentrations and are not the result of systemic distribution.

The ratio of "R" and "S" enantiomers of ketoprofen was determined in skin, fascia, and muscle (Tab. 2). The percentage of "S" enantiomer in the skin was statistically greater than the percentage of "R" enantiometer following both iontophoretic and passive administration. In the fascia and muscle tissues, however, the "R" and "S" enantiomer percentages were equivalent. Thus, transcutaneous delivery of ketoprofen by iontophoresis or passive application was not stereoselective.

Table 2. Comparison ([bar]X [+ or -] SD) of Ratios of "R" and "S" Enantiomers of Ketoprofen at Different Tissue Depths Following Either Cathodic Iontophoresis or Passive Administration(a)

                     Iontophoresis
Tissue          %S                  %R

Skin      53 [+ or -] 0.5   47 [+ or -] 0.5(*)
Fascia    50 [+ or -] 0.4   50 [+ or -] 0.4
Muscle
  1 cm    54 [+ or -] 8     47 [+ or -] 8
  2 cm    55 [+ or -] 9     45 [+ or -] 9
  3 cm    49 [+ or -] 7     51 [+ or -] 7
  4 cm    50 [+ or -] 1     50 [+ or -] 1
  5 cm    51 [+ or -] 2     49 [+ or -] 2

                   Passive Delivery
Tissue          %S                  %R

Skin      53 [+ or -] 0.6   47 [+ or -] 0.6(*)
Fascia    51 [+ or -] 0.7   49 [+ or -] 0.7
Muscle
  1 cm    51 [+ or -] 0.6   49 [+ or -] 0.6
  2 cm    51 [+ or -] 2     49 [+ or -] 2
  3 cm    52 [+ or -] 1     48 [+ or -] 1
  4 cm    43 [+ or -] 18    57 [+ or -] 18
  5 cm    49 [+ or -] 1     51 [+ or -] 1

(a) Statistical comparison of the ratios of "R" and "S" enantiomers was done at each tissue depth and for each delivery modality. Significant differences (*) between the percentage of "S" and "R" isomers at each tissue depth and within a given delivery modality were determined by a 2-tailed paired t test with P [is less than or equal to] .05 (N=5).

Discussion

We examined the in vivo tissue depth for passive and iontophoretic transcutaneous permeation of ketoprofen and whether the transcutaneous permeation was stereoselective. The pig was the animal model selected due to similar iontophoretic permeation properties between porcine and human skin. Riviere et al[32] demonstrated parallel pharmacokinetic variables between human and porcine skin during in vitro cell diffusion experiments. In our investigation, nonstereoselective transcutaneous permeation of ketoprofen was observed for both iontophoretic and passive applications.

Positive clinical outcomes observed following transcutaneous application of anti-inflammatory agents might result from a local effect of a drug or from systemic absorption and redistribution back to a tissue. Investigators have examined the tissue permeation of anionic anti-inflammatory agents following both passive permeation[11,12,16] and iontophoretic permeation.[14,28,33] Local passive transcutaneous permeation has been shown for diclofenac, hydrocortisone, indomethacin, naproxen, piroxicam, and salicylic acid, but ketoprofen tissue permeation following passive or iontophoretic application has not been documented. When these anionic anti-inflammatory drugs were applied passively, the maximal tissue depth for local transcutaneous permeation was down to the superficial muscles.[11,12] Drug concentrations in deeper tissues were equal to or less than systemic blood levels, and the drug probably was delivered to these tissues by the systemic blood flow.

The restrictive properties of the epidermis toward passive transcutaneous permeation of these agents are significant. Singh and Roberts,[11,12,28] using a rat model, obtained 80-[micro]m sections from the epidermal surface with a microtome in order to achieve the passive permeation of anti-inflammatory agents. Piroxicam was also delivered to similar tissue depths in the rat, without epidermal debridement, but with a permeabilizing agent incorporated into the gel.[16] In our investigation with ketoprofen, the epidermis was not sectioned for passive application, but isopropyl alcohol was applied to the surface vigorously as a preparatory step, and the ketoprofen was prepared in a 20% ethanol solution and applied to both the passive and iontophoretic sites. Short-chain alcohols have been shown to reduce the restrictive properties of the epidermis, allowing greater transcutaneous permeation of hydrophilic agents under both passive and iontophoretic conditions.[33-35] Our results are in agreement with previous findings as to the depth of tissue permeation of anionic anti-inflammatory drugs following passive application. In our investigation, passive permeation of ketoprofen to tissue depths previously reported for animals with epidermis removed may be accounted for by animal model differences (ie, the use of rats in previous investigations compared with the use of pigs in our investigation) or the epidermal permeabilizing potential of short-chain alcohols.

In contrast to the multiple anionic anti-inflammatory drugs that have been examined during passive application, limited investigations have examined the depth of tissue permeation following iontophoretic application of these agents. Dexamethasone phosphate iontophoresed from the anode (ie, positive electrode) has been shown to penetrate to deep periarticular structures in several different joints in a single-monkey experiment.[14] In that experiment, the investigators concluded that the presence of dexamethasone in these deep tissues was the result of local tissue permeation during the iontophoresis, not an effect of drug delivery by the systemic circulation. In contrast, cathodic iontophoresis of salicylic acid resulted in local transcutaneous tissue permeation down to superficial muscles.[28] Salicylate reached the deeper tissue structures below the application site only by the systemic bloodstream. In our investigation, iontophoresis delivered more ketoprofen to the first centimeter of muscle than did passive application. Ketoprofen detected in deeper muscle layers may have resulted from passive or iontophoretic transport from the superficial muscle layer, focal circulatory distribution, or other as of yet unexamined experimental variables. The tissue permeation that we found following ketoprofen iontophoresis agrees with the results of a previous investigation of tissue permeation following salicylate iontophoresis.[28] Our results and those of previous salicylate studies contrast with the results of Glass and colleagues' study of dexamethasone iontophoresis.[14]

Methodological variances may account for the differences in the results of Glass and colleagues' study of dexamethasone,[14] and the results of our investigation and Singh and Roberts' research on salicylate iontophoresis.[28] In the study of dexamethasone,[14] the iontophoretic current density was calculated to be 0.94 mA/[cm.sup.2], whereas in our study and in the study of salicylate iontophoresis,[28] the current densities were 0.28 and 0.38 mA/[cm.sup.2], respectively. With commercial electrodes and iontophoretic devices, the current densities obtained during clinical iontophoresis in physical therapy parallel those used in our study and in the study of salicylate iontophoresis.[28] The current density used during clinical iontophoresis may be calculated by dividing the current setting on the iontophoretic device by the active surface area of the electrode. According to Banga and colleagues,[13,36] patients generally can tolerate current densities of less than 0.5 mA/[cm.sup.2]. The aggregate results of our investigation and previous studies of salicylate and dexamethasone suggest that the depth of drug tissue permeation may be positively correlated to the current density. This conclusion is supported by observations of the effect of current density on other in vivo iontophoretic variables. The concentration of total ketoprofen delivered following cathodic iontophoresis in humans appeared to be directly proportional to the current density.[23,24] Iontophoresis of ketoprofen at 4 mA for 40 minutes (0.28 mA/[cm.sup.2]) resulted in approximately twice the ketoprofen delivery when compared with 2 mA for 80 minutes (0.14 mA/[cm.sup.2]). The enhanced iontophoretic delivery of ketoprofen appeared to be dependent on the higher current density (ie, 0.28 versus 0.14 mA/[cm.sup.2]), as the dosage for these 2 iontophoretic regimens remained constant at 160 mA x min. Thus, the high current density used in Glass and colleagues' study of dexamethasone[14] may have resulted in deeper iontophoretic drug permeation, but this density exceeds that normally used during current clinical iontophoresis. We therefore question the value of the study of dexamethasone to current clinical use of iontophoresis.

Regardless of the exact tissue permeation depth of anionic anti-inflammatory drugs following either passive or iontophoretic application, investigators[13,17,25-27] have documented local anti-inflammatory effects when applied transcutaneously. Controversy exists, however, as to whether the anti-inflammatory effects from these transcutaneously administered agents are the result of local tissue permeation or delivery via systemic circulation. Local tissue anti-inflammatory effects have been documented when anionic anti-inflammatory agents were applied passively[21] or by iontophoresis.[37] Locally applied and passively delivered ketoprofen was 3-fold more effective at inhibiting in vivo carrageenan-induced rat paw edema compared with oral administration of the drug.[21] Additionally, in vitro iontophoretic flux of diclofenac across the skin parallels the inhibitory effect of diclofenac, when applied by iontophoresis, on carrageenan-induced rat paw edema.[37] In a study by Byl et al,[8] a single phonophoretic application of dexamethasone demonstrated subcutaneous anti-inflammatory effects at the application site. When submuscular or subtendinous tissue below the application site was examined, no such anti-inflammatory effects were observed. Yet, Byl et al also reported reduced anti-inflammatory effects at sites distal to the phonophoretic application, suggesting systemic distribution of the dexamethasone.

Grahame[17] contended that local application of anionic anti-inflammatory agents may have potential usefulness in the treatment of patients with soft tissue injuries, but not in the treatment of patients with inflammatory arthropathy. The therapeutic range for ketoprofen in the blood has been characterized as 0.4 to 6 [micro]g/mL.[19] The specific density of heart muscle has been documented as 1.05 g/mL,[38] and we believe that skeletal muscle should demonstrate a similar specific density. Based on our data, therapeutic ketoprofen concentrations may occur throughout the first 5 cm of muscle under both passive and iontophoretic applications. Ketoprofen is also highly concentrated within the vasculature.[19] Nonvascular compartments, such as the synovium, may achieve ketoprofen concentrations of only 20% of systemic blood levels.[29] Thus, therapeutic ketoprofen concentrations at nonvascular inflammatory sites may actually be less than the 0.4 to 6 [micro]g/mL observed in the systemic vascular compartment.[19] Radermacher et al,[39] however, documented that transcutaneously applied NSAIDs reached nonvascular compartments, such as the synovium, after systemic absorption from the application site and vascular distribution to the joints.

The clinical implications of our current research and the previous research of other investigators suggest that anionic anti-inflammatory agents, in general, and ketoprofen, in particular, have the potential to penetrate the superficial skeletal muscle tissue at pharmacologically relevant concentrations so as to alleviate inflammation in soft tissue. Under clinical conditions, however, these drugs must be used within the methodologies established in the clinical and experimental research. With our current research, the ketoprofen concentration used was 300 mg/mL in a 20% ethanol solution, and the iontophoretic dosage was 160 mA x min. The extrapolation of these results to the current clinical use of dexamethasone phosphate would be inappropriate. The dexamethasone phosphate concentration available for iontophoresis is either 4 or 24 mg/mL, and device manufacturers recommend a dosage equal to or less than 100 mA x min. The relative mobilities of these 2 drugs in an electrical field are also unknown. Thus, which drug is optimal for iontophoretic applications remains a question. To determine the value of ketoprofen for clinical iontophoresis, further experimental or clinical investigations would be required to document a cause-and-effect relationship between drug concentrations within the tissue and their anti-inflammatory effect on soft tissue injury. Additionally, tissue compartments surrounding joints with a different vascular network may also demonstrate distinctly different drug tissue permeation profiles following iontophoretic or passive anionic anti-inflammatory drug administration. Further pharmacokinetic and pharmacodynamic investigations would be required to document the clinical value of anionic anti-inflammatory drugs at these application sites.

Conclusion

Panus and colleagues[23,24] have previously examined iontophoretic factors that would optimize ketoprofen iontophoresis in vivo, and they documented iontophoretic transcutaneous delivery of clinically relevant ketoprofen concentrations in humans. Our current investigation examined the local tissue permeation depth during in vivo cathodic ketoprofen iontophoresis using previously established optimal variables. The results indicate that transcutaneous permeation of ketoprofen following iontophoretic or passive permeation is not stereoselective at subcutaneous sites. Under the described experimental conditions, iontophoretic and passive application resulted in equivalent permeation of the drug down to the fascia. Iontophoresis delivered greater amounts of ketoprofen to the superficial muscle layer compared with passive delivery. Our results suggest the clinical value of local transcutaneous application of anionic anti-inflammatory agents to muscular sites, with iontophoresis demonstrating superior permeation capabilities over passive application. Following a single application, however, submuscular sites would appear to be beyond the local permeation potential of transcutaneously applied anionic anti-inflammatory drugs.

(*) Sigma Chemical Co, PO Box 14508, St Louis, MO 63178-9916.

([dagger]) Fisher Scientific, PO Box 4829, Norcross, GA 30091.

([double dagger]) El du Pont de Nemours & Co Inc, 1007 Market St, Wilmington, DE 19898.

([sections]) Cole Palmer International, 7425 N Oak Park Ave, Niles, IL 60714.

[parallel] Barnstead Thermolyne, 2555 Kerper Rd, PO Box 797, Dubuque, IA 52007-0797.

(#) Empi Inc, 599 Cardigan Rd, St Paul, MN 55126-3965.

(**) Beckman Inc, PO Box 10200, Pale Alto, CA 94304.

([dagger][dagger]) Waters Inc, 34 Maple St, Milford, MA 01757.

([double dagger][double dagger]) SAS Institute Inc, Box 8000, Cary, NC 27511-8000.

([subsections]) Micron Electronics Inc, 900 E Karcher Rd, Nampa, ID 83687.

References

[1] Benet LZ, Kroetz DL, Sheiner LB. Pharmacokinetics: the dynamics of drug absorption, distribution, and elimination. In: Hardman J, Limbird L, eds. Goodman and Gilman's: The Pharmacological Basis of Therapeutics. 9th ed. New York, NY: McGraw-Hill; 1996:3-28.

[2] Kari B. Control of blood glucose levels in alloxan-diabetic rabbits by iontophoresis of insulin. Diabetes. 1986;35:217-221.

[3] Meyer BR, Kreis W, Eschbach J, et al. Successful transdermal administration of therapeutic doses of a polypeptide to normal human volunteers. Clin Pharmacol Ther. 1988;44:607-612.

[4] Kim A, Green PG, Rao G, Guy RH. Convective solvent flow across the skin during iontophoresis. Pharm Res. 1993;10:1315-1320.

[5] Thysman S, Preat V. In vivo iontophoresis of fentanyl and sufentanil in rats: pharmacokinetics and acute antinociceptive effects. Anesth Analg. 1993;77:61-66.

[6] Li LC, Scudds RA. Iontophoresis: an overview of the mechanisms and clinical application. Arthritis Care Res. 1995;8:51-61.

[7] Miyazaki S, Mizuoka H, Kohata Y, Takada M. External control of drug release and penetration, VI: enhancing effect of ultrasound on the transdermal absorption of indomethacin from an ointment in rats. Chem Pharm Bull (Tokyo). 1992;40:2826-2830.

[8] Byl NN, McKenzie A, Halliday B, et al. The effects of phonophoresis with corticosteroids: a controlled pilot study. J Orthop Sports Phys Ther. 1993;18:590-600.

[9] Mitragotri S, Edwards DA, Blankschtein D, Langer R. A mechanistic study of ultrasonically enhanced transdermal drug delivery. J Pharm Sci. 1995;84:697-706.

[10] Mitragotri S, Blankschtein D, Langer R. Ultrasound-mediated transdermal protein delivery. Science. 1995;269:850-853.

[11] Singh P, Roberts MS. Skin permeability and local tissue concentrations of nonsteroidal anti-inflammatory drugs after topical application. J Pharmacol Exp Ther. 1994;268:144-151.

[12] Singh P, Roberts MS. Deep tissue penetration of bases and steroids after dermal application in rat. J Pharm Pharmacol. 1994;46:956-964.

[13] Banga AK, Panus PC. Clinical applications of iontophoretic devices in rehabilitation medicine. Critical Reviews in Rehabilitation Medicine. 1998;10:147-179.

[14] Glass JM, Stephen RL, Jacobson SC. The quantity and distribution of radiolabeled dexamethasone delivered to tissue by iontophoresis. Int J Dermatol. 1980;19:519-525.

[15] Byl NN. The use of ultrasound as an enhancer for transcutaneous drug delivery: phonophoresis. Phys Ther. 1995;75:539-553.

[16] McNeill SC, Potts RO, Francoeur ML. Local enhanced topical delivery (LETD) of drugs: Does it truly exist? Pharm Res. 1992;9: 1422-1427.

[17] Grahame R. Transdermal non-steroidal anti-inflammatory agents. Br J Clin Pract. 1995;49:33-35.

[18] Insel P. Analgesic-antipyretic and anti-inflammatory agents and drugs employed in the treatment of gout. In: Hardman J, Limbird L, Molinoff PB, et al, eds. Goodman and Gilman's: The Pharmacological Basis of Therapeutics. 9th ed. New York, NY: McGraw-Hill Inc; 1996:617-658.

[19] Ketoprofen--American Hospital Formulary Services: Drug Information. Bethesda, Md: American Society of Health-System Pharmacists Inc; 1996:1416-1422.

[20] Jamali F, Brocks DR. Clinical pharmacokinetics of ketoprofen and its enantiomers. Clin Pharmacokinet. 1990;19:197-217.

[21] Chi SC, Jun HW. Anti-inflammatory activity of ketoprofen gel on carrageenan-induced paw edema in rats. J Pharm Sci. 1990;79:974-977.

[22] Ballerini R, Casini A, Chinol M, et al. Study on the absorption of ketoprofen topically administered in man: comparison between tissue and plasma levels. Int J Clin Pharmacol Res. 1986;6:69-72.

[23] Panus PC, Campbell J, Kulkarni B, et al. Effect of iontophoretic current and application time on transdermal delivery of ketoprofen in man. Pharmaceutical Sciences. 1996;2:467- 469.

[24] Panus PC, Campbell J, Kulkarni SB, et al. Transdermal iontophoretic delivery of ketoprofen through human cadaver skin and in humans. Journal of Controlled Release. 1997;44:113-121.

[25] Matucci-Cerinic M, Casini A. Ketoprofen vs etofenamate in controlled double-blind study: evidence of topical effectiveness in soft tissue rheumatic pain. Int J Clin Pharmacol Res. 1988;8:157-160.

[26] Baixauli F, Ingles F, Alcantara P, et al. Percutaneous treatment of acute soft tissue lesions with naproxen gel and ketoprofen gel. J Int Med Res. 1990;18:372-378.

[27] Airaksinen O, Venalainen J, Pietilainen T. Ketoprofen 2.5% gel versus placebo gel in the treatment of acute soft tissue injuries. Int J Clin Pharmacol Ther Toxicol. 1993;31:561-563.

[28] Singh P, Roberts MS. Iontophoretic transdermal delivery of salicylic acid and lidocaine to local subcutaneous structures. J Pharm Sci. 1993;82:127-131.

[29] Netter P, Bannwarth B, Lapicque F, et al. Total and free ketoprofen in serum and synovial fluid after intramuscular injection. Clin Pharmacol Ther. 1987;42:555-561.

[30] Panus PC, Tober-Meyer B, Ferslew KE. Tissue extraction and high-performance liquid chromatographic determination of ketoprofen entantiomers. J Chromatogr B Biomed Appl. 1998;705:295-302.

[31] Kleinbaum DG, Kupper LL. Applied Regression Analysis and Other Multivariable Methods. North Scituate, Mass: Duxbury Press; 1978.

[32] Riviere JE, Sage B, Williams PL. Effects of vasoactive drugs on transdermal lidocaine iontophoresis. J Pharm Sci. 1991;80:615-620.

[33] Sanderson JE, de Riel S, Dixon R. Iontophoretic delivery of nonpeptide drugs: formulation optimization for maximum skin permeability. J Pharm Sci. 1989;78:361-364.

[34] Srinivasan V, Higuchi WI, Sims SM, et al. Transdermal iontophoretic drug delivery: mechanistic analysis and application to polypeptide delivery. J Pharm Sci. 1989;78:370-375.

[35] Srinivasan V, Su MH, Higuchi WI, Behl CR. Iontophoresis of polypeptides: effect of ethanol pretreatment of human skin. J Pharm Sci. 1990;79:588-591.

[36] Banga AK, Chien YW. Iontophoretic delivery of drugs: fundamentals, developments, and biomedical applications. Journal of Controlled Release. 1988;17:1-24.

[37] Varghese E, Khar RK. Enhanced skin permeation of diclofenac by iontophoresis: in vitro and in vivo studies. Journal of Controlled Release. 1996;38:21-27.

[38] Helak JW, Reichek N. Quantitation of human left ventricular mass and volume by two-dimensional echocardiography: in vitro anatomic validation. Circulation. 1981;63:1398-1407.

[39] Radermacher J, Jentsch D, Scholl MA, et al. Diclofenac concentrations in synovial fluid and plasma after cutaneous application in inflammatory and degenerative joint disease. Br J Clin Pharmacol. 1991;31:537-541.

All animal experimentation in this study received prior approval from the Institutional Animal Care and Use Committee of East Tennessee State University.

This research was supported by East Tennessee State University Research and Development Committee Grants 96-001/GIA and 96-065/MJR) and by endowments from EMPI Inc (Minneapolis, Minn) and Waters Corporation (Milford, Mass).

This article was submitted December 31, 1997, and was accepted September 20, 1998.

PC Panus, PhD, PT, is Assistant Professor, Department of Physical Therapy, College of Public and Allied Health, East Tennessee State University, Box 70624, Johnson City, TN 37614-0624 (USA) ([e-mail]). Address all correspondence to Dr Panus.

KE Ferslew, PhD, is Professor, Section of Toxicology, Department of Pharmacology, James H Quillen College of Medicine, East Tennessee State University.

B Tober-Meyer, DVM, is Director, Division of Laboratory Animal Resources, James H Quillen College of Medicine, East Tennessee State University.

RL Kao, PhD, is Professor, Department of Surgery, James H Quillen College of Medicine, East Tennessee State University.