House dust mite reduction in the management of atopic dermatitis. A critically appraised topic

Does house dust mite (HDM) reduction improve the long‐term disease severity in patients with atopic dermatitis (AD)? A 20‐year‐old patient with known AD since childhood was treated with topicals in an outpatient clinic of dermatology. She experienced an insufficient improvement of her disease activity and she disliked the daily use of greasy topicals. She heard about HDM reduction in patients with asthma and wondered if this might be helpful in the treatment of her AD. AD is an itchy, inflammatory, chronic relapsing eczematous skin disease and it is among the most common dermatological conditions. The prevalence of AD has increased steadily in recent decades and now varies between 1% and 3% in adults and up to 20% in children.1 Topical treatment with corticosteroids and calcineurin inhibitors, such as tacrolimus and pimecrolimus, is often used in combination with emollients to treat the signs and symptoms of AD.2 Unfortunately, topical therapies are sometimes insufficient and therefore other therapies like ultraviolet therapy or systemics are needed. Besides a genetic predisposition, environmental factors such as exposure to HDM might play a role in the course of AD.3 The allergens of HDM (Euroglyphus maynei) are thought to play a role in the sensitization and induction of clinical symptoms of AD.4, 5 The mites that most frequently induce AD are Dermatophagoides pteronyssinus and Dermatophagoides farinae.6 Reduction of these mites, by covering mattresses and pillows, frequent or high‐filtration vacuum cleaning or the use of acaricidal sprays, has been shown to reduce the severity of asthma.7 As this disease is closely related to AD, HDM reduction may play a role in the treatment of established AD.8 A systematic search was performed using PubMed, Embase and the Cochrane Central Register of Controlled Trials (CENTRAL) (March 2012). All articles were screened for relevance regarding our clinical question. Included were studies that have a follow‐up period of at least 6 months and studies that evaluated the effect of HDM reduction on the clinical symptoms of AD, rather than the effect on serological results. A complementary manual search of citations and related articles was used to find missed relevant publications (Fig. 1). The articles were appraised on validity according to predetermined criteria (Table 1). Discordant judgements between the first two authors were resolved by discussion. Particular attention was paid to comparability of baseline characteristics and the description of how HDM levels were reduced. Overall validity was categorized as high, intermediate or low. Studies scoring high on validity were included for further analysis. Critical appraisal for validity HDM, house dust mite; NA, not applicable; RCT, randomized controlled trial; L, low; I, intermediate; H, high. a+, clear and adequate description; −, no clear and adequate description. b+, randomized; +/−, not randomized, equal baseline characteristics or adjustment; −, not randomized, unequal baseline characteristics, not mentioned. c+, standardized measurements and blinding; +/−, standardized measurements, no blinding; −, no standardized measurements. d+, loss to follow‐up < 10%; +/−, loss to follow‐up 10–20%; −, loss to follow‐up > 20%, not mentioned. e+, done; −, not done or not mentioned. f+, all prespecified outcomes have been addressed and reported; −, differences between prespecified methods and final outcomes. g+, standardized therapy and proven HDM reduction; +/−, standardized therapy, but no proven HDM reduction; −, no standardized therapy. Critical appraisal for validity HDM, house dust mite; NA, not applicable; RCT, randomized controlled trial; L, low; I, intermediate; H, high. a+, clear and adequate description; −, no clear and adequate description. b+, randomized; +/−, not randomized, equal baseline characteristics or adjustment; −, not randomized, unequal baseline characteristics, not mentioned. c+, standardized measurements and blinding; +/−, standardized measurements, no blinding; −, no standardized measurements. d+, loss to follow‐up < 10%; +/−, loss to follow‐up 10–20%; −, loss to follow‐up > 20%, not mentioned. e+, done; −, not done or not mentioned. f+, all prespecified outcomes have been addressed and reported; −, differences between prespecified methods and final outcomes. g+, standardized therapy and proven HDM reduction; +/−, standardized therapy, but no proven HDM reduction; −, no standardized therapy. Titles and abstracts of 1079 articles were screened (Fig. 1). After screening the full text of 48 studies, seven were selected for critical appraisal.4, 5, 9–13 No new relevant articles were found by the manual search of related articles and citations. Full‐text copies were obtained of all (potentially) relevant studies. Excluded were studies not in English, German, Dutch or French, studies that were not available as full text, studies with a follow‐up period < 6 months and studies without clinical endpoints. After critical appraisal, the studies of Gutgesell et al.4 (2001), Holm et al.10 (2001), Oosting et al.5 (2002) and Tan et al.13 (1996) were included (Table 2). Although the study of Holm et al. was not randomized, the study was well performed and therefore scored high on validity. None of the articles mentioned an intention‐to‐treat analysis. In the trial of Gutgesell et al.,4 all patients completed the study and therefore we assume that all patients were analysed in the group in which they started. The drop‐outs in the studies of Tan et al.13 and Oosting et al.5 were equally distributed between the active treatment group and the placebo group. Loss to follow‐up seems to be not selective. The results summarized from Gutgesell et al., Holm et al., Oosting et al. and Tan et al. for house dust mite reduction (HDMR) vs. no HDMR AIPE, allergen‐impermeable polyurethane encasings; AS, acaricidal spray; CBE, coated bedding encasement; CE, cotton encasements; CI, confidence interval; CR, carpet removal; GBE, Goretex bedding encasement; HFVC, high‐filtration vacuum cleaning; NR, not reported; NS, not significant (no P‐value reported); RVC, regular vacuum cleaning; SCORAD, Scoring Atopic Dermatitis; SVC, standard vacuum cleaning; WS, water placebo spray. aResults after n months. bSCORAD severity score. cDecrease of median; value derived from small figure; no exact data reported. dComparing the active treatment group and the placebo group after 12 months. eLeicester Sign Score. fComparing the active treatment group and the placebo group after 12 months by analysis of covariance with initial eczema severity score and initial concentration of dust and HDM. The results summarized from Gutgesell et al., Holm et al., Oosting et al. and Tan et al. for house dust mite reduction (HDMR) vs. no HDMR AIPE, allergen‐impermeable polyurethane encasings; AS, acaricidal spray; CBE, coated bedding encasement; CE, cotton encasements; CI, confidence interval; CR, carpet removal; GBE, Goretex bedding encasement; HFVC, high‐filtration vacuum cleaning; NR, not reported; NS, not significant (no P‐value reported); RVC, regular vacuum cleaning; SCORAD, Scoring Atopic Dermatitis; SVC, standard vacuum cleaning; WS, water placebo spray. aResults after n months. bSCORAD severity score. cDecrease of median; value derived from small figure; no exact data reported. dComparing the active treatment group and the placebo group after 12 months. eLeicester Sign Score. fComparing the active treatment group and the placebo group after 12 months by analysis of covariance with initial eczema severity score and initial concentration of dust and HDM. Dust samples from mattresses were collected before and after treatment and a significant reduction in dermatophagoid loads after active treatment was reported in all studies. Gutgesell et al.4 randomized 20 patients to a 1‐year period of either placebo (cotton encasements and water spray containing traces of ethanol) or active treatment (allergen‐impermeable encasings and an acaricide spray). An improvement of disease after treatment was reported in both groups compared with baseline, but there was not a statistically significant difference in the final Scoring AD (SCORAD) score14 (maximum 103 points) between the active treatment group and the placebo group (P =0·901). Holm et al.10 performed a double‐blind, placebo‐controlled trial of 40 patients, divided into two groups that were matched according to sex, eczema severity and sensitization to HDM. In the active treatment group (n =22), water vapour‐permeable cotton encasings were put on the mattresses and pillows of the patient’s and their partner’s bed. Bedroom carpets were removed or vacuumed monthly. In the placebo group (n =18), patients did not take other avoidance measures, except using placebo cotton encasements. The disease severity (according to the SCORAD index) after 12 months was significantly reduced in the active treatment group (from 48 to 26, P <0·001), but also in the control group (from 41 to 27, P <0·001). No between groups comparisons were reported. The randomized, double‐blind trial of Oosting et al. showed that Goretex bedding encasements significantly decreased disease severity, defined by the Leicester Sign Score (LSS, range 0–108 points),15 after 12 months compared with baseline (P =0·017), but found no statistically significant change in final LSS severity or extent between the active treatment group (n =38) and the placebo group (n =35) with cotton encasements only (4 points decrease in the treatment group vs. 2 points decrease in the placebo group).5,15 In the randomized study of Tan et al.13 the active treatment consisted of a Goretex bedding system, benzyl tannate complex spray for carpets and a high‐filtration vacuum cleaner (n =28). The placebo group received light cotton bedcovers, water with a trace of alcohol to spray on carpets and a standard upright vacuum cleaner with a poor filtration performance (n =20). The final LSS score was determined after 6 months. The mean fall in LSS severity score was 12·6 points in the active treatment group vs. 4·2 points in the placebo group. The reduction of LSS, adjusted for initial severity, and HDM and dust loads was significantly greater in the active treatment than in the placebo group (mean difference 4·2 units, 95% confidence interval 1·7–6·7; P =0·008). The mean final area affected was 10% greater in the placebo group (P =0·006). In this critically appraised topic we have to conclude that only one study reported a statistically significant reduction of the clinical symptoms of AD after HDM reduction, compared with placebo. Besides, it should be noted that the reported reductions in disease severity are small and that it is not clear what the clinical value of these minimal reductions will be. The minimal clinically important difference (MCID) of the SCORAD was calculated to be 8·7 by Schram et al.16 Both active treatment and placebo group in the study of Holm et al.10 exceeded the MCID (reduction of 22 points vs. 14 points). The SCORAD decrease reported by Gutgesell et al.4 was ∼9, but conclusions must be drawn cautiously because this value had to be derived from a small figure. The LSS score was investigated by Schmitt et al. and the sensitivity to change (the possibility to detect clinically relevant changes using this measurement) appeared to be acceptable, but the clinical value of the small reduction and the slightly larger reduction in LSS reported by Oosting et al. and Tan et al., respectively, remains unclear.17 Moreover, it must be noted that a reduction of dust loads and the severity of AD were also described in all placebo groups. This may be due to more emphasis on cleaning or due to the fact that placebo covers had an unexpected allergen‐reducing capacity and patients in the placebo groups possibly did not use covers before participating in the study. Furthermore, the four included studies used different ways of reducing HDM and allowed different additional therapies during study treatment, which might influence the results. Moreover, the complex aetiology of AD with both genetic and environmental influences and the fact that HDM is only one of the known flare factors of AD, could be other reasons for the limited success of HDM reduction.8, 18 Finally, a drawback of the study of Tan et al. was that they reported an improvement of AD after only 6 months. It should be noted that this improvement could have been due to seasonal effects.5,13 A limitation of the study of Holm et al. was that, although this study was stated double‐blinded, patients’ blinding was not easy to perform, because carpets were removed in the active treatment group only. A limitation of this critically appraised topic is that pooling of results was not possible, as important differences between the four studies exist. Because of the way data were shown, it was not possible to derive confidence intervals or numbers needed to treat. Some studies have investigated the effect of HDM reduction in preventing AD. Results of these studies are inconclusive and interestingly there is some evidence to suggest that high levels of HDM exposure in early life are associated with a reduction in AD risk.19 We did not include these studies for this critically appraised topic, because we were looking for the effect of HDM reduction in the management of established AD. The strength of this critically appraised topic is our extensive search. We critically appraised seven articles by predefined criteria and selected the four articles that were most relevant and valid according to our clinical question. Because AD is such a prevalent disease, the results of this critically appraised topic might be useful for many dermatologists. Considering the current evidence discussed in this critically appraised topic, we conclude that there is insufficient evidence to support HDM reduction in the management of AD in our patient. Of the four relevant but heterogeneous studies, three studies showed no significant differences in clinical improvement between active HDM reduction and the control groups. Furthermore, the clinical improvement in all groups seems to be small. Better and more homogeneous studies with standardized ways to reduce HDM, standardized and validated outcome measurements and equal study durations with adequate follow‐up periods, should be performed in future. A systematic Cochrane review, including all studies on this subject instead of only studies that are relevant for this clinical case and without any language restrictions, could strengthen and complement our conclusion and is currently being carried out.20 Funding sources None. Conflicts of interest None declared.