Thirty-five healthy volunteers (15 non-atopic and 20 atopic), with no sign of any skin disease participated in this study, which was approved by the Southampton and South West Hampshire Research Ethics Committee and was conducted according to the Declaration of Helsinki. All gave written informed consent to participate in the studies.
At the first visit, the suitability of the subjects to take part in the study was assessed and whether they met the inclusion and exclusion criteria. Exclusion criteria included pregnancy, the presence of skin disease and taking drugs that might interfere with the study, including corticosteroids, antihistamines, antidepressants and psychotropic drugs. Atopic status was determined by the presence of a history of allergic disease, and a positive skin prick test to house dust mite, grass pollen or cat allergen. For the skin prick tests histamine and sterile saline were used as positive and negative controls. A positive response was defined as a wheal of more than 3 mm greater than the negative control.
SCG (a gift from Hewlett Healthcare Ltd, Derby, UK.) was introduced into the skin using iontophoresis (MIC1-e, Moor Instruments Ltd, Axminster, Devon, UK). The chamber was fixed to the skin, filled with a 4% solution of SCG in reverse osmosis purified water and a total charge of 8 mC applied over a period of 40 seconds. It was calculated that this charge would introduce 38.8 μg of SCG into the skin over an area of approximately 0.8 cm2 (1 cm diameter). At control sites, the iontophoresis chamber was filled with reverse osmosis purified water. Readings of blood flux were taken before and 5 minutes after iontophoresis to assess any drug- or iontophoresis-induced dermal vasodilatation.
SCG cutaneous emulsion
Subjects were supplied with 150 ml bottles containing either the cutaneous emulsion base alone (vehicle) or the emulsion with SCG at 1%, 2% or 4% w/w concentration (supplied by Hewlett Healthcare Ltd). The emulsions were identical in appearance and perception after application to the skin.
Histamine (Sigma, Poole, Dorset, UK) dissolved in sterile saline (0.9% NaCl) was injected intradermally (20 μl of 1 μM or 300 nM) into the centre of the iontophoresis sites or sites to which cutaneous emulsion had been applied, 10 minutes after iontophoresis or 10 minutes after the last application of the cutaneous emulsion. As batches of histamine vary in their biological potency, two concentrations of histamine were used so that the most appropriate could be selected.
Scanning Laser Doppler Imaging
Wheal and flare areas and blood flux, indicative of dermal perfusion up to a depth of 1 mm, were assessed in the skin using scanning laser Doppler imaging (Moor LDI, Moor Instruments Ltd, Axminster, Devon, UK). Ten minutes after histamine injection, an area of skin of 5 cm square was scanned giving ~16,000 data points for analysis. Changes in weal and flare areas may be measured to an accuracy of ± 0.05 cm2 and changes in perfusion to ± 5%.
While scanning laser Doppler imaging is ideal for assessing areas of flare and blood flow, it is less accurate in determining the area of a wheal. Consequently, planimetry, in which wheal areas were calculated from traces of their outlines on acetate sheet, was used to confirm wheal size.
Severity of itch
Immediately following histamine injection, subjects recorded the severity of itch on a 100 mm Visual Analogue Scale (VAS) every 20 seconds for 5 minutes.
On the first experimental day, subjects were asked to lie down for 10 minutes prior to the commencement of the study. The volar aspect of both forearms was wiped clean with antiseptic wipes (Sterets) and left to dry. Four sites, two on the volar aspect of each forearm, one for 1 μM histamine and the other for 300 nM histamine, were selected for the study. The VAS score for itch for 1 μM histamine, 33.97 ± 5.35 mm, was greater than that for 300 nM, 28.07 ± 4.25 mm. As a consequence, only the results from the 1 μM histamine sites were used for further calculations.
Three studies were conducted
Study 1: Iontophoresis study
Baseline blood flux was measured by scanning laser Doppler imaging immediately before introduction of 4% sodium cromoglicate in aqueous solution or reversed osmosis purified water (ROW, control), were introduced into the skin using iontophoresis. The order and sites of iontophoresis of sodium cromoglicate and water were randomised with the restriction that the treatments were equally distributed between sites on the upper and lower forearm. Five minutes after iontophoresis, blood flux was reassessed using scanning laser Doppler imaging. Ten minutes after iontophoresis, 20 μl of 1 μM of histamine was injected intradermally and the subjects began to record the severity of itch on a VAS. Ten minutes after histamine injection, the area of the weal was measured using planimetry and the area and intensity of the flare measured using scanning laser Doppler imaging.
The techniques used for this study were similar to those described in the original study with nedocromil sodium .
Study 2: Effect of 4% SCG cutaneous emulsion
Subjects were given 2 bottles of emulsion labelled L1 and R1. Emulsion L1 was used on the left forearm and emulsion R1 on the right. The containers contained either 4% sodium cromoglicate emulsion or the matching vehicle. Subjects were instructed to massage 1 ml of the emulsion into two areas of each forearm, each approximately 10 cm sq, delineated with a marker pen, four times a day for 3 days prior to the next experimental visit. On the experimental day, the last application of the emulsion was made 10 minutes before the injection of 20 μl of 1 μM of histamine. Recordings of itch, wheal and flare made as in study 1. All interventions were administered double-blind and randomised.
Study 3: Comparison of 1%, 2% and 4% SCG cutaneous emulsions
Subjects were supplied with four bottles of emulsions labelled A, B C and D. Subjects were instructed to massage 1 ml of emulsion A into an area of approximately of 10 cm2, delineated with a marker pen, on the upper part of the volar surface of the left forearm four times a day for 3 days prior to the next experimental day. In a similar manner, emulsion B should be applied to the lower part of the left forearm, emulsion C to the upper part of the right forearm and emulsion D to the lower part of the right forearm. On the experimental day, the procedures followed were the same as in Study 2 with the last application of the emulsions being made 10 minutes before the injection of 20 μl of 1 μM of histamine. The containers contained 1%, 2% or 4% SCG emulsion or the matching vehicle. Treatments were randomised and administered double-blind.
All data are shown descriptively as means ± SEM for each treatment. For VAS scores for itch, for each subject the mean of all available values over the 5-minute period was used. A two-sided P-value of 0.05 was taken to indicate statistical significance in any statistical testing.
For study 1, a generalised linear model was fitted to the data incorporating atopic status, subject (within atopic status), treatment and treatment x atopic status interaction. If the interaction was not statistically significant then this was dropped from the model. The least square means for sodium cromoglicate and water are presented together with the difference and 95% confidence interval for the difference.
For study 2, the same method and model was used and the contrast SCG 4% emulsion and vehicle presented.
For study 3, a generalised linear model was fitted to the data incorporating subject and treatment. If the overall treatment effect was statistically significant then pair-wise comparisons of each dose versus vehicle were carried out. In addition, since the treatment groups consisted of increasing concentrations of SCG, concentration (0, 1, 2, 4) was included in the model as a covariate (with the subject factor only) in order to provide a test of linear trend.
Since both studies 2 and 3 contained within subject assessments of 4% concentration of SCG emulsion and vehicle, a combined analysis using only atopic subjects was carried out. A generalised linear model was fitted to the data incorporating study, subject (within study), treatment and treatment x study interaction. If the interaction was not statistically significant then this was dropped from the model. The least square means for SCG emulsion and vehicle are presented together with the difference and 95% confidence interval for the difference.
Checks that the underlying data were normally distributed were undertaken. The formal Shapiro-Wilk's test was satisfactory in all 8 analyses of both itch and flare. Similarly, the distribution of residuals and normal probability plots gave no concern and hence we are confident that the parametric models used for the analyses are appropriate.