The use of ultrasound as an enhancer for transcutaneous drug delivery: phonophoresis.

Key Words: Phonophoresis, Utrasound.

Recently, there has been a growing interest in simulating the advantages of intravenous drug infusion by using the intact skin as the port of drug administration.[1-3] Although topical drugs are usually applied at a convenient site, the target for the drug interaction may be systemic (eg, scopolamine to prevent motion sickness) or local (eg, cortisone creams, counterirritants, analgesics). Regardless of the target, all transcutaneous drugs that penetrate the skin and enter the vascular system will have some systemic effects.

The major barrier to the delivery of transcutaneous drugs is the skin.[1-4] Pharmaceutical companies are continually involved in research to try to find new ways to enhance the delivery of topical drugs.[4] Although complex chemical enhancers have been integrated into some transdermal delivery systems, physical agents such as electricity and ultrasound (US) are becoming increasingly popular as enhancers.[5,6] The use of electricity as an enhancer is referred to as iontophoresis, and the use of US as an enhancer is referred to as phonophoresis or sonophoresis.

Despite the frequency phonophoresis is used in physical therapy clinics,[7] questions remain regarding treatment validity and effectiveness. The purpose of this article is to provide a discussion of the principles of transdermal drug delivery, types of drugs frequently used in phonophoresis, local and systemic effects of phonophoresis, principles of US as an enhancer of topical drug absorption, effectiveness of US as an enhancer of topically applied drugs, clinical implications of research findings, and recommendations for future research.

Transcuteneous Drug Delivery

Phonophoresis and iontophoresis are transdermal drug delivery systems that use physical agents to enhance the delivery of topically applied drugs. Topically applied drugs avoid the risk and inconveniences of intravenous therapy, bypass the liver in terms of elimination, provide less chance of an overdose or underdose, allow easy termination (eg, remove the drug fro the skin), and permit both local and systemic treatment effects.[1-3] The therapeutic effects of topically applied drugs will depend on a variety of factors, including the rate, the amount, and the depth to which the drugs penetrate the skin (the process of entrance of a substance into a layer) and the potential toxicological hazards of the drugs on the tissues.[2,8-10]

Compounds are thought to transfer through the skin by a predictable system of passive diffusion, which is defined by Fick's Law and the rate of permeation.[11,12] Diffusion for most low molecular weight substances (movement of molecules through a membrane) seems to occur uniformly through the stratum corneum over a large fraction of its area.[9] Percutaneous absorption of a topically applied drug (the removal of the penetrant and its metabolites from the dermis via circulation) involves a sequence of individual transport processes.[11] Once diffusion through the stratum corneum is achieved, the molecules permeate the dermis (the migration of a molecule through one of several layers), are absorbed (the sum of all of the processes of molecular movement through a surface, including penetration, permeation, and resorption) into the capillary plexus, and are then transferred to the circulating bloodstream.[11]

Drugs Commonly Used in Transdermal Drug Delivery

Drugs used in transdermal drug delivery systems have a variety of release mechanisms,[12] with a specific target (eg, skin, heart, vagina, uterus, mouth, nose, eyes). Most commonly, the drugs are used to reduce the symptoms of motion sickness, angina pectoris, high blood pressure, and menopause or to provide a method of birth control. The target of most transdermal drug delivery systems is usually systemic. In rehabilitation, however, the target of transdermal drug delivery systems is usually local, with the three most common types of topical drugs used in treatments such as phonophoresis being (1) anesthetics (substances such as lidocaine that block pain receptors by creating numbness), (2) counterirritants (substances such as menthol that cause inflammation of the skin for purposes of relieving pain from stimulation rather than depression of cutaneous sensory receptors), and (3) anti-inflammatories such as nonsteroidal medications (eg, salicylates) or steroidal medications (eg, hydrocortisone, dexamethasone).[5,13]

The Structure and Anatomy of the Skin Relevant to Phonophoresis

The skin is one of the most readily accessible organs of the human body, covering a surface area of approximately 2 [m.sup.2] and receiving about one third of the body's blood circulation.[14,15] The skin is a complex system consisting of the epidermis, the dermis, and the skin appendages interwoven within the two layers (Fig. 1). The outermost layer of the skin, the epidermis, is avascular, receiving nutrients from the underlying dermal capillaries by diffusion through a basement membrane.

The outermost layer of the epidermis is called the stratum corneum, the protective covering that serves as a barrier to prevent desiccation of the underlying tissues and to exclude the entry of noxious substances from the environment, including agents applied to the skin.[14,15] In transcutaneous drug delivery, the stratum corneum is considered the rate-limiting barrier and the target of most enhancers, including US (Fig. 2).[1,2,4]

The skin has selective permeability, which is determined by the nature of its physiochemical condition (eg, thickness, lipid structure, membrane capacity), the viscosity, the extent of the cross-linking of collagen, the extent of the skin appendages (eg, hair follicles, apocrine sweat glands, erector pili muscles, sebaceous glands), the age of the skin, and the state of disease.[11,12,16] The stratum corneum is neither continuous nor homogeneous throughout the body.[9,17] The cells of the stratum corneum are stratified and roughly hexagonal (corneocytes), overlapping at the edges with neighboring cells. The projected areas of the corneocytes vary somewhat in different areas of the body.[14] These cells are continually being transformed from living cells to dead, flattened, scalelike structures (keratinization), which are naturally sloughed. The hairy extensor surface of the skin is generally delicate, thin, and well supplied with sebaceous glands and hair cell follicles. On the glabrous or flexor surfaces, the skin is thicker and more richly supplied with sweat glands. As a consequence of these variations, the effectiveness of transcutaneous drug delivery will depend on the anatomical area treated, the hydration of the skin, the presence of fatty acids, the health or pathological condition of the skin, the state of cutaneous and systemic metabolism, and patient age.[11,18-26]

Hydration of the stratum corneum is critical for effective transcutaneous drug delivery. In asymptomatic individuals, the stratum corneum is always partially hydrated, with an average water concentration of approximately 0.90 [g.sup.-1] of dry tissue.[17] This amount of water increases the permeability of the stratum corneum to hydrophylic drugs approximately tenfold. If the water content is maintained, the stratum corneum could ultimately absorb three to five times its own weight of bound water, resulting in an additional twofold to threefold increase in the permeability to water and other polar molecules. The use of lightly applied occlusive dressings, which seal a wound and prevent moisture from escaping from the wound, further enhances the hydration.

The human skin changes dramatically with age.[11,18-22] The aged stratum corneum is considerably dryer than that of a young adult, contains a lower lipid content, and frequently has reduced microcirculation.[27] Exactly how the penetration barrier of the skin changes with age, however, is not clear. Thus, it should be assumed that the aged skin provides a more variable environment for the diffusion of topically applied drugs[11,18-22] even when patient compliance is high.

Skin denuded of the stratum corneum (as occurs with a minor abrasion, with dry skin, or from shaving) has a greater potential for allowing drug diffusion. Any process that denudes the stratum corneum, however, stimulates the production of keratinized cells, which restores the stratum corneum to almost full protective capacity in 2 to 3 days.[12] On the other hand, pathological conditions of the skin may disrupt the stratum corneum for extended periods of time (eg, a decubitus ulcer),[8] increasing the potential for absorption of topically applied drugs.

Molecules of drugs can penetrate the epithelium transcellularly or interceuularly through channels between cells, but diffusion is easiest through the hair follicles, the sebaceous glands, and the sweat ducts.[10] Given the small cross sections of the sweat and sebaceous pores along with the outward sloughing movement of sweat or sebum, however, the hair follicles serve as the primary means of drug diffusion.

Human skin has a reservoir capacity,[26-29] which was originally demonstrated by Guillot.[23] He showed that topically applied salicylic acid was excreted in the urine for many hours (even days), whereas an intradermal injection of salicylic acid was associated with rapid and complete excretion within a few hours. This reservoir effect is seen when an occlusive bandage is applied following the administration of a topical drug. Guy and Hadgraft,[28] for example, demonstrated that corticosteroid enhanced healing in patients with discoid eczema and psoriasis, but only for 24 hours when the treated area was left uncovered. When the area was covered with an occlusive dressing, healing was facilitated for 48 to 72 hours, even though no additional corticosteroid was added. Apparently, some drug molecules remain in the skin and can be activated by a moist environment.

The state of the vascular network of the dermis will affect transcutaneous drug delivery. Constricted blood flow will limit the opportunity for systemic delivery of a drug. Heating prior to topical application of a drug will dilate the hair follicles, increase kinetic energy and the movement of particles in the treated area, and facilitate drug absorption. Heating the skin after the topical application of a drug will increase drug absorption into the vascular network,20 enhancing the systemic delivery but decreasing the local delivery as the drug molecules are carried away from the local delivery site.

In summary, the diffusion rate of topically applied drugs will vary because of both internal (physiological) and external (environmental) factors. The patient's skin needs to be carefully evaluated to minimize the natural, internal barriers to trancutaneous drug delivery (eg, dry skin, thick skin, dehydration, poor circulation, poor metabolism) and to maximize the natural enhancers (eg, ensuring the patient is well hydrated and selecting an area of skin that is thin, warm, moist, and well perfused). Diffusion of topically applied drugs through the stratum corneum can also be enhanced by preheating the skin to increase kinetic energy and dilate hair follicles, as well as by covering the area with an occlusive dressing after the drug application to maintain moisture and activate the reservoir capacity of the skin.

Local Versus Systemic Effects of Transdermal Drugs

Topically applied drugs are categorized according to their intended site of action: systemic or local.[30-36] Systemic drugs diffuse through the epidermis to the dermis to reach the capillary network. Drugs with local targets difuse into the area immediately below the administration site: subcutaneous tissue, muscle, synovium, ligaments, tendons, and joints. For years it was thought that topically applied drugs all entered the capillary network, became systemic, and then returned to the local area through the bloodstream.[35] Research[30-36] has shown a system of local delivery that is separate from systemic delivery. The local and systemic delivery systems can be distinguished by noting the time required for peak absorption of the drug in the target tissue (drug concentration). Although these studies were carried out using topically applied drugs without US, the findings are equally appropriate for understanding the distribution patterns for topically applied drugs enhanced with US.

McNeill et al[35] used radiolabeled piroxicam to study 36 Sprague-Dawley rats that received doses of 50 [mu]g of 3H-piroxicam by either an intravenous injection or a topical application. Prior to topical application, the hair was clipped with animal shears, taking care not to injure the skin surface. Topical piroxicam was administered in 10 mg of a Carbopol-based gel, with a final concentration of the drug of 0.5%, covering an area about 1 [cm.sup.2] on the shoulder. At preselected times after administering the dose (2, 4, 6, 8, 12, 16, 24, and 48 hours), the rats were lightly anesthetized and blood samples were taken along with samples of skin and muscle from the shoulder.

After topical administration, the plasma reached a maximum drug concentration at 12 hours, then decreased. The muscles excised from the right shoulder showed that the drug level was greatest at the 2-hour time point, implying that the maximum concentration of the drug in the local tissue occurred before the first blood sample was taken. The change in concentration appeared to parallel that of the intravenous plasma, suggesting that the blood and muscles were in a state of pseudo-equilibrium. Two additional distinct peaks in drug concentration were also seen in the deltoid, trapezius, and supraspinatus muscles after topical application. One peak was seen at 4 hours, and a second peak was seen at 12 hours. The increased concentration at 12 hours corresponds to the time at which the maximum drug levels were observed in the plasma. The presence of the earlier peak is not consistent with piroxicam derived from the systemic blood. This study confirms there is a route for topical drug delivery that delivers drug molecules directly to the muscle. The drug levels in the untreated muscles were threefold to fivefold lower than the drug levels observed under the treated site, and the overall shape of the profiles of the untreated muscles resembled that of the topical plasma, with a single peak at 12 hours. The difference in the concentration time profiles and the ratios for the treated and untreated sites supports the hypothesis that topical administration can lead to enhanced local drug concentrations in adjacent tissues.35 it was hypothesized that a "convective" physiological force transported the drug molecules in a relatively short period of time. The cutaneous microvasculature may contribute, at least in part, to this hypothesized convective process.

Most phonophoresis studies focus on the measurement of local effects. Two recent studies, however, measured both local and systemic effects. Bommannan and colleagues[37,38] examined both local (drug penetration in the stratum corneum) and systemic effects (urinalysis) following 5 to 20 minutes of sonication at 10 and 16 MHz. Increased blood levels of the drug were measured at 5 and 20 minutes; however, after 20 minutes of sonication, destructive effects (eg, vacuoles in the stratum corneum) were noted in the local treatment area.

More recently, Wong and Byl[39] reported on the systemic and local effects of phonophoresis with dexamethasone in Sprague-Dawley rats (conference abstract of unpublished study). Subcutaneous collagen deposition and breaking strength of treated incisions were measured for local effects. Systemic effects were measured by a change in body weight, weight of the thymus gland, weight of the adrenal glands, and the amount of collagen deposited in a wound chamber.

In this study, there were systemic and local effects in the three groups that received the topical drug applications. The phonophoresed groups showed more local effects when compared with the group that received the topical application of dexamethasone with sham US. There were no differences in systemic effects among the three treatment groups. In addition, there were also no differences between the two phonophoresed groups (continuous US versus pulsating US). The large size of the treatment area and the initial pretreatment of the skin with a dipilatory (hair-removal cream) may have facilitated both the local and the systemic effects.

Ultrasound as an Enhancer of Transcutaneous Drug Delivery

Enhancers of transcutaneous drugs are used to alter the nature of the stratum corneum to ease diffusion.[6,15,16] This alteration may result from denaturing the structural keratin proteins in the stratum corneum, stripping or delaminating the cornified layers of the stratum corneum, changing cell permeability, or altering the lipid-enriched intercellular structure between the corneocytes. Enhancers may increase the permeation rates fivefold to tenfold or substantially reduce the duration of the time lag prior to drug activation. Enhancers are incorporated into transdermal-controlled drug delivery systems, or they are used prior to, during, or after the topical application of a drug.[4,8] Preferred enhancers allow drugs to diffuse actively and quickly, but do not inactivate the drug molecules, damage healthy epidermis, cause pain, or have "toxicological" side effects.[9,16]

Even though US has been used extensively over the last three decades for medical diagnostics and physical therapy, it has only recently become popular as an enhancer of drug delivery.[5] The first published report on the use of US for increasing drug flux across the skin appeared in 1954.[40] Numerous studies[37,38,41-51] have since demonstrated that US is generally safe, with no negative long- or short-term side effects, but the mechanisms by which US works as an enhancer are less dearly understood.

Ultrasound waves are sound waves that are above the audible limit (>20 kHz).[46,48,49,52,53] The properties of the US are described by the amplitude and the frequency of the longitudinal waves. Similar to audible sound, US waves undergo reflection, refraction, or absorption when they encounter another medium with dissimilar properties.[46,48,53] If the properties of the encountered medium are different from those of the transmitting medium, the acoustic energy of the transmitted US beam is attenuated. The attenuation of US in tissue limits its depth of penetration. This principle would apply both for US administered with conductive gel alone or for US administered with a gel that includes an active drug agent. Typically, using a conductive, water-soluble gel, US at 3 MHz has been reported to penetrate 1 to 2 cm, whereas US at 1 MHz has been been reported to penetrate 2 to 4 cm.[19,52,54]

Both the thermal and nonthermal characteristics of high-frequency sound waves can enhance the diffusion of topically applied drugs. Heating from US increases the kinetic energy of the molecules in the drug and in the cell membrane, dilates points of entry such as the hair follicles and the sweat glands. and increases the circulation to the area sonicated. These physiological changes enhance the opportunity for drug molecules to diffuse through the stratum corneum and be collected by the capillary network in the dermis. Both the thermal and nonthermal effects of US increase cell permeability. The mechanical characteristics of the sound wave also enhance drug diffusion by oscillating the cells at high speed, changing the resting potential of the cell membrane and potentially disrupting the cell membrane of some of the cells in the area.43 There may be some pushing and pulling of the cells with the propogation of the sound wave through incongruous tissues, but it is unlikely that radiation or streaming forces are forceful or consistent enough to push drug molecules into the tissue.

Fellinger and Schmid[40] studied US as an enhancer of drug diffusion by determining whether. the sound waves acted directly on the skin or acted on the release of a drug from a gel formulation. When drug efflux was measured from the gel in the presence or absence of US, no evidence was found that US facilitated the release of the drug from the gel.[40] When the skin was pretreated with US prior to the topical application of a drug, however, enhanced drug transport was measured, supporting the principle that US enhances drug diffusion by acting primarily on the skin.[37-40]

Electron microscopy was used to track the skin permeation of an electrondense tracer (lanthanum nitrate [LN]) to determine the permeation pathway of the diffusing molecules and the morphology of the skin under the influence of US.[38] In this study[,38] (1) LN did not naturally permeate the skin by passive diffusion; (2) LN traversed the skin along an intercellular route under the influence of US; (3) enhancement occurred in only 5 minutes as determined by the detection of a small amount of LN in the dermis; (4) the stratum corneum and the cells of the epidermis did not appear to be adversely affected by either 5 or 20 minutes of US treatment at 2, 10, and 16 MHz; and (5) after 20 minutes of US at 16 MHz, cellular morphology was altered, with structural alterations of the stratum granulosum and stratum basale cells noted as large vacuoles. Bommannon and colleagues[38] concluded that high-frequency US was effective as an enhancer of transcutaneous drugs and could be used safely for short periods of time. When 16-MHz US is used for more extensive periods of time, it is possible that the bubbles of the micronuclei begin to grow, then collapse and self-destruct (cavitation), secondarily enhancing drug penetration.[38]

The simplest explanation for the effectiveness of US as an enhancer of drug delivery is based on its heating effects. Heat increases the kinetic energy of both the drug molecules and the proteins, lipids, and carbohydrates in the cell membrane. The thermal effects of sonic energy are easily monitored through temperature measurements. Although a formula for transient heat flow can be used to estimate the conversion of some of the sonic energy to heat,[38] this formula does not allow the determination of the proportion of the heat that is lost by conduction, convection, and radiation. Temperature changes of approximately 5[degrees]C are necessary to cause measurable changes in cell membrane permeability.[38] This level of increase in heating only predictably occurs when the US intensity is 1.5 W/[cm.sup.2] or higher.[46,53] The thermal elements can be reduced by using a pulsed mode of US or by using a very low intensity (eg, 0.5 W/[cm.sup.2] or less).[46,48,53]

Sonic waves are associated with nonthermal, mechanical effects such as oscillation, radiation pressure, and cavitation of cells as the waves pass through the tissues.[19,41,43] These mechanical effects may facilitate drug diffusion by oscillating the particles in the tissue and drug media, decreasing membrane potential, altering the lipid structure, increasing cell permeability, increasing ion conductance, or disrupting the cell membrane.[37,38,41-51] Some of these effects are similar to those that result from heating.

Mechanically, when US passes through an enclosed medium, the particles in the medium are subjected to periodic oscillations about their resting positions.[38,41-51] The maximum displacement of a particle is proportional to the intensity (energy content of the US beam that is equal to the square root of the incident intensity) modified by the frequency of the wave.[38,41-51] The particles vibrate faster at high frequencies but require the input of more energy to achieve the same maximum displacement amplitude. At any given intensity, the maximum displacement amplitude (A) of the particles of the medium decreases with increasing frequency (f) (A=1/f). A sound-wave frequency of 1 MHz at an intensity of 1.0 W/[cm.sup.2] has an amplitude of 0.018 [mu]m with a maximum particle acceleration equal to 71,000 times the acceleration due to gravity on earth.[48]

For the most part, the particles assume their original position after the wave has departed. Ultrasonic waves, however, do have some radiation pressure forces that could effectively push or pull the cells and gas bodies in a nonhomogeneous acoustic field.[19,52] In addition, when cell membrane permeability is enhanced, some acoustic streaming (steady circulatory flow) is created as cells diffuse through the membrane. Although the streaming only occurs secondary to the change in cell permeability, how the radiation forces are distributed within the irradiated tissues and whether these forces predictably move molecules are not clear.48 Research is needed to determine whether the small amount of sonic radiation "pushes" drug molecules into the tissues. At this time, there is insufficient evidence to claim that sonic radiation pressure pushes the drug molecules through the tissues. The most likely mechanical explanation of increased drug absorption with US is based on the enhanced intercellular diffusion resulting from the near-simultaneous, high-speed vibration of the drug molecules along with the vibration of the cell membrane and its components.

Another important factor that could affect drug diffusion is related to the shear forces (or shock waves) that occur when adjacent portions of the same membranous structures vibrate with different displacement amplitudes.[46] Highly compressible bodies such as gas, vapor-filled "bubbles," or "cavities" are powered by and extract energy from the incident acoustic field.[51] When these bubbles occur in specific cells of the skin, muscle, nerve, artery, or vein, fatigue or rupture of the cells can occur as the bubbles reach an unstable size. Destruction of cells in the transmission path of the US could facilitate intercellular diffusion of drug molecules.

Cavitation can be stable or transient. Stable cavitation occurs when a bubble oscillates in a radial mode about a resonance size for a number of cycles without leaving the field (the original location of the cell).[19] Transient cavitation occurs during the compression phase when media experience tension stress during a portion of the rarefaction phase of the acoustic disturbance, followed by a rapid collapse. This transient cavitation is known to occur in body tissues, but only at the higher frequencies (5-16 MHz).[19,38] Evidence for stable cavitation in tissues is lacking when US is used at therapeutic levels (870 Hz-3 MHz).[49,52,54,55] However, in the blood, a high intensity of US (130-260 W/[cm.sup.2]) can produce such cavitation.[19]

Some of the energy from cavitation (about 10%) is reradiated as an acoustic wave.[19] The remaining energy is transformed into other forms of energy such as heat, shock waves, or hydrodynamic shear fields, which can disrupt biological tissues and facilitate diffusion.[19] The strains that act on these structures are governed by the relation between the size of the structures and the wavelength of US. If the particle size is smaller than the wavelength of the impinging beam, the probability of unequal stress and strain distribution is very low.[37] Ultrasound with a frequency of 10 MHz has been calculated to have a one-half wavelength of 75 [mu]m.[37] Thus, structures having dimensions of this order or larger will be subjected to cylindrical strains equal to twice the maximum particle displacement amplitude. The corneocytes of the stratum corneum, however, are smaller than 75 [mu]m. Therefore, it is unlikely that these structures are affected by cylindrical strains associated with cavitation.37

There are a few in vitro studies that have investigated the effects of different intensities and different frequencies of US on cavitation, but none could be found that directly correlated cavitation with phonophoresis. In one unpublished in vitro study of fibroblasts, a 3-minute continuous US exposure at 0.5 to 1.0 W/[cm.sup.2] at 1 MHz had an explosive effect on the cell colony in a petri dish, destroying the cells and leaving the fragments pushed to the edges of the dish (conference abstract).[56]

High-frequency sonic waves (>3 MHz) deposit more energy in a given volume of tissue than low-frequency sonic waves due to increased particle oscillation,[38] although the frequency also affects the depth of penetration of the sound wave. Minimal research has been done on the unique effects of phonophonoresis on the depth of drug penetration into different tissues as correlated with different frequencies Of US.

Literature Review: Effectiveness of Phonophoresis

Overall, the reviews of the research on the efficacy of phonophoresis are promising. Some of the studies, however, suffered methodological constraints that limit generalizability (eg, no control group, the experimenters were not blinded, the US delivery system was not calibrated, methods of documenting effectiveness were not objective, a small number of subjects were studied, or the topical drug used was not checked to ensure that it transmitted US [transmissivityl). This review will summarize the findings of these research studies by type of drug used.

Phonophoresis With Corticosteroids 

In the majority of the studies on phonophoresis, US was used to enhance the delivery of steroidal anti-inflammatory drugs (eg, hydrocortisone). Ultrasound was used as an enhancer of transcutaneously applied steroids as early as 1954 when Fellinger and Schmid[40] showed that US could carry hydrocortisone across an avascular membrane for the effective treatment of polyarthritis of the hand. In the mid-1960s, Griffin et al[57] published a double-blind study involving hydrocortisone phonophoresis once a week for 3 weeks for patients with osteoarthritis, periarticular arthritis, or joint or muscle pathology. They demonstrated that 68% of the subjects receiving hydrocortisone in addition to US (1.5 W/[cm.sup.2], 1 MHz, continuous, 5 min/25 [in.sup.2] of painful tissue) had a decrease in pain and an increase in range of motion compared with a similar gain for 25% of those receiving a placebo drug in addition to US.

In later animal studies, Griffin and colleagues[58-61] reported that US could enable cortisone to penetrate skeletal muscle and paravertebral nerves of swine under a variety of conditions i.0 or 3.0 W/[cm.sup.2] for 5 minutes, 0.3 W/[cm.sup.2] for 17 minutes, or 0.1 W/[cm.sup.2] for 51 minutes at frequencies of 0.09, 0.25, 0.5, 2.0, or 3.6 MHz). The swine's hair was cut with scissors (not shaved), and 25 mg (at 3.0 W/[cm.sup.2]) or 100 mg (for all other intensities) of cortisol was applied over a 5-[cm.sup.2] area of skin in the paravertebral area. The animals were treated with a stationary sound head. Compared with controls, those animals treated with cortisol plus US had the highest concentration of cortisol. The highest concentrations of cortisol were noted in the neural tissues (146% increase), followed by the muscle (1000/o increase). No increases were measured in systemic cortisol. The concentration of cortisol in muscle was similar for animals treated with 25 mg of cortisol plus US at an intensity of 1.0 W/[cm.sup.2] and for animals treated with 100 mg of cortisol plus US at an intensity of 0.1, 0.3, or 1.0 W/[cm.sup.2]. The amount of cortisol in the muscle was greatest when the animals were treated at the lowest intensity for the longest period. The amount of cortisol in the neural tissue was also highest in animals treated with the low doses (0.1 W/[cm.sup.2]) for 51 minutes. At the time of sacrifice, however, in those animals treated with 3.0 W/[cm.sup.2], second-degree bums up to 1.0 cm were noted in the skin. This tissue damage was thought to be due to excessive heating and excessive cell oscillation, potentially resulting from standing waves created from the use of a stationary sound head. Once the structure of the stratum corneum was destroyed, topically applied cortisol could diffuse easily into the tissue even without the US. The goal is to ensure that similar levels of diffusion could occur without disruption of the stratum corneum. Therefore, the results of this study have questionable clinical applicability. In clinical practice, therapists cannot damage tissues in order to facilitate drug diffusion.

In 1975, Kleinkort and Wood[62] reported a study with subjects with a variety of inflammatory conditions (tendinitis of the elbow, subdeltoid bursitis, plantar fasciitis). They found that a 10% hydrocortisone preparation enhanced with US was more effective in reducing pain from tendinitis or bursitis than a 1% preparation enhanced with US. There was no measurable improvement in patients with plantar fasciitis.

In a controlled study with dogs, Davick and colleagues[63] measured the local effects of US (0.5 W/[cm.sup.2], 870 Hz, continuous, 8 minutes) as an enhancer of radiolabeled cortisol (5% and 100/o). They found that US enhanced the absorption of cortisol through the stratum corneum into the epidermis, but none of the radiolabeled drug was identified in the knee joint or in the muscle. There was no difference between the 5% and 100/o cortisol. In this study, only a single treatment was given, and the were sacrificed immediately after treatment. Thus, the full local effects might not have been achieved, and systemic effects could not have been measured in such a short time frame.[35]

In 1987, McElnay and associates[64] described research on phonophoresis using fluocinolone acetonide (0.25%), a topical corticosteroid. They measured the penetration of the drug into the forearm by a blanching test. The drug was applied for 5 minutes before sonication. Then US was applied as an enhancer for another 5 minutes of sonication (2.0 W/[cm.sup.2], 870 kHz, continuous), followed by a 3-hour period of occlusion with an airtight bandage that maintained hydration by preventing evaporution and drying of the area. Preliminary assessment of the subjects indicated they were responsive to the blanching test. After the US treatment, blanching occurred only when the corticosteroid was topically applied but not when petroleum jelly was applied. in this double-blind, crossover study, the researchers found that US was associated with an increase in drug absorption.[64]

Based on an overview of the research literature on hydrocortisone phonophoresis in 1992,[65] Newman et al suggested that hydrocortisone phonophoresis is a useful modality in the treatment of numerous musculoskeletal injuries. These reviewers interpreted their review broadly by concluding that phonophoresis accelerated healing, increased flexibility, and decreased pain. Newman and colleagues recommended that podiatrists consider phonophoresis more frequently in the care of patients with foot problems.

Kamenskaia and Fedorova[66] studied hydrocortisone phonophoresis and its effects on joints compared with balneotherapy (the use of baths in the treatment of disease), specifically using iodine bromine alone or a combination of the phonophoresis and the balneotherapy. Based on this study of 197 patients with osteoarthrosis, the combined application of the bath and hydrocortisone phonophoresis was associated with the best relief, particularly in the most involved joints, Unfortunately, the variables of the US used in the study were not provided in the English abstract.

Not all of the phonophoresis studies involving topically applied steroids enhanced with US have shown positive effects. in a clinical study of patients with shin splints by Smith et al,[67] the effects of adjunctive ice, ice massage, US, iontophoresis, or phonophoresis added to a program of gait training and strengthening exercises were compared. For iontophoresis and phonophoresis, a mixture of 33 mg of Decadron with 16 mL of 2% lidocaine gel in 60 mg of water-soluble base was applied. Ultrasound was administered at 1.5 W/[cm.sup.2] at 1 MHz for 5 minutes. These researchers found that all of the groups had a reduction in symptoms of pain compared with a control group, but there were no differences among the various treatment groups. No objective measurements of drug penetration were made in this study, and information on transmissivity of the drug to US was not reported.

This review of studies using US to enhance the diffusion of hydrocortisone is somewhat confusing, given recent studies[68-70] indicating that some premixed solutions of hydrocortisone (5% or 10%) do not trasmit US. If the sound waves are not transmitted, then phonophoresed hydrocortisone and topically applied hydrocortisone should have the same effect. If hydrocortisone is the preferred drug in terms of effect, then a more conductive carrying agent is needed to allow the transmission of US.

Phonophoresis With Salicylates

Salicylate creams are commonly purchased over the counter to help reduce inflammation and pain. Salicylate is reported to penetrate the skin following topical application.[71-73] In combination with US, it is thought that salicylate could be moved into the deeper, subdermal tissues. In one study of seven asymptomatic human volunteers,[71] three phonophoresis treatments were administered using 10% tolamine salicylate (Myoflex) as the coupling agent. Continuous or pulsed (1:1) US was used at an intensity of 1.5 W/[cm.sup.2] at 1 MHz. A sham US treatment group was also included. All subjects received all treatments with the order of treatments randomized. All treatments were performed 1 week apart. The US machine was calibrated, but the transmissivity of the salicylate was not reported. An area on the right anterior forearm was marked (8.0 cm in diameter), and 1 tablespoon (12-13 g) of salicylate cream was applied. A sound head with a diameter of 5 cm was circled slowly for 5 minutes over the target area. Before treatment and 2 hours following each treatment, a 7.0-mL blood sample was drawn from the left arm. Peak salicylic acid levels were known to occur 2 hours following an oral dose of 1,000 mg of aspirin.[71]

Although no differences were found in salicylate blood levels before or after the different treatments, no explanation was given for the noticeable variability in baseline salicylate levels for the three groups before the treatment (61.3, 84.4, and 100.4 mg/[mL.sup.-1], respectively, for the groups receiving continuous, sham, and pulsed US).[73] These unusual baseline measurements were correlated with a large increase in the group treated with continuous US (79.6 ng/[m.sup.-1]), a small increase after sham US (89 ng/[mL.sup.-1], and a reduction in drug levels in the group treated with pulsed US (84.1 ng/ [mL.sup.-1]). The investigators did not check the salicylate for transmissibility.[64]

In 1991, Ciccone et al[74] reported on a study to evaluate US as an enhancer of topically applied salicylates (trolamine salicylate, 10%, a nonsteroidal anti-inflammatory drug). These researchers measured the clinical effects (delayed-onset muscle soreness and range of motion) of the drug as a confirmation of drug penetration. Treatments were given for 5 minutes at 1 MHz, 1.5 W/[cm.sup.2] (continuous). Bilateral, repeated eccentric contractions of the elbow flexors using a barbell weight equivalent to a one repetition maximum was the protocol used to induce the muscle soreness. Three sets of 10 contractions were performed with a 2-second rest. The effect of the treatment was measured using a visual analog scale to assess muscle soreness and a goniometer to measure range of motion (elbow flexion).

Peak values of soreness were reported at 3 days, and peak values of decreased range of motion were noted around days 2 and 3. In the control group, there were no differences in pain between the two arms, whereas in the other groups variations in pain and range of motion were reported between the two arms. In the US group, the soreness was greatest in the treated arm on day 3. Compared with the other groups, between days 2 and 4 the phonophoresed group showed the greatest reduction in muscle soreness in the treated arm compared with the control arm (a 45.8% reduction on day 4). There were variations in the range of motion for almost all of the treated and control arms in each of the groups.

The findings from this study are somewhat confusing. Ultrasound alone at a moderately high intensity was associated with increased pain compared with the control arm and the control group. When the salicylates were added to the sonication, the pain appeared to be controlled compared with the US group alone and similar to the salicylate group, but the salicylates did not seem to affect flexibility. The authors concluded that salicylates could offset the potential negative effects of US-induced pain when this modality is used to treat an acute "injury." Because no blood studies were done, however, it could not be determined whether the difference between the two sonicated groups (with and without salicylates) was due to the absorption of salicylates facilitated by US or to the reduced transmissivity of the salicylates, which would have blocked some of the US to the tissues.

Two important articles on sonophoresis were published by Bommannan and colleagues in 1992.[37,38] These investigators defined sonophoresis as the enhanced flux of molecules due to ultrasonic perturbation, the equivalent of phonophoresis. The investigators reported on the results of the absorption of salicylates topically applied to the flank of hairless guinea pigs when enhanced with US.[37] These researchers also analyzed the physiological mechanisms involved.[38] They mixed their radioactively labeled salicylic acid with 0.5 g of Carbopol and 0.5 g of sodium hydroxide (leading to a completely ionized solution at a pH of 5.5). The mixture was checked for transmissivity of US. They used US at 0.2 W/[cm.sup.2] for 5 or 20 minutes at three different frequencies (2, 10, and 16 MHz). They did not indicate whether the sound head was mobile or stationary. All animals received all treatments, with a 2-week interval between treatments. Drug penetration was quantified in two ways: (1) Sequential tape strips of the salicylate cream were removed using adhesive tape, and the amount of radioactivity recovered from each strip was determined, and (2) urine was collected periodically and monitored for radioactivity for 1 to 3 days postsonication. The flank was treated.

The investigators found no differences in penetration as measured by the tape or the urine in those treatment sessions with 2 MHz at 0.2 W/[cm.sup.2]. There were differences in the urine and skin between the treatment and control groups at 10 and 16 MHz. There was a difference in penetration between those subjects treated for 20 minutes compared with those treated for 5 minutes with either the 10- or 16-MHz frequency. Unfortunately, when the stratum corneum was analyzed after treatment with a 10- or 16-MHz frequency for 20 minutes, large vacuoles were noted in the stratum corneum. This finding was thought to represent a burning of the tissues. At this time, US above 3 MHZ is not available to the clinical practitioner. These investigators are working with industry to develop therapeutic devices with the capability to deliver high-frequency US (>10 MHz). Even at very low intensities, however, phonophoresis at a frequency of 10 or 16 Mhz appears to be therapeutically safe only if the application is less than 20 minutes and the intensity is very low (0.2 W/[cm.sup.2]).

Phonophoresis With Anesthetics

The effectiveness of phonophoresis has been studied extensively with the application of anesthetics such as lidocaine. In 1964, using the quadriceps femoris muscle of rabbits, Novak[75] reported that US (2.0 W/[cm.sup.2] for 5 minutes) could increase the absorption of lidocaine (4%) even without injection. Unfortunately, he failed to include a control group that received a topical application of the drug without US or a group that received US alone. information on the frequency and the mode of the US was not provided. Moll[76] studied the effects of topically applied lidocaine (140.7 mg) and Decadron (3.75 mg) enhanced with US (six treatments, two times a week for 3 weeks, with either sham US or US at 1.5 W/[cm.sup.2], 870 kHz, continuous for 5 minutes). She reported that 88% of the patients receiving phonophoresis with Decadron and lidocaine (group A) obtained relief from their trigger-point pain. Of those receiving US alone (group B), 560/o reported symptom relief compared with those receiving the topically applied drug with sham US (group C), where 23% reported some relief. The differences between groups A and C and groups B and C were statistically significant. This positive response to trigger-point phonophoresis with lidocaine is interesting based on the physiology of trigger points and the absorption of topical lidocaine. A trigger point is usually a point at which the circulation is disrupted anoxic). In a study by Hua et al,[77] topically applied lidocaine (100 mg) was rapidly absorbed (5-15 minutes) when the circulatory system was intact. When the arterial vessels were surgically severed and reanastomosed, however, the time required for absorption was increased and the amount of the drug absorbed was decreased. Although this study was done with rabbits, it does raise concern about the effectiveness of absorption of a topical drug with or without the enhancement of US if there is spasm or other type of disruption of the circulation to the area.

In 1985, McElnay and associates[78] described a phonophoresis study using lignocaine. Pricking of the skin in the treated area was the qualitative method used for assessing skin anesthesia. Two treatments were given, 7 days apart, on the forearm of human subjects, with random asignment of 2.0 or 0.0 (W/[cm.sup.2 on the first treatment, crossed over for the second treatment (pulsed, 20%, 870 kHz, 5 minutes). Ultrasound was not associated with an increase in the percutaneous absorption of lignocaine, as measured by the response of the subject to a sharp stimulus.

Williams[79] quantified the physical effects of three different commercially available topical anesthetic preparations on superficial sensory cells (pain receptors) on the skin in human subjects. Following a low-intensity treatment to minimize heating (0.25 W/[cm.sup.2], 1.1 MHz, continuous), US had no detectable effects on the rate of penetration of any of the three anesthetic preparations. Unfortunately, Williams did not document whether any of the anesthetics had been checked for transmissivity to US.

Although anesthetics are frequently mixed with anti-inflammatory drugs, none of the investigators in phonophoresis studies using anesthetic drugs discussed the potential risk for muscle necrosis following repeated exposure to local anesthetics such as lidocaine.[80] Although most of this research has been carried out following needle injections, if US enables absorption of an anesthetic drug to the muscle, then it could also be considered a risk factor with phonophoresis.

Phonophoresis With Dimethyl Sulfoxide 

Romanenko and Araviiskii[81] applied dimethyl sulfoxide (DMSO) to the skin followed with the application of amphotericin B (an antifungal agent). When the amphotericin ointment was used in combination with US, the content of amphotericin B in the skin and subcutaneous fatty tissue 1, 3, 24, 48, and 72 hours after application was much higher than that measured after a local application of the ointment alone. When the amphotericin B ointment was applied to the skin after preliminary treatment with DMSO, 3 hours after the application the maximum content of the antifungal agent in the skin and subcutaneous fatty tissue was higher than the content observed after the ointment application, which was sonicated without pretreatment with DMSO. At 48 and 72 hours, the highest concentrations of the antifungal ointment were in the skin and subcutaneous fatty tissue for the groups treated with the ointment enhanced by US. These researchers concluded that both US and DMSO were enhancers of transcutaneous drug delivery, with DMSO serving as a quick but short-lived enhancer and US serving as a more long-lasting enhancer. This delayed elevation of the drug at 48 and 72 hours suggests a systemic effect and not just a local effect.[35] Unfortunately, the variables of the US were not included in the English abstract.

Matinian et al[82] also studied the effect of papain and DMSO phonophoresis on inflammation and the healing of purulent wounds in a small group of patients. A 1% papain solution together with DMSO enhanced with US was effective for the treatment of purulent wounds and inflammatory infiltrates, as measured by the time required to restore the structure of the injured tissues. Restoration in the treatment group was reportedly reduced by almost twice that of controls. Further details of the methodology of this study were not available in the translated abstract; therefore, claims must be viewed with caution.

Phonophoresis With Other Drugs

Tyle and Agrawala[83] reviewed the effects of drug delivery by phonophoresis in 1990 and concluded that US effectively enhanced localized and systemic drug delivery. Several pharmaceutical companies are now researching the use of US to deliver drugs that in the past have not been suitable for transdermal delivery (mannitol, insulin, and physostigmine). Tyle and Agrawala recommended that further research be conducted to more carefully identify the critical physicochemical properties of the drug molecule that could be influenced by US, the equilibrium partition coefficient between the skin and the drug in the coupling agent, whether US causes any type of chemical changes in the drug molecules themselves, what methods should be used to measure changes in the absorption of drugs, and whether continuous or pulsed US should be used.

McElnay,[84] in 1993, studied the effect of US on methyl nicotinate (a vasodilator) in 10 asymptomatic volunteers in a double-blind, placebo, crossover study. A group receiving US at 3.0 Mhz and 1.0 W/[cm.sup.2] (continuous mode) was used compared with a placebo treatment group receiving US treatment alone (same variables of US as for the treatment group). Methyl nicotinate was checked for transmissivity. The methyl nicotinate was applied in an adhesive, slow-release package before the US treatment and then was left on the skin for 15 seconds, 1 minute, and 2 minutes after US. The dependent variable was blood flow (measured with a Doppler flow meter). These researchers found that pretreatment of the skin with US prior to the topical application of methyl nicotinate led to enhanced percutaneous absorption of the methyl nicotinate, as determined by an increased blood flow. The rate of absorption time to peak blood flow, however, was not enhanced. At 2 minutes, there was an increase in the peak blood flow and percutaneous absorption in the area under the Us-treated skin, suggesting that US affected the skin structure to provide penetration enhancement. Ultrasound treatment increased flux on the order of 42% (63% at the 15-second interval, 4% at the 1-minute interval, and 680/o at the 2-minute interval). Perhaps, to maximally enhance diffusion, US should be provided both before a topical drug is applied and as part of the treatment.

Benson and colleagues[85'86] studied US as an enhancer of benzydamine hydrochloride (3%), a nonsteroidal anti-inflammatory drug. The drug was mixed in a gel and validated for transmission at the three different frequencies of US used in the study. The investigators used several different US protocols. The mode of US was continuous or pulsating (1:1 with 2 milli-seconds on and 2 milliseconds off). The intensity was set at 1.0 W/[cm.sup.2], and the following frequencies were used: 0.875, 1.5, and 3.0 Mhz. Three treatment sessions were arranged 7 days apart. This was a crossover, double-blind design. All of the subjects were sonicated with all three frequencies. A control group received US at 0.0 W/[cm.sup.2]. The drug was applied 5 minutes before the US treatment. Drug assays were performed using high-performance liquid chromatography and ultraviolet detection. No differences were found between the US-treated groups and the shamtreated group. These researchers concluded that US energy did not influence percutaneous absorption, but they did raise some concern that bendydamine in the formulation used in this study potentially made it unsuitable for enhanced percutaneous absorption because it may not have transmitted the US.

Kost et al[87] studied the phonophoresis of D-Mannitol (a drug used to remove extracellular fluid in rats and guinea pigs). The skin was treated with 1.5 W/[cm.sup.2], continuous mode, for 3 minutes or 3.0 W/[cm.sup.2], pulsating mode, for 5 minutes at 1.0 MHz. Permeability experiments were performed during the first 2 hours postsonication. Ultrasound had an effect on the diffusion of D-Mannitol through both the rat skin and the guinea pig skin. These investigators also topically applied D-Mannitol to a shaved site on the upper back of Sprague-Dawley rats. Ultrasound was applied for 2 minutes (continuous, 1-2 W/[cm.sup.2 870 kHz). Control rats received the same treatment with a moving ultrasonic applicator, but the US was turned off. The skin showed no damage in either condition, but D-Mannitol showed increased diffusion only in the US-treated rats.

Romanenko[88] used US with topically applied amphotericin B, a drug used to treat systemic fungal infections. After 1, 3, 24, 48, and 72 hours, there was a higher content of the drug detected in the skin and the subcutaneous fatty tissue in phonophoresed subjects as compared with subjects who received a topical application. The drug contents in the sonicated samples increased by the third hour after the application. For those receiving topical amphotericin B without sonication, the maximum values were noted after 1 hour. After 24, 48, and 72 hours, amphotericin B concentration gradually decreased, but remained higher in the sonicated samples than in the nonsonicated samples. The researchers concluded that the topical application of amphotericin B in combination with US was an appropriate treatment for patients with candidosis. In this English-translated abstract, the variables of US were not provided.

Voloshin and Meshchishena[89] studied patients with anococcygeal pain syndrome. The treatment consisted of an intrarectal administration of anti-inflammatory reparative anesthetic aerosols in association with phonophoresis of the same drug on the lumbosacral region. Recurrence of the pain syndrome was 3.5 times less frequent when the aerosols were combined with the phonophoresis. Because only an English-translated abstract was available for review, no additional details of this study were available and the results should be considered with caution until the article can be more closely examined.

Phonophoresis Compared With Injections or Iontophoresis

Pratzel et al[90] carried out two studies of sonophoresis with indomethacin (15 mg), one with pigs and one with humans. These researchers compared the differences in blood and urine concentrations of indomethacin following either topical application alone or topical application enhanced with electrical current or US. Only one treatment was given. In the pigs, the US was applied at an intensity of 2.0 W/[cm.sup.2] for 30 minutes and then for 10 minutes every 30 minutes for up to 5 hours, with the area remaining covered between treatments. A 150-[cm.sup.2] area of the back was treated. in human subjects, the indomethacin was in contact with 1,380 [cm.sup.2] of Skin for 1 hour, but the details of the US were not described. All assays were done 5 hours at the US. Iontophoresis was associated with an increased diffusion of indomethacin in both the blood and the urine in the human subjects and the pigs, but there were no differences in blood or urine measurements following US in either human or animal subjects. These investigators did not mention whether they had evaluated the indomethacin to ensure it transmitted US waves. As a follow-up to this review article, I evaluated the transmission characteristics of indomethacin (10%) premixed by a pharmacist with a conductive gel. The pharmacist reported that indomethacin was not water soluble and the mixture was cloudy, similar to what is seen with a mixture of hydrocortisone. The indomethacin preparation transmitted only 31% of the US energy, based on measurements with an Ohmic Calibration Unit.(*) This finding could explain why US was not effective as an enhancer.

In 1958, Newman and associates[91] reported that injected hydrocortisone (25 mg for 4 treatments) or hypospray (50 mg for 12 treatments) followed by treatment with US (1 MHz) was superior to injection alone for the treatment of subdeltoid bursitis. unfortunately, the other variables of US were not described in this report. Muir et al[92] compared the effectiveness of sonicated (2.75 W/[cm.sup.2], 1 MHz, continuous) hydrocortisone (10%) and injected hydrocortisone (10%) in greyhound dogs over a 25-[cm.sup.2] area of the infrapatellar region. They compared the intra-articular levels of hydrocortisone 10 minutes after treatment. Intra-articular hydrocortisone levels obtained with phonophoresis were extremely low in comparison with the levels measured following injection. There were no differences in intra-articular hydrocortisone levels between the animals receiving phonophoresis with hydrocortisone, those receiving topical application alone, or controls. These authors concluded that phonophoresis was an ineffective method of facilitating hydrocortisone penetration. These researchers did not indicate whether they had evaluated the transmissivity of the hydrocortisone. This has been reported as a limitation by other investigators.[63,93]

Using an animal model (mini Yucatan pigs), Byl et al[93] compared the effect of dexamethasone (0.334%) and hydrocortisone (10%) on collagen deposition,[94] as measured by the amount of hydroxyproline deposited in polytet-rafluroethylene (ePTFE) tubes following topical application, injection, or phonophoresis. Hydroxyproline is the precursor to collagen and can be used to estimate collagen. In addition, the ePTFE tubes can easily be threaded and removed from subcutaneous tissue without sacrificing the skin. One treatment was given. The phonophoresed group was treated with US at 1.0 W/[cm.sup.2], 1 MHz, continuous. There was a similar reduction in collagen in both the sonicated and injected ePTFE tubes treated with dexamethasone, which was less than in the controls. There were no measurable reductions in collagen deposition in the submuscular or subtendinous areas. For the ePTFE tubes treated with hydrocortisone, only those injected with hydrocortisone showed a reduction in collagen. These researchers found that hydrocortisone did not transmit the US. This study did not measure systemic effects. One potential confounding variable in this study was that the pigs were shaved prior to the implantation and sonication of the ePTFE tubes. If the shaving abraded the stratum corneum, it would have been easier for the topically applied drugs to diffuse through the skin. This ease of diffusion could limit the generalizabilty of findings only to those situations in which the skin is pretreated with shaving.

Summary

The purposes of this article were to build a theoretical foundation for understanding phonophoresis and to review the efficacy of phonophoresis as a clinical treatment in physical therapy. Although US is not specifically approved by the Food and Drug Administration for purposes of drug enhancement,[95] phonophoresis is commonly used in the practice of physical therapy, representing up to 30% of the physical therapy visits in some sites.[7] Even though phonophoresis is not listed as a specific procedure code on most medical fee schedules, approximately 75% of the research studies reviewed in this article indicated some level of effectiveness of US as an enhancer of topically applied drugs. As noted, however, many of these studies lacked important elements that would have enhanced their credibility,

Anti-inflammatory agents were most frequently used as the topical drug agent for the phonophoresis studies. The studies using corticosteroid drugs reported positive drug enhancement with US, but questions were raised about whether hydrocortisone transmits ultrasonic energy. The nonsteroidal anti-inflammatories of benzydamine and indomethacin were not enhanced by US, and the enhancement of salicylates was questionable. Although the transmissivity of benzydamine was not reported, indomethacin provided limited transmission of US. In some of the salicylate studies, the transmissivity of US was not evaluated. The diffusion of analgesics such as lidocaine (lignocaine) appeared to be enhanced by US.

Most studies documented that phonophoresis was associated with an increase in drug diffusion at the subcutaneous level. There were inconsistent findings for successful drug delivery beyond the subcutaneous level. Most studies did not find increased penetration at the level of the muscle or the joint. The studies that both included a control group and directly measured drug concentration levels provided the most convincing evidence that US enhanced drug diffusion.

In physical therapy, the preferred target for most topically applied drugs has been local. It is important to recognize, however, that if the drug diffuses through the epidermis and the dermis into the capillary network, some systemic effects can be expected. The maximum local effects can be measured in about 2 hours, whereas the systemic effects can be measured at 12 and 24 hours. The systemic effects may also augment the local effects.

Some investigators did not measure the US transmission characteristics of the drug selected for the phonophoresis treatment. This limitation was particularly common in those studies in which phonophoresis was not found to be effective. Unfortunately, some of the studies that showed increases in drug diffusion with US were correlated with damage to the stratum corneum. This damage was either part of the pathology designated for treatment (eg, psoriasis, dermatitis, ulcer)[80] or it was an outcome of the US treatment itself (eg, either a high-frequency US outside the therapeutic range [10-16] Mhz for 20 minutes][37,38] or the use of a stationary sound head[59-61]).

Based on the research reviewed, it appears that to maximize the effectiveness of phonophoresis treatments, clinicians should (1) select only topical agents that transmit US; (2) check the skin carefully for moistness and hydration; (3) pretreat the skin with US, heating, or shaving; (4) position the patient to maximize circulation to the area being treated; (5) use an intensity in the thermal range (1.5 W/[cm.sup.2] or higher), unless there are contraindications for heating based on the patient's condition (eg, acute injury, open wound); (6) use pulsating US with an intensity of 0.5 to 1.0 W/[cm.sup.2] when treating an open wound; and (7) leave the drug on the skin with an occlusive dressing after treatment. When treating an open wound, the rate of diffusion can be expected to be higher than usual. All patients should be monitored for systemic drug side effects, especially when a large area of the skin is phonophoresed or there is a loss of continuity of the stratum corneum.

The use of US as an enhancer of transcutaneous drug delivery is still a relatively new concept in pharmacology. Phonophoresis is now mentioned in some pharmaceutical textbooks on topical drug delivery,[1,5] but more research is needed. There will continue to be new drugs on the market that have the potential for transdermal delivery, and, undoubtedly, new types of ultrasonic instruments will be developed and marketed specifically for use as enhancers. A critical variable to consider when selecting US as an enhancer, however, is transmissivity of the topical drug. Both the active drug and the carrying agent should transmit US.

Physical therapists can contribute to the body of knowledge by carrying out controlled clinical phonophoresis studies. In preparation for research in this area, therapists must review not only the literature on phonophoresis, but also on sonophoresis. Good controlled research on phonophoresis should include (1) two control groups, (2) a group that receives US alone, and (3) a group that receives the topical application of the drug in addition to sham US. Crossover studies can be used that allow the investigator to cross over the control groups to the treatment group after a defined study period. Given the systemic nature of transcutaneous drug delivery, the patient's unaffected side should not serve as the control. Where possible, measurements should be blinded. Neither the therapist nor the patient should know whether the medication is in the gel. This means the bottles of conductive gels must look similar, be stored and managed in a similar way, and be coded for later identification. For the most part, relatively small, circumscribed areas comparable to the size of the sound head should be targeted for phonophoresis treatment. Measurement tools to reliably document a reduction in signs and symptoms of the pathology (eg, inflammation, infection, muscle spasm, collagen deposition) must be developed for use in the clinical setting. The research studies should include the study of new and commonly used drugs that are likely to be used by therapists in clinical practice (eg, pharmaceuticals appropriate for the treatment of decubitus ulcers, bursitis, capsulitis, epicondylitis, nerve entrapment).

Studies on the physics of US as an enhancer of transcutaneous drug delivery must also be continued. This research should be closely paralleled with research on the variables of US delivery, the length of a treatment session, and the duration of phonophoresis treatments in terms of effectiveness of diffusion at the local as well as systemic level. This same research should be carried out with asymptomatic subjects as well as those with pathology.

The goal of aggressive research in phonophoresis is to improve our ability to offer effective treatment. Ultimately, these research studies will not only enable therapists to more carefully match treatment variables (US and topical agent) with specific patient conditions but should also encourage therapists to stop using treatment procedures that are not effective.

(*)Ohmic Instruments Co, 102 Chew Ave, Saint Michaels, MD 12663.

References

[1] Bronaugh RL, Maibach HI. In Vitro Percutaneous Absorption: Principles, Fundamentals, and Applications. Ann Arbor, Mich: CRC Press; 1991:280. [2] Chien YW. Advances in transdermal systemic medications. In: Chien YW, ed. Transdermal Controlled Systemic Medications Volume 31. New York, NY: Marcel Dekker Inc; 1987: chap 12(1-24). [3] Chien YW. Developmental concepts and practice in transdermal therapeutic systems. In: Chien YW, ed. Transdermal Controlled Systemic Medications: Volume 31. New York, NY: Marcel Dekker Inc; 1987: chap 2(25-82). [4] Cooper E, Dinesh C. Practical considerations for topical drug formulations with and without enhancers. in: Osborne DW, Amann AH, eds. Topical Drug Delivery Formulations: Volume 42. New York, NY: Marcel Dekker Inc; 1990: chap 1(1-12). [5] Kost J, Levy D, Langer R. Ultrasound as a transdermal enhancer. In: Osborne DW, Amann AH, eds. Topical Drug Delivery Formulations: Volume 42. New York, NY: Marcel Dekker inc; 1990: chap 34(603-632). [6] Maleche AN, DelTerza S, Higuchi WI, Srinivasan T. Iontophoretic drug delivery: effect of physicochemical effectors on skin uptake of drugs. In: Osborne DW, Amann AH, eds. Topical Drug Delivery Formulations: Volume 42. New York, NY: Marcel Dekker Inc; 1990: chap 35(633-652). [7] Pottenger FJ, Karalfa GL. Utilization of hydrocortisone phonophoresis in United States Army physical therapy clinics. Milw Med. 1989;154:355-358. [8] Ward AJL, Tallon R. Penetration enhancer incorporation in bilayers. In: Osborne DW, Amann AH, eds. Topical Drug Delivery Formulations. Volume 42. New York, NY: Marcel Dekker Inc; 1990: chap 4(47-68). [9] Friberg SE, Ibrashum K, Margosiak M, et al. Stratum corneum and transport properties. In: Osborne DW, Amann AH, eds. Topical Drug Delivery Formulations: Volume 42. New York, NY: Marcel Dekker Inc; 1990: chap 3(1-12). [10] Bronaugh RL. A flow-through diffusion cell. In: Bronaugh RL, Maibach HI, eds. In Vitro Percutaneous Absorption: Principles, Fundamentals, and Applications. Ann Arbor, Mich: CRC Press; 1991; chap 3(17-24). [11] Webster RC, Maibach HI. Individual and regional variation with in vitro percutaneous absorption. In: Bronaugh RL, Maibach HI, eds. In Vitro Percutaneous Absorption Principles, Fundamentals, and Applications. Ann Arbor, Mich: CRC Press; 1991: chap 4(26-30). [12] Scott R. In vitro absorption through damaged skin. In: Bronaugh RL, Maibach HI, eds. In Vitro Percutaneous Absorption: Principles, Fundamentals, and Applications. Ann Arbor, Mich: CRC Press; 1991: chap 20(129-156). [13] Henley EL. Iontophoresis and phonophoresis transcutaneous drug delivery. Presented at the 65th Annual Conference of the American Physicalt Therapy Association; Anaheim, Calif; 1990. [14] Murphy GH, Kwan TH, Mihm MC. The skin. In: Robbins SL, Ctran RS, Jumar V, eds. Robbins Pathological Basis of Disease. 4th ed. Philadelphia, Pa: WB Saunders Co; 1984:1273-1301. [15] MacKie RM. Clinical Dermatology: An Illustrated Textbook. 3rd ed. New York, NY: Oxford University Press; 1991:162. [16] Cooper E. Alterations in skin permeability. In: Chien YW, ed. Transdermal Controlled Systemic Medications: Volume 31. New York, NY: Marcel Dekker Inc; 1987: chap 3(83-92). [17] Elias PM. The importance of epidermal lipids for the stratum corneum barrier. In: Osborne DW, Amann AH, eds. Topical Drug Delivery Formulations: Volume 42. New York, NY: Marcel Dekker Inc; 1990:1-12. [18] Tagami H. Functional characteristics of aged skin. Acta Dermatol (Kyoto). 1991;66: 19-21. [19] Lehman JF, DeLateur BJ, Therapeutic heat. in: Lehman JF, ed. Therapeutic Heat and Cold. 3rd ed. Baltimore, Md: Williams Wilkins; 1982:486-552. [20] Huang Y-C. In vitro evaluations of transdermal drug delivery. In: Chien YW, ed. Transdermal Controlled Systemic Medications: Volume 31. New York, NY: Marcel Dekker Inc; 1987:159-178. [21] Dohner JW. Development of processes and equipment for rate-controlled transdermal therapeutic systems. In: Chien YW, ed. Transdermal Controlled Systemic Medications: Volume 31. New York, NY: Marcel Dekker Inc; 1987:349-364. [22] Webster R, Maibach HI. Short-term in vitro skin exposure. In: Bronaugh RL, Maibach HI, eds. In Vitro Percutaneous Absorption: Principles, Fundamentals, and applications. Ann Arbor, Mich: CRC Press; 1981: chap 18(231-242). [23] Gilchrest BA. Skin and Aging Processes. Boca Raton, Fla: CRC Press Inc; 1984:73. [24] Christophers E, Kligman AM. Percutaneous absorption in aged skin. In: Montagna W, ed. Advances in Biology of Skin, Volume VI: Aging. New York, NY: Pergamon Press Inc; 1964:163-175. [25] Flynn GL. Mechanisms of percutaneous absorption from physicochemical evidence. In: Bronaugh RL, Maibach HI, eds. Percutaneous Absorption. 2nd ed. Nye York, NY: Marcel Dekker Inc; 1989:270. [26] Guillot M. Urinary leak of salicylic acid after topical application. J Physiol (Paris). 1954;46:31-37. [27] Guy RH, Hadgraft J, Hinz RS, et al. In vivo evaluations of transdermal drug delivery. In: Bronaugh RL, Maibach HI, eds. In Vitro Percutaneous Absorption: Principles, Fundamentals, and Applications. Ann Arbor, Mich: CRC Press; 1991: chap 8(179-216). [28] Guy RH, Hadgraft J. The effect of penetration enhancers on the kinetics of percutaneous absorption. J Control Rel. 1987;5:43-53. [29] Vickers CF. Reservoir effects of human skin: pharmacological speculation. In: Mauvais-Javis P, Vickers CFH, Wepierre J, eds. Percutaneous Absorption of Steroids. San Francisco, Calif: Academic Press Inc; 1990:19-30. [30] Torrent J, Izquierdo I, Barbanoj MJ, et al. Anti-inflammatory activity of piroxicam after oral and topical administration of an ultraviolet-induced erythema model in man. Curr Ther Res. 1988;44:340-347. [31] Rabinowitz JL, Feldman ES, Weinberger A, Schumacher HR. Comparative tissue absorption of oral aspirin and topical triethanolamine salicylate in human and canine knee joints. J Clin Pharmacol, 1988;22:42-48. [32] Wada Y, Etoh Y, Ohira A, et al. Percutaneous absorption and anti-inflammatory activity of indomethacin in ointment. J Pharm Pharmacol. 1982;34:467-468. [33] Guy RH, Maibach HI. Drug delivery to local subcutaneous structures following topical administration. J pharm Sci. 1983;72:1375-1380. [34] Riviere JE, Williams PL. On the pharmacokinetic implications of changing blood flow in the skin. J Pharm Sci. In press. [35] McNeill SC, Potts RO, Francoeur ML. Local enhance topical delivery (LETD) of drugs: Does it truly exist? Pharmaceutical Research. 1992;9:1422-1427, [36] Anjo DM, Fedlmann RJ, Maibach HI. Methods of predicting percutaneous penetration in man. In: Mauvais-Javis P, Vickers CFH, Wepierre J, eds. Percutaneous Absorption of Steroids. San Francisco, Calif: Academic Press Inc; 1980: chap 3(31-52). [37] Bommannan D, Okuyama H, Stauffer P, Guy RH. Sonophoresis, I: the use of high-frequency ultrasound to enhance transdermal drug delivery. Pharmceutical Research. 1992; 9:559-564. [38] Bommannan D, Menon GK, Okuyama H, et al. Sonophoresis, II: examination of the mechanisms(s) of ultrasound-enhanced transdermal drug delivery. Pharmaceutical Research. 1992;9:1043-1047. [39] Wong T, Byl N. Phonophoresis: the thermal and nonthermal effects of ultrasound. In: Proceedings of the joint Congress of the American Physical Therapy Association and the Canadian Physiotherapy Association, Toronto, Ontaho, Canada; June 1994. Abstract. [40] Fellinger K, Schmid J. Klinik und Therapie des Chronischen Gelenkbuematismum. Vienna, Austria: Maudrich (Austrian); 1954:549-552. [41] Nyborg WL. Mechanisms. In: Nyborg WL, Ziskin MC, eds. Biological Effects of Ultrasound. New York, NY: Churchill Livingstone Inc; 1985:23-33. [42] Antich TJ. Phonophoresis:the principles of ultrasonic driving force and efficacy in treatment of common orthopedic diagnoses. J Orthop Sports Phys Ther. 1982;4:99-102. [43] Dinno MA, Crum LA, Wu J. The effect of therapeutic ultrasound on the electrophysiologic parameters of frog skin. Med Biol. 1989:25;461-470. [44] Dinno MA, Dyson M, Young SR, et al. The significance of membrane changes in the safe and effective use of therapeutic and diagnostic ultrasound. Ultrasound Med Biol. 1989;34: 1543-1552. [45] Potts RO, Francoeur ML. Lipid biophysics of water loss through the skin. Proc Natl Acad Sci USA. 1990;87:3871-3873. [46] Frizzell LA, Dunn F. Biophysics of ultrasound. In: Lehman JF. Therapeutic Heat and Cold. 3rd ed. Baltimore, Md: Williams & Wilkins; 1982: chap 8(353-385). [47] Kost J, Chen R, Langer R. Membrane permeability enhancement by ultrasound. In: Proceedings of the Third World Biomaterials Congress, Japan, 1988. [48] Williams AR. Ultrasound: Biological Effects and Potential Hazards. San Francisco, Calif: Academic Press Inc; 1983. [49] Miller MW, Kaufman GE, Cataldo FL, Cartensen EL. Absence of mitotic reduction in regenerating rat livers exposed to ultrasound. J Clin Ultrasound. 1976;4:169-172. [50] Miller MW. Comparison of microneuclei induction for x-ray and ultrasound exposures of vicia fibia root meristern cells. Ultrasound Med Biol. 1978;4:263-267. [51] Miller DL, Nyborg WL, Whitcomb CC. Platelet aggregation induced by ultrasound under specialized conditions in vitro. Science. 1979;205:505-507. [52] Ziskin MC, McDiarmid T, Michlovitz SL. Therapeutic ultrasound. In: Michlovitz SL, ed. Thermal Agents in Rehabilitation. Philadelphia, Pa: FA Davis Co; 1990: chap 7(1-16). [53] Lehmann JF, DeLateur BJ, Stonebridge JB, Warren CG. Therapeutic temperature distribution produced by ultrasound as modified by dosage and volume of tissue exposed. Arch Phys Med. 1967;48:662-666. [54] Hayes KW. Manual for Physical Agents. 4th ed. Ultrasound. East Norwalk, Conn: Appleton and Lange: 1993: chap 6(37-42). [55] Dyson M, Suckling J. Stimulation of tissue repair by ultrasound: a survey of the mechanisms involved. Physiotherapy. 1978;64:105-108. [56] Sitton P, Toulouse L, Byl N. Movement of fibroblasts with ultrasound. In: Proceedings of the Annual Meeting of the California Chapter of the American Physical Therapy Association; October 10, 1993. Abstract. [57] Griffin JE, Echternach JL, Price RE, et al. Patients treated with ultrasonic driven hydrocortisone and with ultrasound alone. Phys Ther. 1967;47:594-601. [58] Griffin JE, Touchstone JC. Ultrasonic movement of cortisol into pig tissues:, I: movement into skeletal muscle. Am J Phys Med. 1963;42:77-85. [59] Griffin JE, Touchstone JC, Liu AC-Y. Ultrasonic movement of cortisol into pig tissue, 11: movement into paravertebral nerve. Am J Phys Med. 1965;44:20-25. [60] Griffin JE, Touchstone JC. Low-intensity phonophoresis of cortisol in swine. Phys Ther. 1968;48:1336-1344. [61] Griffin JE, Touchstone JC. Effects of ultrasound frequency on cortisone into swine tissue. Am J Phys Med, 1972;41:62-78. [62] Kleinkort JA, Wood AF. Phonophoresis with 1% versus 100/o hydrocortisone. Phys Ther. 1975;55:1320-1324. [63] Davick JP, Martin RK, Albright JP. Distribution and deposition of tritiated cortisol using phonophoresis. Phys Ther. 1988;68:1672-1675. [64] McElnay JC, Kennedy TA, Harland R. The influence of ultrasound on the percutaneous absorption of fluocinolone acetonide. International Journal of Pharmaceutics. 1987;40:105-110. [65] Newman JT, Nellermoe MD, Carnett JL. Hydrocortisone phonophoresis: a literature review. J Am Podiatr Med Assoc. 1992:82:432-435. [66] Kamenskaia NS, Fedorova NE. The therapeutic use of iodide-bromide sodium choride baths combined with hydrocortisone phonophoresis in patients with osteoarthrosis and gout. Vaprosy Kurortologii, Fizioterapii i Lechebnoi Fizicheskoi Kultury. 1990;6:47-50. [67] Smith W, Winn F, Parette R. Comparative study using four modalities in shin splints treatments. J Orthop Sports Phys Ther. 1986;8: 77-80. [68] Cameron MH, Monroe LG. Relative transmission of ultrasound by media customarily used for phonophoresis. Phys Ther. 1992;72: 142-148. [69] Cameron MH. Effect of phonophoresis on serum salicylates levels. Med Sci Sports Exerc. 1991;23:1213. [70] Byl N, McKenzie A, Haliday B, et al. The effects of phonophoresis with corticosteroids: a controlled pilot study. J Orthop Sports Phys Ther. 1993;18:590-600. [71] Hill DW, Richardson JD. Effectiveness of 10% trolamine salicylate cream on muscular soreness induced by a reproducible program of weight training. J Orthop Sports Phys Ther. 1989;22:19-23. [72] Oziomek RS, Perrin DH, Herhold DA, Denegar CR. Effect of phonophoresis on serum salicylate levels. Med Sci Sports Exerc. 1991;23:397-401. [73] Hollister GE, Kanter SL. Studies of delayed action medication, IV: salicylates. Clin Pharmacol Ther. 1985;6:5-13. [74] Ciccone CD, Leggin BG, Callamaro J. Effects of ultrasound and trolamine saticylate phonophoresis on delayed-onset muscle soreness. Phys Ther. 1991;71:666-678. [75] Novak EJ. Experimental transmission of lidocaine through intact skin by ultrasound. Arch Phys Med Rehabil. 1964:45:231-232. [76] Moll MJ. A new approach to pain: lidocaine and Decadron with ultrasound. USAF Med Serv Digest. 1979;30:8-11. [77] Hua S-M, Liu T-K, Yu H-Y. Absorption of lidocaine following topical application in microvascular procedures in rabbits. J Orthop Res. 1991;9:545-549. [78] McElnay JC, Kennedy TA, Harland R, et al. The effect of ultrasound on the percutaneous absorption of lignocaine. Br J Clin Pharmacol. 1985;20:421-424. [79] Williams AR. Phonophoresis: an in vivo evaluation using three topical anesthetic preparations. Ultrasonics. 1990;28:137-141. [80] Foster AH, Carlson BM. Myotoxicity of local anesthestics and regeneration of the damaged muscle fibers. Anesth Analg. 1980;59: 727-736. [81] Romanenko IM, Araviiskii RA. Comparative levels of amphotericin B in the skin and subcutaneous fatty tissue after cutaneous application of amphotericin ointment by phonophoresis and with preliminary treatment by dimethyl sulfoxide. Antibiotiki i Khimioterapiia. 1991;36:29-31. [82] Matinian IA, Nagapetian KHO, Amirian SS, et al. Papain phonophoresis in the treatment of suppurative wounds and inflammatory processes. Khirurgiia. 1990;9:74-76. [83] Tyle P, Agrawala P. Drug delivery by phonophoresis. Pharmaceutical Research. 1989;6: 355-361. [84] McElnay JC, Benson FIA, Harland R, Hadgraft J. Phonophoresis of methyl nicotinate: a preliminary study to elucidate the mechanisms of action. Pharmaceutical Research. 1993;10:1726-1731. [85] Benson HAE, McElnay JC. A high-performance liquid chromatography assay for the measurement of benzydamine hydrochloride in topical pharmaceutical preparations. J Chromatogr, 1987; 394:395-399, [86] Benson HAE, McElnay JC, Harland R. Use of ultrasound to enhance percutaneous absorption of Benzydamine. Phys Ther. 1989;69: 113-118. [87] Levy D, Kost J, Meshulain Y, Langer R. Effect of ultrasound on transdermal drug delivery to rats and guinea pigs. J Clin Invest. 1989;83:2074-2080. [88] Romanenko IM. A comparative evaluation of amphotericin pharmacokinetics in the skin and subcutaneous fatty tissue when used dermally and by phonophoresis. Farmakologiia i Toksikologiia. 1992;54:50-51. [89] Voloshin AI, Meshchishena NI. The treatment of patients with anococcygeal pain syndrome. Vrachebnoedelo. 1990;10:99-101. [90] Pratzel H, Dittrich P, Kukovetz W. Spontaneous and forced cutaneous absorption of indomethacin in pigs and humans. J Rheumatol. 1986;13:1122-1125. [91] Newman MK, Kill M, Frampton G. Effects of ultrasound alone and combined with hydrocortisone injections by needle or hypospray. Am J Phys Med. 1958;37:206-209. [92] Muir WS, Magee EP, Longo JA, et al. Comparison of ultrasonically applied vs intraarticular injected hydrocortisone levels in canine knees. Orthop Rev. 1990;29:351-356. [93] Byl N, McKenzie A, West J, et al. Low-dose ultrasound effects on wound healing: a controlled study with Yucatan pigs. Arch Phys Med Rehabil. 1992;73:565-664. [94] Hydrocortisone. In: Physican Desk Reference. Oradell, NJ: Medical Economics Co; 1991:1390-1391. [95] Shah VP, Vinod P, Skelly JP. Regulatory considerations in transdermal drug delivery systems in the United States. In: Chien YW, ed. Transdermal Controlled Systemic Medications: Volume 31. New York, NY: Marcel Dekker Inc; 1987: chap 16(399-410).

NN Byl, PhD, is Associate Professor, Graduate Program in Physical Therapy, School of Medicine, University of California, San Francisco, 374 Parnassus Ave, Box 0736, San Francisco, CA 94122 (USA).