Microsoft word - iontopath.vs.ionto.doc.doc

Evaluation of Anderson et al., (2003) Effects of iontophoresis current magnitude
and duration on dexamethasone deposition and localized drug retention. Physical
Therapy 83:161-170
The Anderson et al. (2003) paper may be broken into two sections. The first part evaluates the efficiency of dexamethasone delivery using a model of the human body. The second part, unrelated to the first part, evaluates in vivo skin reactions to the I. Evaluation of the efficiency of dexamethasone delivery into a model of human
Evaluation of iontophoretic delivery is only as good as the model mimics human tissue. Sage and Riviere (1992) point out that “to be useful a model system should allow prediction of the therapeutic utility of a dosage in man”. The iontophoretic efficiency chamber is required to mimic or model 1) the permeability of the upper layer of skin, the stratum corneum, 2) the electrical resistance of the skin, 3) and the mechanisms that remove a drug, such as enzymes and vasculature, in the subcutaneous tissue. The Iontophoretic Efficacy Chamber
Anderson et al., (2003) used a cellulose ultrafiltration membrane placed on agar, a gelatinous hydrogel containing sodium chloride. The cellulose membrane mimics the selective permeability of the stratum corneum with its restricted pore size. However, the membrane does not reproduce the negative charge of the stratum corneum, the electrical resistance of the stratum corneum, nor the enzymes that alter dexamethasone found in the stratum corneum. The agar model of the subcutaneous tissue lacks a vascular uptake mechanism provided in the human case by capillaries, as well as enzymes. Electrical resistance of the iontophoretic chamber was not explicitly adjusted to approximate human The iontophoretic efficacy chamber had a surface on which the active electrode was placed with an area of 1.76 cm2. A silver-silver chloride active electrode with this surface area was placed over the chamber and the return electrode consisted of a silver wire. Current was generated by a constant current Iontophor delivery system. The Birch
Point IontoPatch was not used in this experiment. The IontoPatch is a voltage
device and does not control current like the Iontophor iontophoretic generator used.
The IontoPatch has a surface area of 12 cm2, much larger than the area of the tested
electrode of 1.76 cm2. The return electrode of the IontoPatch is not a silver wire.
The active electrode was loaded with 1.5 ml 4 mg/ml dexamethasone phosphate. After the iontophoresis session, the amount of dexamethasone in the agar was measured by spectrophotometry. To obtain concentration measurements as a function of depth, the agar was cut in 2 mm slices, parallel to the surface of the active electrode. The amount of dexamethasone in each slice was thus measured. The Experimental Comparison
The purpose of this experiment is to compare the amount of dexamethasone delivered by two iontophoresis paradigms. The first paradigm approximates the traditional clinic based iontophoresis application in which a total dose of 40 mA*min is delivered over ten minutes by passing a controlled current of 4 mA. This experimental arm is called HCSD for “high current, short duration”. The second paradigm approximates the newer paradigm of “low current, long duration” or LCLD iontophoresis in which a total dose of 40 mA*min is delivered over 400 minutes by passing a controlled current of 0.1 mA. Both experimental arms were controlled by also testing electrically inactive systems to evaluate the amount of dexamethasone delivered by diffusion without the aid of electrical The Experimental Results
The comparative efficacy of drug delivery by means of HCSD and LCLD is the important finding of this paper. To date, this is the only comparison of these two iontophoresis modalities in the literature. For the total amount of dexamethasone delivered, the results quoted from the legend of Figure 2 are as follows: “Total and passive drug delivery measured for 4.0-mA iontophoresis were 0.42+0.05 and 0.04+0.003, respectively. Total and passive drug delivery measured for 0.1-mA iontophoresis were 0.75+0.40 and 0.51+0.35, respectively Are there significant differences in the total amount of dexamethasone delivered by the two methods, given above? The values are summarized using Gaussian-based descriptions using standard deviations, so the assumption of a normal distribution was made by Anderson et al. (2003). While an n=3 is a low number of experimental replications, this is all the data we have to evaluate. So, for an n=3 and an alpha=0.5 a two-tailed t-test can be used to test if the total amount of dexamethasone delivery is truly The cardinal rule of statistical comparison: if no statistically significant difference
exists, one cannot even begin to compare the relative magnitudes of the compared values. For example, the measurement of two experimental groups results in means of 2 and 4. When statistically tested, the difference turns out not to be significant (p>0.5). At no point can it be stated that the first group resulted in twice the value of the second. Statistics point to no difference between the values. In the modern peer reviewed literature, no discussion of statistically insignificant trends is generally allowed. For the HCSD case, the difference between active and passive dexamethasone delivery is significant (p<0.05). Thus, we can safely say due to the significance of the difference, that active HCSD transmits ten times more than when the device is shut off. For the case of LCLD, the difference between active and passive conditions is not significant (p>0.05). It can only be stated, given the statistics, that LCLD does not pass any more dexamethasone than passive diffusion alone. There is no difference whether the LCLD device is electrically active or shut off. Can anything be made out of a difference between 0.75 and 0.51, such as “one mean is almost 50% larger than the other”? No. See the cardinal rule of statistical comparison, above. Can the amount actively delivered by HCSD and LCLD be compared? The answer is no for two reasons. The standard deviations are greatly different between the two cases, making comparison problematic, and even if this provision is ignored, the t-test indicates that this difference is not significant (p>0.05). So, statistically, no statements can be made about the relative differences of active HCSD and LCLD. There is no difference between the means of The preceding has dealt with the comparison of total amounts of dexamethasone into the agar of the chamber. Are there any differences in the subsets of dexamethasone in the layers of agar as shown in Figure 2? Statistical differences were not tested by the authors. However, the distributions in Figure 2B and C, given the size of the standard From this data, we can ask some pertinent questions. Is the amount of dexamethasone delivered by active HCSD iontophoresis any more than that with the device shut off? For the conventional short term iontophoresis (HCSD) the answer is yes. Does LCLD iontophoresis result in any more dexamethasone delivery when the device is on relative to off? The answer is “we can’t tell”. Does LCLD deliver more dexamethasone actively than HCSD? The answer is again “we can’t tell, given the reported results”. The only conclusion allowable is that HCSD delivers ten times more dexamethasone actively than passively. The only rationale for turning on an iontophoresis device is provided only for There is of course more to evaluating a scientific paper than looking at the reliability of the data. Also of concern is validity. How closely does the experiment resemble the real world in the parameters studied and controlled? The iontophoretic chamber does not model the human tissue well. It does not incorporate an approximate skin electrical resistance. The electrodes are too small to be realistic. The experimental current- controlled LCLD iontophoretic devices are not used clinically. The IontoPatch is a controlled-voltage device. The difference is marked in the face of finite, variable, and nonlinear human electrical tissue resistance (Prausnitz, 1996). II. Evaluation of skin blanching and cutaneous thermal changes in humans during
iontophoresis of dexamethasone
Anderson et al (2003) perform a largely unrelated second experiment evaluating the vascular reaction to the iontophoresis of dexamethasone into humans. Again the HCSD and LCLD iontophoresis paradigms are compared. However, different devices are used in this second part. For the HCSD experimental condition, a Life-Tec controlled-current generator was used. For the LCLD condition, a controlled-voltage Birch Point IontoPatch was used. (Note: in the previous delivery experiment, a controlled-current Iontophor was used for the LCLD condition. Controlled-current and controlled-voltage technologies are not comparable in the face of the appreciable electrical resistance of human tissue.) In both cases, 1.5 ml of 4 mg/ml of dexamethasone was used and a total dose of 40 mA*min was delivered. The cases differed in terms of duration of treatment: HCSD was delivered at varying current levels from 1.5 to 4 mA, thus duration ranged from 35 to 10 minutes. The LCLD was delivered at current levels from 0.05 to 0.16 (this is hypothetical as current was not measured during the experiment. Since the current output of voltage-controlled device is highly dependent on skin resistance, the value of delivered current could be zero for large skin impedances). The duration was fixed at 12 hours. The passive control was performed for both HCSD and LCLD cases, but for different periods of time. Since diffusion is a time dependent process, a control for duration is necessary to directly compare passive diffusion between cases. Based on McKenzie et al (1962) and McKenzie (1962) it is known that passively applied topical steroids produce blanching of the human skin. This blanching assay has only been validated for the passive case of no current. Current itself causes erythema, either due to direct vascular effects, neural effects, or histamine release which should interact in some unspecified way with the blanching induced by contact with a steroid. The end result of the iontophoresis was measured by a) human observation of the duration of visually identified skin blanching and b) by the use of a temperature sensor. Following completion of iontophoresis (after minutes for HCSD and after hours for LCLD) cutaneous temperature was sequentially measured and graphed in Figures 3 and 4. The HCSD condition resulted in skin cooling over 300 minutes. The LCLD condition resulted in an apparent warming of the skin over 300 minutes. Whatever the direction of the temperature deviation, baseline values are achieved in both conditions after 300 minutes. Subjective visual inspection of the duration of blanching was performed. Blanching after LCLD resulted in a mean value of 984 minutes, while HCSD produced a mean of 415 minutes. The blanching was also qualitatively different for the two conditions. For the HCSD, it did not appear for 165-195 minutes after iontophoresis, unlike the LCLD condition in which blanching appeared instantly. Why this is so is unexplained. The pertinent question is; why did the subjectively observed blanching time course not show up in the presumably more quantitatively measured temperature time course of Figures 3 and 4? The course of blanching after 180 minutes for HCSD does not register as a temperature difference in Figure 3. For the LCLD condition, the visually observed blanching for 900 minutes does not correspond to the temperature time course, which returns to baseline after 300 minutes. Anderson et al. (2003) provide two outcome measures of vasoreactivity, temperature and visually determined blanching. The results of the two measures are contradictory, another indication of a poorly validated assay. Why does HCSD result in erythema (presumably vasodilatation) and LCLD result in blanching (presumably vasoconstriction)? This is not answerable from the data of Anderson et al. (2003). The processes of vasodilation and vasoconstriction may be independent and additive. More probably, they are related in a nonlinear and unpredictable way. Once the relationship between vasoreaction, the action of steroids, and action of current is known, the test will be validated. Until validation, the conclusion “Finally, we provide evidence to suggest that comparable iontophoretic doses delivered by low currents over several hours are more effective than those delivered by higher currents over 10 to 30 minutes in the creation of a localized physiologic effect for DEX/DEX-P, based on the magnitude and duration of local cutaneous vasoconstriction.” cannot be made. It is equally realistic to postulate that the vasoactivity differences between HCSD and LCLD are due to a failure of the voltage- controlled Birch Point IontoPatch to deliver an appreciable current through the skin resistance. Vasodilatation associated with current is found for the HCSD case, but not the LCLD case. No measurement of current delivered by either HCSD or LCLD was Conclusions
1. A high current, short duration (HCSD) iontophoresis device delivers ten times more dexamethasone into a model of human tissue than it would if it were turned OFF.
2. A low current, long duration (LCLD) iontophoresis device delivers no more dexamethasone when turned ON as when turned OFF.
3. No differences in terms of amount of drug delivery can be ascertained between HCSD and LCLD from the data given. The delivery study results in comparisons that are statistically significant only for the HCSD iontophoresis. 4. The validity of the study is also questionable. A commercially available LCLD iontophoresis device (Birch Point IontoPatch) was not used in the drug delivery study
(part 1). Rather, a controlled current generator (Iontophor) coupled to very small nonstandard electrodes was used. No electrical resistance mimicking that of the human skin was incorporated into the test chamber. For the vasoreactivity study (part 2), a voltage controlled Birch Point IontoPatch was used. Thus when the two experiments are
compared, the constant current device and a voltage-controlled device are discussed as if 5. Furthermore, the unvalidated vasoreactivity assay resulted in conflicting results depending on the measurement system used. All one can say is that dexamethasone alone has an effect on the vasculature. The amount and time of current may change this effect. The interaction of the effects of dexamethasone and current is unknown. References
1. Sage, B.H. and Riviere, J.E. (1992) Model systems in iontophoresis – transport efficacy, Advanced Drug Delivery Reviews, 9:265-287. 2. Prausnitz, M.R. (1996) The effects of electric current applied to the skin: A review for transdermal drug delivery. Advanced Drug Delivery Reviews, 18:395-425.

Source: http://sad.prim.es/documentacion/documentacion/prim_hospitales/cuidados_intensivos/articulos_tecnicos/dolor/iontoforesis/iontopath.vs.ionto.doc.pdf