Modulation of the atopy patch test reaction by topical corticosteroids and tar

Transcription

1 Dermatologic and ocular diseases Modulation of the atopy patch test reaction by topical corticosteroids and tar Elisabeth G. Langeveld-Wildschut, MD, PhD, a Hannes Riedl, MD, a Theo Thepen, PhD, a Ilse C. Bihari, a Piet L. B. Bruijnzeel, MD, PhD, b and Carla A. F. M. Bruijnzeel- Koomen, MD, PhD a Utrecht and Rijswijk, The Netherlands Background: Pharmacologic studies in atopic eczema (AE) are difficult to standardize. Patients with AE differ in the stage of their skin disease (acute, subacute, chronic). Objective: This study was designed to assess macroscopic and microscopic effects of pretreatment with topical glucocorticosteroids (GCSs) and tar on the atopy patch test (APT) reaction in patients with atopic eczema. Methods: Nonlesional skin of the back of patients with AE (n = 6) was treated for 3 weeks at 3 different sites with triamcinolonacetonide 0.1% in cetamacrogol ointment (GCSs), pix liquida 10% in cetamacrogol ointment (tar), and cetamacrogol ointment (vehicle), respectively. APTs were performed, and biopsy specimens were taken from all these sites (time = 0 and 24 hours) for immunohistochemical analysis. Results: Treatment with both GCSs and tar was able to reduce the macroscopic outcome of the APT reaction. Furthermore, both treatment modalities had an almost equally inhibiting effect on the influx of T cells, eosinophils, and CD1 +, RFD1 +, IFN-γ +, and IL-4 + cells, as well as on the percentage of vessels expressing the adhesion molecules vascular cell adhesion molecule 1 and E-selectin in response to epicutaneous aeroallergen challenge. Conclusion: Although both treatments significantly reduced the various cellular constituents of allergic inflammation, all cell types remained present. In addition, this study shows that the APT can be used to evaluate the effect of topical antiinflammatory treatments on allergic inflammation in patients with AE. (J Allergy Clin Immunol 2000;106: ) Key words: Atopic eczema, atopy patch test, tar, glucocorticosteroids, therapy From a the Department of Dermatology/Allergology, University Medical Center Utrecht, Utrecht; and b the Department of Pharmacology, T.N.O., Rijswijk. Reprint requests: Carla A. F. M. Bruijnzeel-Koomen, MD, PhD, Department of Dermatology/Allergology, University Medical Center Utrecht, PO Box 85500, 3508 GA Utrecht, The Netherlands. Copyright 2000 by Mosby, Inc /2000 $ /1/ doi: /mai Abbreviations used AE: Atopic eczema APT: Atopy patch test GCS: Glucocorticosteroid GP: Grass pollen HDM: House dust mite LPR: Late-phase reaction VCAM: Vascular cell adhesion molecule Epicutaneous application of aeroallergens induces eczematous skin reactions in about 50% of aeroallergensensitized patients with atopic eczema (AE). This procedure is called the atopy patch test (APT). The APT is a reproducible test, with reactions that do not occur in nonatopic control persons. 1 Macroscopically, APT reactions resemble AE lesions by showing erythema, induration, papules, and/or vesicles. Microscopically, the APT reaction is characterized by acanthosis, spongiosis, and a dermal infiltrate that consists of predominantly CD4 + T cells, dendritic cells, activated eosinophils, and CD1 + cells. 2-4 The APT is proposed as a model for the study of inflammatory reactions in AE by reason of its close macroscopic and microscopic similarities with lesional skin of patients with AE. 3 The subject of the present study was to assess how topical antieczema treatment of the skin influences the macroscopic and microscopic outcome of the APT reaction. Triamcinolonacetonide 0.1% ointment and pix liquida 10% ointment, a pine tar, were compared to determine whether the APT would reflect clinical benefit for one or the other therapy. These specific topical treatments and concentrations were chosen because they are both frequently used in The Netherlands, and their clinical effectiveness in patients with AE is evident. Topical glucocorticosteroids (GCSs) have been used for 4 decades to treat AE and other inflammatory skin diseases. They are the most frequently prescribed of all dermatologic drug products. 5 The major actions of GCSs in eczema are the modulation of the infiltrating inflammatory cells, as well as the spectrum of mediators, including cytokines, chemical modulators, and enzymes, which are released in the tissue. 6 Tar treatment has a long-standing history but is still actively used in modern practice. It is effective in the treatment of chronic lesions on patients with AE. 7,8 Literature concerning the mechanism of action of tar is very scarce. There are probably hundreds of different chemical compounds in tar, but the specific chemicals that are therapeutic in eczema are unknown. Tar treatment is not popular in patients because it stains and smells. However, it 737

2 738 Langeveld-Wildschut et al J ALLERGY CLIN IMMUNOL OCTOBER 2000 may be used as an alternative for, or as an intermittent therapy during, long-term treatment with topical GCSs to reduce the side effects of the latter drug. Various studies have been performed in which the latephase reaction (LPR) after intracutaneous allergen injection was used as a test system to evaluate the effects of topical and systemic GCSs. It was shown that prolonged topical treatment with GCSs (0.05% clobetasol 17-propionate) could block the LPR and also the immediate allergen-induced wheal-and-flare response. 9,10 Oral prednisolone (20 mg daily for 5 days) significantly inhibited both the size of the LPR at 24 hours and the accompanying local eosinophil infiltration. 11 The LPR is macroscopically characterized by swelling and erythema and histologically characterized by edema and a dermal infiltrate consisting of mononuclear cells, eosinophils, neutrophils, and basophils. 12 This does not correspond with the macroscopic and microscopic features of lesional AE. Therefore the APT is considered to be a better model for the study of AE than the LPR. 3 It thus may also be a more appropriate instrument for evaluating the effects of antieczema therapies than the LPR. In the current study nonlesional skin of the back of patients with AE (n = 6) was pretreated for 3 weeks at 3 different sites with triamcinolonacetonide 0.1% in cetamacrogol ointment (GCS), pix liquida 10% in cetamacrogol ointment (tar), and cetamacrogol ointment (vehicle), respectively. The effects of these different treatments on clinical outcome of the APT, cellular infiltrate, expression of IL-4 and IFN-γ, and the adhesion molecules E-selectin and vascular cell adhesion molecule (VCAM) 1 during the APT reaction were studied. Both treatments reduced the macroscopic and microscopic outcome of the APT reaction. This study shows that the APT may be used to test the efficacy of topical anti-inflammatory drugs in the treatment of AE. METHODS Patients Six patients with AE, diagnosed according to the criteria of Hanifin, 13 were selected on the basis of demonstration of a positive APT reaction at 24 hours for house dust mite (HDM), grass pollen (GP), or both. Patients characteristics are shown in Table I. During the study, all patients were hospitalized. Their backs were treated according to the schedule described in the study design. The clinical activity of the skin was established according to the Costa score. 14 The Costa score at admission varied between 35 and 69 (mean, 46.6; Table I). At the time of patch testing (3 weeks later), the Costa score varied between 9 and 27 (mean, 19). The patients ranged in age from 19 to 32 years (mean, 26.5 years). Systemic GCSs were not administered. Oral antihistamines were discontinued for at least 7 days before testing. None of the patients was treated with UV therapy before testing. At least 14 days before the procedure was started, topical therapy on the back was discontinued. Total serum IgE levels of all patients were determined by using the paper radioimmunosorbent test (PRIST; Pharmacia, Uppsala, Sweden) according to the manufacturer s instructions (mean, 8069 ku/l). Specific serum IgE levels for HDM (being Dermatophagoides pteronyssinus) and GP of all patients were measured by using the CAP method (Pharmacia, Uppsala, Sweden) according to the manufacturer s instructions. All participating patients gave their informed consent. Study design The back of all patients was divided in 3 equal parts and marked with a skin marker. In some patients (n = 3) there was very mild eczema on the skin of the back before treatment was started. However, after 3 weeks of treatment, no visible eczema was seen. The 3 different parts of the back were treated for 3 weeks with cetamacrogol ointment (Genpharma, Maarssen, The Netherlands), triamcinolonacetonide 0.1% in cetamacrogol ointment (Genpharma), and pix liquida 10% in cetamacrogol ointment (OPG Pharma, Utrecht, The Netherlands), respectively. The topical preparations were applied every morning as a thin film on the marked areas by a qualified nurse. After application, the areas were covered with Tricofix (Beiersdorf, Hamburg, Germany), a fixating dressing to minimize contamination by contact with clothing. The last application was given 24 hours before APTs were performed with either HDM or GP (see below). Three hours before the APTs were performed, the rest of the ointments were removed with oleum arachidis (OPG), followed by washing with water and carefully drying of the skin with a clean towel. The investigator who performed the APTs did not know which treatment was applied at each site. Atopy patch tests After 3 weeks of treatment, APTs were performed in duplicate in the center of the 3 marked areas on the back, always between 11 and 12 AM. In two patients APTs were also performed after a period of 1 and 2 weeks of treatment. APTs were performed as described previously. 1 The distance between the test sites was 3 cm. Commercially available allergen preparations provided by HAL (0.08 ml per test; Haarlem s Allergenen Laboratorium, Haarlem, The Netherlands) were used, which contained HDM (D pteronyssinus) or GP (a mixure of Agrostis stolonifera, Anthoxanthum odoratum, Dactylis glomerata, Lolium perenne, Arrhenatun elatuis, Festuca rubra, Poa pratenis, Secale cereale, Holcus lanatus, and Phleum pratense) at a concentration of 10,000 AU/mL. APTs were performed with the allergen (either HDM or GP) for which the patient previously showed a positive APT reaction. If previous APT reactions were positive for both HDM and GP, then APTs were performed with HDM. APT reactions were recorded after 24 hours as follows: +, erythema with slight induration; 2+, erythema with papules; and 3+, erythema, papules, and vesicles. 1 In all patients the diluent (provided by HAL) alone was tested as a negative control. Tissue processing and immunohistochemical staining All patients gave us permission to take punch biopsy specimens (3 mm, 1% lidocaine for anesthesia) from APT sites at 0 and 24 hours from the 3 different sites on the back. Both single-staining and double-staining procedures were performed, as described earlier. 3,4 The monoclonal mouse antibodies used in single-staining experiments were anti-cd3 (1:50, Leu-4; Becton Dickinson, San Jose, Calif), anti-eg2 (1:400; Pharmacia, Uppsala, Sweden), anti- CD1 (1:50; Dako M-0731, Glostrup, Danmark), anti-rfd1 (1:700; kindly provided by Dr L. W. Poulter, The Royal Free Hospital and school of Medicine, London, UK), and anti-il-4 (1:20) and anti- IFN-γ (1:40; both kindly provided by Dr F. Kalthoff, Sandoz, Vienna, Austria). To determine adhesion molecule expression on endothelial cells, sections were double stained with biotinylated Ulex Europaeus agglutinin (UEA-1, 1:200; Sigma, St Louis, Mo) as a panendothelial cell marker and anti-cd62e (staining E-selectin, 1:100; Genzyme, Cambridge, Mass) or anti-cd106 (staining VCAM-1, 1:100).

3 J ALLERGY CLIN IMMUNOL VOLUME 106, NUMBER 4 Langeveld-Wildschut et al 739 TABLE I. Patients characteristics Patient No. Sex Age (y) Costa score * Total IgE (ku/l) Specific/IgE (IU) 1 F (GP) 2 M >100 (HDM) 3 M ,598 >100 (HDM) 4 F ,250 >100 (GP) 5 M (HDM) 6 M >100 (HDM) * At admission. TABLE II. Macroscopic APT reactions (time = 24 hours) after treatment of the skin for 3 weeks with cetamacrogol ointment (vehicle), triamcinolonacetonide 0.1% in cetamacrogol ointment (GCS), and pix liquida 10% in cetamacrogol ointment (tar) Patient No. Allergen APT 24-hour vehicle APT 24-hour GCS APT 24-hour tar 1 GP 2+ Negative Negative 2 HDM 1+ Negative Negative 3 HDM 2+ Negative Negative 4 GP 2+ Negative 1+ 5 HDM 2+ Negative Negative 6 HDM Quantification of staining Biopsy specimens were coded, and the number of infiltrating cells in each section was assessed in blinded fashion at 400 magnification. In the epidermis the numbers of antibody-staining cells overlying 200 basal cells were counted. In the dermis all cells between the dermal-epidermal junction to the base of the hair follicles were counted, and the number of positive cells was calculated per square millimeter. In fields containing sweat ducts and hair shafts, only intervening dermal regions were counted. The percentage of dermal vessels expressing E-selectin and VCAM-1 was assessed by calculating the ratio of vessels staining positive for both Ulex and one of the adhesion molecules and cells staining positive for Ulex only. Cells staining for IL-4 and IFN-γ were counted in the dermis and not in the epidermis. Statistical analysis Statistical differences were evaluated by using the Wilcoxon signed-rank sum test for paired data. A P value of less than.05 was considered significant. RESULTS Comparison of macroscopic reactions All patients with AE had positive APT reactions (time = 24 hours) on the vehicle-treated skin (Table II). These reactions resembled lesional AE skin and were characterized by erythema, induration, papules, and sometimes vesicles. The scores of the APT reactions varied from 1+ to 3+. In 2 patients APTs were performed after 1, 2, and 3 weeks of treatment with GCSs and tar. A complete inhibition of the APT reaction was only observed after a treatment period of 3 weeks. In all patients the macroscopic APT reactions at 24 hours at either GCS- or tar-treated skin (treatment period of 3 weeks) were less in magnitude compared with the APT reactions at the vehicle-treated skin (Table II). At the GCS-treated skin, only 1 of 6 patients showed a positive APT reaction, which was less in magnitude compared with the APT reaction performed on vehicle-treated skin (Table II, patient 6). At the tar-treated skin, 2 of 6 patients showed a positive APT reaction, both of which were less in magnitude compared with the APT reactions at vehicle-treated skin (Table II, patients 4 and 6). In all patients patch testing with the diluent did not induce a clinical reaction at any of the sites. Comparison of cellular composition in vehicle-treated, GCS-treated, and tar-treated skin at 0 hours Epidermis. At all 3 treatment sites, very few CD3 cells were present in the epidermis at time 0, and there appeared no significant difference between the different sites. No EG2 cells were observed at either of the treatment sites. There was a slight but not significant decrease in CD1 and RFD1 cells in the epidermis of both GCS- and tar-treated skin compared with vehicle-treated skin (Fig 1, A). Dermis. The number of CD3, CD1, and RFD1 cells was lower in both GCS- and tar-treated skin compared with vehicle-treated skin (P <.05). No significant difference between GCS- and tar-treated skin was seen. No EG2 cells were seen at either of the sites. The number of cells staining with IL-4 was low. At the vehicle-treated skin, this number was slightly but not significantly higher than at the GCS- and the tar-treated skin. There were slightly higher numbers of cells expressing IFN-γ at the vehicle-treated skin compared with the GCS- and tartreated sites (P <.05). The percentage of vessels staining for E-selectin and VCAM-1 at the vehicle-treated skin was significantly higher compared with the GCS- and tar-treated skin (P <.05; Fig 1, A).

4 740 Langeveld-Wildschut et al J ALLERGY CLIN IMMUNOL OCTOBER 2000 A FIG 1. APTs (time = 0 and 24 hours) performed on vehicle-, GCS-, and tar-treated skin, respectively. A, Numbers of CD3 +, EG2 +, CD1 +, and RFD1 + cells in the epidermis (per 200 basal membrane cells) and dermis (per square millimeter). B, Numbers of cells staining with anti-il-4 and anti-ifn-γ. C, Percentage of vessels double staining for both Ulex and E-selectin and both Ulex and VCAM-1.

5 J ALLERGY CLIN IMMUNOL VOLUME 106, NUMBER 4 Langeveld-Wildschut et al 741 B C FIG 1. Continued Comparison of cellular composition of the APT reaction at 24 hours (vehicle-treated skin) with that at 0 hours (vehicle-treated skin) Epidermis. There was an increase in CD3, EG2, CD1, and RFD1 cells at 24 hours compared with at 0 hours (P <.05; Fig 1, A). Dermis. An increase in CD3, EG2, CD1, and RFD1 cells was observed at 24 hours compared with at 0 hours (P <.05). There was also a significant increase of cells staining for IL-4 and IFN-γ and an increase in the percentage of vessels expressing E-selectin and VCAM-1 at 24 hours (Fig 1, B and C). Comparison of cellular composition in the APT reaction (time = 24 hours) at vehicletreated skin with that at GCS-treated skin and tar-treated skin Epidermis. At the vehicle-treated site, the numbers of CD3, EG2, CD1, and RFD1 cells were significantly higher than at the GCS- and tar-treated sites (P <.05; Fig 1, A). There was no significant difference between GCSand tar-treated skin. Dermis. At the vehicle-treated site, the numbers of T cells, eosinophils, CD1 cells, and RFD1 cells were significantly higher than at the GCS- and tar-treated sites (P <.05; Fig 1, A). Furthermore, there was a higher number of cells staining with IL-4 and IFN-γ at the vehicle-treated site (Fig 1, B). The percentage of vessels expressing E-selectin, VCAM, or both was significantly higher at the vehicle-treated skin compared with the GCS- and tartreated skin (P <.05; Fig 1, C). There was no significant difference in phenotypic markers between GCS- and tartreated skin. DISCUSSION The APT is a reproducible model for the study of allergic inflammatory reactions in AE. 2,3 The aim of this study was to investigate in what way topical antieczema therapies influence the allergic inflammation during the APT reaction. For this purpose, macroscopic, as well as microscopic, effects of topical treatments with triamcinolonacetonide 0.1% in cetamacrogol ointment (GCS) and pix liquida 10% in cetamacrogol ointment (a pine tar) on the APT reaction were evaluated. Andersson and Pipkorn 9 and Pipkorn et al 10 pointed to the treatment time as a crucial factor for the expression of some GCS

6 742 Langeveld-Wildschut et al J ALLERGY CLIN IMMUNOL OCTOBER 2000 effects. In one of these studies 10 topical GCS treatment of the skin (0.05% clobetasol-17-propionate) was performed for 1 week, and this caused a 30% to 40% reduction in the allergen-induced flare response. By extending the treatment time to 4 weeks, an almost complete reduction in the reaction could be accomplished. We chose a treatment period of 3 weeks because in 2 of 2 tested patients a complete inhibition of the clinical APT reaction was seen after a treated period of 3 weeks that was not present after a treatment period of 1 or 2 weeks. Macroscopically, there was a clear influence of both treatments on the APT reaction. In all patients the macroscopic magnitude of the APT reactions performed on both GCS- and tar-treated skin was diminished compared with APT reactions performed on skin that was treated with the vehicle only (Table II). Nevertheless, in one patient (Table II, patient 6) who showed the strongest APT reaction on vehicle-treated skin (3+), a clinical positive APT reaction was still present on both GCS-treated (1+) and tartreated (1+) skin, and in another patient (Table II, patient 4) a positive APT reaction appeared on tar-treated skin (1+), although it was less in magnitude than that on vehicle-treated skin (2+). It is possible that the required treatment time for a complete inhibition of the clinical APT reaction varies between patients, possibly being dependent on the extent of inflammation present at the moment at which the treatment was started. Presumably, the treatment time of 3 weeks was not sufficient for these 2 patients. Unfortunately, we were not able to repeat the APTs in these patients after a period of 4 weeks of treatment. Further studies should therefore be performed to investigate the time course of the inhibitory effect from both GCS and tar treatment on the outcome of the APT reaction. Microscopically, the effects of both treatments on (1) the numbers of CD3 +, EG2 +, CD1 +, and RFD1 + cells (Fig 1, A), (2) the numbers of cells staining for the cytokines IFN-γ and IL-4 (Fig 1, B), and (3) the numbers of blood vessels expressing the adhesion molecules E-selectin and VCAM-1 (Fig 1, C) were studied. After treatment for 3 weeks with GCS and tar, respectively, a significant decrease was seen at 0 hours in the dermal numbers of CD3 +, CD1 +, and RFD1 + cells and in the percentage of blood vessels expressing the adhesion molecules E-selectin and VCAM-1 compared with skin that was treated with the vehicle only. However, all the above-mentioned cells were still present in both GCSand tar-treated skin at time 0, although in lesser numbers compared with those found in the vehicle-treated skin. Patients 4 and 6 showed higher numbers at time 0 on the GCS- and tar-treated skin for almost all cell types (Fig 1, A). This may explain why in these patients positive APT reactions developed even on skin pretreated with GCSs (patient 6) or tar (patients 4 and 6). In these patients the application of aeroallergens might have triggered the higher numbers of remaining inflammatory cells, which were able to heighten the inflammatory cascade, resulting in a positive APT reaction. An exacerbation of eczema after topical treatment, which frequently occurs in clinical practice, may be explained by this mechanism. The APT (time = 24 hours) performed on vehicletreated skin was associated with a significant increase in all phenotypic markers when compared with time 0, confirming our previous studies. 2-4 The inflammatory mechanism involved in the APT was considerably inhibited by treatment with GCSs and tar. First, the numbers of eosinophils, T cells, CD1 + cells, and RFD1 + cells in the APT reactions (time = 24 hours) on both GCS- and tartreated skin were decreased compared with those in APT reactions performed on vehicle-treated skin (Fig 1, A). Second, it was shown that both treatment modalities had an inhibitory effect on the allergen-induced increase of cells staining for IFN-γ and IL-4 (Fig 1, B). Third, the percentage of blood vessels expressing the adhesion molecules E-selectin and VCAM-1 was decreased in APT reactions at the GCS- and tar-treated skin compared with APT reactions at the vehicle-treated skin (Fig 1, C). We conclude that the effect of both treatment modalities on the allergic inflammation during the APT is not specifically directed to one cell type. Furthermore, cytokine production and adhesion molecule expression are influenced by the two therapies. A striking finding of this study is that triamcinolonacetonide 0.1% and pix liquida 10% had similar effects on both the macroscopic and microscopic APT reaction. To our knowledge, there is only one other article in which topical GCS treatment is compared with topical tar treatment. In that study the effect of 1% hydrocortisone cream was compared with that of 1% Stantar (purified coal tar cream), and these two treatments were found not to differ in their efficacy in reducing the symptoms of AE. 15 However, it is not possible to compare the latter study with our study because of differences in study design and the use of topical GCSs and tar. The results of this study are important for the technical procedure of performing APTs. In clinical practice it is possible that patients with AE have applied topical GCSs or tar to their skin shortly before the patch tests are performed, which may result in false-negative APT reactions. Therefore studies should be performed on the time course of the inhibitory effects of different topical treatments. At this moment, we recommend that patients attending for patch testing with aeroallergens withdraw the application of any anti-inflammatory topical drug on the back for at least 2 weeks. Classical allergic patch test reactions can also become false negative when performed on GCS-treated skin. Recently, Green 16 showed that pretreatment with betamethasone dipropionate 0.05% (a potent corticosteroid) two times a day for 3 days caused total or partial suppression of the contact allergic reaction in 8 of 10 patients. Topical hydrocortisone acetate (a very mild corticosteroid) was found to have no effect on allergic patch test reactions. 17 The effect of tar treatment on the outcome of classical allergic patch test reactions has, to the best of our knowledge, never been studied. In conclusion, this study shows that topical GCS and tar therapy are both able to decrease the outcome of the

7 J ALLERGY CLIN IMMUNOL VOLUME 106, NUMBER 4 Langeveld-Wildschut et al 743 macroscopic and microscopic APT reaction. Significant differences concerning clinical and histologic outcome between the two treatments were not detected. Furthermore, this study indicates that the APT may be a useful model with which to evaluate the efficacy of topical antieczema treatments in suppressing the allergic inflammation. This system could have various advantages: (1) the APT is a standardized and reproducible procedure; (2) the system is easy to perform, and only small parts of the skin have to be treated with the concerning topical therapy; (3) multiple therapies, varying in concentration or active components, can be compared at the same time in the same patient; and (4) fewer patients are necessary to compare topical preparations because the patient serves as his or her own intrinsic control. Taking into account these advantages, we suggest use of the APT as an instrument to evaluate the efficacy of (new) topical anti-inflammatory treatments on allergic inflammation in patients with AE. REFERENCES 1. Langeveld-Wildschut EG, van Marion AM, Thepen T, Mudde GC, Bruijnzeel PLB, Bruijnzeel-Koomen CAFM. Evaluation of variables influencing the outcome of the atopy patch test. J Allergy Clin Immunol 1995;96: Bruijnzeel-Koomen CAFM, Van Wichen DF, Spry CJF, Venge P, Bruijnzeel PLB. Active participation of eosinophils in patch test reactions to inhalant allergens in patients with atopic dermatitis. Br J Dermatol 1988;118: Langeveld-Wildschut EG, Thepen T, Bihari IC, Van Reijsen FC, De Vries JM, Bruijnzeel PLB, et al. Evaluation of the atopy patch test and the cutaneous late-phase reaction as relevant models for the study of allergic inflammation in patients with atopic eczema. J Allergy Clin Immunol 1996;98: Thepen T, Langeveld-Wildschut EG, Bihari IC, Van Wichen DF, Van Reijsen FC, Mudde GC, et al. Biphasic response against aeroallergens in atopic dermatitis showing a switch from an initial Th2 response to a Th1 response in situ: an immunocytochemical study. J Allergy Clin Immunol 1996;97: Surber C, Itin PH, Bircher AJ, Maibach HI. Topical corticosteroids. J Am Acad Dermatol 1995;32: Schleimer RP. Effects of glucocorticosteroids on inflammatory cells relevant to their therapeutic applications in asthma. Am Rev Respir Dis 1990;141:S Van der Valk PGM, Snater E, Verbeek-Gijsbers W, Duller P, van de Kerkhof PCM. Out-patients treatment of atopic dermatitis with crude coal tar. Dermatology 1996;193: Przybilla B, Eberlein Konig B, Rueff F. Practical management of atopic eczema. Lancet 1994;343: Andersson M, Pipkorn U. Inhibition of the dermal immediate allergic reaction through prolonged treatment with topical steroids. J Allergy Clin Immunol 1987;79: Pipkorn u, Hammarlund A, Enerback L. Prolonged treatment with topical glucocorticoids results in an inhibition of the allergen-induced wealand-flare response and a reduction in skin mast cells numbers and histamine content. Clin Exp Allergy 1989;19: Taborda-Barata L, Jacobson M, Walker S, Njuki F, Ying S, Randev P, et al. Effect of cetirizine and prednisolone on cellular infiltration and cytokine mrna expression during allergen-induced late cutaneous responses. Clin Exp Allergy 1996;26: Charlesworth EN, Hood AF, Soter NA, Kagey Sobotka A, Norman PS, Lichtenstein LM. Cutaneous late-phase response to allergen. Mediator release and inflammatory cell infiltration. J Clin Invest 1989;83: Hanifin JM. Diagnostic features of atopic dermatitis. Acta Derm Venereol Suppl Stockh 1980;92: Costa C, Rilliet A, Nicolet M, Saurat JH. Scoring atopic dermatitis: the simpler the better? Acta Derm Venereol Stockh 1989;69: Munkvad M. A comparative trial of clinitar versus hydrocortisone cream in the treatment of atopic eczema. Br J Dermatol 1989;121: Green C. The effect of topically applied corticosteroids on irritant and allergic patch test reactions. Contact Dermatitis 1996;35: Haxthausen H. Studies on the local effect of various corticosteroids on experimentally produced eczematous reactions. Acta Derm Venereol 1956;36:381-8.