Electrical impedance as a potential tool to distinguish between allergic and irritant contact dermatitis

Transcription

1 Electrical impedance as a potential tool to distinguish between allergic and irritant contact dermatitis Miruna Nyrén, MD, Natalia Kuzmina, MD, and Lennart Emtestam, MD Stockholm, Sweden Background and Objective: The allergic contact reaction is a model reaction for studying the cellmediated immune system of the skin. In this study we use a noninvasive method, electrical impedance (IMP), to compare nickel (Ni) allergic contact reactions with an irritant contact reaction induced by sodium lauryl sulfate, which has already been carefully evaluated with this method. For this purpose, we included only Ni- and sodium lauryl sulfate induced reactions of very similar appearance. Methods: Various concentrations of Ni sulfate in distilled water were applied on the volar aspect of the forearms of 33 adult women who were allergic to Ni. Assessments were made using visual scoring, a new IMP technique, and transepidermal water loss 3 and 7 days later. Results: In the 19 patients who completed the study, 3 of the 4 impedance indices were significantly lower at the sodium lauryl sulfate sites than at the Ni sites on day 3; ie, the mean magnitude (MIX) (P.001), imaginary part (IMAX) (P.001), and real part (RIX) indices (P.01). Unlike the irritant reactions, no significant increases in transepidermal water loss occurred in the allergic contact reactions. This may be because, in reactions of the studied magnitude, an allergic contact reaction does not significantly affect the epidermis because the inflammatory process is located deeper in the dermis than an irritant reaction. Conclusion: This study suggests that IMP, as used herein, is suitable for distinguishing between contact reactions of allergic and irritant nature. Although pathophysiologic events in the tissue studied significantly modify impedance patterns, little is known about how to interpret the structural and chemical changes underlying these patterns. Studies are needed to determine the relation between anatomic or pathophysiologic parameters, and the findings using IMP and other established methods, such as chemical extraction and histopathology. (J Am Acad Dermatol 2003;48: ) The patch test is the only practical method for verifying contact sensitization. It was first devised by Jadassohn in 1896, later brought into general use by Bloch, and during the past few years, has been undergoing further development. There are still many uncertainties about patch tests such as the exposure time, vehicle, test site, false-negative From the Department of Medicine, Section of Dermatology and Venereology, Karolinska Institutet at Huddinge University Hospital. Funding sources: Supported by grants from the Swedish Council for Work Life Research, Swedish Society of Medicine, Karolinska Institutet, Edvard Welander Foundation, Finsen Foundation, Cancer and Allergy Foundation, Åke Wiberg Foundation, and Tore Nilson Foundation for Medical Research. Conflict of interest: None identified. Accepted for publication September 22, Reprint requests: Miruna Nyrén, MD, Department of Medicine, Section of Dermatology and Venereology, I 43, Karolinska Institutet at Huddinge University Hospital, SE Stockholm, Sweden. Miruna.Nyren@dermat.hs.sll.se. Copyright 2003 by the American Academy of Dermatology, Inc /2003/$ doi: /mjd tests despite positive clinical findings/history, and interpretation. A correct evaluation of a patch-test reaction in clinical practice is important because the medical prognosis of contact dermatitis depends on its cause and the possibility of avoiding repeated or continued exposure to the causal allergen. A positive test may be interpreted as allergic or irritative. Positive results are graded on a quantitative scale. 1 However, the interpretation is subjective and the distinction between allergic and irritant reactions requires experience. The clinical criteria for distinguishing between irritant and allergic patch-test reactions are useful when the patch test is strongly positive, but in many cases they are of little value when the reaction is weak, 2 because it may be impossible on clinical examination to distinguish between these reactions. There is, therefore, a need for a sensitive, noninvasive method for assessing the patch test in dermatology clinics and research laboratories. An objective method for evaluating the patch test would also permit comparison with results obtained by others. 394

2 J AM ACAD DERMATOL VOLUME 48, NUMBER 3 Nyrén, Kuzmina, and Emtestam 395 Several bioengineering methods, on the basis of various physical principles, have been used in research on patch tests and have been reviewed elsewhere. 3,4 These include transepidermal water loss (TEWL) measurements, laser Doppler flowmetry and ultrasound. 5-7 Invasive techniques have also been used, such as light and electron microscopy, immunohistochemistry, and ribonuclease protection assay, 8-12 to distinguish between allergic and irritant reactions. However, no method can be regarded as entirely satisfactory for this purpose. Biologic tissues have complex electrical impedance (IMP), which is a function of frequency because tissues contain components that have both resistive and charge storage (capacitive) properties, and pathophysiologic events are related to changes of the IMP of the tissue. 13 For example, some data suggest that impedance can be used to detect dental decay 14 and that there may be a relation between electrical current flow and the severity of cervical dysplasia. 15 A new method for noninvasive study of the skin on the basis of IMP has been described elsewhere. 16,17 In a recent open pilot study on 8 patients with nickel (Ni) allergic contact reactions, a particular pattern was noted that differed markedly from that in control patients 18 and from that in previous studies of irritant contact reactions. This study aims to compare the IMP and TEWL of allergic and irritative contact dermatitis lesions that are clinically similar, that is, in relevant clinical setting. MATERIAL AND METHODS Patients A total of 33 women with a history of metal allergy and a mean age of 39 years (range: years) were recruited from the department of dermatology and venereology at Huddinge University Hospital where they had previously had or reactions scored according to International Contact Dermatitis Research Group (ICDRG) recommendations. 1 To avoid patients with the angry back syndrome, 19 we included only patients who had reacted on standard epicutaneous testing to 1 or 2 patches. All measurements were made the same year, during late fall and winter, between September and December. During the study the relative humidity varied between 17% and 44%, and the temperature in the examination room was maintained at 22.3 C 1.04 C. The study was approved by the local ethics committee and informed consent obtained from all patients. Procedure Contact allergic reactions were produced in normal-appearing skin using Ni sulfate in distilled water at concentrations of 0.15%, 0.31%, 0.62%, 1.25%, and 2.5% (Chemotechnique, Tygelsjö, Sweden). The irritant, sodium lauryl sulfate (SLS) (approximately 99%) (Henkel, Düsseldorf, Germany) was dissolved in distilled water at a concentration of 2% by weight. Absorbent paper disks and 50 L of the substances were applied for 24 hours in 12-mm Finn Chambers (Epitest Ltd Oy, Hyrylä, Finland), and sealed to the skin with tape (Scanpor, Norgeplaster A/S, Oslo, Norway) using Pirilä s method. 20 The test sites were randomized on the midvolar aspect of both forearms; all Ni tests on one arm, and SLS and the reference on the other. An untreated site, without a chamber or tape, was used as the control. Ink marks were made on the skin close to the test sites to ensure correct placement of the instrument probes. Readings were made before applying the substances (day 0), and on days 3 and 7 in the following order: visual assessment, TEWL, and IMP. Visual assessment We used the visual scoring system for contact allergy reactions recommended by the ICDRG, where 1 represents a weak (nonvesicular) reaction, 2 a strong (edematous or vesicular) reaction, and 3 a very severe (bullous or ulcerative) reaction. 1 Similar guidelines were used to assess the irritant skin reactions. 21 Transepidermal water loss TEWL was measured (Evaporimeter EPI, Servomed AB, Kinna, Sweden). Readings were made when the level had stabilized, 30 to 60 seconds after applying the probe, and the values expressed as g/m 2 /h. TEWL is useful for quantifying the barrier status of the skin. 22,23 Measurements were made in accordance with the recommendations of the European Society of Contact Dermatitis standardization, apart from no use of a drought shield. 23 Electrical impedance We used a skin IMP spectrometer (developed by Dr Stig Ollmar at Karolinska Institutet, SciBase AB, Huddinge, Sweden) for this purpose. It records magnitude and phase spectra of impedance at 31 logarithmically distributed frequencies between 1 khz and 1 MHz in a defined volume below the probe. The volume can be adjusted to record spectra, stepwise, at 5 skin depths. 16 We used impedance spectra at all 5 settings at each test site. The data are condensed by using the magnitude and phase values at only 2 frequencies, 20 khz and 500 khz, and expressed as 4 indices, representing changes with frequencies along the 4 major aspects of IMP space as described in detail previously. 17,24

3 396 Nyrén, Kuzmina, and Emtestam JAM ACAD DERMATOL MARCH 2003 Statistical analysis Results were analyzed using a 2-way analysis of variance with repeated measures of 3 factors. 25 The repeating factor was Condition with the levels of reference, Ni, and SLS. The findings on each day were analyzed separately. If the correlation between each pair of repeated measures was not the same (test of sphericity), the degrees of freedom of the F test were reduced by multiplying each of the degrees of freedom by the Greenhouse and Geisser. When the F ratio for the factor condition was significant and the sphericity assumptions were met, the univariate test of significance for planned comparisons was used to make all pairwise comparisons among means. Otherwise planned comparisons were made between the repeated measures and the P values were corrected according to the Bonferroni procedure. 25 RESULTS Because our aim was to compare Ni- and SLSinduced reactions that were as similar in appearance and severity as possible, 14 participants were excluded on the first test reading in accordance with our criteria; that is, the Ni test was negative (2 patients), too weak (little, if any, erythema) (4), or too strong (confluent vesiculobullous) (8). Fig 1 shows typical clinical photographs of patients who were included or excluded from the study. Visual scoring On day 3, 5 of the 19 remaining patients had positive reactions to Ni graded 1, 5 to grade 2, and 7 to grade 3 ; 9 patients showed a 1 reaction tosls,8a2, and2a3. On day 7 the scores decreased. The results of the clinical readings are given in Table I. Transepidermal water loss Analysis of variance revealed significantly higher (P.001) mean values in the SLS-induced reactions on day 3, but not on day 7, than in the control patients (Fig 2). The changes in Ni-induced reactions were not significant on day 3 or 7. Electrical impedance The mean values and standard deviation were calculated for the 4 indices in all test sites. The measurements using IMP were made at 5 skin depths, but only the results from depth 5 are shown. Depth 5 includes information from all the more superficial depth settings. Analysis of variance showed significant changes in the following indices: magnitude (MIX) (Fig 3, A), imaginary part (IMIX) (Fig 3, B), and real part (RIX) (Fig 3, D). On day 3, the mean indices MIX and IMIX in the SLS-induced reactions were significantly lower than the Ni-induced ones (P.001) and in control patients (P.001). On day 7 the SLS-induced reactions caused significantly lower MIX and IMIX indices than the Ni-induced reactions (P.01). Similarly, on day 3, the mean RIX index was significantly lower in the SLS-induced reactions than in Ni-induced reactions and control patients. The RIX index was also lower in the SLS-induced reactions than the Ni-induced reactions on day 7. DISCUSSION Although various studies have shown good agreement among dermatologists evaluating patchtest reactions (eg, 5 dermatologists read 82% of all tests similarly), the patch assessment remains subjective, and depends on the reader s knowledge and experience. 26 It is mainly on the basis of scoring visible erythema, palpable edema, and the presence or absence of vesicles. Regardless of whether the skin reaction is of allergic or irritant nature, one of the typical signs of an inflammatory response is the formation of edema. 2 Measurements of TEWL reflect epidermal waterbarrier function and they have been used to evaluate allergic and irritant skin reactions before. 6 Unlike the irritant skin reaction, in the allergic reaction, no increases in TEWL occurred after 24 hours, suggesting that the inflammatory process, at least in the initial phase, does not disturb water loss throughout the epidermis. 6 In this study we found that, on day 3, TEWL increased in the irritant reactions also, but not significantly in the Ni sites. This accords with our previous findings. 16,27 Measurement of skin thickness by high-frequency ultrasound is another noninvasive method, which has been used to assess patch-test reactions. Serup and Staberg 5 showed that allergic and irritant patchtest reactions of similar severity on clinical examination differ in skin thickness. By applying the substances for a 24-hour period and making measurements on the following day, one can distinguish between allergic and irritant reactions at least by the amount of edema formed. 5 However, in clinical work, to avoid missing a significant number of positive reactions, 48 hours of exposure are usually recommended. 2 In this study, we selected the Ni site in each patient that most resembled the irritant site induced by SLS. Unlike in our previous study on Ni contact reactions, 18 in which we did not compare the findings with an irritant, in this study we compared the clinical appearance of the reactions to Ni, but did not take into account the concentration of Ni at the

4 J AM ACAD DERMATOL VOLUME 48, NUMBER 3 Nyrén, Kuzmina, and Emtestam 397 Fig 1. A, Patient with nickel (Ni) allergy showing location of test sites on volar aspect of forearm (day 3). B, Typical sodium lauryl sulfate induced reaction on day 3. Typical Niinduced reactions on day 3 are shown that were included in study (C), and too weak (D) and too strong for inclusion (E). site. Therefore, 15 of the enrolled 34 patients had to be excluded from further analysis because the reactions induced in those patients did not fulfill our criteria of close resemblance, that is, too weak (little, if any, erythema) or too strong (confluent vesiculobullous). Our main finding was that the IMP patterns of the 4 indices permitted us to distinguish between the 2 types of reactions. On day 3, the impedance indices MIX, IMIX, and RIX were lower in the SLS-induced reactions than in the Ni reactions. This difference tended to normalize on day 7. Unlike in our pilot

5 398 Nyrén, Kuzmina, and Emtestam JAM ACAD DERMATOL MARCH 2003 Fig 2. Transepidermal water loss (TEWL) expressed in grams per square meter per hour showing mean values (n 19), 95% confidence interval, in skin reactions to nickel sulfate (Ni) and sodium lauryl sulfate (SLS), and in normal skin (REF) as measured on days 0 (before application of substances), 3, and 7. Data were analyzed using 2-way analysis of variance. P values refer to comparisons of Ni and SLS (*** P.001). Table I. Results of visual scoring of patch tests to sodium lauryl sulfate and dose response to nickel sulfate on days 3 and 7 Patient No. SLS (2%) Nickel (%) /1 0/1 1/1 1/2 1/2 ND/1 2 2/1 0/0 0.5/0 0.5/0 0.5/0 2/1 3 1/1 0/0 1/0 1/1 1/1 2/1 4 1/1 1/1 1/0 1/1 2/1 1/1 5 1/1 3/2 3/2 3/2 3/2 3/2 6 2/2 3/2 3/2 0/2 3/2 3/2 7 1/1 0/0 0/0 1/0 1/1 1/1 8 2/1 ND/ND 2/1 2/1 2/1 2/1 9 2/1 1/1 1/1 0/ND 0/1 2/2 10 2/1 0/0 0/0 2/1 2/1 2/1 11 1/1 1/0 1/0 0/0 0/0 1/1 12 1/1 2/1 0/0 0/1 2/1 2/2 13 2/1 3/2 3/3 3/3 3/3 3/ND 14 3/1 3/2 3/2 2/0 3/2 3/2 15 1/1 1/2 3/1 3/3 3/3 3/3 16 3/1 0/0 2/0 1/0 0/0 2/1 17 2/1 3/2 3/3 3/ND 3/ND 3/ND 18 1/1 3/2 3/2 3/2 3/ND 3/ND 19 2/2 3/3 3/1 3/2 3/3 3/3 ND, No data; SLS, sodium lauryl sulfate. Nickel scores in bold are sites chosen for comparisons. study, where we found significant increases mainly in the MIX and IMIX indices in certain Ni concentrations, 18 in this study the changes in the impedance indices were similar to those in normal skin. At least 4 reasons may account for this. First, in the pilot study, the analysis was made by grouping the reactions according to the concentrations of Ni; in this study we used the macroscopic appearance for inclusion. Second, this study was larger (19 vs 8 patients). Third, the proportion of patients with Ni reactions of 1 and 2 was higher in this study because we aimed at simulating the difficult clinical problem of distinguishing between an irritant and an allergic reaction. Fourth, in this study Ni sulfate and

6 J AM ACAD DERMATOL VOLUME 48, NUMBER 3 Nyrén, Kuzmina, and Emtestam 399 Fig 3. Four electrical impedance indices MIX (a), IMIX (b), phase (PIX) (c), and RIX (d) show mean values (n 19), 95% confidence interval, for measurements of skin reactions to nickel sulfate (Ni), sodium lauryl sulfate (SLS), and of normal skin (REF) on days 0 (before application of substances), 3, and 7. Data were analyzed using 2-way analysis of variance. P values refer to comparisons of Ni and SLS (**.001 P.01; *** P.001). SLS were dissolved in distilled water, the same as for SLS, but in petrolatum in our pilot study. It is known that irritant reactions usually have faster kinetics than type IV allergic reactions, and on day 3 the allergic reaction is at its peak whereas the irritant one is already declining. The IMP method has also been found useful for assessing other immunologic and nonimmunologic skin reactions, including wheals of Gell and Coombs type I, tuberculin reactions, and irritant reactions induced by substances other than SLS as shown in studies from our laboratory. 16,28-30 The IMP indices are derived from 4 values: the magnitude and phases at 20 khz and at 500 khz. No assumptions about any tissue model are needed; the indices contain most of the information from a physical point of view in the selected frequency range. The conversion of the measured values into indices also entails normalization, which reduces the number of artefacts caused by individual variations. A comprehensive textbook on the basic elements of bioimpedance and bioelectricity has recently been published. 13 In conclusion, to our knowledge this is the first time that findings using IMP instrumentation indicate that one can distinguish between allergic and irritant contact reactions. This view is on the basis of patterns of changes in the impedance indices. In this study we determined the mean values in 19 patients, but the diagnostic accuracy of these indices alone is not sufficient to evaluate epicutaneous tests in a given patient. However, the use of all the information in the impedance spectra might make this possible. 31 Although the actual measurement is simple, more improvements in the technique are needed. The instrument provides a large amount of data, which complicates the interpretation, but an improved version of the software is in progress. Further automation of the data analysis and, probably, one-purpose-built variants of the instrument (eg, for a patch-test unit) would facilitate its use by clinicians. Although pathophysiologic events in the tissues studied significantly changed the impedance patterns, little is known about how to interpret the patterns from a structural and chemical point of

7 400 Nyrén, Kuzmina, and Emtestam JAM ACAD DERMATOL MARCH 2003 view. In studies with SLS-induced irritant contact reactions, the changes in the impedance parameters are partly correlated to the amount of extracellular edema. 16,27 One main criticism concerns the fact that different irritants produce different skin changes and our experience with SLS should not, perhaps, be uncritically extended to all irritants. Furthermore, although we clearly made great efforts to match the intensity of reactions to SLS and Ni, we had a mean score for SLS reactions of 1.8 on day 3, as compared with 2.2 for Ni at the same time point. This may have resulted in differences in the extent of the edema produced, which could account, to some degree, for the differential readings of the 2 chemicals. It is known that irritant reactions usually have faster kinetics than type IV allergic reactions; on day 3 the allergic reaction is at its peak whereas the irritant one is already declining. For an ideal comparison of the 2 reactions, they should both be analyzed at their peak activity, but this fit neither with the study protocol nor the clinical setting. More studies are clearly needed to determine the relation between anatomic or pathophysiologic parameters and the findings using IMP and other established methods, such as chemical extraction and histopathology. We thank Professor Jan E. Wahlberg and Dr Stig Ollmar for valuable comments on the manuscript. The technical assistance of Gun-Britt Karlberg, RN, is gratefully acknowledged. We also thank Drs Zoe and Francis P. Walsh for linguistic revision and Elisabeth Berg, section of statistics, Department of Learning, Informatics Management and Ethics (HIS), Karolinska Institutet, for statistical analysis. REFERENCES 1. Wilkinson DS, Fregert S, Magnusson B, Bandmann HJ, Calnan CD, Cronin E, et al. Terminology of contact dermatitis. Acta Derm Venereol 1970;50: Wahlberg JE. Patch testing. In: Rycroft RJG, Menné T, Frosch PJ, editors. Textbook of contact dermatitis. 2nd ed. Berlin: Springer; p Serup J, Jemec G. Handbook of non-invasive methods and the skin. Boca Raton (FL): CRC Press; Berardesca E, Maibach HI. Bioengineering and the patch test. Contact Dermatitis 1988;18: Serup J, Staberg B. Ultrasound for assessment of allergic and irritant patch test reactions. Contact Dermatitis 1987;17: Serup J, Staberg B. Differentiation of allergic and irritant reactions by transepidermal water loss. Contact Dermatitis 1987;16: Staberg B, Serup J. Allergic and irritant skin reactions evaluated by laser Doppler flowmetry. Contact Dermatitis 1988;18: Medenica M, Rostenberg A Jr. A comparative light and electron microscopic study of primary irritant contact dermatitis and allergic contact dermatitis. J Invest Dermatol 1971;56: Lachapelle JM. Histopathological and immunohistopathological features of irritant and allergic contact dermatits. In: Rycroft RJG, Menné T, Frosch PJ, editors. Textbook of contact dermatitis. 2nd ed. Berlin: Springer; p Vestergaard L, Clemmensen OJ, Sorensen FB, Andersen KE. Histological distinction between early allergic and irritant patch test reactions: follicular spongiosis may be characteristic of early allergic contact dermatitis. Contact Dermatitis 1999;41: Morhenn VB, Chang EY, Rheins LA. A noninvasive method for quantifying and distinguishing inflammatory skin reactions. J Am Acad Dermatol 1999;41: Hoefakker S, Caubo M, van t Erve EH, Roggeveen MJ, Boersma WJ, van Joost T, et al. In vivo cytokine profiles in allergic and irritant contact dermatitis. Contact Dermatitis 1995;33: Grimnes S, Martinsen ØG. Bioimpedance and bioelectricity basics. San Diego (CA): Academic Press; Longbottom C, Huysman M-CDNG, Pitts NB, Los P, Bruce PG. Detection of dental decay and its extent using a.c. impedance spectroscopy. Nat Med 1996;2: Brown BH, Tidy JA, Boston K, Blackett AD, Smallwood RH, Sharp F. Relation between tissue structure and imposed electrical current flow in cervical neoplasia. Lancet 2000;355: Nicander I, Ollmar S, Lundh Rozell B, Eek A, Emtestam L. Electrical impedance measured to five skin depths in mild irritant dermatitis induced by sodium lauryl sulphate. Br J Dermatol 1995; 132: Nicander I, Ollmar S, Eek A, Lundh Rozell B, Emtestam L. Correlation of impedance response patterns to histological findings in irritant skin reactions induced by various surfactants. Br J Dermatol 1996;134: Nicander I, Ollmar S, Lundh Rozell B, Emtestam L. Allergic contact reactions in the skin assessed by electrical impedance a pilot study. Skin Res Technol 1997;3: Mitchell JC. The angry back syndrome: eczema creates eczema. Contact Dermatitis 1975;1: Pirilä V. Chamber test versus patch test for epicutaneous testing. Contact Dermatitis 1975;1: Willis CM, Stephens JM, Wilkinson JD. Experimentally-induced irritant contact dermatitis: determination of optimum irritant concentrations. Contact Dermatitis 1988;18: Blichmann CW, Serup J. The effects of skin moisturizers determined by the measurement of transepidermal water loss. Ugeskr Laeger 1987;149: Pinnagoda J, Tupker RA, Agner T, Serup J. Guidelines for transepidermal water loss (TEWL) measurement: a report from the standardization group of the European society of contact dermatitis. Contact Dermatitis 1990;22: Nyrén M. Skin sensitivity testing a biophysical approach [thesis]. Stockholm: Karolinska Institutet; Kirk RE. Experimental design: procedures for the behavioral sciences. Pacific Grove (CA): Brooks/Cole Publishing Co; Bruze M, Isacsson M, Edman B, Björkner S, Möller H. A study on expert reading of patch test reactions: inter-individual accordance. Contact Dermatitis 1995;32: Ollmar S, Nyrén M, Nicander I, Emtestam L. Electrical impedance compared with other non-invasive bioengineering techniques and visual scoring for detection of irritation in human skin. Br J Dermatol 1994;130: Nicander I, Ollmar S. Mild and below threshold skin responses to sodium lauryl sulphate assessed by depth-controlled electrical impedance. Skin Res Technol 1997;3: Emtestam L, Nyrén M. Electrical impedance for quantification and classification of experimental skin reactions. Am J Contact Dermat 1997;8: Nyrén M, Hagströmer L, Emtestam L. Instrumental measurement of the mantoux test: differential effects of tuberculin and sodium lauryl sulphate on impedance response patterns in human skin. Dermatology 2000;201: Ollmar S. Making electronic biopsies into a viable future for noninvasive diagnostics with electrical impedance. Med Biol Eng Comput 1999;37:116-7.