University of Groningen Hand Eczema Christoffers, Wianda

University of Groningen

Hand eczema Christoffers, Wianda

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record

Publication date: 2014

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA): Christoffers, W. (2014). Hand eczema: interventions & contact allergies. [S.n.].

Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment.

Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

Download date: 10-10-2021 Hand eczema interventions and contact allergies

Wietske Andrea Christoffers

ISBN: 978-90-367-7355-3 (printed version) ISBN: 978-90-367-7354-6 (e-version)

Financial support for the publication of this thesis was provided by:

AbbVie, Afdeling Dermatologie UMCG, ALK-Abelló BV, Almirall BV, BAP Medical BV, Combicare, Eucerin, Fagron BV, Galderma, GlaxoSmithKlineBV, La Roche-Posay, LEO Pharma, Novartis Pharma BV, Stichting Milieu en Arbeidsdermatologie, Studiefonds Dermatologie, Van Der Bend Medical Supplies, Waldmann BV, Will Pharma

Cover design: Id Graficus, Assen, www.idgraficus.com, The Netherlands Layout: Id Graficus, Assen, www.idgraficus.com, The Netherlands Printing: GVO drukkers & vormgevers B.V., Ede, The Netherlands

Hand eczema interventions & contact allergies

Proefschrift

ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen op gezag van de rector magnificus prof. dr. E. Sterken en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

maandag 24 november 2014 om 12:45 uur

door

Wietske Andrea Christoffers

geboren op 29 april 1987 te Assen Promotor Prof. dr. P.J. Coenraads

Copromotor Dr. M.L.A. Schuttelaar

Beoordelingscommissie Prof. dr. T. Agner Prof. dr. T. Rustemeyer Prof. dr. W.J.C. Uter Paranymfen Jorinde Dries-Talman Tatiana Alexandra Vogel

Contents

Chapter 1 Introduction 8

Chapter 2 Cochrane review: Interventions for hand eczema (Short version) 30

Chapter 3 Evidence based dermatology – Hand eczema 104

Chapter 4 Drug survival of cyclosporine in the treatment of hand eczema: 134 a multicenter, daily use study

Chapter 5 Patch test results of hand eczema patients: relation to clinical types 148

Chapter 6 Severe bullous allergic contact dermatitis caused by glycidyl 166 methacrylate and other acrylates

Chapter 7 Two decades of occupational (meth)acrylate patch test results and 174 focus on isobornyl acrylate

Chapter 8 Co-sensitization to ascaridole and tea tree oil 190

Chapter 9 The optimal patch test concentration for ascaridole as a sensitizing 198 component of tea tree oil

Chapter 10 Discussion How to treat hand eczema in daily practice? 216

Summary 240

Nederlandse samenvatting 246

Dankwoord 254

Bibliografie 260

Curriculum Vitae 264

7 Chapter

8 chapter 1

Introduction Hand eczema: interventions & contact allergies

Wietske Andrea Christoffers

Department of Dermatology, University of Groningen, University Medical Center Groningen, Groningen, Netherlands

9 chapter A few million years ago, our ancestors decided to start walking on two legs, which made it easier for them to collect food from high branches and to use tools. Moreover, humans 1 standing upright appeared longer and more intimidating. Over the years, the proximal extremities developed for other functions, making it easier to conduct more precise tasks. Nowadays hands are an important part of the human body and play a prominent role in social interaction: they enable us able to greet each other, shake hands, and hug or caress each other. Hands appear even in proverbs: “Many hands make light work”, “One hand washes the other” or “The hand that rocks the cradle rules the world”. Hands are indispensable to our daily lives. Therefore, diseases affecting the hands, such as hand eczema, can have a tremendous impact on daily activities and the quality of life. Hand eczema, also called hand dermatitis, is an inflammatory skin reaction confined to the hands. It often manifests as a chronic and relapsing condition, involving pruritus and painful fissures. Although many patients suspect that allergies play a dominant role in their hand eczema, other factors like exposure to irritants such as water, detergents or oils are just as important. This thesis discusses the treatment options for patients with hand eczema; it features a Cochrane review regarding interventions for hand eczema and a practical therapeutic plan. We also discuss the relationship between hand eczema and allergic contact dermatitis. Finally, it highlights two relatively unknown contact allergens which can play a role in hand eczema: ascaridole and isobornyl acrylate.

Epidemiology Hand eczema is a common condition. The 1-year prevalence in the general Swedish population was estimated at 4%, though when mild cases were also included the 1-year prevalence was as high as 10%.1 The incidence was around 5.5 per 1,000 persons. Approximately one third of patients with hand eczema developed symptoms before the age of 20.1 A Swedish population-based study gave a self-reported 1-year prevalence of hand eczema of 8% (females 10%, males 6%).2 The highest prevalence was reported among females aged between 19 and 40 years; this seemed to be related to an increased frequency of hand washing, nickel contact allergy and childhood atopic dermatitis.1 In a cross-sectional study among 1,501 Danish school children (aged 12-16 years), the self-reported life time prevalence for hand eczema was 9.2% and the 1-year prevalence was 7.3%, with a preponderance in girls.3 Eczema was defined as itching, erythema, vesicles and/or papules and scaling localized to the fingers or finger webs, backs of hands or palms, and with a duration of at least two days. The clinically evaluated point-prevalence in these school children was 3.2%, with a borderline significant sex difference.1 The majority of epidemiology studies of hand eczema have been conducted in Scandina- vian countries and most are questionnaire based. Due to underestimation of mild cases or

10 recall bias, questionnaires might underestimate the true prevalence. However, investigating chapter large cohorts of “healthy” people for the diagnosis of hand eczema is expensive and very time- and labor consuming. Moreover, a recent study of Danish hairdressers’ apprentices 1 concluded that self-reporting of hand eczema is a valid method for estimating its prevalence with good sensitivity and very high specificity.4 Meding et al. validated the self-reported 1-year prevalence of hand eczema in car mechanics, dentists and office workers, based on questionnaires, interviews and clinical examinations.5 Meding also found a high specificity (96–99%), but a lower sensitivity (53–59%) for the use of a self-reporting questionnaire. Occupational hand eczema is one of the most frequently recognized occupational skin diseases.6 The incidence of notified hand eczema in occupational related cases (and thus probably more severe cases) is above 0.5-1.9 per 1,000 workers a year.7 In a large database of 1,504 Danish cases of occupational contact dermatitis, irritant contact dermatitis accounted for 70% of all cases; 68% of these were caused by wet work.8 Occupations that involve exposure to water, irritants or trauma are especially at risk for developing hand dermatitis. Cleaners, bakers, hairdressers and caterers are exposed to a lot of wet-work and are therefore prone to develop hand eczema, but one of the most investigated occupational risk groups are health care workers. In the Netherlands the self-reported 1-year prevalence of hand eczema among 1,232 health care workers was 12% (95% confidence interval 11–14) with a point-prevalence of 4.9% (95% confidence interval 3.8–6.3).9 Almost half of all questioned health professionals reported at least one symptom of hand eczema over the last three months. In this study population 1.7% of workers with hand eczema reported sick because of hand eczema in the past three months. These results were reproduced in 525 Taiwanese nurses: 75.6% of them reported symptoms of hand eczema, and 31.0% suffered from self-reported hand eczema.10 Young age, history of atopic dermatitis, frequent hand washing (>20 times/day) and wearing gloves longer than 5 minutes a day were risk factors for hand eczema, while frequent use of hand moisturizer (>3-4 times/day) had a protective effect. Other employees at risk for developing hand eczema are those at risk of occupational sensitization because they work with well known allergens. Examples of these are hairdressers11 with contact allergies for p-phenylenediamine (PPD), p-toluenediamine (PTD), and ammonium persulfate;12 dental workers for (meth)acrylates13 or line assembly workers for different allergens in glues or coatings.8,14

Clinical presentation of hand eczema and diagnostic strategy Hand eczema or hand dermatitis is primary an inflammatory skin reaction of the hands, although the feet and other body parts may also be affected. Clinically, the condition is characterized by erythema, vesicles, papules, edema, scaling, fissures, erosions and/or hyperkeratosis, signs which may appear at different times. Acute hand eczema often presents with erythema and vesicles, while in the chronic variants scaling, fissures and

11 chapter hyperkeratosis are more apparent. Chronic hand eczema is defined as eczema of the hands which lasts for more than three months, or relapses at least twice within 12 months despite 1 adequate treatment.15 Itch is a common feature and can be very severe, resulting in loss of sleep. Fissures and vesicles can be painful and cause bleeding of the skin. Eczema can also become superimposed with a bacterial infection, resulting in oozing and crusting and, although less likely, with viral infections such as herpes. A thorough history of the patient’s hand eczema is crucial. The history should focus on exposure to potential irritants and allergens at home, during hobbies and in occupational settings. Frequent exposure to water, detergents, frequent hand washing and the use of occlusive gloves are well known risk factors. The course of the disease should be discussed, including the age of onset and aggravating factors. Improvement of the eczema during holidays or sick leave might point to an occupational dermatitis. A history of atopic dermatitis predisposes for hand eczema.16-18 One should discuss a personal and family history of atopic dermatitis, rhino conjunctivitis or asthma.19 Furthermore, the family history should take into account alternative diagnoses such as psoriasis or hereditary palmoplantar keratodermas. Dermatological examination of the hands is the main pillar in the diagnostic procedures. One should investigate the different morphological features, the extent of the hand eczema and the localization on the hand. In addition, one should inspect other body sites like the feet to confirm or exclude other diagnoses such as psoriasis or dermatomycosis. After history taking and dermatological examination, further diagnostics can be issued. Patch tests can be performed to rule out the possibility of an allergic contact dermatitis. This will be discussed in a separate section. Prick-testing and a serum radioallergosorbent (RAST) test to determine an IgE-mediated type I allergy is not routinely recommended, except in the case of protein contact dermatitis and recurrent contact urticaria.20 Neither is the determination of allergen-specific IgE levels in the routine work-up of a patient with hand eczema, though it can help to establish the atopy status.15 Laboratory evaluations are required only when systemic therapy is considered. Histopathological examination of hand eczema lesions is rarely necessary in daily practice, since the diagnosis is based on clinical findings.21 However, when the hand eczema is refractory to therapy or when other diagnoses need to be excluded one can consider a biopsy.15 Different stages and morphologies of hand eczema result in different histopathological features. In general, histological biopsies of hand eczema demonstrate dermatitis with spongiosis and varying degrees of acanthosis.22 Superficial perivascular infiltrate of lymphocytes and histiocytes can be seen.23,24 Acute vesicular hand eczema is characterized by intra-epidermal spongiotic vesicles or bullae that do not involve the intra-epidermal portion of the eccrine sweat duct (acrosyringium).23-25 Other changes include a sparse, superficial perivascular infiltrate of lymphocytes with some exocytosis. To exclude a derma-

12 tomycosis one can perform a periodic acid–Schiff (PAS) staining, particularly if neutrophils chapter are present within the vesicles or stratum corneum. Hyperkeratotic hand eczema consists of chronic spongiotic dermatitis with marked hyperkeratosis, compact orthokeratosis and 1 possible small foci of parakeratosis. Psoriasiform hyperplasia of the epidermis can be preent, although the elongation of the rete ridges is usually less regular than in psoriasis. In addition, the amount of spongiosis and the absence of psoriasis spongiform pustules or microabscesses and neutrophils in the epidermis help to distinguish hyperkeratotic hand eczema from psoriasis.22,24

Differential diagnosis The clinician should consider alternative diagnoses before diagnosing hand eczema. Hand eczema is most frequently confused with psoriasis, dermatomycosis, scabies and lichen planus. Classic psoriasis consists of erythematosquameous plaques, which are sharply demarcated and especially affect the extensor sites of elbows, knees and scalp, though the palms and knuckles of the hands may be affected as well. Vesicles are absent in psoriasis. A positive family history of psoriasis and the presence of arthritis may help in making the diagnosis. Moreover, the presence of nail pits, without involvement of the nail fold and without itching, are suggestive for psoriasis. However, the distinction between hyperkeratotic palmar eczema and psoriasis can be difficult, and definite cases of hand eczema may over time be labeled as psoriasis and vice versa. Dermatomycosis (fungal infection, tinea) is a scaling skin condition which mimics hand eczema. Fungal infections should be ruled out, especially when one hand is more prominently affected than the other or when the lesion is more active in the borders. Moreover, the feet, especially the interdigital spaces, or the groin can be affected. Diagnosis in these cases is confirmed by fungal examination of skin surface scrapings. Scabies is a skin condition caused by mites; it usually presents with small erythematous papules, associated with variable numbers of excoriations; vesicles, indurated nodules, eczematous dermatitis and secondary bacterial infection may also be present. The pathog- nomonic sign is a burrow, representing the 1-10 mm long tunnel that a female mite excavates to lay eggs. Usually the hands are affected, and often the web-spaces and volar wrist sites, as well as the trunk and limbs. Multiple family members might be affected and (nocturnal) itch can be a prominent future. Lichen planus is an idiopathic, inflammatory disease which affects the skin and mucosa. Classical lichen planus is characterized by pruritic, sharply demarcated violaceous papules, which mainly affect the extremities. These polygonal papules may be flat and slightly shiny or transparent, with a network of white lines (Wickham’s striae). The papules can develop into plaques and hyperkeratotic lichen rubber planus can resemble hyperkeratotic hand

13 chapter eczema; well demarcated margins and typical lichen planus lesions may also be present. Examples of other conditions which might resemble hand eczema are granuloma annulare, 1 keratolysis exfoliativa (dyshidrosis lamellosa sicca), adverse drug reactions, keratoderma palmare et plantare or porphyria cutanea tarda, though this list is by far not extensive.26,27

Etiology of hand eczema The etiology of hand eczema is often unclear. Multiple risk factors may play a role. The most important is probably onset early in life, even if only a single episode. In one third of the patients the disease starts before the age of 20.1 It affects more women than men and its highest incidence is among young women. Another important risk factor is atopic dermatitis, especially a history of childhood atopic dermatitis. In a population-based study, childhood atopic dermatitis was found to be more important than female gender or occupational exposure.28 Between one third and half of hand eczema patients are atopic, which is defined by the World Allergy Organization as: “a personal and/or familial tendency, usually in childhood or adolescence, to become sensitized and produce IgE antibodies in response to ordinary exposures to allergens, usually proteins.”29 Subjects can develop typical symptoms of asthma, rhino conjunctivitis, or atopic dermatitis. Atopic dermatitis is associated with a worse prognosis for hand eczema; patients with atopic dermatitis, hand eczema and filaggrin null mutations had an earlier onset of the disease and a more persistent form in adulthood.30 Filaggrins are a key component of the epidermal differentiation complex of the stratum corneum. Loss-of-function mutations in the filaggrin gene cause an impaired skin barrier function, which can result in ichthyosis vulgaris and an increased risk of atopic dermatitis. Thyssen et al. demonstrated that filaggrin loss-of-function mutations not only increase the risk of atopic dermatitis and dry skin, but also the risk of fissured hand eczema in subjects without atopic dermatitis.31 Other studies concluded that atopic dermatitis was a major confounder in cases of more severe hand eczema accompanied by mutations within the filaggrin gene,32 but the debate on the role of filaggrin mutations in hand eczema is ongoing. Allergic contact dermatitis seemed to be associated with filaggrin null mutations in the presence of atopic dermatitis.33 Despite the important role of genetic factors in hand eczema, external factors such as exposure to wet work are often involved as well. Lerbaek et al. conducted an epidemiological twin study on hand eczema and concluded that, after correcting for atopic dermatitis, 59% of hand eczema could be attributed to external factors.34 The most common external cause of hand eczema is probably exposure to irritants or toxic agents, such as water and soap, which result in irritant contact dermatitis. Repeated and prolonged exposure to irritants impairs the skin barrier function and results in inflammation. Patients often present with a history of exposure to wet work, contact with soaps, solvents, or prolonged use of occlusive gloves. Patch tests are negative or demonstrate irritant reac-

14 tions, but a simple diagnostic procedure to establish irritant contact dermatitis in daily clinic chapter does not exist. 1 Although allergic contact dermatitis is probably less common than irritant contact dermatitis patients often assume that contact allergy is the cause of their hand eczema. Allergic contact dermatitis (ACD) is a delayed-type hypersensitive immune response, also known as Gell and Coombs type IV hypersensitivity. The adaptive T-cell-mediated immune response is one of the key players in allergic contact dermatitis, though the innate immunity also plays a roll. The immune response can be divided into two distinct phases: the sensitization (induction) phase and the elicitation (effector) phase. The sensitization phase is initiated by contact allergens. Contact allergens (also known as haptens) such as nickel, are natural or synthetic low-molecular-weight chemicals, which are usually lipophilic. Because of these qualities, haptens can easily penetrate the stratum corneum and diffuse into the skin. In the epidermis, haptens react with various intra- or extracellular endogenous proteins such as major histocompatibility complex proteins (MHC), resulting in a hapten-carrier complex. This complex activates Langerhans cells and other dendritic cells in the epidermis. The activated Langerhans cells then migrate in less than 24 hours from the epidermis through the afferent lymphatic vessels to the regional lymph nodes. In the paracortical areas of the draining lymph node the Langerhans cell presents the hapten to naive T-cells. Subsequently, the naive T-cells with matching T-cell receptors are activated by the presence of a specific hapten and start to proliferate. These activated T-cells start to produce IL-2 and other interleukins. After this, specific memory T-cells are formed, which target only the specific hapten. The activated T-cells migrate through the blood and start their surveillance. This sensitization phase requires from three days to several weeks; during this phase patients are usually asymptomatic. The elicitation phase (effector) follows, wherein the allergic contact dermatitis is now established and on each subsequent contact with the hapten, however small, the patient can develop a rash. After hapten penetration of the skin, a hapten-carrier complex comes in contact with memory T-cells in the epidermis, whereupon the T-cell produces pro-inflammatory cytokines. Because of the release of cytokines, more T-cells start to migrate to the contact site and further stimulate the inflammatory response; the result is an eczematous reaction. The skin rash continues to develop and reaches its maximum after 18-48 hours.35-38 After elimination of the hapten the rash gradually disappears within a few days. However, since hand eczema is rarely caused by a single hapten, elimination of one hapten does not necessarily cure the hand eczema. Allergic contact dermatitis is diagnosed by means of an epicutanous patch test. A fixed amount of contact allergen is placed in a patch test chamber and applied to the subject’s upper back for 48 hours. This activates the elicitation phase; if a subject is sensitized to the specific hapten, a skin reaction can be expected. The results of the patch test are read 72

15 chapter hours after application. An additional reading on day 7 can preempt late reactions. A combination of erythema and infiltration covering the whole patch test area is considered 1 as a positive patch test reaction. The strength of the patch test reaction is scored according to the guidelines of the International Contact Dermatitis Research Group (ICDRG). The reactions are quantified as negative (-), doubtful (?+) , weak positive (+), strong positive (++), extremely positive (+++) or irritant (IR).

Symbol Morphology ICDRG score - No reaction Negative ?+ Erythema only, no infiltration, or erythema and infiltration not covering the Doubtful reaction whole area + Erythema, infiltration, possibly discrete papules Weak positive reaction ++ Erythema, infiltration, papules, vesicles Strong positive reaction +++ Erythema, infiltration, confluent vesicles Extremely positive reaction IR Different types of reactions (silky skin/ cigarette paper skin, blisters, ero- Irritant reaction sions, pustels, necrosis)

Table 1: Scoring of patch test reactions according to ICDRG recommendations

Although patch testing is the gold standard for diagnosing sensitization, its sensitivity and specificity are not optimal.35 Moreover, its interobserver variation is considerable, even among experts.39 For these reasons a patch test calibration protocol was developed. 40,41 This protocol contains a scoring list with standardized criteria. Calibration of the patch test grading system improved the quality of patch testing and thus its diagnostic accuracy. However, since a positive patch test reaction is in itself not necessarily relevant for the current complaints of the patient a repeated open application test (ROAT) with a suspected product can further establish the clinical relevance of the patch test reaction. During a ROAT the product is repeatedly applied to the volar side of the forearm. If a rash develops within two weeks, the ROAT is positive.42,43 Protein contact dermatitis is a rare a subtype of allergic contact dermatitis, which can affect workers in the food industry. The initial reaction to proteins is urticarial (contact urticaria), but over time eczema may develop. In this specific subtype of hand eczema determination of IgE levels of specific proteins is of additional value. Therefore prick tests or serum levels should be evaluated.

Classification of hand eczema The classification of hand eczema is complex. Some physicians even wonder if the term hand eczema is not used for different entities. An accurate diagnosis of hand eczema is,

16 however, crucial for proper registration of the disease and its clinical course. A uniform chapter disease classification is also essential for conducting clinical trials. Although the debate on the classification of hand eczema is ongoing, it can be classified according to its etiology or 1 morphology or a combination of both. Etiological classifications include: • Irritant contact dermatitis • Allergic contact dermatitis • Atopic hand eczema • Hybrid hand eczema • Protein contact dermatitis

Recently, the Danish Contact Dermatitis Group developed another classification for hand eczema.20 They applied to every patient one clinical diagnosis and one or more etiological diagnoses. The following clinical subtypes were distinguished, which included morphological features: • Chronic ﬁssured hand eczema: dry eczema, usually with scaling and possibly with hyperkeratotic areas and fissures, with a limited number of vesicles on the sides and palmar aspects of the fingers or on the palmar aspects of the hands. This morphology is typically seen in hand eczema that lasts from months to years (Fig. 1a). • Recurrent vesicular hand eczema: recurrent eruptions of vesicles on the palms and/ or on the sides of the fingers, and possibly also on the palmar aspects of the fingers and around the fingernails. Eruptions may occur at intervals of weeks or months or be ongoing, mimicking a chronic eczema. The plantar aspects of the feet may be involved. The history of the patient provides information about the eruptive nature of this type of hand eczema (Fig. 1b). The recurrent vesicular hand eczema subtype is also called ‘dyshidrotic’ hand eczema (although not related to sweat glands) and ‘pompholyx’. • Hyperkeratotic palmar eczema: well demarcated hyperkeratosis on the palms, possibly extending to the palmar aspects of the fingers. There may be fissures. This type of eczema is distinguished from psoriasis by not being inflammatory and by having no psoriasiform scaling. It does not evolve into psoriasis. There are no accompanying nail changes. There are no vesicles at any time. This eczema can also be seen on the plantar aspects of the feet. This type of eczema is most common among middle-aged men (Fig. 1c). It is also called hyperkeratotic-rhagdiform hand eczema. • Pulpitis: hyperkeratotic eczema on the fingertips, possibly with fissures extending under the nails, especially on the thumbs and middle fingers, but it may affect all fingers; vesicles are occasionally seen (Fig. 1d). This subgroup is also called chronic finger tip dermatitis. • Interdigital eczema: eczema in the proximal part of the interdigital spaces with erythema and scaling. Vesicles are rarely seen (Fig. 1e).

17 chapter • Nummular hand eczema: located on the dorsal aspects of the hands or fingers. The well-circumscribed lesions are characterized by erythema, keratosis, vesicles, and pos- 1 sibly oozing. Nummular eczema can secondarily become infected with Stafylococcus Aureus (Fig. 1f). • Non-classifable: contains the remaining groups of hand eczema, such as apron hand eczema, acute red hand eczema or other unclassifiable groups.

a b c

d e f

Fig. 1. (a) Chronic fissured hand eczema, (b) Recurrent vesicular hand eczema, (c) Hyperkeratotic palmar eczema, (d) Pulpitis, (e) Interdigital eczema, (f) Nummular hand eczema Adapted from the Danish Hand Eczema Guidelines20 and reprinted with permission of John Wiley & Sons Ltd

Johansen et al. studied this classification in a prospective cohort of 710 Danish hand eczema patients, although they eliminated the ‘interdigital eczema’ and divided the ‘vesicular hand eczema’ into two subgroups: ‘repeated eruptions’ and ‘few eruptions’.44 Johansen did not identify a simple relationship between the etiology and morphology of hand eczema, but

18 they did establish some clinical patterns. Irritant contact dermatitis was, for example, most chapter frequently the diagnosis in cases of chronic, dry fissured hand eczema (44.3%), pulpitis (41.7%), and nummular hand eczema (40.9%), whereas allergic contact dermatitis dominated 1 as etiological diagnosis for vesicular types of hand eczema. Hyperkeratotic palmar hand eczema was most frequently diagnosed in older men and showed a trend towards nonspecific dermatitis. Diepgen et al. studied the etiology and morphology of hand eczema and came up with a mixture of etiological and morphological diagnoses: allergic contact dermatitis, irritant contact dermatitis, atopic hand eczema, discoid hand eczema, vesicular hand eczema and hyperkeratotic hand eczema.45 The majority of patients were diagnosed with either irritant (21.5%) or allergic (15.2%) contact dermatitis or a combination of both (15.2%), while diagnoses of vesicular hand eczema (9.3%) and hyperkeratotic hand eczema (5.3%) were less frequent. The diagnosis of atopic hand eczema was applied especially to younger age groups and in combination with irritant contact dermatitis. However, with this mixed classification system a single hand eczema patient can be diagnosed in different groups; for example an atopic nurse with vesicular hand eczema due to allergic contact dermatitis caused by nickel can be classified as irritant, allergic, atopic and/or vesicular hand eczema. Molin et al. formulated an algorithm for the diagnosis of hand eczema.46 Based on features such as irritant skin damage, contact allergies, atopy and the involvement of the feet, they defined eight different subtypes of chronic hand eczema. When these were combined with the morphologic type (hyperkeratotic-rhagdiform or dyshidrotic or mixed), certain types of hand eczema seemed to fit a certain morphology. With the use of this algorithm chronic hand eczemas are easy to diagnose, although multiple diagnoses of the same patient remain possible. Since the Danish guideline clearly distinguishes an etiological and a clinical subtype and the other classification systems face difficulties with overlapping diagnoses, we will use the Danish classification system in this thesis.

Outcome measurements In daily practice, grading the severity of hand eczema helps to assess disease severity and the impact of therapeutic interventions. In clinical studies a severity score is necessary to demonstrate the effectiveness of an intervention and to compare different interventions. Either the patient or the physician can grade the severity of the hand eczema, using various scoring systems. In 2010 Weistenhöfer et al. reviewed the skin scores used for the quantification of hand eczema.47 A total of 45 different scoring methods were used in 69 studies; for only three of these methods the inter- and/or intra-observer variability were studied. The most extensively studied scoring systems are the Hand Eczema Severity Index (HECSI)48 and the Osnabrück hand eczema severity index (OHSI)49.

19 chapter The Hand Eczema Severity Index (HECSI) assesses the clinical severity of hand eczema; it scores the intensity of six clinical signs (erythema, induration/ papules, vesicles, fissures, 1 scaling and oedema) on a scale from 0 to 3 in five different locations on both hands (fingertips, fingers [except the tips], palms, back of hands and wrists), combined with the extent of the affected area. The score ranges from 0 to 360. This objective severity assessment based on clinical symptoms has an excellent inter- and intra-observer reliability48 and has been used in various studies.50-52 In the Osnabrück hand eczema severity index (OHSI) the clinical characteristics of six morphological criteria (erythema, infiltration, papules, vesicles, fissures and scaling) are scored in eight different areas (two palmar areas, the backs of both hands, and the dorsal and palmar aspects of the fingers of each hand).49 The extent is assessed by the area of the skin of the hands affected by one or more of the morphological characteristics, which are added up and finally divided by 8. This results in a score between 0 and 18; a score above 7 represents severe hand eczema. The interobserver variability is good, but the intra-observer reliability has not been investigated. The OHSI’s validity and responsiveness to changes make it suitable for monitoring the severity of hand eczema and the effects of treatment in a clinical trial.53 The intensity of lesions is, however, not included in the OHSI, contrary to the HECSI. In addition, the HECSI score seems more sensitive to minimal changes. Both scores can be performed on an average of less than three minutes.54 Each of the mentioned scoring systems has its (dis)advantages. Most of them take several minutes, which is too long in everyday clinical practice. In daily practice physicians use less complicated scoring systems such as the Physician Global Assessments (PGA) to determine the overall severity of hand eczema on a scale from 0 to 4 or 5. The current development of easy-to-use tools will also contribute to the ongoing debate as to a feasible scoring system for hand eczema.55

Primary prevention Adequate education and the use of personal protective equipment like emollients, barrier creams and gloves are widely recommended for people in professions involving skin hazards. A Cochrane review concluded that the protective effects of barrier creams, moisturizers, after work creams, and complex educational interventions had been demonstrated with regard to occupational related irritant hand eczema, but without statistical significance.56 For the protective effect of gloves no randomized controlled trials were identified. Baueret al. concluded that much larger randomized controlled trials over extended time periods (up to one year) are needed to determine the effectiveness of interventions to prevent occupational irritant hand eczema.56 Education of apprentices in occupations at risk, such as nurses57 and hairdressers58, has been found to be worthwhile, for though adherence to prevention measures diminishes over time, those educated about prevention were less likely to develop hand eczema.

20 Management chapter Various treatment regimes and different guidelines have been developed to treat hand eczema.15,19,20,59 All guidelines agree on the importance of prevention and education. 1 All guidelines encourage education and the use of emollients. Topical corticosteroids are a next important step; if topical corticosteroids are insufficient one can consider other options like phototherapy or systemic treatment. These treatment options and the evidence supporting their effectiveness are discussed in this thesis.

Implications for daily life Hands are important organs in communication and expression. Therefore hand eczema has a tremendous impact on a person’s health related quality of life. Hand eczema can cause itching, pain, discomfort, embarrassment and social stigmatization.60 The burden of this disease is comparable to that of eczema on the whole body61, psoriasis and asthma.62 One explanation for this level of distress is that hand eczema afflicts visible parts of the body; Picardi et al. demonstrated that especially female outpatients with lesions in the face or hands were more likely to suffer from psychiatric co-morbidity.63 Another explanation is the level of itching and pain suffered, often resulting in sleep deprivation or even in loss of function. The latter can lead to absence from work or even change or loss of occupation. Cvetkovski reported that among 612 cases of severe occupational hand eczema, 22% disclosed that they had in the past year lost their jobs at least once due to their hand eczema.64 Especially patients in food-related occupations reported job loss. In 120 German patients with chronic occupational related hand eczema and threatened with job loss, three-quarters expressed moderate to very severe impairments in their health related quality of life (HRQoL), both their physical and mental health were impaired.65 A multicenter study by Agner et al. also found that frequent eruptions of eczema and long sick leave contributed negatively to the quality of life.66 Although females were often less severely affected with hand eczema, they experienced an impairment in their quality of life comparable to that of severely affected men. One in four women experienced impairment in doing (wet) work in private and occupational settings.62 Another factor, which may especially be important for women, was the cosmetic appearance. Psychiatric co-morbidity is a neglected issue in patients with skin disorders.67 Hand eczema can lead to or influence mental disorders; especially female patients with occupation related hand eczema were more likely to have higher anxiety scores, correlating with higher scores for VAS pruritus and sleeplessness.65 In addition, 13.5% of the hand eczema patients had a positive depression scale. However, this high depression score was not reproduced by Cvetkovski et al, who found a moderate-to-severe depression rate in 9% of patients with newly-diagnosed hand eczema, as compared with the general Danish population.68 One

21 chapter year after diagnosis this number increased only slightly, although 31% of the participants no longer had any signs of hand eczema. 1 One unexpected complication of hand eczema is the change of the fingerprint or even its complete disappearance (adermatoglyphia). Fingerprints are increasingly used for identification to access buildings, phones, computers or bank accounts and are nowadays registered in the biometric passport. A recent study of Lee et al. demonstrated that 27% of hand eczema patients failed to identify themselves by means of a thumb print, versus 2% of the healthy controls.69 The main explanations for this were dystrophy due to scaling and abnormal white lines due to fissures and wrinkling. Another hypothesis is that treatment with topical corticosteroids causes loss of fingerprints.70,71 Impaired fingerprint recognition leads to less secure forms of identification and might even result in economic loss. One case report described an employee unable to identify himself at work by using a fingerprint. As a result he was unable to clock-in properly, unfairly resulting in absence from work.72 Another patient with vesicular hand eczema was repeatedly put in detention by the immigration department when travelling to the USA because he was unable to identify himself by his fingerprints.71 The failure to provide proper fingerprint identification can in the future result in unforeseeable problems in the light of the increasing demand for biometric authenti- cation techniques.

Prognosis After 1 and 5 years following initial presentation at the dermatologist, respectively 41%73 and 70%74 of the patients reported improvement. During a 15 year period two-thirds of the patients reported periods of hand eczema and 44% reported hand eczema in the past year.75 And although after 15 years 74% of the patients reported that the hand eczema was less severe, the disease has a chronic and relapsing course and its prognosis is poor.76 The mean disease duration is more than 10 years. Unfortunately, patients wait an average of 3 months before consulting a physician.77 Longer delay is associated with a poorer prognosis, and the more chronic and severe the hand eczema becomes, the more difficult will be the treatment. It is therefore important to start treatment in an early phase.

Aims of this thesis The aims of this thesis are: • To describe and compare different pharmacological interventions for hand eczema. • To study the drug survival of cyclosporine for patients with hand eczema in a daily use setting. • To evaluate the diagnostic value of patch testing in patients with different clinical subtypes of hand eczema.

22 • To investigate whether two relatively unknown chemicals (ascaridole in household and chapter cosmetic products and isobornyl acrylate in an occupational setting) are sensitizers and whether they play a role in hand eczema. 1

This thesis contains a Cochrane review (chapter 2), which studies different pharmacological interventions for hand eczema. It also includes a review which aims to answer questions a clinician might encounter in daily practice when dealing with hand eczema patients (chapter 3). In addition, we studied the drug survival of cyclosporine which is prescribed off-label to treat hand eczema (chapter 4). Based on this we composed a practical tool to treat clinical subtypes of hand eczema (chapter 10). The second part of this thesis focuses on allergic contact dermatitis, a well-known cause of hand eczema. We investigated the relationship between different clinical subtypes of hand eczema and sensitization to contact allergens (chapter 5). Finally, the thesis focuses on two relatively unknown contact allergens: isobornyl acrylate (chapter 7) and ascaridole (chapter 8 and 9). Isobornyl acrylate is, as its name implies, an acrylate and (meth)acrylates are a well-known cause of occupational-related hand eczema. Ascaridole is a degradation product of tea tree oil. Tea tree oil can be used as alternative treatment for hand eczema, but is also used in household or cosmetic products. The degradation products of tea tree oil can be sensitizers.

23 chapter References 1. Meding B, Jarvholm B. Incidence of hand eczema-a population-based retrospective study. J Invest 1 Dermatol 2004:122:873-877. 2. Meding B, Liden C, Berglind N. Self-diagnosed dermatitis in adults. Results from a population survey in Stockholm. Contact Dermatitis 2001:45:341-345. 3. Mortz CG, Bindslev-Jensen C, Andersen KE. Prevalence, incidence rates and persistence of contact allergy and allergic contact dermatitis in The Odense Adolescence Cohort Study: a 15-year follow-up. Br J Dermatol 2013:168:318-325. 4. Bregnhøj A, Søsted H, Menné T, Johansen JD. Validation of self-reporting of hand eczema among Danish hairdressing apprentices. Contact Dermatitis 2011:65:146-150. 5. Meding B, Barregard L. Validity of self-reports of hand eczema. Contact Dermatitis 2001:45:99-103. 6. Diepgen TL. Occupational skin diseases. J Dtsch Dermatol Ges 2012:10:297-313; quiz 314-5. 7. Diepgen TL. Occupational skin-disease data in Europe. Int Arch Occup Environ Health 2003:76:331-338. 8. Carøe TK, Ebbehøj N, Agner T. A survey of exposures related to recognized occupational contact dermatitis in Denmark in 2010. Contact Dermatitis 2014:70:56-62. 9. van der Meer EW, Boot CR, van der Gulden JW, Jungbauer FH, Coenraads PJ, Anema JR. Hand eczema among healthcare professionals in the Netherlands: prevalence, absenteeism, and presenteeism. Contact Dermatitis 2013:69:164-171. 10. Lee SW, Cheong SH, Byun JY, Choi YW, Choi HY. Occupational hand eczema among nursing staffs in Korea: Self-reported hand eczema and contact sensitization of hospital nursing staffs. J Dermatol 2013:40:182-187. 11. Lysdal SH, Søsted H, Andersen KE, Johansen JD. Hand eczema in hairdressers: a Danish register- based study of the prevalence of hand eczema and its career consequences. Contact Dermatitis 2011:65:151-158. 12. Uter W, Lessmann H, Geier J, Schnuch A. Contact allergy to hairdressing allergens in female hairdressers and clients--current data from the IVDK, 2003-2006. J Dtsch Dermatol Ges 2007:5:993-1001. 13. Aalto-Korte K, Alanko K, Kuuliala O, Jolanki R. Methacrylate and acrylate allergy in dental personnel. Contact Dermatitis 2007:57:324-330. 14. Kiec-Swierczynska M, Krecisz B, Swierczynska-Machura D, Zaremba J. An epidemic of occupational contact dermatitis from an acrylic glue. Contact Dermatitis 2005:52:121-125. 15. Diepgen TL, Elsner P, Schliemann S, Fartasch M, Kollner A, Skudlik C, et al. Guideline on the management of hand eczema ICD-10 Code: L20. L23. L24. L25. L30. J Dtsch Dermatol Ges 2009:7 Suppl 3:S1-16. 16. Johannisson A, Pontén A, Svensson A. Prevalence, incidence and predictive factors for hand eczema in young adults -- a follow-up study. BMC Dermatol 2013:13:14.

24 17. Meding B, Swanbeck G. Predictive factors for hand eczema. Contact Dermatitis 1990:23:154-161. chapter 18. Veien NK, Hattel T, Laurberg G. Hand eczema: causes, course, and prognosis II. Contact Dermatitis 2008:58:335-339. 1 19. Lynde C, Guenther L, Diepgen TL, Sasseville D, Poulin Y, Gulliver W, et al. Canadian hand dermatitis management guidelines. J Cutan Med Surg 2010:14:267-284. 20. Menné T, Johansen JD, Sommerlund M, Veien NK, Danish Contact Dermatitis Group. Hand eczema guidelines based on the Danish guidelines for the diagnosis and treatment of hand eczema. Contact Dermatitis 2011:65:3-12. 21. Posada C, Garcia-Doval I, de la Torre C, Cruces MJ. Value of palmar and plantar biopsies of hyperkeratotic and vesicular pustular lesions: a cross-sectional study. Actas Dermosifiliogr 2010:101:103-105. 22. Hersle K, Mobacken H. Hyperkeratotic dermatitis of the palms. Br J Dermatol 1982:107:195-201. 23. Barnhill RL. Dermatopathology. 3rd edition: New York,McGraw Hill Medical,2010. 24. Weedon D. Weedon’s skin pathology. 3rd edition: China,Churchill Livingstone, Elsevier,2010. 25. Kutzner H, Wurzel RM, Wolff HH. Are acrosyringia involved in the pathogenesis of “dyshidrosis”? Am J Dermatopathol 1986:8:109-116. 26. Chang YY, van der Velden J, van der Wier G, Kramer D, Diercks GF, van Geel M, et al. Keratolysis exfoliativa (dyshidrosis lamellosa sicca): a distinct peeling entity. Br J Dermatol 2012:167:1076-1084. 27. Weisshaar E, Kallen U, Weiss M. “The itching hand”- important differential diagnoses and treatment. J Dtsch Dermatol Ges 2013:11:31-42. 28. Meding B, Swanbeck G. Predictive factors for hand eczema. Contact Dermatitis 1990:23:154-161. 29. Johansson SG, Bieber T, Dahl R, Friedmann PS, Lanier BQ, Lockey RF, et al. Revised nomenclature for allergy for global use: Report of the Nomenclature Review Committee of the World Allergy Organization, October 2003. J Allergy Clin Immunol 2004:113:832-836. 30. Thyssen JP, Carlsen BC, Menné T, Linneberg A, Nielsen NH, Meldgaard M, et al. Filaggrin null mutations increase the risk and persistence of hand eczema in subjects with atopic dermatitis: results from a general population study. Br J Dermatol 2010:163:115-120. 31. Thyssen JP, Linneberg A, Engkilde K, Menné T, Johansen JD. Contact sensitization to common haptens is associated with atopic dermatitis: new insight. Br J Dermatol 2012:166:1255-1261. 32. Landeck L, Visser M, Skudlik C, Brans R, Kezic S, John SM. Clinical course of occupational irritant contact dermatitis of the hands in relation to filaggrin genotype status and atopy. Br J Dermatol 2012:167:1302-1309. 33. Uter W. Allergic contact dermatitis. In: Thyssen JP, Maibach HI, editors. Filaggrin, basic science, epidemiology, clinical aspects and management. 1st ed. Heidelberg: Springer; 2014. p. 251-257. 34. Lerbaek A, Kyvik KO, Ravn H, Menné T, Agner T. Incidence of hand eczema in a population-based twin cohort: genetic and environmental risk factors. Br J Dermatol 2007:157:552-557.

25 chapter 35. Frosch PJ, Menné T, Lepoittevin JP. Contact Dermatitis. 4th edition: Berlin Heidelberg, Springer, 2006. 1 36. Rietschel RL, Fowler JF. Fisher’s Contact Dermatitis 6. 6th edition: Hamilton,BC Decker Inc,2008. 37. Martin SF, Esser PR, Weber FC, Jakob T, Freudenberg MA, Schmidt M, et al. Mechanisms of chemical-induced innate immunity in allergic contact dermatitis. Allergy 2011:66:1152-1163. 38. Iwasaki A, Medzhitov R. Regulation of adaptive immunity by the innate immune system. Science 2010:327:291-295. 39. Bruze M, Isaksson M, Edman B, Bjorkner B, Fregert S, Moller H. A study on expert reading of patch test reactions: inter-individual accordance. Contact Dermatitis 1995:32:331-337. 40. Svedman C, Isaksson M, Björk J, Mowitz M, Bruze M. ‘Calibration’ of our patch test reading technique is necessary. Contact Dermatitis 2012:66:180-187. 41. Bruze M, Svedman C, Andersen KE, Bruynzeel D, Goossens A, Johansen JD, et al. Patch test concentrations (doses in mg/cm2 ) for the 12 non-mix fragrance substances regulated by European legislation. Contact Dermatitis 2012:66:131-136. 42. Farm G. Repeated open application tests (ROAT) in patients allergic to colophony--evaluated visually and with bioengineering techniques. Acta Derm Venereol 1998:78:130-135. 43. Hannuksela M, Salo H. The repeated open application test (ROAT). Contact Dermatitis 1986:14:221-227. 44. Johansen JD, Hald M, Andersen BL, Laurberg G, Danielsen A, Avnstorp C, et al. Classification of hand eczema: clinical and aetiological types. Based on the guideline of the Danish Contact Dermatitis Group. Contact Dermatitis 2011:65:13-21. 45. Diepgen TL, Andersen KE, Brandao FM, Bruze M, Bruynzeel DP, Frosch P, et al. Hand eczema classification: a cross-sectional, multicentre study of the aetiology and morphology of hand eczema. Br J Dermatol 2009:160:353-358. 46. Molin S, Diepgen TL, Ruzicka T, Prinz JC. Diagnosing chronic hand eczema by an algorithm: a tool for classification in clinical practice.Clin Exp Dermatol 2011:36:595-601. 47. Weistenhofer W, Baumeister T, Drexler H, Kutting B. An overview of skin scores used for quantifying hand eczema: a critical update according to the criteria of evidence-based medicine. Br J Dermatol 2010:162:239-250. 48. Held E, Skoet R, Johansen JD, Agner T. The hand eczema severity index (HECSI): a scoring system for clinical assessment of hand eczema. A study of inter- and intraobserver reliability. Br J Dermatol 2005:152:302-307. 49. Skudlik C, Dulon M, Pohrt U, Appl KC, John SM, Nienhaus A. Osnabrueck hand eczema severity index--a study of the interobserver reliability of a scoring system assessing skin diseases of the hands. Contact Dermatitis 2006:55:42-47. 50. Agarwal US, Besarwal RK. Topical clobetasol propionate 0.05% cream alone and in combination with azathioprine in patients with chronic hand eczema: an observer blinded randomized comparative trial. Indian J Dermatol Venereol Leprol 2013:79:101-103.

26 51. Bauer A, Lange N, Matterne U, Meurer M, Braeutigam M, Diepgen TL. Efficacy of pimecrolimus chapter 1 % cream in the long term management of atopic hand dermatitis. A double-blind RCT. J Dtsch Dermatol Ges 2012:10:426-433. 1 52. Yousefi M, Barikbin B, Kamalinejad M, Abolhasani E, Ebadi A, Younespour S,et al. Comparison of therapeutic effect of topical Nigella with Betamethasone and Eucerin in hand eczema. J Eur Acad Dermatol Venereol 2013:27:1498-1504. 53. Dulon M, Skudlik C, Nubling M, John SM, Nienhaus A. Validity and responsiveness of the Osnabruck Hand Eczema Severity Index (OHSI): a methodological study. Br J Dermatol 2009:160:137-142. 54. Baumeister T, Weistenhofer W, Drexler H, Kutting B. Spoilt for choice--evaluation of two different scoring systems for early hand eczema in teledermatological examinations. Contact Dermatitis 2010:62:241-247. 55. van der Valk PG, van Gils RF, Boot CR, Evers AW, Donders R, Alkemade HA, et al. A simple tool with which to study the course of chronic hand eczema in clinical practice: a reduced-item score. Contact Dermatitis 2013:69:112-117. 56. Bauer A, Schmitt J, Bennett C, Coenraads PJ, Elsner P, English J, et al. Interventions for preventing occupational irritant hand dermatitis. Cochrane Database Syst Rev. 2010:16:CD004414. 57. Visser MJ, Verberk MM, van Dijk FJ, Bakker JG, Bos JD, Kezic S. Wet work and hand eczema in apprentice nurses; part I of a prospective cohort study. Contact Dermatitis 2014:70:44-55. 58. Bregnhøj A, Menné T, Johansen JD, Søsted H. Prevention of hand eczema among Danish hairdressing apprentices: an intervention study. Occup Environ Med 2012:69:310-316. 59. English J, Aldridge R, Gawkrodger DJ, Kownacki S, Statham B, White JM, et al. Consensus statement on the management of chronic hand eczema. Clin Exp Dermatol 2009:34:761-769. 60. Fowler JF, Ghosh A, Sung J, Emani S, Chang J, Den E, et al. Impact of chronic hand dermatitis on quality of life, work productivity, activity impairment, and medical costs. J Am Acad Dermatol 2006:54:448-457. 61. Zachariae R, Zachariae C, Ibsen HH, Mortensen JT, Wulf HC. Psychological symptoms and quality of life of dermatology outpatients and hospitalized dermatology patients. Acta Derm Venereol 2004:84:205-212. 62. Moberg C, Alderling M, Meding B. Hand eczema and quality of life: a population-based study. Br J Dermatol 2009:161:397-403. 63. Picardi A, Abeni D, Renzi C, Braga M, Puddu P, Pasquini P. Increased psychiatric morbidity in female outpatients with skin lesions on visible parts of the body. Acta Derm Venereol 2001:81:410- 414. 64. Cvetkovski RS, Rothman KJ, Olsen J, Mathiesen B, Iversen L, Johansen JD, et al. Relation between diagnoses on severity, sick leave and loss of job among patients with occupational hand eczema. Br J Dermatol 2005:152:93-98.

27 chapter 65. Boehm D, Schmid-Ott G, Finkeldey F, John SM, Dwinger C, Werfel T, et al. Anxiety, depression and impaired health-related quality of life in patients with occupational hand eczema. Contact 1 Dermatitis 2012:67:184-192. 66. Agner T, Andersen KE, Brandao FM, Bruynzeel DP, Bruze M, Frosch P, et al. Hand eczema severity and quality of life: a cross-sectional, multicentre study of hand eczema patients. Contact Dermatitis 2008:59:43-47. 67. Picardi A, Abeni D, Melchi CF, Puddu P, Pasquini P. Psychiatric morbidity in dermatological outpatients: an issue to be recognized. Br J Dermatol 2000:143:983-991. 68. Cvetkovski RS, Zachariae R, Jensen H, Olsen J, Johansen JD, Agner T. Quality of life and depression in a population of occupational hand eczema patients. Contact Dermatitis 2006:54:106-111. 69. Lee CK, Chang CC, Johar A, Puwira O, Roshidah B. Fingerprint changes and verification failure among patients with hand dermatitis. JAMA Dermatol 2013:149:295-299. 70. Gean CJ, Hiatt GF, Maibach HI. Complete eradication of fingerprints associated with topical corticosteroids. Semin Dermatol 1983:2:257-61. 71. Sergeant A, McPhee N, Holme SA. Acquired loss of fingerprints: do topical corticosteroids play an aetiological role? Clin Exp Dermatol 2012:37:679-680. 72. Ramam M, Krishna SG. A novel cause of economic loss due to hand dermatitis. Arch Dermatol 2011:147:753. 73. Cvetkovski RS, Zachariae R, Jensen H, Olsen J, Johansen JD, Agner T. Prognosis of occupational hand eczema: a follow-up study. Arch Dermatol 2006:142:305-311. 74. Rosen RH, Freeman S. Prognosis of occupational contact dermatitis in New South Wales, Australia. Contact Dermatitis 1993:29:88-93. 75. Meding B, Wrangsjo K, Jarvholm B. Fifteen-year follow-up of hand eczema: persistence and consequences. Br J Dermatol 2005:152:975-980. 76. Mälkönen T, Alanko K, Jolanki R, Luukkonen R, Aalto-Korte K, Lauerma A, et al. Long-term follow-up study of occupational hand eczema. Br J Dermatol 2010:163:999-1006. 77. Hald M, Agner T, Blands J, Johansen JD, Danish Contact Dermatitis Group. Delay in medical attention to hand eczema: a follow-up study. Br J Dermatol 2009:161:1294-1300.

28 chapter 1

29 Chapter

30 chapter 2 Cochrane review: Interventions for hand eczema Re-submitted to the Cochrane Database of Systematic Reviews June 2014

Pieter-Jan Coenraads1, Wietske Andrea Christoffers1, Åke Svensson2, Thomas L Diepgen3, Janine L Blok1, Uwe Matterne3, Hywel C Williams4

1 Department of Dermatology, University of Groningen, University Medical Center Groningen, Groningen, Netherlands. 2 Department of Dermatology, Skåne University Hospital, Malmö, Sweden. 3 Department of Clinical Social Medicine, Heidelberg University Hospital, Heidelberg, Germany. 4 Centre of Evidence Based Dermatology, The University of Nottingham, Nottingham, UK

Note to the reader: In this thesis we only included the conventional systemic interventions for the treatment of hand eczema. Therefore we re-wrote the original review. The original abstract and discussion are still in situ.

Key words: hand eczema, interventions, evidence-based dermatology, alitretinoin, corticosteroids

31 Abstract Background: Hand eczema describes an inflammation of the skin of the hands and tends to run a chronic relapsing course. The condition is common and is often associated with itch, social stigma and impairments in employment, which are major burdens for patients. Many chapter different interventions, some of unknown effectiveness, are used to treat hand eczema. Objective: To assess the effects of systemic interventions for patients with hand eczema. 2 Methods: We searched the following databases up to February 2014: the Cochrane Skin Group Specialised Register, CENTRAL in The Cochrane Library (2014, Issue 1), MEDLINE (from1946), EMBASE (from 1974), AMED (from 1985), LILACS (from 1982) and GREAT (from inception). We also searched trials registers for ongoing and unpublished studies, and checked the reference lists of published studies for further references to relevant studies. All randomized controlled trials (RCT) of systemic interventions for hand eczema were included regardless of hand eczema type, other affected sites and placebos or active comparators. Data was extracted according to standard procedures expected by The Cochrane Collaboration. The following primary outcomes were included: 1. Percentage of participants with self-rated good/excellent control of symptoms 2. Percentage of participants with investigator-rated good/excellent control of symptoms 3. Adverse events And the risk of bias for each study was assessed. Results: Seven RCTs regarding systemic therapy were included with a total of 1,661 participants. Three studies investigated alitretinoin, one acitretin, one azathioprine, one cyclosporine and one study compared alitretinoin to cyclosporine. Five studies included our primary outcome good/ excellent control either participant- or investigator-rated and all studies monitored adverse events. The studies were of good quality with overall low risks of bias in the different domains, although the larger studies on alitretinoin were sponsored by pharmaceutical companies (funding). There was substantial clinical heterogeneity in design, outcome measures and timing of the outcome assessments and only for alitretinoin 30 mg we were able to cluster the results. The retinoid alitretinoin was the most investigated treatment and seemed safe and effective. Conclusions: The results of this review cannot be used to inform clinical practice with regard to the best way of managing hand eczema, especially in the long-term, due to the diversity and quality of the included studies. Besides head-to-head trials were lacking. Alitretinoin was more effective than placebo in two large (industry funded) RCTs. However, the comparative advantage of alitretinoin over other treatments needs further evaluation.

32 Abstract complete review Background: Hand eczema describes an inflammation of the skin of the hands and tends to run a chronic relapsing course. The condition is common and is often associated with itch, social stigma and impairments in employment, which are major burdens for patients. In chapter practice, chronic hand eczema is due to a combination of external and internal factors. Many different interventions, some of unknown effectiveness, are used to treat hand eczema. Objectives: To assess the effects of topical and systemic interventions for people with hand 2 eczema. Search methods: We searched the following databases up to February 2014: the Cochrane Skin Group Specialised Register, CENTRAL in The Cochrane Library (2014, Issue 1), MEDLINE (from1946), EMBASE (from1974), AMED (from1985), LILACS (from1982) and GREAT(from inception). We also searched trials registers for ongoing and unpublished studies, and checked the reference lists of published studies for further references to relevant studies. In an earlier phase, we hand searched 21 dermatology journals, and conference proceedings of the European Academy of Dermatology and Venereology. Selection criteria: We included randomized controlled trials (RCT) of interventions for hand eczema regardless of hand eczema type, other affected sites and placebos or active comparators. We excluded studies which focused on prevention of hand eczema or integrated care programs. Data collection and analysis: We used standard procedures expected by The Cochrane Collaboration and three authors screened titles and abstracts. Potential suitable full text articles were assessed independently by two authors using a data extracting form which judged the risk of bias. Disagreement was solved by means of discussion. We grouped trials into the following intervention categories: skin protection measures, topical treatments, systemic treatments, and extracted the following outcomes 1. Percentage of participants with self-rated good/ excellent control of symptoms 2. Percentage of participants with investigator-rated good/ excellent control of symptoms 3. Adverse events Secondary outcomes: 1. Reduction in severity by a participant-rated scoring system 2. Reduction in severity by an investigator-rated scoring system 3. Time until relapse 4. Dose reduction We composed risk of bias tables for each study and judged the overall result in the risk of bias assessment. Main results: We identified 55 RCTs. These covered a large variety in treatments and reporting outcomes. In total, 4,619 participants were enrolled. There was substantial clinical

33 heterogeneity in interventions, outcome measures and timing of the outcome assessments. Only five studies compared two different classes of interventions. Statistical pooling was not possible for the studies on corticosteroids, UV phototherapy and ionizing radiation/ X-rays (six to eight studies in each category), because they were considered too heterogeneous. chapter Many studies were at unclear risk of bias in one or more components of trial design. For the topical treatments the overall quality of the body of evidence the outcome can not be 2 given, because the variability of the treatments was too diverse. For oral treatments overall this quality was higher. About half of the studies (n = 30) included our primary outcome good/ excellent control either participant- or investigator-rated. The majority of the studies included our primary outcome adverse events (n = 52): most recorded adverse events were mild (local irritation with stinging, erythema and burning), except for the adverse events due to systemic therapies such as oral metal chelating agents. Authors’ conclusions: The results of this review cannot be used to inform clinical practice with regard to the best way of managing hand eczema, especially in the long-term, due to the diversity and quality of the included studies. Topical corticosteroids and UV-phototherapy appear to be the major standard treatment categories, but there is little evidence of a comparative advantage of one specific treatment within these categories, and also little evidence for a comparative advantage with other types of treatment. The effect of topical calcineurin inhibitors is unclear. An oral retinoid for people with severe chronic hand eczema seemed more effective than placebo in two large (industry funded) RCTs. The comparative advantage over other treatments needs further evaluation.

Summary of findings for the main comparison Interventions for hand eczema Patient or population: Participants (male and female, adults and children) with hand eczema, regardless of the underlying cause Settings: Secondary setting (hospital) Intervention: Topical and systemic interventions for the treatment of hand eczema Comparison: Another topical or systemic intervention, the same intervention but in another dose or placebo Comments: This Cochrane review covers a wide variety of different interventions for hand eczema. Due to the heterogeneity of the included studies with regards to interventions, design and outcomes, we were unable to pool the data. Adding a single summary of findings table would therefore mean replicating the entire review, since almost all recommendations are based on a single study result and findings have not been replicated elsewhere.

34 Background Description of the condition Defnition and epidemiology ’Hand eczema’ is an inflammation of the skin (dermatitis) that is confined to the hands. chapter With a point prevalence of 1 to 5% among adults in the general population, and a one-year prevalence of up to 10%, depending on whether the disease definition includes more pronounced or mild cases1, hand eczema can be considered as a common condition. The 2 incidence is estimated to be about 4 per 1,000 in men and about twice as common in women with an estimated incidence of 9 per 1,000 women.1-3 Reasons for the sex difference are unknown, although greater exposure of women to wet work is a likely contributory factor.4 The incidence of notified (i.e. usually more severe) occupation related cases is estimated to be above 0.7 per 1,000 people per year, with much higher incidences (up to 1 in 100) in high-risk populations such as hairdressers.5 A decreased prevalence has been observed in Swedish adults, which was attributed to a decline in occupational exposure to irritants.6 Over the years several authors have proposed a workable definition of hand eczema, whereby different subtypes are recognized.7 Different types of hand eczema have been recognized and hand eczema can be classified according to etiological (causative), clinical- morphological typology or a combination of both.8,9 However, due to multicausality it is difficult to assess the influence of each causative factor, and therefore only one etiological diagnosis might be insufficient.

Impact Itch is common among those with hand eczema. The itch caused by hand eczema can be intense, leading to sleep loss in the sufferer and other family members. A vicious cycle of symptoms causing skin damage can develop, the so-called itch/scratch/itch cycle. Cracks and blisters can be painful. A visible skin disease can be a great burden and lead to a social stigma. The hands are important organs of communication and expression, and therefore any visible skin disease on the hands may result in major psychosocial problems, e.g. anxiety, low self-esteem and social phobia. Painful cracks and blisters, besides their negative effect on daily life outside work, can prevent the ability to carry out manual work, leading to significant disability and huge economic loss to both individuals and society.10 Hand eczema accounts for an estimated 90% of occupational skin disease. Quality of life assessments have shown an impact on daily life and on the employment.11,12 A comparison between the generic quality of life instrument SF-36 and the skin related DLQI showed a slightly higher impact of hand eczema on women compared to men for specific sub-items.13 A comparison of physician-rated with patient-rated assessments of severity showed a poor correlation,

35 indicating that patients may evaluate several aspects of their hand eczema (including the degree of erythema, vesicles and fissures) differently from physicians.14

Prognosis chapter Previous studies have suggested that hand eczema tends to run a chronic relapsing course, 2 with the vast majority of people experiencing negative psychosocial consequences.15-17 Description of the intervention There are many diverse therapies being used to control the disease, such as: 1. Skin protection measures, including gloves; 2. Topical treatments (bland emollients, corticosteroid creams/ ointments, calcineurin inhibitors, coal tar and derivatives, irradiation with UV-light or X-rays); 3. Systemic treatments (oral corticosteroids, other immunosuppressants such as cyclosporine, retinoids). The Cochrane review covers all these different outcomes, however, in this thesis only the systemic treatment options are displayed.

Why it is important to do this review The high prevalence of hand eczema, along with its poor prognosis and associated disability with economic losses and the impairment of quality of life, makes hand eczema an important disease to study from an individual as well as from a societal perspective. This, coupled with the large list of diverse treatments of unknown effectiveness and several conflicting studies18,19, suggests that a systematic review is needed. Even if methodological constraints do not permit sufficient clarification of existing conflicts to provide clear guidance in clinical practice, the review will be an important step in identifying the research gaps and consequently providing directions for future research.

Objectives To assess the effects of systemic interventions for hand eczema in adults and children.

Methods Criteria for considering studies for this review Types of studies We included randomized controlled trials of interventions for hand eczema regardless of hand eczema type and other affected localizations.

36 Types of participants People (adults and children, occupational and non occupational) with the diagnoses of hand eczema, regardless of the underlying assumed cause were eligible. We also included participants were other parts of the body were affected in addition to hand eczema. The terms ’eczema’ and ’dermatitis’ were acceptable, wherever it referred to the hands. chapter Other terms as ’pompholyx’, ’dyshidrosis’ and ’pulpitis’ were also deemed acceptable. Studies which included participants with other diagnoses besides hand eczema were only 2 included in the analyses when we were able to obtain separate data for the hand eczema participants.

Types of interventions Only studies comparing the intervention with no treatment, placebo, vehicle or other active treatments were included. All types of interventions were considered, except for interventions to prevent hand eczema (primary prevention). We excluded studies which focused on prevention of hand eczema and studies which investigated integrated care programs or education programs (non-pharmacologic interventions). When a study reported on treatment during a remission- or clearance-induction phase for participants before they were randomized to a follow-up or maintenance phase, only the latter (randomized) phase was considered for this review.

Types of outcome measures We extracted the following primary and secondary outcomes from the included studies:

Primary outcomes • Percentage of participants with self-rated good/excellent control of symptoms • Percentage of participants with investigator-rated good/ excellent control of symptoms • Adverse events: adverse effects (long and short-term) of the intervention

Secondary outcomes • Reduction in severity participant-rated • Reduction in severity investigator-rated • Time until relapse • Dose reduction: reduction in treatment dose per time unit or cumulative prescribed treatment dose. For example: decrease in daily topical medication, or decrease in weekly photo irradiation

37 Search methods for identification of studies We aimed to identify all relevant randomized controlled trials (RCTs) regardless of language or publication status (published, unpublished, in press, or in progress). chapter Electronic searches We searched the following databases up to 4 February 2014: 2 • The Cochrane Skin Group Specialised Register • The Cochrane Central Register of Controlled Trials (CENTRAL) 2014 • MEDLINE via OVID (from 1946) • EMBASE via OVID (from 1974) • AMED via OVID, Allied and Complementary Medicine (from 1985) • LILACS, Latin American and Caribbean Health Science Information database (from 1982) • The Global Resource of Eczema Trials. Centre of Evidence Based Dermatology. Accessed at http:// www.greatdatabase.org.uk on 4 February 2014, using the following terms in the title of the records: hand* or finger* or palm or palms.

Trials Registers We searched the following trials registers on the 19 February 2014 using the following search terms: hand and (eczema or dermatitis): • The metaRegister of Controlled Trials www.controlledtrials.com. • The U.S. National Institutes of Health Ongoing Trials Register www.clinicaltrials.gov. • The Australian and New Zealand Clinical Trials Registry www.anzctr.org.au. • The World Health Organisation International Clinical Trials Registry platform www.who. int/trialsearch. • The Ongoing Skin Trials Register on www.nottingham.ac.uk/ongoingskintrials (which has been transferred to the Australian and New Zealand Clinical Trials Registry (ANZCTR) www.nottingham.ac.uk/ongoingskintrials/ www.anzctr.org.au.

Searching other resources Correspondence with authors If needed for clarifications, we contacted the authors by means of the correspondence options stated in the articles for articles published since 1999. Whenever email addresses did not work, we tried to find recent publications of the same corresponding author for more recent contact data, or we searched Facebook, LinkedIn and the WorldWideWeb to connect with the authors. In addition, we tried to contact all the authors of studies which included other dermatoses among hand eczema to obtain separate data for the hand eczema participants.

38 Unpublished literature We have contacted the authors and pharmaceutical industries of the ongoing trials which were recently completed according to the five trial registers mentioned under “Trial Registers”. When results were published on the trial register web sites, we included these in the results and we tried to contact the author for additional information. We did not search chapter other sources for grey literature or other unpublished trials. 2 Conference proceedings We searched the conference proceedings of the annual conferences of the European Academy of Dermatology and Venereology (EADV) from 2000 to 2011 for further RCTs. Some were available from the JEADV, however, some had to be obtained from the organization itself from whom we requested the material on CDROM.

Hand searching We hand searched using the terms eczema, dermatitis, hand(s), palmoplantar, inflammatory in 16 English journals, 2 German, 1 Italian, 1 French and 1 Dutch dermatology journal (all journals 1977 through 2003).

Data collection and analysis Selection of studies Four authors (PJC, JLB, UM and WAC) independently checked titles and abstracts identified from the searches. In addition, three authors (PJC, TD, ÅS) conducted a additional hand search. If it was clear that the study did not refer to a randomized controlled trial on hand eczema, we excluded it. We retrieved all those that might be trials as full text articles for further independent examination by two other authors (TD and ÅS). These two authors decided which trials conformed to the inclusion criteria. Excluded studies and the reason for exclusion are listed in Characteristics of excluded studies (data not shown). The original protocol stated that these reviewers would also assess ’methodological quality’ (risk of bias), but this was done only on the items ’type of control’ and ’randomization’. In a subsequent step, we performed a thorough assessment of risk of bias (authors TD, PJC and ÅS). We resolved discrepancies by discussion between the two authors. In case of duplicate publications of the same trial, we used the paper with the most relevant data.

Data extraction and management Three authors (PJC, TD and ÅS) extracted data independently, using a data extraction form. We resolved discrepancies and uncertainties in a series of consensus meetings. The outcome data extracted from the included studies were checked and entered into Review Manager (RevMan) by three other authors (JLB, WAC and UM).

39 Assessment of risk of bias in included studies Two review authors (ÅS and TD) independently assessed the risk of bias in the included studies following the domain-based evaluation described in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions.20 Whenever there was a disagreement chapter between the assessments of risk of bias, this was resolved in a consensus meeting with a third reviewer. The completed risk of bias forms were assessed by two authors (JB and WAC) 2 and entered in RevMan.

Randomization procedure The procedure was judged as adequate (low risk of bias) when the allocation sequence was able to protect against biased allocation of the comparison groups. If no details were given about the methods of sequence generation, i.e. if there was doubt about the adequacy of sequence generation, we judged studies as ’unclear’. We considered systematic methods which allow biased allocations, such as alternation or assignment based on day of admis- sion as inadequate (high risk of bias).The references to a lottery system, throwing dices or a computer program was considered as adequate (low risk of bias).

Concealment of allocation We judged this as adequate (low risk of bias) when clinicians and participants are unaware of future allocations before the participant gave consent to the study. Examples of these are randomization by a third party or the use of sequentially numbered, opaque, sealed envelopes. We judged this as ’unclear’ if insufficient details are given about methods of allocation concealment. We judged the allocation inadequate (high risk of bias) when there was a possibility of knowledge of the next assignment, so when investigators could have successfully guessed the allocation before the participant gave consent.

Blinding Performance bias refers to systematic differences between groups in the care that is provided, or in exposure to factors other than the interventions of interest.20 After enrolment into the study, blinding of participants and site staff can reduce the risk that knowledge of which intervention was received, rather than the intervention itself, affects outcomes. Effective blinding can also ensure that the compared groups receive a similar amount of attention, ancillary treatment and diagnostic investigations. With regards to performance bias, the patient and the study personnel had to be blinded to judge a low risk. The use of identical looking study and control drug (placebo) is considered as an adequate method of blinding if the study is double blind. When for example the radiographer is the only person aware of the treatment allocation, however the study could not be done in another way, we did judge this as a low risk of bias, since we considered this as the best possible way to minimize the

40 risk of performance bias. Detection bias refers to systematic differences between groups in how outcomes are determined. Blinding of outcome assessors may reduce the risk that knowledge of which intervention was received, rather than the intervention itself, affects outcome measurement.20 The procedure was judged as low risk of bias for detection when the outcome assessor was unaware of the allocation. When an article only states that the study chapter was investigated blinded or double-blinded, we considered this as too few information to judge the risk of bias and concluded “unclear”. We judged low risk of bias for detection 2 bias for studies were independent observers were used, the study drugs were packed and dispensed by a third party or were an adequate other method to blind the observer was described.

Loss to follow up and intention-to-treat analysis We judged the risk of attrition bias (incomplete outcome bias) as adequate (low risk of bias) when more than 80% of participants are followed up and analyzed in the groups to which they were originally randomized. In addition, an intention-to-treat analysis was considered as a low risk of bias for the attrition bias. When more than 20% of the participants dropped-out and no intention-to-treat analysis was carried out, this is considered as a high risk of bias.

Other risk of bias Baseline comparison for severity of the disease We judged this as adequate (low risk of bias) when a comparison between the groups was made and did not show a statistically significant difference.

Certainty that participants have hand eczema We judged this as adequate (low risk of bias) when the diagnosis was made or confirmed by a physician or if the participant was seen in a hospital setting.

Trial conditions In addition to the components stated in the original protocol, we assessed the items: • rationale of sample size described; • primary/major outcome measure defined; • inclusion and exclusion criteria for participants stated; • duration of study specified. Each component was characterized as adequate (low risk of bias), unclear or inadequate (high risk of bias).

41 Funding Systematic bias might favor products that are made by the company funding the research. Therefore studies funded by pharmaceutical industries were judged as high risk of bias. Studies which were sponsored by universities, hospitals and governmental organizations chapter were judged as low-risk of bias.

2 Measures of treatment effect To measure the effect of a treatment, risk ratios (RR) with 95% confidence intervals (CI) were used. This gave an accurate estimation of the relative differences between comparison groups, as in some studies the proportion of outcome events might be close to one in one of both groups. Results from analyses of continuous data are expressed as mean differences (MD), including the CIs and the respective P value. If insufficient data is available for any of the two analyses we will sum up the available data from the respective study including the stated P value. Numeric data in charts and tables are interpreted. We tried to extract numeric data from graphical presentations by the use of a ruler or contacted the authors for recent trials if the data are unclear. For data that was extracted from a graph, we always added this as remark. In studies where exclusively the median values were presented for a particular outcome, the median was substituted for the mean, provided that data were not too skewed. Whenever standard deviations were not available from a paper, we tried to calculate it from other available data. When confidence intervals were provided, the formula given in chapter 7.7.3.2 of the Cochrane handbook was used.20

Unit of analysis issues Cluster randomized trials Cluster randomized trials (groups of individuals randomized to intervention or control instead of individuals) were checked for unit of analysis errors based on the advice provided in Section 16.3.4 of the Cochrane Handbook for Systematic Reviews of Interventions.20

Cross-over studies In cross-over studies (allocate each participant to a sequence of interventions instead of to only one intervention) unit of analysis issues can arise when participants have been randomized to multiple treatments in multiple periods or where there has been an inadequate wash-out period. Cross-over studies were dealt with by analyzing the first treatment period only as a simple parallel group study.

Within-patient studies (self-controlled, left-right designs) Since the analysis of paired data was not possible with RevMan after the data from within- patient studies have been entered into analysis tables, the results will be summarized.

42 The unit of analysis in within-patient studies was one hand per participant, whereas in parallel group studies the unit of analysis was per participant.

Dealing with missing data We tried to minimize the amount of missing data by contacting all authors from1999 and chapter forward. We contacted authors by personal communication by email, letters or social media features such as LinkedIn. Questions were asked with regard to uncertainty in the assess- 2 ment of risks of bias or trial design. Another way to deal with missing data was searching for other sources such as trial registers, which may provide additional information with regards to the study design, or comparing the study to similar studies done by the same authors.

Assessment of heterogeneity Clinical heterogeneity (or clinical diversity) was considered as the variability in the participants, interventions and outcomes. We planned to assess clinical heterogeneity by examining the characteristics of the studies, the similarity between the types of participants, the interventions and the outcomes. If studies were sufficiently similar, statistical pooling would be done using a weighted treatment effect. Random effects analysis will be used because of the anticipated differences across studies in, amongst other things, the patient base included. Statistical heterogeneity will be investigated with the I2 test. Reasons for heterogeneity in studies will be explored and subgroups can be studied further (for example children or hyperkeratotic hand eczema). If there is considerable variation in the design, meta-analyses will not be performed.

Assessment of reporting biases We tried to encounter reporting bias by searching databases of ongoing trials, and searching the LILACS (Latin American and Caribbean Health Science Information) database. Whenever a trial was completed according the trial register, but was not yet published, we contacted the authors for clarification. We planned on including statistical methods for detecting publication bias (e.g. Begg’s funnel plots). However, funnel plots are recommended by the Cochrane hand book when at least a substantial amount of studies (ten or more) is included in the meta-analysis. This was not feasible due to the heterogeneity of the included studies. For reporting bias we therefore studied the authors and institutions involved (pharmaceutical companies or not), funding, sponsorship of commercially available supplements and finally conflict of interests, which is listed under funding in the Risk of Bias tables.

Data synthesis According to the protocol we planned on statistical pooling, using a weighted treatment effect, whenever studies appeared sufficiently similar. If pooling seemed feasible, random

43 effects analysis would have been used because of the anticipated differences across studies in, amongst other things, the number of participants included. Heterogeneity statistics were reported for the pooled analyses. In the process of the review it was realized that there were hardly any grounds for data pooling because there were considerable differences between chapter the studies. Pooling was deemed possible for the results of only two studies (Ruzicka 2004; Ruzicka 2008). Studies with high risks of bias in the majority of the trial design will be 2 excluded from analyses.

Subgroup analysis and investigation of heterogeneity We planned sensitivity analyses to examine the effects of excluding study subgroups (for example children versus adults, or recurrent vesicular versus hyperkeratotic hand eczema), which were defined a priori, or those studies with lower reported methodological quality (’Inadequate’). Unfortunately, we could not perform sensitivity analysis because pooling was deemed feasible for the results of only two studies (Ruzicka 2004; Ruzicka 2008), and therefore heterogeneity and the effects of excluding subgroups could not be studied in other studies.

Sensitivity analysis We intended to perform sensitivity analyses for pooled data.

Results Description of studies We included 55 randomized controlled trials on different interventions for hand eczema.

Results of the search The initial searches until February 2014 resulted in 165 trials after removal of duplicate records. Of these, 93 were identified as controlled trials. After a second round of assessment of their design and methodological quality, 15 were rejected because they did not meet our inclusion criteria. Among the remaining controlled trials, a further 17 trials were not randomized, leaving 61 RCTs to be included in this review of which 55 full articles contained information regarding participants with hand eczema, of which seven RCTs included conventional systemic interventions (please see Fig. 1.). Two publications were articles that were based on the same group of participants from the same study (Granlund 1996).

44 chapter 2

Fig. 1. Study flow diagram.

45 Included studies Details of the 55 included studies with a total of 4,619 participants. We included studies published from May 1967 to March 2013 and unpublished data form studies registered in trial registers up to February 2014. The seven studies on systemic interventions included chapter 1,661 participants and were conducted between 1996 and 2013.

2 Design All seven studies used a parallel group design. One parallel group study used a cross-over design (Granlund 1996), but was parallel prior to crossover. None of the seven RCTs included a within-patient design (i.e. having a left-right design, comparing one hand with the other).

Participants A total of 4,619 participants were enrolled. Most studies were relatively small (12 to 158 participants) and sample size calculations were often not stated. A large proportion of the participants were included in two trials (Ruzicka 2004; Ruzicka 2008) with enrolment of 319 and 1032, respectively. The seven studies on systemic interventions included 1,661 participants, varying between 15 (Schmitt 2013) and 1,032 (Ruzicka 2008) participants. The original protocol stipulated diagnosis by a physician. Although only one of the studies identified by us stated this explicitly, all studies were based on participants being outpatients at hospitals. Therefore, we assumed that the diagnosis was established by a physician in all participants. Some studies include a specific subgroup of hand eczema, while others excluded these subgroups, for example, atopic dermatitis and atopic dermatitis on the hands: one study targeted atopic eczema in specific (Schmitt 2013), while others studies excluded participants with characteristics of atopic eczema. Overall children were not included in the study population. Older participants were excluded by a few studies. An upper limit of 65 years of age was used by one study (Agarwal 2013), 70 years of age by two studies (Granlund 1996; Ruzicka 2004) and two studies applied an upper limit of 75 years of age (Ruzicka 2008; Schmitt 2013). The studies included both female and male participants. Pregnant and/or lactating women were excluded in about half of the studies. Overall, participants were in general good health and often systemic diseases such as diabetes, renal or hepatic diseases were excluded.

Setting None of the studies was conducted in a primary care setting. As far as we know, all studies have been conducted in secondary care with outpatients from hospitals. About half of the studies were conducted as multicenter study, usually within the same country. Three studies were international multicenter studies (Bissonnette 2010; Ruzicka 2004; Ruzicka 2008). Although most studies did not declare the country in which the study was conducted, we

46 assumed them to be conducted in the hospitals of the investigators. Based on this assump- tion, the majority of the studies was conducted in the North-America and Europe.

Duration In general, the studies in the Cochrane review were of relatively short duration. The duration chapter of studies on systemic interventions was longer than studies on topical interventions. The treatment episode was less than two months (eight weeks) in two studies (Granlund 2 1996; Thestrup-Pedersen 2001). Five studies had duration of more than four months of active treatment (Agarwal 2013; Bissonnette 2010; Ruzicka 2004; Ruzicka 2008; Schmitt 2013). The study of Granlund 1996 with a cross-over design had an active treatment phase of six weeks for both drugs. The majority of the studies did not include a follow-up period. Only four studies included a follow-up period (Granlund 1996; Ruzicka 2004; Ruzicka 2008; Schmitt 2013). This period varied from a week to several months and existed of scheduled visits or just a single follow-up questionnaire. One study was ended prematurely (Schmitt 2013).

Interventions and Comparisons In most studies an active intervention was compared to no treatment, variants of the same medication, or placebo (or vehicle). There were very few studies comparing two different classes of interventions: one study comparing cyclosporine with a topical steroid (Granlund 1996), and one study compared oral cyclosporine to oral alitretinoin (Schmitt 2013). We organized the remaining trials into the categories as described below and provide details of the various dose regimens. Full details of interventions and comparisons for each included study are given in the Characteristics of included studies.

I Skin protection measures, including gloves Not included in this thesis.

II Topical treatments Studies on bland emollients, corticosteroid creams or ointments, coal tar and derivates, irradiation with UV light, irradiation with X-rays (ionizing radiation), topical calcineurin inhibitors and other topical interventions are not printed in this thesis, but are included in the full Co- chrane review.

III Systemic treatments A Oral corticosteroids We identified no randomized controlled trials addressing oral corticosteroids.

47 B Immunosuppressants There were two publications on cyclosporine (Granlund 1996), but these studies were based on the same trial that had three phases. Oral cyclosporine 3 mg/kg/day and placebo cream for 6 weeks was compared with topical betamethasone dipropionate 0.05% cream and chapter placebo capsules identical to cyclosporine. This was a cross-over trial, where participants who failed to respond to their intervention in phase I, were crossed over to the alternative 2 intervention. The use of own emollients was allowed in both groups. One study (Agarwal 2013) investigated a low dose of azathioprine combined with topical clobetasol 0.05% cream compared to topical clobetasol 0.05% cream alone during 24 weeks. One study compared the effect of oral cyclosporine 2.7 to 4.0 mg/ kg to alitretinoin 30 mg a day during 24 weeks (Schmitt 2013). Please see Table 8.

Study Comparison Good/excellent control Comments Granlund 1996 Oral cyclosporine versus topical Both patient-rated and betamethasone investigator rated: higher rate in cyclosporine group but no significance tests provided Agarwal 2013 Oral azathioprine and clobetasol In the control group 39.13% Investigator rated: statisti- propionate 0.05% cream twice achieved complete response cally significant difference daily versus topical clobetasol (75%reduction in severity propionate0.05% cream twice HECSI) After 24 weeks 91.1% daily had complete response in the azathioprine group RR 2.28 (95% CI 1.58 to 3.30, P<0.001) Schmitt 2013 Oral cyclosporine versus alit- Study terminated prema- retinoin turely and only included 15 participants

Table 8. Overview of studies on immunosuppressants

C Oral retinoids We identified five studies evaluating oral retinoids (Bissonnette 2010; Ruzicka 2004; Ruzicka 2008; Schmitt 2013; Thestrup-Pedersen 2001). Ruzicka 2004 had previously been presented in part as a conference abstract (Larsen 2003). In three studies the effect of 10 mg oral alitretinoin was investigated: Bissonnette 2010; Ruzicka 2004 and Ruzicka 2008 Two large multicenter studies (Ruzicka 2004; Ruzicka 2008) compared a total of four different oral doses of a novel retinoid (alitretinoin) with placebo capsules. In Ruzicka 2004 three groups, each receiving respectively, 10, 20 or 40 mg per day, were compared with a placebo group. The trial lasted 12 weeks. The other study (Ruzicka 2008, also known as Benefit of Alitretinoin in Chronic Hand eczema or BACH study) compared two groups, receiving respectively oral alitretinoin 10 or 30 mg once daily, with placebo up to 24 weeks. In both studies, participants

48 were allowed to use a standard emollient. Schmitt 2013 aimed to compare the effectiveness and safety of 30 mg alitretinoin to cyclosporine during 24 weeks in 78 participants. In the study by Bissonnette 2010, 117 participants suffering from chronic hand eczema were included who were successfully treated with alitretinoin in an earlier study (Ruzicka 2008) and who had relapsed within the 24-week observation period after treatment. These 117 relapsed chapter participants were randomized to receive their previous treatment or placebo in a 2:1 ratio. 73 participants were included who had been treated with 30 mg alitretinoin in the previous 2 BACH study (Ruzicka 2008). No other topical or systemic medication for hand eczema was allowed during the treatment period. Dose reductions of study medication were not allowed. In Thestrup-Pedersen 2001 acitretin orally 30 mg daily for eight weeks was compared to placebo capsules for eight weeks. Both groups were allowed to use topical emollients. For an overview see Table 9.

Study Comparison Good/excellent control Comments Thestrup- Oral acitretin versus No data Pedersen 2001 placebo Ruzicka 2004 Oral alitretinoin (20 40 mg: Participant rated: RR Alitretinoin significantly better for all and 40 mg) versus 3.51 (1.80-6.82) Investigator: doses. placebo RR 1.97 (1.3-3.0) 20 mg: Participant rated: RR 2.74 (1.37-5.46) Investigator: RR 1.49 (0.94-2.34) Ruzicka 2008 Oral alitretinoin versus 30 mg: Participant rated: RR Alitretinoin significantly better placebo 2.87 (2.08-3.97) Investigator: RR 2.64 (1.87, 3.72) Ruzicka 2004 / Oral alitretinoin versus 10 mg: Participant rated: RR Alitretinoin significantly better. Ruzicka 2008 placebo 1.73 (1.25-2.40) Investigator: RR 1.63 (1.23-2.16) Bissonnette Re-treatment with oral 30 mg: Participant rated: 30 mg: Participant rated: Alitretinoin 2010 alitretinoin (30 and 76% Investigator rated: RR better than placebo, however, percent- 10 mg) versus placebo 9.55 (2.51, 36.27) age of participants with good/ excellent 10 mg: Participant rated: control in the group treated with placebo 38% Investigator rated: RR not stated. Investigator rated: Alitretinoin 4.76 (0.70, 32.25) significantly better 10 mg: Participant rated: Alitretinoin 10 mg better than placebo, however, percentage of participants with good/ excellent control in the group treated with placebo not stated. Investigator rated: Alitretinoin not significantly better

Table 9. Overview of studies with oral retinoids

49 D Other oral interventions There were six studies on rather unconventional oral interventions. The studies included one study on triethylenetetramine (Burrows 1986), two on disulphiram (Kaaber 1983; Sharma 2006), a low-nickel diet versus oral treatment with disodiumcromoglycate (Pigatto 1990), oral chapter ranitidine (Veien 1995) and evening primrose oil (Whitaker 1996). These will not be presented in 2 this thesis, but are available in the full Cochrane review. Outcomes The included 55 RCTs reported diverse outcomes. About half of the studies included our primary outcome good/ excellent control either participant- or investigator-rated. Almost all studies reported our primary outcome “adverse events”. There was substantial heterogeneity between the studies in terms of interventions, outcome measures and timing of the outcome assessments. Of the seven studies on systemic therapy, five included primary outcome good/ excellent control either participant- or investigator-rated (Bissonnette 2010; Granlund 1996; Ruzicka 2004; Ruzicka 2008; Schmitt 2013). All seven studies included adverse events as outcome.

The majority of the studies used a scale to score the (change in) severity of hand eczema or the rate of clearance. However; many different scoring systems were composed scoring different items on different scales and some did, while others did not include the affected area. Most scoring systems were unnamed, non-validated, self-created and combined objective and subjective scores, in these cases we provided a narrative account of the study results and did not attempt quantitative analyses. The Hand Eczema Severity Index (HECSI) is an assessment of the clinical severity of hand eczema which included the extent and the severity of hand eczema. The HECSI score ranges from 0 to 360. It is a validated scoring system with excellent agreement existed for both inter- and intraobserver reliability.21 Two studies used the HECSI (Agarwal 2013; Schmitt 2013) and three ongoing studies included this outcome parameter (EUCTR2005-005793-75-DE; IRCT201112018263N1; NCT01950494). The Dyshydrotic Eczema Area and Severity Index (DASI) is an assessment of severity combining objective (vesicles, erythema and desquamation) and subjective (itch) evaluations. The DASI was first described by Odia.22 The DASI is regularly used, but this instrument is not validated. None of the studies on systemic treatment used this outcome. The Total Lesion Symptom Score (TLSS) is the sum of seven items (erythema, edema, vesicles, desquamation, hyperkeratosis, fissures and pruritus/ pain) scored on a 4-point scale (0 = absent, 1 = mild, 2 = moderate, 3 = severe). The TLSS was used in Ruzicka 2004 and afterwards slightly modified (modified Total Lesion Symptom Score, mTLSS) in the studies of Ruzicka 2008 and Bissonnette 2010 in which the seven items were: erythema, oedema, vesiculation, scaling, lichenification/ hyperkeratosis, fissures and pruritus/pain. The mTLSS

50 relates to the Physician Global Assessment and a photographic guide is developed to train observers. The mTLSS is also used in two ongoing studies (NCT00817063; ISRCTN18213910). Scoring systems derived and validated for atopic dermatitis were used by some studies. The study of Schmitt 2013 included the validated SCORing Atopic Dermatitis (SCORAD) next to the HECSI for participants with atopic hand eczema. Investigator and Physician Global chapter assessments (PGA and IGA) or variants of this scoring system were used in different studies, on a four or five point scale for both hands overall (Bissonnette 2010; Ruzicka 2004; Ruzicka 2 2008; Schmitt 201). PGA scores have been extensively studied and compared to for example the HECSI and HEAS.23,24 Itch was scored as subjective parameter in the majority of the studies (Agarwal 2013; Granlund 1996; Ruzicka 2004; Ruzicka 2008; Thestrup-Pedersen 2001). Three studies included quality of life as outcome parameter, and especially more recent studies and studies in trial registers included quality of life. The extensively studied and validated Dermatology Life Quality Index (DLQI) was used in one study (Ruzicka 2004). Other quality of life questionnaires used were the Eczema Disability Index (EDI) (Granlund 1996) and the skindex-17 (Schmitt 2013). Economic losses such as sick-days of out-of-pocket expenses were rarely registered as outcome parameter and were not included as outcome parameter before 2004. The study of Schmitt 2013 contained extensive cost-effectiveness analyses by means of the EQ-5D and in addition out-of-pocket expenses were registered. One study included the influence of hand eczema on work impairment with the Work Productivity and Activity Impairment Questionnaire (WPAI) (Schmitt 2013).

Excluded studies The excluded studies comprised 32 studies, which were excluded for different reasons such as: • Study on ’slightly irritated hands’ in employees which we did not accept as being hand eczema (Berndt 2001); • Quasi-randomized studies or unclear whether the study was randomized (Aertgeerts 1985; Güler Özden 2004; HogenEsch 1998; Petering 2004; Rosén 1987; Zimmerman 1967) Excluded studies are not presented in this thesis.

Ongoing studies The search yielded the 14 ongoing studies. Ongoing studies are not presented in this thesis.

Risk of bias in included studies Many studies were at high or unclear risk of bias in one or more components of trial design. One study was assessed as low-risk of bias in all components of trial design (Schmitt 2013).

51 Two studies had deficiencies in only two components of trial design (Ruzicka 2004; Ruzicka 2008). Further information can be found in the risk of bias tables for each included study and Fig. 2 and Fig. 3. chapter 2

Fig. 2a. Risk of bias graph: review authors’ judgments about each risk of bias item presented as percentages across all included studies.

Fig. 2b. Risk of bias graph for studies on systemic interventions: review authors’ judgments about each risk of bias item presented as percentages across all included studies.

52 chapter 2

Fig. 3. Risk of bias summary: review authors’ judgments about each risk of bias item for each included study.

Allocation Randomization procedure There were uncertainties in the method of randomization with uncertainties about allocation concealment. The randomization procedure was unclear in three studies. In three studies we judged the randomization procedure as adequate based on the article (Ruzicka 2004; Ruzicka 2008; Schmitt 2013). In three studies we were unable to base the judgment on the article, but personal communication with the authors clarified that the randomization pro-cedure was adequate (Agarwal 2013; Bissonnette 2010; Thestrup-Pedersen 2001). In total we judged the randomization procedure as adequate in six studies.

53 Concealment of allocation Of the above mentioned studies with an appropriate randomization procedure, the concealment of allocation was adequate too (Agarwal 2013; Ruzicka 2004; Ruzicka 2008; Schmitt 2013; Thestrup-Pedersen 2001). In one study the concealment of allocation was clear; chapter however the randomization procedure was unclear (Granlund 1996). In total in six studies the 2 method to conceal the allocation was judged as adequate. Blinding Performance bias Performance bias refers to systematic differences between groups in the care that is provided, or in exposure to factors other than the interventions of interest. After enrolment into the study, blinding (or masking) of study participants and personnel may reduce the risk that knowledge of which intervention was received, rather than the intervention itself, affects outcomes. Effective blinding can also ensure that the compared groups receive a similar amount of attention, ancillary treatment and diagnostic investigations. In four studies both the participants and the staff were blinded in an adequate manner (Bissonnette 2010; Ruzicka 2004; Ruzicka 2008; Schmitt 2013). One study was only patient-blinded (Granlund 1996) and in one study no blinding of participants was attempted (Agarwal 2013).

Detection bias Detection bias refers to systematic differences between groups in how outcomes are determined and can be minimized by blinding of the outcome assessors (the observer). The observer was blinded in an adequate manner in six studies (Agarwal 2013; Bissonnette 2010; Granlund 1996; Ruzicka 2004; Ruzicka 2008; Schmitt 2013). In the remaining of the studies it was unclear whether the observer was truly blinded.

Incomplete outcome data Loss to follow up and intention-to-treat (ITT) analysis In one study there were no drop-outs (Thestrup-Pedersen 2001). Despite loss of participants during follow up, data were analyzed according to the intention-to-treat analysis principle in four studies (Bissonnette 2010; Granlund 1996; Ruzicka 2004; Ruzicka 2008). The study of Schmitt 2013 included an intention-to-treat analysis, but only included 14 of the 78 planned participants due to early termination. In the remaining study at least 80% of the participants was followed up and included in the analyses (Agarwal 2013).

Selective reporting Many, especially older studies did not register prior to commencement of the trial, so that the correspondence between actually reported outcomes and outcomes intended to be

54 reported could not be assessed for most included studies. We found trial registers for four studies (Bissonnette 2010; Ruzicka 2004; Ruzicka 2008; Schmitt 2013). Discrepancies were found in the study of Ruzicka 2008 with regards to additional or missing outcome parameters. No major discrepancies between protocol and report were found in three studies (Bissonnette 2010; Schmitt 2013). For the other studies the discrepancies were judged chapter between the material& methods section and the result section. No major discrepancies were found in the majority of the studies (Agarwal 2013; Granlund 1996; Thestrup-Pedersen 2001). 2

Other potential sources of bias Baseline comparison for severity of the disease We assumed that this component was not relevant for within-patient studies. For the parallel group studies disease severity at baseline was reported to not significantly differ between the groups in three studies (Ruzicka 2004; Ruzicka 2008; Schmitt 2013). In one study dis- similarities in disease severity or patient characteristics were observed at baseline (Granlund 1996). Disease severity and other baseline parameters were provided in one study, but the authors did not provide a statistical significance test (Bissonnette 2010). In two studies it was unclear whether a comparison had been made and whether the groups were sufficiently comparable (Agarwal 2013; Thestrup-Pedersen 2001).

Certainty that participants had hand eczema Only one study (Ruzicka 2008) clearly stated a specific procedure to confirm whether the participants had hand eczema. The original protocol stipulated diagnosis by a physician. However, none of the studies identified by us stated this criterium explicitly. Since all studies were based on participants attending a specialized clinic, we therefore assumed that in all participants the diagnosis was established by a physician. In a number of studies, the entry criteria stipulated additional criteria delineating subcategories of hand-eczema, however without describing the diagnostic procedures in detail. A clear description of the disease was given in five studies (Bissonnette 2010; Granlund 1996; Ruzicka 2004; Ruzicka 2008; Thestrup-Pedersen 2001).

Rationale of the sample size Only three studies (Ruzicka 2004; Ruzicka 2008; Schmitt 2013) mentioned how the number of participants to be enrolled was worked out and included a clear sample size rationale. In the remaining studies it was unclear how the sample size was calculated, except for the study of Bissonnette 2010 which included all the relapsed participants of the earlier study. Overall the studies were small, though the studies on systemic interventions included more participants than studies on topical interventions. Two studies included less than 50 participants (Granlund 1996; Thestrup-Pedersen 2001). Between 100 and 500 participants were included

55 in three studies (Agarwal 2013; Bissonnette 2010; Ruzicka 2004). One study included more than 500 participants (Ruzicka 2008). The study of Schmitt 2013 aimed to include 78 participants based on a sample size calculation, however the study was ended prematurely and only included 15 participants. chapter Funding 2 For many older studies it was unclear who funded the study. More recent studies did often declare the funding for the study or clearly stated their relationship with pharmaceutical companies. In total four studies were funded by pharmaceutical industries or (co-) authored by employees of pharmaceutical companies and thus at high risk of bias (Bissonnette 2010; Granlund 1996; Ruzicka 2004; Ruzicka 2008). One study was sponsored by governmental organizations, universities or hospitals and was judged as low risk of bias (Schmitt 2013).

Defnition of a major outcome measure Five studies (Bissonnette 2010; Granlund 1996 (publication 1997); Ruzicka 2004; Ruzicka 2008; Schmitt 2013) stated a major/ primary outcome measure. The remaining studies did not clearly state a primary outcome.

Reporting of inclusion and exclusion criteria Of the included studies, all stated that participants with hand eczema (sometimes specifying a particular subtype) were enrolled. Some studies specified more extensive inclusion criteria, with regard to age, gender or disease severity. The majority of the studies also listed exclusion criteria and some studies even gave an extensive list of exclusion criteria (Bissonnette 2010; Ruzicka 2004; Ruzicka 2008; Schmitt 2013).

Duration of study specifed All studies clarified the duration of the study or the length of follow-up.

Publication bias Funnel plots were not a feasible option for the risk of reporting bias, due to the heterogeneity of the included studies. We did try to contact the authors of trials registered in the trial registers of which the status was “finished” or of which the status was unclear. Especially with regards to studies on emollients, we did not receive a response from the investigators. Some authors responded that the study results were marginal and would not be published; others were not at liberty to disclose the results due to confidentiality issues. For one study on alitretinoin, the authors were not at liberty to disclose the results due to confidentiality issues and they referred us to the pharmaceutical company, however, they remained unresponsive after multiple attempts (NCT00817063). Due to the search in clinical trial

56 registers, we located one additional study which was not (yet) published (Schmitt 2013).

Agarwal 2013 Methods Parallel group, randomized controlled study. This study was probably carried out in secondary care setting as a single-centre study in a department of Dermatology in India. Observers were blinded, chapter participants were not blinded. Intention-to-treat analysis was not conducte. Participants 108 participants included with clinically diagnosed hand eczema, 91 completed the study. 2 Drop-outs: 17 Inclusion criteria of the trial • Clinically diagnosis of hand eczema with a duration of more than 6 months Exclusion criteria of the trial • Pregnancy • Lactating mothers • Age less than 18 years or more than 65 years • Any associated systemic disease (diabetes, hypertension, thyroid disorders, any renal or liver disease, malignancy, etc) • Hypersensitivity to azathioprine Interventions Intervention A: Topical clobetasol propionate 0.05% cream twice daily with oral azathioprine 50 mg daily in 46 participants for 24 weeks. Control intervention B: Topical clobetasol propionate 0.05% cream twice daily alone in 45 participants for 24 weeks. Participants were instructed to use the topical clobetasol intermittently and stop application whenever the signs and symptoms disappeared and restart when the complaints returned. Outcomes Primary outcomes of the trial Not defined. Other outcomes 1. Reduction in severity, investigator-rated scoring measured by the Hand eczema scoring index (HECSI) at 2, 4, 8, 12 and 24 weeks 2. Reduction of severity of itch, patient-rated measured on a numerical scale from 0-10 3. Number of exacerbations 4. Adverse events Notes The total amount of corticosteroids used was not registered. Authors were contacted on 27 February 2014 by email and responded 1 March 2014.

Bias Risk of Bias Support for judgment Random sequence Low “Patients were randomized into 2 groups using block randomization”. generation (selection Unclear from the article how these blocks were generated, however contact bias) with the authors clarified that a valid computer generated table was used. Allocation conceal- Low No information on how allocation was concealed from participants and ment (selection bias) investigators is provided in the article, but personal communication revealed that randomization was done by a third person.

57 Blinding of partici- High Participants were not blinded, since they had to buy there own study drugs. pants and personnel (performance bias) Blinding of outcome Low Authors state that the study was observer blinded, but no details as to how chapter assessment (detec- this was achieved are given in the article. Contact with authors clarified tion bias) that the observers were independent and not involved in the treatment or 2 dispensation of the study drugs. Incomplete outcome 91 of the 108 (84.3%) included participants were analyzed. No intention-to- data (attrition bias) treat analysis. Drop-outs were not evenly distributed between the 2 groups. However, since at least 80% was analyzed, this was considered as a low-risk of bias. Selective reporting Low No trial registration found. However, all mentioned outcomes are described (reporting bias) in the result section. Other bias Unclear No sample size rationale provided. No baseline comparisons conducted or reported. Diagnostic certainty: yes. Funding Unclear Not stated.

Bissonnette 2010 Methods Parallel group, placebo-controlled, randomized study with participants who initially responded in a previous BACH study (Ruzicka 2008) but relapsed in the observational period. Participants were retreated with the same dose as they had received in the previous study or they were treated with placebo This study was carried out in the secondary care setting. It was a multicenter study, conducted in Canada, France and Germany. The study was blinded and an intention-to-treat analysis was conducted. Participants 117 participants with chronic hand eczema who relapsed after they were successfully treated with alitretinoin 30 mg, 10 mg or placebo. Drop-outs: 24. Inclusion criteria of the trial • Participants who relapsed after successful treatment with alitretinoin or placebo in a previous trial (Ruzicka 2008) Exclusion criteria of the trial • Well defined Interventions Intervention A. Alitretinoin 10 mg once daily for 12-24 weeks in 21 participants B. Alitretinoin 30 mg once daily for 12-24 weeks in 49 participants Control intervention C. Placebo for 12-24 weeks in 47 participants Participants, who were initially treated with placebo in the first trial, were treated with placebo again. No other topical or systemic therapy for hand eczema was allowed. Outcomes Primary outcomes of the trial 1. Physician’s Global Assessment (PGA): whereby physician global assessment is categorized in clear, almost clear, mild, moderate, and severe. Responders are defined as clear or almost clear at week 12 or last evaluation

58 Other outcomes 1. Patient’s Global Assessment (PaGA) 2. Modified Total Lesion Symptom Score (mTLSS) 3. Extent of disease 4. Time to response chapter 5. Adverse events Notes Study included participants who relapsed after successful treatment in a previous study Ruzicka 2 2008. No other active treatment as comparator. Analysis of efficacy based on intention-to-treat principle. Study included a safety assessment by careful medical and laboratory monitoring. Authors were contacted and referred us for further information to GSK, who provided additional information.

Bias Risk of Bias Support for judgment Random sequence Low ”Patients were randomized to the same dose they received in the BACH generation (selection study or to placebo“. No further details given. In personal communication bias) the authors clarified: ”The study was a follow on from the BACH study BAP00089 which was a randomized double blind placebo controlled study of subjects in a 2:2:1 randomization of treatment to alitretinoin 30 mg, alitretinoin 10 mg or placebo respectively. Patients who responded in study BAP00089 and relapsed during the post treatment observation period were assigned to the same dose they had received or to placebo in a 2:1 ratio. Responding patients who had received placebo in study BAP00089 were assigned to continue receiving placebo. Each patient was assigned a coded allocation of study drug containing either placebo or a dosage of active drug. The randomization was computer generated.” Allocation conceal- Unclear No details about how allocation was concealed from the participants and ment (selection bias) investigators. In personal communication the authors stated: “It is unclear how this knowledge was imparted but it is clear from the proto-cols that those subjects who had received placebo in the original trial BAP00089, and who had been successfully treated but had subsequently relapsed would upon entering this study be given placebo again as it was considered unethical to expose them unnecessarily to drug.” Blinding of partici- Low Authors state to have conducted double blind study: “The investigator, pants and personnel sponsor and all participants remained blinded throughout the course of the (performance bias) BACH and retreatment studies.” No statements have been made how this was conducted. ”The placebo and active drug were indistinguishableand packaged in the same way “Identical packages is a sufficient form of blinding, however it is unclear whether site staff was also blinded, since the study was a follow-up study in which the majority of the participants received the same treatment as in the previous study Personal communication clarified: “A list of treatment assignments was sealed and kept in a central repository by the Biometrics Departmentand by the Drug Safety Department. No open key to the code was available at the Study Center, or to monitors or members of the project team.”

59 Blinding of outcome Low Investigators were blinded throughout the study; however no details assessment (detec- regarding blinding were given. The observers might have been the same tion bias) as in the study of Ruzicka 2008. Personal communication The investigator had access to coded, sealed envelopes for each patient to be used in an emergency that would have required knowledge of the study medication chapter to manage the emergency. If the investigator wished to know the identity of the treatment given to study subjects for any other purpose, this request 2 was first to be discussed with Basilea Pharmaceutica.” Incomplete outcome Low Intention-to-treat analysis. data (attrition bias) Selective reporting Low The trial was registered at clinicialtrials.gov (NCT00124436 and BAP00091 (reporting bias) and EUCTR2004-000432-85-HU). No major changes in primary or secondary outcomes. The only difference is that the trial registration states enrolment of 300 participants, while the actual trial only included 117 participants, but this was done to be able to include all subjects who relapsed from the BACH study. Other bias Unclear No sample size rationale provided (“The sample size was not pres-pecified, and all relapsing patients from the BACH trial were eligible for trial screening.”). Baseline comparison of disease severity in table, but no significance of difference tests provided, personal communication clarified that there were no statistical significant differences. Diagnostic certainty: Yes. Funding High The study was supported and funded by Basilea Pharmaceutica Inter- national Ltd, Basel, Switzerland, the manufacturer of alitretinoin The authors were either investigator in Basilea clinical trials, consultant or employee of Basilea Pharmaceutica International Ltd.

Granlund 1996 Methods Randomized parallel group design, with a partial cross-over design in 2nd phase. Randomization procedure unclear. The study was carried out in secondary care setting. It was a mono-centre study, conducted in Finland. It was a double-blind study (although it was unclear how the observers were blinded) with an intention-to-treat analysis. Participants 41 participants with hand eczema, continuously for 6 months, significant disability, inadequate response to conventional treatment, confirmation by histopathology. Drop-outs: 6 in the first phase, 1 in second phase. Inclusion criteria of the trial • 18-70 years old • Continuously hand eczema for at least 6 months • Causing significant disability Inadequate response to conventional treatment Exclusion criteria of the trial • Other skin disorders • Treatment with systemic corticosteroids within 4 weeks or topical steroids or UV radiation within 2 weeks before the study • Standard exclusion criteria for patients undergoing cyclosporine treatment

60 Interventions Intervention A. Oral cyclosporine 3 mg/kg/day and placebo cream for 6 weeks in 17/20 participants B. Topical betamethasone dipropionate 0.05% cream and placebo capsules identical to cyclosporine in 19/21participants At week 6, cross-over of those who had treatment failure in the first 6 week phase: 8 participants chapter switched to betamethasone, and 6 to cyclosporine. In third phase a 24-week follow-up period without intervention. The total maximum duration of the study was 36 weeks. The use of own emollients was allowed in both groups. 2 Outcomes Granlund 1996: Primary outcomes of the trial Not defined Other outcomes 1. Participant-rated overall assessment of efficacy (1 = very good, 2 = good, 3 = moderate, 4 = slight, 5 = none) 2. Observer-rated overall assessment of efficacy (1 = very good, 2 = good, 3 = moderate, 4 = slight, 5 = none) 3. Observer-rated disease activity score: grading 0 to 3 (0 = none, 1 = mild, 2 = moderate, 3 = severe) on erythema, scaling, infiltration, excoriation, crusting, vesicles 4. Observer-rated extent of disease 5. Use of emollients 6. Participant-rated itch and sleep disturbances for the final 2 weeks on a VAS 7. Treatment success, defined as decrease in disease activity score (see 1) above) to <5 0% of baseline score 8. Adverse events Granlund 1997: Primary outcomes of the trial 1. Quality of life assessed by the Eczema Disability Index (EDI) at week 6 and week 12 Other outcomes 1. Observer-rated disease activity score: grading 0 to 3 (0 = none, 1 = mild, 2 = moderate, 3 = severe) on erythema, scaling, infiltration, excoriation, crusting, vesicles for both hands 2. Observer-rated extent of the disease 3. Use of emollients 4. Participant-rated itch and sleep disturbances for the final 2 weeks on a VAS Notes Granlund 1996: Study had 3 phases, which were partially overlapping. The 2nd phase dealt with participants who had treatment failure in phase 1. In this 2nd phase, participants were switched over to the alternative intervention. The 3rd phase includes only participants who had treatment success in phase 1. Outcome assessment based on intention-to-treat analysis. This review deals only with phase 1 and phase 3. Granlund 1997: Paper is based on the same trial (same participants) as Granlund 1996, but only dealing with phase 1 and 2. The study had 3 phases, which were partially overlapping. The 2nd phase dealt with participants who had treatment failure in phase 1. In this 2nd phase, participants were switched over to the alternative intervention. The 3rd phase includes only participants who had treatment success in phase 1. In this review only the results of the 1st phase will be discussed. Outcome assessment based on intention-to-treat analysis

61 Bias Risk of Bias Support for judgment Random sequence Unclear “Patients were given numbers 1-41 in consecutive order, which had been generation (selection reassigned to treatment with...”. However, neither article clarifies how this bias) reassignment was done. chapter Allocation conceal- Low The study used identical placebos for topical and oral treatment and ment (selection bias) participants were given consecutive numbers. The codes were not opened 2 until all patients had finished. Blinding of partici- Low Authors stated double-blinded design. Sufficient information about pants and personnel provided how patient blinding was achieved and used identical tubes and (performance bias) placebos: “Soft gelatine capsules containing 25, 50 or 100 mg and identical placebo capsules were supplied by ...” “Identical 100 tubes were used for the creams.” Blinding of outcome Low Double-blind design, since identical placebos were used and the random- assessment (detec- ization code was not broken before the end of the study, it is unlikely that tion bias) the assessors were aware of the treatment group. Incomplete outcome Low Intention-to-treat analysis. data (attrition bias) Selective reporting Low No trial registration found, however all the relevant outcomes are (reporting bias) addressed in the material and methods of the two articles and are depicted in graphs in the result section. Other bias Unclear No sample size rationale provided. Baseline comparisons revealed a significant difference between the groups in terms of prestudy antibioics treatment. Diagnostic certainty: Yes. Funding High The study was supported by Sandoz Pharmaceuticals, Switzerland (manufacturer of the study drug) and Finland and by a grant from Finska Läkaresällskapet, Finland.

Ruzicka 2004 Methods Randomized, parallel group design of 1 placebo group and 3 treatment groups of different doses of same (oral) medicament This study was carried out in secondary care setting. It was a multicenter study involving 43 clinics in Belgium, Denmark, Finland, France, Germany, Holland, Hungary, Poland, Switzerland and the United Kingdom Participants were blinded, it was unclear whether observers were blinded. Intention-to-treat analysis was carried out. Participants 319 participants (235 male, 84 female) with moderate or severe chronic hand dermatitis of at least 3 months duration and refractory to standard therapy. All types of hand dermatitis. Drop-outs: 75 Inclusion criteria of the trial • Moderate or severe hand eczema for at least 3 months • Refractory to standard therapy • Aged between 18 and 70 years Exclusion criteria of the trial • Well defined

62 Interventions Intervention A. Oral alitretinoin 10 mg daily for 12 weeks in 62/80 participants B. Oral alitretinoin 20 mg/day in 67/80 participants for 12 weeks C. Oral alitretinoin 40 mg/day in 63/81 participants for 12 weeks Control intervention chapter D. Placebo capsules in 62/78 participants for 12 weeks Standard emollients were allowed in all treatment groups. The responders were followed-up for 3 months. 2 Outcomes Primary outcomes of the trial 1. Responders according to physician global assessment of overall severity, whereby physician global assessment is categorized in clear, almost clear, mild, moderate, severe. Responders are defined as clear or almost clear at week 12 or last evaluation Secondary outcomes of the trial 1. Observer-rated total lesion symptom score: sum of scores (0 = absent, 1 = mild, 2 = moderate, 3 = severe) for erythema, oedema, vesicles, desquamation, hyperkeratosis, fissures, pruritus/ pain 2. Participant-rated global assessment: clearing or almost clearing (>90% clearing of signs and symptoms compared with baseline), marked improvement (>75%), moderate improvement (>50%), mild improvement (>25%), no change, worsening. 3. Observer-rated extent of the disease: total percentage involvement of palms and dorsum of both hands 4. Dermatological life quality index (DLQI) 5. Adverse events Notes No other active treatment as comparator. More males enrolled because of exclusion of women of childbearing potential. Study included a safety assessment by careful medical and laboratory monitoring. Analysis of efficacy based on intention-to-treat principle. Of the 127 responders 117 were followed up for another 12 weeks after end of treatment; only summary data about this extra follow-up are presented.

Bias Risk of Bias Support for judgment Random sequence Low Computer-generated randomization codes. generation (selection bias) Allocation conceal- Low Randomization codes were provided by the study sponsor and ment (selection bias) incorporated into double-blind coded drug packaging. The site staff had no direct access to the randomization codes. Blinding of partici- Low Patient blinding: “Placebo and active drug (as soft gelatin capsules) and pants and personnel packaging were indistinguishable.” The double-blind coded packagings (performance bias) were provided by the sponsor, which blinded the site personnel and the participant sufficient.

63 Blinding of outcome Low Observers were blinded during the study, since the identical looking pack- assessment (detec- age of study drugs were provided by a third party. One might argue that tion bias) the observer could have guessed the treatment group due to headache and dry mucosa; however this was also seen in the control group and therefore not conclusive. The trial was designed in such a way to minimize chapter the risk of bias and we agree that this could not have been done in a better 2 way. Incomplete outcome Low Intention-to-treat analysis data (attrition bias) Selective reporting Low No trial register found. No major discrepancies between methods and (reporting bias) result section. Other bias Low Sample size rationale is provided. Baseline comparisons: No significant difference between the groups in demographic or disease characteristics. Diagnostic certainty: Yes. Funding High The study was supported and funded by Basilea Pharmaceutica Ltd, Basel, Switzerland, manufacturer of the study drug Some of the authors were employee, received grants or had received consultancy fees from Basilea Pharmaceutica.

Ruzicka 2008 Methods Randomized, parallel group design of 1 placebo group and 2 treatment groups of different doses of same (oral) medicament This study was carried out in a secondary care setting. It was a multicenter study in 111 clinics in Europe and Canada The study was double-blind and intention-to-treat analysis was carried out. Participants 1,032 participants (582 male, 450 female) with severe chronic hand dermatitis of at least 6 months duration and refractory to standard therapy. All types of hand dermatitis. Drop-outs: 273 Inclusion criteria of the trial • Severe chronic hand eczema refractory to standard therapy • Between18 and 75 years of age Exclusion criteria of the trial • Well defined Interventions Intervention A. Oral alitretinoin 10 mg once daily for 12 or 24 weeks (depending on moment of response according to the PGA) in 319/418 participants B. Oral alitretinoin 30 mg/day in 303/409 participants for 12 or 24 weeks Control intervention A. Placebo capsules in 137/205 participants for 12 or 24 weeks Standard emollient in all treatment groups. All participants were followed-up for 4 weeks and responders were observed for relapses for 24 weeks after the end of treatment.

64 Outcomes Primary outcomes of the trial 1. Responders according to physician global assessment of overall severity, whereby physician global assessment is categorized in clear, almost clear, mild, moderate, severe. Responders are defined as clear or almost clear at week 12 or last evaluation. Other outcomes chapter 1. Time to response 2. Partial response (PGA assessment of clear, almost clear, or mild) 3. Observer-rated modified total lesion symptom score: sum of scores (0 = absent, 1= mild, 2 2 = moderate, 3 = severe) for erythema, oedema, vesicles, desquamation, hyperkeratosis, fissures, pruritus/pain 4. Participant-rated global assessment: clearing or almost clearing (>90% clearing of signs and symptoms compared with baseline), marked improvement (>75%), moderate improvement (>50%), mild improvement (>25%), no change, worsening 5. Time to relapse 6. Observer-rated extent of the disease: total percentage involvement of palms and dorsum of both hands 7. Adverse events Notes No other active treatment as comparator. Analysis of efficacy based on intention-to-treat principle. Study included a safety assessment by careful medical and laboratory monitoring. More males enrolled because of exclusion of women of childbearing potential.

Bias Risk of Bias Support for judgment Random sequence Low Computer-generated randomization codes. generation (selection bias) Allocation conceal- Low Codes provided by study sponsor. Sequentially numbered packages of the ment (selection bias) different treatment modalities of identical appearance were used. Blinding of partici- Low Double-blinded: “Placebo, active drug and packaging were pants and personnel indistinguishable.” The identical looking packages were provided by the (performance bias) sponsor and the site staff and participants were unaware of the treatment. Blinding of outcome Low Observers were blinded during the study, since the identical looking pack- assessment (detec- age of study drugs were provided by a third party. One might argue that tion bias) the observer could have guessed the treatment group due to headache and dry mucosa; however this was also seen in the control group and therefore not conclusive. The trial was designed in such a way to minimize the risk of bias and we agree that this could not have been done in a better way. Incomplete outcome Low Intention-to-treat analysis. data (attrition bias) Selective reporting Low Trial registration on clinicaltrials.gov (NCT00124475). No differences in (reporting bias) primary outcome, however small discrepancies in other outcomes between trial registration and article.

65 Other bias Low Sample size rationale provided. Baseline comparisons: No significant difference between the groups in demographic or disease characteristics. Diagnostic certainty: Yes. Funding High The author were employee, received grants or consultancy fees from chapter Basilea Pharmaceutica, manufacturer of the study drug.

2 Schmitt 2013 Methods Randomized, parallel group design Participants 78 patients with severe atopic hand dermatitis. Included: 15. Drop-out: 6 Inclusion criteria of the trial • Male and female Age >18 years and ≤ 75 years Body weight 50 to 100 kg • Chronic hand dermatitis (duration >6 months) • Atopic constitution according to Erlanger Atopiescore 1 and/or positive personal history for atopic eczema, allergic rhinitis, allergic asthma and/or elevated serum IgE • Severe hand dermatitis not responding to treatment with potent topical steroids for at least 4 weeks within the past 6 months due to IGA Written informed consent Exclusion criteria of the trial • Well defined Interventions Intervention A. Oral cyclosporine depending on body weight: 50-74.9 kg: daily dosage 200 mg; 75-100 kg: daily dosage 300 mg B. Oral alitretinoin 30 mg once daily Outcomes Primary outcomes of the trial 1. Proportion of patients with complete or almost complete clearance according to the Investigator Global Assessment (IGA) within 24-week active therapy in both groups Secondary outcomes of the trial 1. Time to complete or almost complete clearance according to IGA in both groups 2. Proportion of patients with complete or almost complete clearance according to the Patients Global Assessment (PGA) within 12 weeks and 24 weeks of active therapy 3. Mean relative change in objective disease severity by means of the Hand Eczema Severity Index (HECSI) between baseline and week 4, 8, 12, 16, 20, 24 in both groups 4. Mean relative change in quality of life (Skindex 17) between baseline and week 24 in both groups 5. Cost-effectiveness of the studied treatment options (cost / QALY gained; assessed by means of the (EQ-5D) 6. Mean relative change in work productivity (assessed by means of the work limitations questionnaire (WLQ) in both groups 7. Mean utilization of topical steroids within the follow-up period in both groups 8. Patient satisfaction with treatment in both groups (assessed using a 100mm VAS Scale) 9. Proportion of patients with relapse (≥ 75% of baseline HECSI) within 24-week follow-up after previous complete/almost complete clearance

66 10. In patients with atopic dermatitis on the body: measured percentage of patients with at least 50% improvement in disease severity with the active therapy using the SCORAD. 11. Tolerability and safety in both study groups Notes The study was ended prematurely. Results are not yet published. The author released the preliminary

study results in personal communication. chapter

Bias Risk of Bias Support for judgment 2 Random sequence Low Participants were randomized by means of a computer generated generation (selection randomization table with constant length of blocks and stratification bias) according to body weights (50-74.9 kg versus 75-100 kg) with equal allocation to cyclosporine and alitretinoin. The random sequence generation was done by the KKS Dresden with the trial software MACRO 3. Allocation conceal- Low The authors declare that they were blinded during allocation: the clinical ment (selection bias) trials pharmacist stored the randomization list and prepared and labeled the treatment packs. New participants were allocated by consecutively numbered packs. Blinding of partici- Low Drug dispensation was done by a third party (the pharmacist). Unclear pants and personnel. whether drugs were identical in appearance. (performance bias) Blinding of outcome Low The article claims that the observer had no access to the randomization assessment (detec- list and a third party was used for drug dispensation. Unblinding tion bias) occurred before data analyses. Incomplete outcome Low Intention-to-treat analysis. The studied aimed to include 78 participants data (attrition bias) but was terminated early due to an inability to include the number of participants. Finally, 15 participants were randomized and one withdrew before the study drug was used. The intention-to-treat analysis included all participants that received the study drug (14 of 15 participants). Selective reporting Low The trial was registered under NCT01231854 prior to the start of the trial. (reporting bias) We found no major discrepancies between the trial register and the final study report. Other bias Low Sample size rationale: clearly stated. Baseline comparison: the groups were comparable at baseline with regards to disease severity. Diagnostic certainty: Yes. Funding Low TU Dresden, Germany.

Thestrup-Pedersen 2001 Methods Randomized, parallel group study. This study was conducted in secondary care setting in 4 dermatology departments or clinics in Denmark The study was observer-blinded (because several patients in the active group suffered from dryness of the lips), All included participants were analyzed.

67 Participants 29 participants (21 males, 8 females) with hyperkeratotic eczema on palms, patch-test negative or irrelevant. No drop-outs. Inclusion criteria of the trial • Hand eczema based on clinical diagnosis Exclusion criteria of the trial chapter • Allergic contact dermatitis 2 Interventions Intervention A. Acitretin orally 30 mg daily for 8 weeks in 14/14 participants Control intervention B. Placebo capsules in for 8 weeks15/15 participants Both groups were allowed to use topical emollients. Outcomes Primary outcomes of the trial 1. Not defined Other outcomes 1. Mean observer-rated severity scores (0 = absent, 1 = slight, 2 = moderate, 3 = severe) combined of the signs: hyperkeratosis, fissures, scaling, itch, redness, vesicles at week 4 and week 8 2. Change in biochemical parameters (Hb, hepatic function, Cholesterol, Triglyceride) 3. Adverse events Notes We contacted the author for additional information by letter; however he was unable to respond to all of our questions. Randomization unclear from the paper, but was confirmed after writing to the 1st author. No overall scores presented as outcome. Details of bio-chemical parameters not given A proper between group comparison was not conducted, only within group comparisons with Wilcoxon-rank sum test were conducted.

Bias Risk of Bias Support for judgment Random sequence Low From the article it is unclear whether the study was randomized at all since generation (selection it only states “they were asked to take three 10-mg capsules of acitretin bias) once daily for 8 weeks or identically looking placebo capsules.” Personal communication with the authors clarified that the randomization was done by a third party according to a predefined randomization list. Allocation conceal- Low No details about how allocation was concealed from the participants and ment (selection bias) clinicians. Personal communication clarified that the sponsor shipped four identical boxes to all participating centres which could ad random be dispensed to participants. The investigators were unaware of the contend of the boxes and therefore we judged this as a low risk. Blinding of partici- Low The study contained an identical looking placebo in attempt to blind pants and personnel participants and the randomization and dispensation of drugs was done at (performance bias) a remote site by a third party. Therefore we judged this as adequate blinding of participants. Blinding of outcome Low The authors declared this a single-blind study, since the observers could assessment (detec- have guessed the acitretin group due to side effects of acitretin. tion bias)

68 Incomplete outcome Low None of the participants dropped-out and all participants were included in data (attrition bias) the analyses. Selective reporting Low No trial registration found. However we found no major discrepancies (reporting bias) between the methods and result section.

Other bias Low No sample size rationale provided. Baseline comparisons: Not stated chapter whether there was a significant difference in disease severity between both groups. Diagnostic certainty: Yes. 2 Funding Low Roche A/S, Copenhagen supplied the study drug free of charge, but the investigators did not receive financial support or consultant fee from Roche. We expect that this did not influence the study, however we cannot exclude this, therefore the risk of bias was judged as unclear.

Effects of interventions In the text below, where it has been possible to calculate an effect size, we have reported these with 95%confidence intervals. Where the calculated effect size is statistically significant (P<0.05), we state whether the result favors the intervention or control condition. If only insufficient data was available for these analyses we summed up the available data from the respective study including the stated P. We considered statistical pooling, but the studies were too heterogeneous in terms of design, types of particular treatment compared, assessment of outcome, duration of trial and presentation of data. The only exceptions were two studies, both comparing 10 mg alitretinoin as treatment versus placebo (Ruzicka 2004; Ruzicka 2008). Beside the results of the pooled analysis, heterogeneity statistics are also reported for these two studies.

I Skin protection measures, including gloves Not included in this thesis

II Topical Treatments Not included in this thesis

III Systemic treatments Oral corticosteroids We identified no randomized controlled trials on oral corticosteroids. Oral immunosuppressants: oral cyclosporine versus topical betamethasone dipropionate Primary outcome: percentage of participants with self-rated good/excellent control This was addressed by one study (Granlund 1996). Overall assessment of efficacy rated as good/very good in the cyclosporine group 60%, and 48% in the betamethasone group; the difference was not statistically significant according to the authors (Analysis 14.1).

69 Primary outcome: percentage of participants with investigator-rated good/excellent control Overall assessment of efficacy good/very good in the cyclosporine group was 60%, and 31% in the betamethasone group. The difference was reported not to be statistically chapter significant (Analysis 14.1).

2 Primary outcome: adverse events In the cyclosporine group one participant experienced dizziness, vomiting and facial oedema. “Some kind of adverse event” occurred in 19 of 28 participants on cyclosporine. In the betamethasone group one participant had insomnia. “Some kind of adverse event” in 15 of 27 participants on betamethasone (Table 11).

Secondary outcome: reduction in severity, participant-rated scoring No statistically significant differences in occurrence of itch between the two groups (shown as a graphical presentation in the report of the study). Sleep disturbance was statistically significantly reduced in the group treated with cyclosporine, but not in the group treated with betamethasone-dipropionate cream (graphic presentation).

Secondary outcome: reduction in severity, investigator-rated total disease activity score after 6 weeks of treatment For this outcome we only found one relevant trial (n = 35) (Granlund 1996). There were two papers (Granlund 1996) dealing with different aspects of the same trial: one paper reported on the effect of the intervention on the disease activity, while the other reported on an analysis of the quality of life in the same participants. Treatment success (defined as >50% decrease in disease activity score) in the cyclosporine group in 50% of the participants, and in the betamethasone group in 32%; the difference was not statistically significant according to the report (Analysis 14.1). The mean total decrease in total disease activity score was six in the cyclosporine group and 5.7 in the betamethasone group (mean difference 0.30, 95% CI -2.46 to 3.06, P = 0.83; see our Analysis 14.2). There was no difference in extent of disease (graphic presentation).

Secondary outcome: time until relapse Not stated in this phase I. The cumulative relapse rates in the participants who had successfully responded in phase I were comparable according to a graphic presentation.

Secondary outcome: dose reduction No difference between the groups in emollient use (graphic presentation). In a separate paper (Granlund 1996 study 2 from 1997), quality of life was assessed with the eczema

70 disability index (EDI) at baseline, at the end of phase I, i.e. at week six, and at the end of phase II. At the end of phase II this assessment was restricted to those whose eczema had failed to respond in the preceding first phase. The total EDI score decreased statistically significantly to the same degree in both groups, i.e. according to the abstract from a mean value of 30.5 to 20.9 in the cyclosporine group (mean decrease 10.4 points) and from 27.2 to chapter 18.9 (mean decrease of 8.4 points) in the betamethasone group. Irrespective of the dimen- sion of the EDI (daily activity, school/work, personal relationship, leisure, treatment), the two 2 treatment groups did not statistically significantly differ in EDI scores at the end of the first treatment period (week six) according to the authors (stated P = not statistically significant). Whether this comparison accounted for baseline variation could not be established.

Study Cyclosporine Betamethasone DP Granlund 1996 Patient-rated very good or good efficacy in 60%. Exact Patient-rated very good or good numbers are difficult to reconstruct. No significant differ- efficacy in 48%. ence with betamethasone (P-value not given, Fisher’s exact test or Mann-Whitney U-test) Granlund 1996 Investigator-rated very good or good efficacy in 60%. Exact Investigator-rated very good or numbers difficult to reconstruct. No significant difference good efficacy in 31%. with betamethasone (P-value not given, Fisher’s exact test or Mann- Whitney U-test) Granlund 1996 Investigator-rated treatment success in 50%. Difference Investigator-rated treatment with betamethasone not significant (P = 0.233) success in 32%. Granlund 1996 No difference in mean disease activity score, extent, itch or use of emollients. Only bar-graphs, no exact data.

Analysis 14.1. Primary and secondary outcomes: investigator- and patient-rated good/ excellent control of symptoms after 6 weeks of treatment.

Analysis 14.2. Secondary outcome: investigator-rated reduction in severity in total disease activity score after 6 weeks of treatment.

Oral immunosuppressants: oral azathioprine and topical clobetasol propionate versus topical clobetasol propionate Agarwal 2013 compared oral azathioprine with topical clobetasol propionate 0.05% cream to topical clobetasol propionate 0.05% cream alone in 108 participants; 91 participants completed the trial.

71 Primary outcome: percentage of participants with investigator-rated good/excellent control After eight weeks 36.95% had a good response (defined as 75% improvement in signs and symptoms) in the control group versus 73.3% in the intervention group. After 24 weeks chapter 39.13% improved in the control group and 91.1% in the intervention group. RR 0.43 (95% CI 2 0.30 to 0.62, P<0.0001). Primary outcome: adverse events No adverse events were reported that would require reduction of dosage or discontinuation of treatment (Table 11).

Secondary outcome: reduction in severity, investigator-rated scoring Measured by the Hand eczema severity index (HECSI): after 24 weeks there was an improvement of 64.66% in the control group and of 91.29% the intervention group (stated P = 0.001).

Secondary outcome: reduction of severity of itch, participant-rated measured on a numerical scale from 0-10 After 24 weeks the itch score difference was 6.0444 (SD 2.35445) in the intervention group and 4.5652 (SD2.26718) in the control group (stated P = 0.003).

Oral immunosuppressants: cyclosporine versus alitretinoin The study of Schmitt 2013 compared cyclosporine to alitretinoin, but was ended prematurely due to the inability to include the total number of participants. According tot the sample size calculation 78 participants should have been included, however only 15 participants were included and 14 were analyzed. Therefore, results are given in a descriptive way and no statistical significance tests are applied.

Primary outcome: percentage of participants with investigator-rated good/excellent control (IGA) after 24 weeks In the cyclosporine group 4 out of 7 participants (57.1%) reached a complete or nearly complete clearance of hand eczema according to the Investigator Global Assessment (IGA) and 2 out of 7 participants (28.6%) in the alitretinoin group. According to the authors this was not statistically significant (chi-square-test, stated P = 0.592); however this group was too small for reliable statistical analyses.

Primary outcome: percentage of participants with participant-rated good/excellent control (PGA) after 24 weeks In the cyclosporine group 4 out of 7 participants (57.1%) reached a complete or nearly

72 complete clearance of hand eczema according to the Patient Global Assessment (PGA) and 2 out of 7 participants (28.6%) in the alitretinoin group.

Primary outcome: adverse events Six adverse events were documented of which 2 possible were related to the use of chapter cyclosporine (fatigue, bone-ache, dry lips in one participant and exacerbation of atopic eczema in one participant). No serious adverse events were recorded throughout the trial 2 (Table 11).

Secondary outcome: reduction in severity, investigator-rated HECSI scoring after 24 weeks After 24 weeks both groups showed a decrease in the HECSI with a mean reduction of 80.9% in the cyclosporine group and 69.8% in the alitretinoin group.

Secondary outcome: time until relapse None of the participants relapsed during the 24 weeks of follow-up (0 of 7 versus 0 of 7).

Oral retinoids Oral retinoids: acitretin versus placebo Oral acitretin was compared with placebo capsules in a study which enrolled 29 participants (21 men and 8 women, age range 30 to 76 years), with hyperkeratotic dermatitis of the palms (Thestrup-Pedersen 2001). The authors called it a single-blind study, because participants were aware of the active medication due to the mild adverse events (dry lips). Fourteen participants were allocated to 30 mg acitretin once daily for eight weeks, and 15 participants to identically looking placebo capsules. Because the statistical analyses were insufficient the authors’ conclusion of acitretin to be efficacious is invalid.

Primary outcome: adverse events No adverse events were reported and all biochemical parameters were within normal limits in both groups (Table 11).

Secondary outcome: reduction in severity, participant-rated scoring The percent reduction in the mean itch score in the acitretin group was 41% (from 1.43 to 0.85), and in the placebo group 19% (from 1.40 to 1.13). It is not stated whether the difference was statistically significant. It appears as if no between-group analysis was conducted.

Secondary outcome: reduction in severity, investigator-rated scoring In the acitretin group there was a 51% reduction in overall score, and in the placebo group

73 a 9% reduction. In the acitretin group there were statistically significant reductions in scores for hyperkeratosis (Wilcoxon rank test for paired data): 50% reduction in the acitretin group from 2.14 to 1.08 and 14% reduction in the placebo group from 2.40 to 2.07. The authors did not state whether this is significantly different from placebo group. The acitretin group chapter reduced for fissures 67%, from 1.64 to 0.54% and for scaling 48% and redness 43%. It was not stated whether this was statistical significantly different from the score reductions in the 2 placebo group (hyperkeratosis 14%; fissures 10%; scaling 0%, redness 7%). It appears as if no between-group comparison was conducted.

Oral retinoids: alitretinoin versus placebo In three studies the effect of 10 mg oral alitretinoin was investigated: Bissonnette 2010; Ruzicka 2004; Ruzicka 2008.

Primary outcome: percentage of participants with self-rated good/excellent control (clear or almost clear) with PaGA at week 12 or end of treatment Ruzicka 2004 shows that for all doses of alitretinoin statistical significantly more participants rated their eczema as clear or almost clear compared to placebo: 12% in the placebo group rated their eczema as clear or almost clear, compared to 34% in the 20 mg group and 43% in the 40 mg group.

Alitretinoin 40 mg versus placebo In this subgroup we only found one relevant trial (n = 147) (Ruzicka 2004). There was a statistically significant difference: 32 out of the 74 participants in the 40 mg group judged a clear or almost clear status for PaGA, compared to 9 of 73 in the placebo group (RR 3.51, 95% CI 1.80 to 6.82, P = 0.0002) (see our Analysis 15.1).

Alitretinoin 30 mg versus placebo at day 200 (end of treatment) In this subgroup we only found one relevant trial (n = 614) (Ruzicka 2008). There was a statistically significant difference: 163 out of the 409 participants in the 30 mg group judged a clear or almost clear status for PaGA, compared to 31 of 205 in the placebo group, RR 2.64 (95% CI 1.87 to 3.72, P = 0.00001) (see our Analysis 15.1).

Alitretinoin 20 mg versus placebo In this subgroup we only found one relevant trial (n = 147) (Ruzicka 2004). There was a statistically significant difference: 25 out of the 74 participants in the 20 mg group judged a clear or almost clear status for PaGA, compared to 9 of 73 in the placebo group (RR 2.74, 95% CI 1.37 to 5.46, P = 0.004 Analysis 15.1).

74 Alitretinoin 10 mg versus placebo In this subgroup we found two relevant trials (n = 765). Both studies (Ruzicka 2004; Ruzicka 2008) found that 10 mg alitretinoin was statistically significantly more effective (respectively 29% and 24% clear or almost clear) than placebo. Pooling these data for 10 mg alitretinoin (heterogeneity statistics: Chi-square test = 0.89, P = 0.35) gives an overall RR of 1.73 (95% CI chapter 1.25 to 2.40, P = 0.001) (see our Analysis 15.1). 2 Primary outcome: percentage of participants with investigator-rated good/excellent control (clear or almost clear) In Ruzicka 2004 clearance or almost clearance of eczema occurred statistically significantly more often in all the groups treated with alitretinoin compared to placebo: 27% of the subjects in the placebo group were rated as clear or almost clear, compared to 41% in the 20 mg group and 53% in the 40 mg group (see our Analysis 15.2). Ruzicka 2008 also demonstrated statistically significantly better results with all doses of alitretinoin compared to placebo: 17%of the subjects in the placebo group were rated clear or almost clear, compared to 48% in the 30 mg group According to both studies (Ruzicka 2004; Ruzicka 2008) alitretinoin 10 mg was more effective for this outcome (respectively 39% and 28%) compared to placebo.

Alitretinoin 40 mg versus placebo In this subgroup we only found one relevant trial (n = 159) (Ruzicka 2004). There was a statistically significant difference: 43 out of the 81 participants in the 40 mg group judged a clear or almost clear status for PaGA, compared to 21 of 78 in the placebo group, RR 1.97 (95%CI 1.30 to 3.00, P = 0.001) (see our Analysis 15.2).

Alitretinoin 30 mg versus placebo In this subgroup we only found one relevant trial (n = 614) (Ruzicka 2008). There was a statistically significant difference between oral retinoids: alitretinoin and placebo (RR 2.87 CI 2.08 to 3.97, P = 0.0001 Analysis 15.2).

Alitretinoin 20 mg versus placebo In this subgroup we only found one relevant trial (n = 158) (Ruzicka 2004). There was no statistically significant difference: 32 out of the 80 participants in the 20 mg group judged a clear or almost clear status for PaGA, compared to 21 of 78 in the placebo group, RR 1.49 (95% CI 0.94 to 2.34, P = 0.09, Analysis 15.2).

75 chapter 2

Analysis 15.1 Primary outcome: patient-rated investigator-rated good/ excellent control of symptoms

76 Alitretinoin 10 mg versus placebo In this subgroup we found two relevant trials (n = 781). There was a statistically significant difference between oral retinoids: alitretinoin and placebo. According to both studies (Ruzicka 2004; Ruzicka 2008) alitretinoin 10 mg was more effective for this outcome (respectively 39% and 28%) compared to placebo. Pooling these data for 10 mg alitretinoin (heterogeneity chapter statistics: Chi square test = 0.48, P = 0.49) gives an overall RR of 1.63 (95% CI1.23 to 2.16, P = 0.0006) (see our Analysis 15.2). 2

Primary outcome: percentage of participants with investigator-rated good/excellent control (clear or almost clear) at different time points Ruzicka 2008 included a time to response curve, which we translated into RR’s at different time points.

Alitretinoin 30 mg versus placebo at day 50 In the 30 mg group 26 of 409 participants were clear or almost clear, compared to 2 of 205 in the placebo group RR 6.52 (95% CI 1.56 to 27.19, P = 0.01).

Alitretinoin 10 mg versus placebo at day 50 In the 10 mg group 7 of 418 participants were clear or almost clear, compared to 2 of 205 in the placebo group RR 1.72 (95% CI 0.36 to 8.19, P = 0.50).

Alitretinoin 30 mg versus placebo at day 100 In the 30 mg group 124 of 409 participants were clear or almost clear, compared to 8 of 205 in the placebo group RR 7.77 (95% CI 3.88 to 15.57, P<0.0001).

Alitretinoin 10 mg versus placebo at day 100 In the 10 mg group 45 of 418 participants were clear or almost clear, compared to 8 of 205 in the placebo group RR 2.76 (95% CI 1.33 to 5.74, P = 0.0067).

Alitretinoin 30 mg versus placebo at day 150 In the 30 mg group 163 of 409 participants were clear or almost clear, compared to 21 of 205 in the placebo group RR 3.89 (95% CI 2.55 to 5.94, P<0.0001).

Alitretinoin 10 mg versus placebo at day 150 In the 10 mg group 84 out of 418 participants were clear or almost clear, compared to 21 of 205 in the placebo group RR 1.96 (95% CI 1.25 to 3.07, P = 0.0032).

77 chapter 2

Analysis 15.2 Primary outcome: investigator-rated good/ excellent control of symptoms

78 Primary outcome: adverse events The study listed in detail the adverse events that were observed; headache was the most frequent event (22 in 40 mg group, eight in 20 mg group; four in 10 mg group), leading to six withdrawals. In the placebo group headache was reported by seven participants (Table 11). chapter Secondary outcome: reduction in severity, participant-rated scoring Ruzicka 2004 reported improvement in DLQI in all groups but no details were given. For 2 Ruzicka 2008 this outcome was not stated.

Secondary outcome: reduction severity, investigator rated in total lesion symptom score In Ruzicka 2004 a statistically significant higher median % reduction in total lesion symptom score was observed for all doses of alitretinoin compared to placebo: 25% in the placebo group (stated 95%CI -42 to -14) versus 52%in the 20 mg group (stated 95% CI -73 to -42) and 71% in the 40 mg group (stated 95% CI -80 to -44 (Analysis 15.3). The median reduction in the modified Total Lesion Symptom Score (Ruzicka 2008) was 75% in the 30 mg group and 56% in the 10 mg group, compared to 39% in the placebo group. The difference between alitretinoin and placebo was statistically significant for both doses according to the authors (Analysis 15.3). There was also reporting of a decrease of extent of disease in all groups, but no details were given. The 40 mg group, however, was stated to be statistically significantly different from the placebo group according to the authors. We have plotted these data (Analysis 15.4).

Study Dosage Reduction in severity Ruzicka 2004 40 mg Median of % change from baseline: -70.5% (95% CI -44 to -80) Significant more reduction than placebo (P<0.001; Kruskal-Wallis test) Ruzicka 2008 30 mg Median of % change from baseline: -75% Significant more reduction than placebo (P<0.001; Kruskal-Wallis test) Ruzicka 2004 20 mg Median of % change from baseline: -52 (95% CI -42 to -73) Significant more reduction than placebo (P<0.01; Kruskal-Wallis test) Ruzicka 2004 10 mg Median of % change from baseline: -25% (95% CI -14 to -42) Significant more reduction than placebo (P<0.01; Kruskal-Wallis test) Ruzicka 2008 10 mg Median of % change from baseline: -56% Significant more reduction than placebo (P<0.001; Kruskal-Wallis test)

Analysis 15.3 Secondary outcome: Investigator-rated reduction in severity in (modified) total lesion symptom score

79 Alitretinoin 40 mg versus placebo In this subgroup we only found one relevant trial (n = 159) (Ruzicka 2004). There was a statistically significant difference between oral retinoids: alitretinoin and placebo (MD 45.5 CI 23.04 to 67.96, Analysis 15.4). chapter Alitretinoin 20 mg versus placebo 2 In this subgroup we only found one relevant trial (n = 158) (Ruzicka 2004). There was a statistically significant difference between oral retinoids: alitretinoin and placebo (MD 27.0 CI 4.54 to 49.46, Analysis 15.4).

Alitretinoin 10 mg versus placebo In this subgroup we only found one relevant trial (n = 158) (Ruzicka 2004). There was a statistically significant difference between oral retinoids: alitretinoin and placebo (MD 34.0 CI 9.96 to 58.04, Analysis 15.4).

Secondary outcome: reduction in severity, participant-rated scoring Ruzicka 2004 reported improvement in DLQI in all groups but no details were given. For Ruzicka 2008 this outcome was not stated.

Secondary outcome: time until relapse At the end of the study period, i.e. at the end of treatment, the majority of the participants who had cleared or almost cleared were followed up. This was done for up to 12 weeks, probably until relapse. Details on this follow-up were not given, except that 26% required treatment at some time during this period because of relapse.

Oral retinoids: re-treatment with alitretinoin versus placebo Primary outcome: percentage of participants with investigator-rated good/excellent control (clear or almost clear) In the study by Bissonnette 2010, 117 participants suffering from chronic hand eczema were included who were successfully treated with alitretinoin in an earlier study (Ruzicka 2008), 24 withdrew.

Alitretinoin 30 mg versus placebo In this subgroup we only found one relevant trial (n = 73) (Bissonnette 2010). The authors used an overall Physician’s Global Assessment (PGA) rating “clear” or “almost clear” hands at the end of therapy. In the 30 mg alitretinoin group, 39 of 49 participants (80%) who were re-treated with 30 mg alitretinoin were rated as “clear” or “almost clear” according to the PGA, compared to two of 24 participants (8%) who were retreated with placebo (RR 9.55, 95% CI 2.51 to 36.27, P = 0.0009) (see our Analysis 16.1).

80 Alitretinoin 10 mg versus placebo In this subgroup we only found one relevant trial (n = 31) (Bissonnette 2010). In the group who was treated previously with 10 mg alitretinoin (n = 31), ten were re-treated with placebo and 21 with 10 mg alitretinoin. Ten out of 21 participants cleared or almost cleared again under re-treatment with 10 mg alitretinoin (48%) in comparison to one out of ten participants chapter retreated with placebo (10%), RR 4.76 (95% CI 0.70 to 32.25, P = 0.11, see our Analysis 16.1). In the group who was re-treated with placebo, nine out of 13 participants (69%) responded 2 again.

Analysis 16.1 Primary: Investigator-rated good/excellent control of symptoms

Primary outcome: percentage of participants with self-rated good/excellent control In the alitretinoin 30 mg group, 37 participants (76%) who were again treated with alitretinoin at 30 mg reported clear or almost clear hands. The authors state this was statistically significantly different from the placebo group. However, it is not stated how many participants who were treated with placebo in this group reported clear or almost clear hands. In the 10 mg group eight participants (38%) reported clear or almost clear hands. This was not statistically significantly different from the placebo group.

Primary outcome: adverse events Headache was the most frequent reported adverse event and all reports occurred in the 30 mg group (n = 7, 14%). Other events did not occur more often in the active treatment groups than in the placebo arm. There were three serious adverse events: one case of acute cardiac failure with fatal outcome in the 10 mg group which was not related to the study drug, one case of aortic aneurysm and one case of coronary artery disease (both in the 30 mg group) were both assessed as having a remote relationship to the study drug. Six participants withdrew due to adverse events, two in each of the three treatment groups. One participant withdrew in the alitretinoin 30 mg group due to headache. Observed biochemical abnormalities were typical effects of retinoids (high cholesterol, high triglycerides, low TSH) (Table 11).

81 chapter 2

Analysis 15.4 Secondary outcome: Investigator-rated reduction in severity in total lesion symptom score

82 Secondary outcome: reduction in severity, investigator-rated scoring The median reduction in the modified Total Lesion Symptom Score was 92% for participants receiving alitretinoin 30 mg, 71% for those treated with alitretinoin 10 mg and 43% in the placebo group (data obtained from a figure, no significance tests conducted). chapter Study Treatment Comparator 1 Comparator 2 Comparator 3 Agarwal Azathioprine with Topical clobetasol 2 2013 topical clobetasol propionate: no adverse propionate: no adverse events reported which events were reported required a dose reduc- which required a tion, other adverse dose reduction, other events not stated adverse events were not stated Bissonnette Alitretinoin 30 mg: any Alitretinoin 10 mg: Placebo: any adverse 2010 adverse event in 22 any adverse event in event in 12 (26.1%), (44%), treatment-relat- 9 (42.9%), treatment- treatment-related ed adverse events in 16 related adverse events adverse events in 8 (32%): nasopharyngitis in 5 (23.8%): nasophar- (17.4%): nasopharyn- 4%, influenza 2%, rhini- yngitis 4.8%, erythema gitis 4.3%, Influenza tis 2%, bronchitis 2%, 4.8%, dermatitis 4.8%, 2.2%, eczema 2.2%, upper respiratory tract dry mouth 4.8%, blood dermatitis 2.2%, dry infection 2%, erythema triglyceride elevated lips 2.2%, cheilitis 12%, eczema 2%, dry 4.8%, cholesterol high 2.2%, dry mouth 2.2%, skin 2%, headache 11.8%, triglycerides cholesterol high 3%, 14%, dry lips 2%, nau- high 5.9% 1 serious triglycerides high 3% sea 2%, dry mouth 4%, adverse event: cardiac blood creatine phos- arrest, not related phokinase elevated 2%, TSH high 2%,TSH low 8%, thyroxine low 4%, cholesterol high 21.1%, triglycerides high 13.2%, AST high 4%, bilirubin high 2% Serious adverse events: aortic aneurysm in 1, coronary artery disease in 1, both remotely related

83 Granlund Cyclosporine: 19 of 28 Topical corticosteroids: 1996 experienced adverse 15 of 27 experienced events. Discontinuation adverse events Severe due to dizziness, vomit- insomnia in 1. ing and facial oedema chapter in 1. Serum creatinine levels increased to 30% 2 above baseline value in 2 but normalized before the doses were reduced. Hypertension was not recorded. Ruzicka Alitretinoin 40 mg: Alitretinoin 20 mg: Alitretinoin 10 mg: Placebo: adverse 2004 adverse events in 43 adverse events in 28 adverse events in 28 events in 27 (35%): (53%): headache in (35%): headache in (35%): headache in headache in 7 (9%), dry 22 (27%), dry lips in 8 (10%), dry lips in 4 (5%), flushing in 1 lips in 4 (5%), flushing 5 (6%), flushing in 5 4 (5%), flushing in 1 (1%), dry mouth in 1 in 1 (1%), dry mouth in (6%), dry mouth in 2 (1%), dry mouth in 3 (1%), erythema in 1 1 (1%), erythema in 1 (2%), erythema in 3 (4%), eczema in 2 (3%), (1%), eczema in 1 (1%), (1%), eczema in 1 (1%), (4%), eczema in 3 (4%), eye pruritus in 2 (3%), conjunctivitis in 1 (1%), conjunctivitis in 2 (3%), conjunctivitis in 1 (1%), fatigue in 2 (3%), phar- eye pruritus in 1 (1%), eye pruritus in 1 (1%), fatigue in 1 (1%), pharyngitis in 1 (1%) fatigue in 1 (1%), rigors tonsillitis in 2 (3%), yngitis in 1 (1%). in 2 (3%), pharyngitis in pharyngitis in 1 (1%). 2 (3%). Ruzicka Alitretinoin 30 mg: Alitretinoin 10 mg: Placebo: nasopharyngi- 2008 nasopharyngitis in 24 nasopharyngitis in 22 tis in 14 (7%), influenza (6%), influenza in 6 (5%), influenza in 10 in 4 (2%), respiratory (2%), respiratory tract (2%), respiratory tract tract infection in 4, infection in 9 (2%), infection in 5 (1%), rhinitis in 4, skin and rhinitis in 2 (1%), skin rhinitis in 2 (1%), skin subcutaneous tissue and subcutaneous and subcutaneous disorders in 20 (10%), tissue disorders in 60 tissue disorders in 36 headache in 13 (6%), (15%), headache in 81 (9%), headache in 45 dry lips in 4 (2%), (20%), dry lips in 15 (11%), dry lips in 9 (2%), nausea in 3 (2%), dry (4%), nausea in 14 (3%), nausea in 10 (2%), dry mouth in 2 (1%), flush- dry mouth in 10 (2%), mouth in 10, flushing in ing in 2 (1%), elevated flushing in 18 (4%), 5 (1%), elevated blood blood CPK in 4 (2%). elevated blood CPK in CPK in 8 (2%) and 13 (3%), elevated trig- elevated triglycerides lycerides in 12 (3%) in 3 (1%).

84 Schmitt Cyclosporine: one Alitretinoin: one par- 2013 participant reported ticipants suffered from fatigue, bone-ache, cystitis (relatedness: dry lips (relatedness: non) and one from possible), two had an tonsillitis (relatedness: chapter exacerbation of hand unlikely) eczema (relatedness: possible/ not) and 2 onesuffered from a viral infection, headache, exhaustion (relatedness: not reported). Thestrup- Acitretin: no adverse Placebo: no adverse Pedersen events reported; events reported. 2001 especially biochemical investigations were within normal limits

Table 11. Adverse events

Discussion Summary of main results of the full review Hand eczema is a common condition. Considering the high prevalence of hand eczema, it is striking that the results of all 55 identified RCTs are based on approximately 4,619 participants, whereby about half of them (n = 2,297) were enrolled in four recent RCTs: two on the oral retinoid alitretinoin (Ruzicka 2004; Ruzicka 2008) and two on the topical calcineurin inhibitor pimecrolimus (Belsito 2004; Hordinsky 2010). Although many systematic reviews focus on a single treatment modality or its closely related variants, we have tried to include all interventions in this review in an attempt to determine which therapy would reflect current standard treatment, and to which extent there is evidence for its effectiveness. The wide range of available treatments underlines the fact that there does not seem to be a single candidate for standard therapy. Major textbooks mention topical corticosteroids and UV-phototherapy as the major treatment options for chronic hand eczema. Though, in this review we found little strong or consistent evidence that one intervention for hand eczema should be recommended over the other. Of all treatment categories, the biggest number of studies was on topical steroids (nine RCTs) and UV-therapy (eight RCTs). Therefore, calling topical corticosteroids and UV-phototherapy the first-choice treatment options may be supported by this review. Nevertheless, most trials do not have one of these treatments as comparator. In fact, most trials have either placebo, vehicle or a variant of its intervention as comparator, making it difficult to

85 draw conclusions on the comparative advantage of the different treatments. Of the nine trials on topical corticosteroids, each dealt with a different type of steroid. The duration of six studies was rather short, namely one week (Gupta 1993), two weeks (Fowler 2005; Kircik 2013; Lodén 2012; Uggeldahl 1986) and three weeks (Uggeldahl 1986). Three of the trials chapter compared two different corticosteroids (Bleeker 1989; Fowler 2005; Möller 1983). The comparators of the remaining trials were the same cortico-steroids as the ones used for the 2 intervention, while the comparator was either in a different vehicle, a different dosage, applied in a different frequency or combined with zinc sulphate (Gupta 1993; Uggeldahl 1986; Veien 1999; Faghihi 2008; Lodén 2012) or only the vehicle (Kircik 2013). Due to the differences in design between the trials, no generalization or recommendation for practice can be made regarding the use of topical corticosteroids, although one trial showed a positive tendency towards the application of corticosteroids three times weekly rather than twice weekly (Veien 1999). The eight trials on UV-phototherapy were too heterogeneous for pooling. Three studies had UVB as the main intervention, two had UVA-1 and five had PUVA as main intervention or comparator. Various trails used placebo-irradiation, while a statistical significance between different interventions was not often found. Two obvious questions for clinicians would be whether PUVA is more effective than UVB treatment, in addition to whether PUVA and UVB phototherapy are more effective than corticosteroids. One study (Sezer 2007) compared the effectiveness of local narrowband- UVB with topical PUVA and found no difference between them. There were no trials that compared phototherapy with topical corticosteroids. One study demonstrated equal efficacy of oral PUVA in the hospital versus topical PUVA at home, and assumed that home PUVA would be cheaper (van Coevorden 2004). However, in another study no difference between topical bath PUVA and oral PUVA could be demonstrated (Tzaneva 2009). The topical immunomodulators were studied in nine RCTs and almost all studies compared tacrolimus or pimecrolimus to vehicle. Only two studies (Schnopp 2002, Katsarou 2012) compared tacrolimus with mometasone furoate, but did not find evidence for a comparative advantage of tacrolimus over mometasone. Pacor 2006 demonstrated a statistically significant advantage of tacrolimus over placebo. Another study on tacrolimus (Krejci-Manwaring 2008) found no statistically significant difference in time to relapse, no statistically significant difference in investigator- rated severity, but a greater participant-rated improvement in severity. Belsito 2004 and Hordinsky 2010, both comparing pimecrolimus with vehicle, failed to find an advantage for the former. Similarly, Bauer 2012 could not detect any statistically significant advantage of pimecrolimus over vehicle in delaying relapse. Two conference abstracts (Cherill 2000; Baskan 2005) showed statistically significant superiority of pimecrolimus in investigator-rated severity reduction. The studies of Baskan 2005; Bauer 2012; Pacor 2006 had, compared to all other studies, an overall low risk of bias.

86 A relatively new treatment option is oral alitretinoin, which has been studied in two large trials (Ruzicka 2004; Ruzicka 2008), with a total enrolment of 1,351 participants. Both of these trials investigated, in addition to other dosages, a daily dosage of 10 mg. The studies were considered sufficiently equivalent to pool the data (765 and 781 participants) of daily 10 mg and showed that alitretinoin was more effective compared to placebo Higher dosages of chapter alitretinoin (30 and 40 mg), however, were statistically significantly more effective than 10 mg. One study (Bissonnette 2010) showed that participants with relapsed hand eczema after 2 treatment with alitretinoin responded to a second course of treatment with alitretinoin and that it was well tolerated. Oral alitretinoin has not yet been compared to other treatment modalities such as corticosteroids or UV-phototherapy and a study comparing alitretinoin to cyclosporine was ended prematurely (Schmitt 2013). In many trials there was also improvement of the hand eczema in the comparison group that received placebo, vehicle or the same emollient as the intervention group; apparently the emollient, vehicle or factors related to being a participant have beneficial therapeutic effects. Improvement in the comparison group could also be due to the natural chronic relapsing course of hand eczema. About half of the studies (n = 30) included our primary outcome good/ excellent control either participant- or investigator-rated. The definition of good/ excellent control varied per study since a wide variety of outcome measures was used. The majority of the studies included the primary outcome adverse events (n = 52). None of the adverse events were life-threatening; most of them were mild (local irritation with stinging, erythema and burning). Overall, we can conclude that there is little high-level and consistent evidence for the treatment of hand eczema and we can not make a judgment whether one treatment option should be preferred over the other.

Overall completeness and applicability of evidence We included 55 RCT with a total of 4,619 participants in the complete review and seven studies on systemic interventions with a total of 1,661 participants in this thesis. The majority of the studies had no drop-outs (n = 13), included an intention to treat analyses (n = 13) or analyzed at least 80% of the included participants (n = 21). Overall, the studies included adults of both genders with general good health, which is in our opinion applicable to an important part of the hand eczema population, since hand eczema can be related to occupation. There was too much heterogeneity in the majority of the treatment categories to attempt consistent pooling and meta-analysis. The studies on ionizing radiation might be eligible for pooling, but for a critical evaluation of this therapeutic modality, we felt that a long follow-up period (longer than in these studies) should be taken into consideration in view of the discussions around the long term consequences of exposure to ionizing radiation. Furthermore, modern insights into potential short-term benefit versus long-term risk have made such therapies obsolete. One of the major drawbacks in this review was the

87 definition of hand eczema. We stipulated in our protocol that the hand eczema had to be diagnosed by a physician; however this was often not stated in the included studies. Further- more, the definition of hand eczema was different in almost all of the trials. Studies defined “chronic hand eczema” either as duration longer than 6 months or longer than 3 months, chapter or did not include a minimal duration of disease at all. We intended to conduct subgroup analyses, but a minority of the trials defined subgroups. It was not clear which participants 2 had hyperkeratotic hand eczema or vesicular hand eczema; while clinical experience suggests that the clinical subtype might influence treatment success. Without logical and comprehensive definitions of hand eczema with clear diagnostic criteria for hand eczema and its subgroups, RCTs are seriously flawed, which is one of the main pitfalls of this review. In this review, we found a wide range of different severity scoring systems for hand eczema. Only four studies used the Hand Eczema Severity Index (HECSI), four the Dyshydrotic Eczema Area and Severity Index (DASI), two studies used the Hand Eczema Area and Sever- ity Score (HEAS), one the Hand Eczema Extent Score (HEES). Two studies derived a scoring system from atopic dermatitis and used the (adjusted) eczema area and severity index (EASI) and one the SCORing Atopic Dermatitis (SCORAD), while the rest of the studies used a self- created, not-validated scoring system. This is in line with the findings of a review regarding the quantification of hand eczema25; it reported 45 different methods of which for only three scoring systems (including the HECSI) the inter and/ or intra-observer variability was studied. We agree with Weistenhöfer et al. that there is an obvious need for consensus regarding the scoring of the severity and extent of hand eczema. We would like to recommend the same procedure as is currently ongoing in atopic dermatitis: Harmonising Outcome Measures for Eczema (HOME).26 The HOME group is a worldwide initiative with the aim to develop a consensus-based set of core outcome domains for trials and clinical record keeping in atopic dermatitis. This is important to allow comparison of data across trials, one of the difficulties which we encountered in this review on interventions for hand eczema.

Quality of the evidence We found some serious limitations in the quality of reporting and aimed to discuss these according the GRADE considerations (study limitations, consistency of effect, imprecision, indirectness and publication bias).

Limitations in the study design and implementation: We only included randomized controlled trials in this review. Overall, the older studies had more shortcomings with regard to risk of bias and were judged as “high” or “unclear” risk of bias with regard to allocation concealment, blinding, and/ or a loss to follow-up. Frequent shortcomings were: missing information on randomization and blinding, no justification of the number of participants, and no analysis of drop-outs. The studies that were conducted more recently had overall a

88 low or unclear risk of bias for the majority of the risks (allocation concealment, blinding, intention-to-treat analysis, and selective reporting of outcomes), although they were often sponsored by pharmaceutical industries. Over a third of the studies had a within patient design (left-right studies). Although these studies show strengths in terms of power to obtain statistically significant results with small numbers of participants, this is at the expense chapter of problems in interpreting studies where no difference in effect is found, which might be a consequence of cross contamination of the topical interventions, the possible systemic 2 effects of topical preparations, or both. In general, we consider the quality of body of evidence in this review as an “unclear risk of bias”.

Indirectness of evidence: This review analyzed the effectiveness of many different interventions. Thus, the evidence is not only restricted to indirect comparisons between two different interventions, but the review contains two major groups (topical and systemic interventions), which all have different subgroups. Therefore all the evidence is very indirect. Not only the different interventions are indirect, also the included population and outcomes are indirect. Some studies included patients with specific subtypes of hand eczema such as recurrent vesicular hand eczema, whereas others excluded this subgroup. As mentioned above, the definition of “chronic hand eczema” was not always clear and could be defined as a minimal duration of six weeks to a minimal duration of six months. A wide range of outcome parameters was presented, most of which were however not validated. Some studies used a validated outcome measure such as the Hand Eczema Severity Index (HECSI), whereas most studies created their own non-validated, non-named scoring system. Another indirectness of evidence arose from the comparators used: most interventions were compared to an inactive placebo, which is less effective than standard treatment in most settings. In addition, only 30 studies assessed our primary outcome good/ excellent control either participant- or investigator-rated. Almost all studies did include our primary outcome with regards to adverse events (n = 52). Finally, we encountered various studies which included participants who suffered from different dermatoses and combined in their results with hand eczema other skin diseases such as atopic dermatitis, psoriasis or diaper dermatitis. These studies addressed outcomes of interest, but did not report the data for hand eczema participants separately and we were unable to obtain these from the authors. Because of the very high level of indirectness of these studies, these studies were not included in our analyses, but listed in the section “studies awaiting classification”.

Consistency of results: It is difficult to judge the consistency of the results, since we did not pool the study results. We were unable to pool data due to heterogeneity in interventions (and co-interventions), duration, comparison groups, and outcomes measured or reported. Therefore, most of the results are derived from single studies and the findings have not been

89 replicated elsewhere. The heterogeneity in study designs was so overwhelming, that we can not make a single recommendation, but only interpret this review as a scoping review.

Imprecision of results: Overall the included studies were relatively small and of short dura- chapter tion. Although hand eczema usually has a chronically relapsing course, less than half of the studies had a duration of more than four months, which in our opinion is the minimum 2 duration required to document important data such as duration and frequency of disease relapse. Therefore the rating of the quality of the evidence should be lower. Since we did not cluster studies and all the analyses are based on a single study result, we cannot give overall confidence intervals.

Probability of publication bias: We did not conduct a funnel plot, due to the inability to pool studies. Publication bias may especially be present in the wide range of studies on different moisturizers to treat hand eczema: we did find various registered trials in trial registers which were not (yet) published. Contact with the authors revealed in some cases that the results were minimal and would not be published or authors did not respond to our writings at all. In some cases the authors were not at liberty to disclose results, but referred us to pharmaceutical sponsors, who often remained unresponsive. Whenever a pharmaceutical company decides not to publish the results of a new potential moisturizer because the results are disappointing, we do not expect this to influence the level of evidence, since this is only a single study on a single treatment option (NCT00843466). However, we do think that the lack of publication of some studies might influence the level of quality of evidence, especially with regards to the study of Laumann on alitretinoin (NCT00817063). All of the remaining studies on alitretinoin are (large, well-designed) studies sponsored by pharmaceutical companies. Unfortunately we were unable to obtain additional information on the study of Laumann despite our efforts, and it is unclear whether it found a benefit of alitretinoin over placebo. The study of Schmitt 2013 also studied alitretinoin and was not (yet) published; personal communication with the author clarified that the study was ended prematurely.

Potential biases in the review process We acknowledge that there was potential for bias at all stages in the reviewing process, but we made various attempts to restrict the level of bias. We comprehensively searched for randomized controlled trials from a wide range of databases in order to avoid the risk of publication bias and we used clinically relevant outcome measures. We tried to compare the respective trial registrations with the published trial to ascertain whether there was lack of correspondence between what was intended to be an outcome and the actually reported outcomes. We attempted to be as inclusive as possible in the search strategy and included

90 studies reported in languages other than English. The different language backgrounds of the authors enabled us to include Dutch (Kemper 1998) and German (Adams 2007; Bayerl 1999) articles. We translated a Turkish article (Baskan 2005) to minimize language bias. Nevertheless, the studies included in this review were predominantly conducted in European or North American countries and published in European or American journals. chapter The authors of this review independently assessed eligibility of studies for inclusion in this review, two other authors extracted the data and assessed the risks of bias to minimize the 2 potential for additional bias beyond that detailed in the ’Risk of bias in included studies’ tables. Discrepancies between authors were resolved by discussion to reach consensus. However, we acknowledge that our assessments may occasionally have been subjective, for example in the case of the not-blinded radiographers (Cartwright 1987; Fairris 1984; Fairris 1985; King 1984). Therefore, another review team may not agree with all of our decisions. Authors that were involved as trialist for certain studies, such as for example TD in Bauer 2012, were not involved in the selection, assessment and data extraction of those studies. A lot of studies were judged to have an unclear risk of bias, especially with regards to selection bias since a substantial amount of studies did not describe the way of allocation concealment and sequence generation. In order to obtain more clarity on this matter, we contacted all authors from studies published after 1999 by email or other forms of social media such as LinkedIn. Authors with a personal or professional relation to one of the authors of this review may have been reached easier and might have been more prone to respond to our requests. Therefore, these studies may have been judged more often as a low risk of bias. For studies pre-1999 and reports of which the authors were unresponsive, we had less information and had to deal with more ambiguities which we were unable to resolve; this may have contributed to some bias in the assessments of these studies and these studies were more often judged as having an unclear risk of bias. The time-frame of the studies covered in this review inevitably shows that there is a time trend in the treatments that are evaluated: earlier studies tend to focus on corticosteroids, UV-phototherapy, or X-rays, while more recent trials evaluate the effect of novel medicaments such as oral retinoids and topical calcineurin inhibitors.

Agreements and disagreements with other studies or reviews This Cochrane review studies a wide range of treatments that have been evaluated by randomized controlled trials (RCTs) since 1967. Within the same time-frame, many uncontrolled and non-randomized controlled studies have been published. Van Coevorden conducted a review to describe study designs and the quality of studies on hand eczema, covering the time period from 1977 up to 2003. They included 90 studies of which 44 were case series, 15 non-randomized controlled trials and 31 RCTs. In total 11 different categories of treatment were found and the majority of the trials studied ultraviolet irradiation (n = 32)

91 or corticosteroids (n = 13). The review concluded that the overall quality of reporting on hand eczema was poor and most studies of hand eczema trials were not considered a dequate to guide clinical practice. Since the current Cochrane review was conducted by partly the same authors and incorporated the same RCTs, it is not surprising that the results chapter and conclusions of both reviews overlap. However, since we tried to obtain more information by contacting authors in this Cochrane review and sought additional published and un- 2 published data, we had fewer uncertainties with regard to the quality of evidence. We also included RCTs which were conducted after 2003 and these studies were in general of better quality than the older studies. We maintain that the overall quality of reporting in hand eczema is low and that there is a need for well-designed head-to-head studies of adequate duration, reported according to the CONSORT-guidelines. Over the years, various groups composed guidelines for the management of (chronic) hand eczema.18,27-31 These guidelines have in comment that they all acknowledge the lack of randomized controlled trials. All guidelines recommend topical corticosteroids as one of the first steps in pharmacological treatment for all types of hand eczema. Thereafter, the guidelines recommend different steps, in which the subtype of hand eczema can be a leading factor, usually starting with topical treatments (corticosteroids or calcineurin inhibitors). For severe hand eczema, unresponsive to topical treatment, basically all guidelines recommend a step-up regime with tar, phototherapy and systemic (oral) treatment (acitretin, alitretinoin, cyclosporine, corticosteroids or others). Since alitretinoin was recently licensed for the treatment of hand eczema in Europe and Canada (not yet in the United States), the more recent guidelines include this treatment option for severe chronic hand eczema.27-29,31 English et al. published a consensus statement on the management of chronic hand eczema in the view of general practitioners and dermatologists. They did not conduct a systematic review, but based their statement on a mix of clinical experience and a variety of RCT and non-RCTs in hand eczema and atopic dermatitis. In general, they advise a skin-protection program and topical treatment with corticosteroids or calcineurin inhibitors in a primary care setting whenever possible. For referrals to secondary care (dermatologist), the treatment options are PUVA, cyclosporine, azathioprine and alitretinoin as preferred treatment options for hyperkeratotic and vesicular hand eczema, with emphasis on the importance of the patient’s choice and the local availability. Furthermore, PUVA (also in our review a well studied intervention) is recommended for hyperkeratotic hand eczema. Methotrexate and mycophenolate are recommended after failure of other systemic interventions; however, this recommendation is not supported because of a lack of RCTs for these interventions. The German Dermatologic Society stresses the importance of education and prevention.31 Topical corticosteroids, topical calcineurin inhibitors and iontophoresis are the first treatment steps. For moderate to severe hand eczema high potent corticosteroids, UV therapy and alitretinoin are recommended, while other systemic treatment options such as cyclosporine are the final resort. This recom-

92 mendation is largely based on the fact that alitretinoin is registered for the treatment of hand eczema, while cyclosporine is an off label therapy. The Canadian guideline states that the treatment of hand eczema can be difficult and unsatisfactory.28 They distinguish three important clinical types: irritant contact dermatitis, allergic contact dermatitis and atopic hand eczema. The guideline provides a clear flow diagram for acute hand eczema and chapter chronic hand eczema, which is divided in mild, moderate and severe. Topical corticosteroids are the mainstream of treatment, with phototherapy recommended for moderate hand 2 chronic hand eczema unresponsive to topical corticosteroids. For the severe cases which are unresponsive to potent topical corticosteroids, phototherapy and alitretinoin are recommended. When this is insufficient as well, cyclosporine can be considered. Because the comparative study of Schmitt 2013 was ended prematurely, we do not know whether alitretinoin should be preferred over cyclosporine. The Danish guidelines state that the treatment of hand eczema should be tailored to the individual patient and skin care education is very important.29 They classify hand eczema in six different clinical types: chronic fissured hand eczema, recurrent vesicular hand eczema, hyperkeratotic palmar hand eczema, pulpitis, interdigital eczema, nummular hand eczema. Furthermore they distinguish mild/ moderate hand eczema and severe hand eczema. Mild/moderate hand eczema should be treated with topical corticosteroids, potentially in rotation with calcineurin inhibitors. For severe hand eczema a step-up with topical corticosteroids and “possibly potassium perman- ganate baths” for vesicular hand eczema and “silver nitrate solutions” for hyperkeratotic eczema is recommended. However, our review did not find evidence for these treatment options. If topical treatment is insufficient, a further step-up regime is recommended with tar, phototherapy and systemic treatment (acitretin, alitretinoin, cyclosporine, corticosteroids or others), although the guideline does not given an order of priority and does not make further recommendations regarding the different subtypes of hand eczema. The American Academy of Dermatology has published guidelines on the use of topical glucocorticoids, the mainstay of treatment for hand eczema.32 The British Photodermatol- ogy Group developed a guideline on phototherapy and included a comment on the use of phototherapy in hand eczema.33 Although the evidence for topical PUVA over oral PUVA is scarce, they suggest a common sense approach, which is not contradictory to the findings of this review. Overall, we can conclude that most guidelines do not give a single recommendation based on the current literature, which is in line with the main finding of this review.

93 Authors’ conclusion Implications for practice The results of this review cannot be used to inform clinical practice with regard to the best way of managing hand eczema, especially in the long-term. Until such data are forthcom- chapter ing, physicians will be tempted to use an array of treatments. Topical corticosteroids and UV-phototherapy appear to be the major standard treatment categories, and most of these 2 interventions showed a beneficial effect. We do not know whether there is a clear comparative advantage of topical steroids over UV-phototherapy or vice versa. There is also little evidence of a comparative advantage of topical steroids or UV-phototherapy over other types of treatment. The oral immunosuppres- sant cyclosporine has been compared with a topical steroid, suggesting an advantage of cyclosporine, but the difference was not statistically significant (Granlund 1996 study 1 from 1996). The effect of topical calcineurin inhibitors is unclear, and their comparative advantage to other established treatments (especially to the above mentioned topical corticosteroids or UV-phototherapy) has not yet been demonstrated. A relatively new systemic treatment with an oral retinoid for participants with severe chronic hand eczema showed clearance or almost clearance of about half the participants in two large RCTs. The comparative advantage over other treatments needs further evaluation.

Implications for research The most important implication from this review is the need to conduct high-quality RCTs of people with hand eczema comparing commonly used interventions using simple outcome measures that can be understood by participants and clinicians. • E (Evidence): The current evidence for managing hand eczema is of poor-quality and especially head-to-head trials are missing. • P (Population): People with chronic (longer than six months), moderate to severe hand eczema should be included in future trials. Subgroup analyses on participants with different variants of hand eczema are recommended, although the lack of consensus regarding the classification of hand eczema is a major limitation. In addition, the multifactorial origin of hand eczema is a potential limitation. Finally, a standard severity scale is lacking, so therefore international consensus should be sought. • I (Intervention): All sources of treatment can be included, although we would recommend including the main interventions (topical corticosteroids, UV-therapy, topical calcineurin inhibitors, acitretin, alitretinoin and cyclosporine). Studies should be of an adequate duration (greater than six months) in order to capture the effect of interventions on long term disease control as well as short-term relief of symptoms. • C (Comparison): Head-to-head trials, in which different groups of commonly used

94 interventions are compared, are highly desirable. For example cyclosporine versus alitretinoin or UVA therapy versus topical corticosteroids. If a RCT has placebo (or vehicle or inactive treatment) as the only comparator instead of an established treatment modality, this should be clearly and convincingly justified. • O (Outcome): There is currently no consensus on a standard severity scale for hand chapter eczema. A validation of commonly used scoring systems or of simple global ratings with for example photographic anchors is needed.25,34 Duration of remission, the way the dis- 2 ease is brought under control, adverse events and simple outcome measures applicable to all participants are preferable. Hand eczema is known to influence quality of life, and therefore quality of life should be an important outcome. In addition, trials should focus on economic consequences, since hand eczema is a common occupational disease. A major limitation of almost all reviewed trials is that no measure of effect size including precision is given. This is necessary in order to be able to judge whether advantages of treatments are not only statistically significant but also meaningful. • T (Time stamp): Our latest search was conducted in February 2014. Older studies mainly focused on topical corticosteroids, UV-therapy and irradiation, while more recent (namely industry funded) studies focus on topical immunomodulators (pimecrolimus and tacrolimus) and alitretinoin. The included studies were predominately of short duration, though studies on systemic interventions were more extensive. Future studies should have adequate treatment durations, preferably more than 4 months, which is in our opinion the minimum duration required to document important data such as duration and frequency of disease relapse. Furthermore, studies should include a follow-up period. It was obvious in many of the reviewed trials that the approach to the statistical analyses left a lot to be desired. Several parametric and non-parametric statistical procedures that are able to model both within (person and/or time) and between subject (treatment) factors simultaneously have been offered by most statistical packages for many years. A major limitation of many of the treatment comparisons was that they did not control for baseline variation. In addition, omnibus factorial designs (allowing contrasts to be specified a priori) reduce the type 1 error rate because they test several hypotheses at the same time. Post- hoc comparisons would only be necessary should the data reveal surprising results. These analyses would of course, have to be viewed in an explorative fashion. Future studies need to overcome said limitations. Many deficiencies in trial reporting thus far can be avoided if all specialist dermatology journals adopt the CONSORT guidelines35, especially since many of the “unclear” risks, turned out to be based on missing information in the report instead of flaws in the study design. All future studies should adhere to these guidelines.

95 References References to studies included in the review in this thesis Agarwal 2013 {published and unpublished data} Agarwal US, Besarwal RK. Topical clobetasol propionate 0.05% alone and in combination with azathioprine in patients with chronic hand eczema: an chapter observer blinded randomized comparative trial. Indian Journal of Dermatology, Venerology, & 2 Leprology 2013;79:101–3. Bissonnette 2010 {published and unpublished data} Bissonnette R, Worm M, Gerlach B, Guenther L, Cambazard F, Ruzicka T, et al. Successful retreatment with alitretinoin in patients with relapsed chronic hand eczema. British Journal of Dermatology 2010;162:420–6. Granlund 1996 {published data only} Granlund H, Erkko P, Eriksson E, Reitamo S. Comparison of cyclosporine and topical betamethasone-17,21-dipropionate in the treatment of severe chronic hand eczema. Acta Dermato-Venereologica 1996;76(5):371–6.[MEDLINE: 8891011] Granlund H, Erkko P, Reitamo S. Comparison of the influence of cyclosporin and topical betamethasone-17,21-dipropionate treatment on quality of life in chronic hand eczema. Acta Dermato-Venereo- logica 1997;77:54–8. Ruzicka 2004 {published and unpublished data} Larsen F, Galewicz, Ruzicka T, Horváth A, Coenraads PJ, Zouboulis CC, et al.BAL40/9 (alitretinoin, 9-cisretinoic acid) is effective and well tolerated in the treatment of refractory hand dermatitis. Journal of the European Academy of Dermatology & Venereology 2003;17:158. Ruzicka T, Larsen FG, Galewicz D, Horváth A, Coenraads PJ, Thestrup-Pedersen K, et al. Oral alitretinoin (9-cisretinoic acid) therapy for chronic hand dermatitis in patients refractory to standard therapy. Results of a randomized, double-blind placebo-controlled, multicenter trial. Archives of Dermatology 2004;140:1453–9. Ruzicka 2008 {published and unpublished data} Ruzicka T, Lynde CW, Jemec GB, Diepgen T, Berth-Jones J, Coenraads PJ, et al. Efficacy and safety of oral alitretinoin (9-cis retinoic acid) in patients with severe chronic hand eczema refractory to topical corticosteroids: results of a randomized, double-blind, placebo-controlled, multicentre trial. British Journal of Dermatology 2008;158:808–17. Schmitt 2013 {unpublished data only} Schmitt J. Ciclosporin versus alitretinoin for severe atopic hand dermatitis (TocyDD). clinicaltrials.gov/show/ NCT01231854 (accessed 27 February 2013). Thestrup-Pedersen 2001 {published and unpublished data} Thestrup-Pedersen K, Andersen KE, Menné T, Veien N. Treatment of hyperkeratotic dermatitis of the palms (eczema keratoticum) with oral acitretin. A single-blind placebo controlled study. Acta Dermato-Venereologica 2001;81: 353–5.

References to studies included in this review, not included in this thesis Adams 2007 Adams S, Bayerl C. Medium-dose-UVA-1 irradiation - and topical PUVA - therapy in chronic dyshidrotic hand dermatitis - a prospective randomized study. [Mittel–Dose–UVA–1– und lokale

96 PUVA–Therapie beim dyshidrosiformen Handekzem– eine prospective randomisierte Studie]. Aktuelle Dermatologie 2007;33: 142–5. Baskan 2005 Baskan EB, Kaçar SD, Tunali S. The efficacy of topical pimecrolimus cream 1% in hand dermatitis. Journal of the European Academy of Dermatology & Venereology 2005;19 Suppl(2):267. Baskan EB, Kaçar SD, Turan A, Tunali S. A New Alternative in the Treatment of Hand Eczema: Topical chapter Pimecrolimus. Turkderm- Archieves of the Turkish Dermatology and Venerology 2007;41:125–8. Bauer 2012 Bauer A, Lange N, Matterne U, Meurer M, Braeutigam M, Diepgen TL. Efficacy of 2 pimecrolimus 1% cream in the long term management of atopic hand dermatitis. A double-blind RCT. Journal der Deutschen Dermatologischen Gesellschaft 2012;10:426–33. Bayerl 1999 Bayerl C, Garbea A, Peiler D, Rzany B, Allgäuer T, Kleesz P, et al. Pilot study for the treatment of occupational hand dermatitis with a new portable UVB irradiation unit [Pilotstudie zur Therapie des beruflich bedingten Handekzems mit einer neuen tragbaren UVB–Bestrahlungseinheit]. Aktuelle Dermatologie 1999;25:302–5. Belsito 2004 Belsito DV, Fowler JF Jr, Marks JG Jr, Pariser DM, Hanifin J, Duarte IA,et al. Pimecrolimus cream 1%: a potential new treatment for chronic hand dermatitis. Cutis 2004;73:31–8. Bleeker 1989 Bleeker J, Anagrius C, Iversen N, Stenberg B, Cullberg Valentin K. Double-blind comparative study of Corticoderm® cream + unguentum Merck® and Betnovate® cream + unguentum Merck® in hand dermatitis. Journal of Dermatological Treatment 1989;1:87–90. Burrows 1986 Burrows D, Rogers S, Beck M, Kellet J, McMaster D, Merrett D, et al. Treatment of nickel dermatitis with trientine. Contact Dermatitis 1986;15:55–7. Cartwright 1987 Cartwright PH, Rowell NR. Comparison of Grenz rays versus placebo in the treatment of chronic hand eczema. British Journal of Dermatology 1987;117:73–6. Cherill 2000 Cherill R, Tofte S, MacNaul R, Maher T, Abrams B, Graeber M, et al. SDZ ASM981 is effective in the treatment of chronic irritant hand dermatitis: a 6-week randomized, double-blind, vehicle-controlled, single center study. Contact Dermatitis 2000;42 Suppl:16–7. Chia Yu 2009 Chai Yu C. A 4-Week Randomized, Double-Blind, Placebo-Controlled, Right-Left Comparison Study with E-DO® in Chronic Hand Dermatitis [Comparison Study With E–DO in Chronic Hand Dermatitis]. Clinical Study Report Jan 5,2009. Faghihi 2008 Faghihi G, Iraji F, Shahingohar A, Saidat A. The efficacy of ’0.05% Clobetasol + 2.5% zinc sulphate’ cream vs. ’0.05% Clobetasol alone’ cream in the treatment of the chronic hand eczema: a double-blind study. Journal of European Academy of Dermatology & Venereology 2008;22:531–6. Fairris 1984 Fairris GM, Mack DP, Rowell NR. Superficial X-ray therapy in the treatment of constitutional eczema of the hands. British Journal of Dermatology 1984;111:445–9. Fairris 1985 Fairris GM, Jones DH, Mack DP, Rowell NR. Conventional superficial X-ray versus Grenz ray therapy in the treatment of constitutional eczema of the hands. British Journal of Dermatology 1985;112:339–41. Fowler 2005 Fowler JF Jr, Fransway AF, Jackson JM, Rohowsky N. Hydrocortisone butyrate 0.1% cream in the treatment of chronic dermatitis. Cutis 2005;75:125–31.

97 Fredriksson 1975 Fredriksson T, Gip L. Urea creams in the treatment of dry skin and hand dermatitis. International Journal of Dermatology 1975;14:442–4. Grattan 1991 Grattan CE, Carmichael AJ, Shuttleworth GJ, Foulds IS. Comparison of topical PUVA with UVA for chronic vesicular hand eczema. Acta Dermato-Venereologica 1991; 71:118–22. chapter Gupta 1993 Gupta AK, Shear NH, Lester RS, Baxter ML, Sauder DN. Betamethasone dipropionate polyacrylic film-forming lotion in the treatment of hand dermatitis.International Journal of Dermatology 2 1993;32:828–9. Hanifn 2004 Hanifin JM, Stevens V, Sheth P, Breneman D. Novel treatment of chronic severe hand dermatitis with bexarotene gel. British Journal of Dermatology 2004;150:545–53. Hill 1998 {published data only (unpublished sought but not used)} Hill VA, Wong E, Corbett MF, Menday AP. Comparative efficacy of betamethasone/clioquinol (Betnovate-C) cream and betamethasone/ fusidic acid (Fucibet) cream in the treatment of infected hand eczema. Journal of Dermatological Treatment 1998;9:15–9. Hordinsky 2010 Hordinsky M, Fleischer A, Rivers JK, Poulin Y, Belsito D, Hultsch T. Efficacy and safety of pimecrolimus cream 1% in mild-to-moderate chronic hand dermatitis: a randomized, double-blind trial. Dermatology 2010;221: 71–7. Jowkar 2011 Jowkar F, Jamshidzadeh A, Mirzadeh Yazdi A, Pasalar M. The effects of fumaria parviflora L extract on chronic hand eczema: a randomized double-blind placebo controlled clinical trial. Iranian Red Crescent Medical Journal 2011;13:824–8. Kaaber 1983 Kaaber K, Menné T, Veien N, Hougaard P. Treatment of nickel dermatitis with Antabuse; a double blind study. Contact Dermatitis 1983;9:297–9. Katsarou 2012 Katsarou A, Makris M, Papagiannaki K, Lagogianni E, Tagka A, Kalogeromitros D. Tacrolimus 0.1% vs mometasone furoate topical treatment in allergic contact hand eczema: a prospective randomized clinical study. European Journal of Dermatology 2012;22:192–6 Kemper 1998 Kemper MFP, Van der Valk PGM, Arnold WP. Effectivity of coal tar paste versus betamethasone valerate ointment 0.1% in hand eczema [Effectiviteit van koolteerpasta versus betamethasonvaleraatzalf 0.1%bij handeczeem]. Nederlands Tijdschrift voor Dermatologie en Venereologie 1998;8:289–91. King 1984 King CM, Chalmers RJ. A double-blind study of superficial radiotherapy in chronic palmar eczema. British Journal of Dermatology 1984;111:451–4. Kircik 2013 Kircik LH, Eastman WJ, Gwazdauskas J. A randomized, double-blind phase 4 study of the efficacy and safety of ethanol-free clobetasol propionate foam, 0.05%, vs vehicle foam in the treatment of chronic hand dermatitis. Journal of Drugs in Dermatology 2013;12:328–34. Krejci-Manwaring 2008 Krejci-Manwaring J, McCarty MA, Camacho F, Manuel J, Hartle J, Fleischer A Jr, et al. Topical tacrolimus 0.1% improves symptoms of hand dermatitis in patients treated with a prednisone taper. Journal of Drugs in Dermatology 2008;7:643–6. Kucharekova 2003 Kucharekova M, van de Kerkhof PC, van der Valk PG. A randomized comparison of an emollient containing skin related lipids with a petrolatum-based emollient as adjunct in the

98 treatment of chronic hand dermatitis. Contact Dermatitis 2003;48:293–9. Lauriola 2011 Lauriola MM, Pigatto PD, Pedrelli VF. A single-center, randomized, perspective, investigator blinded, controlled trial to examine efficacy and safety of a Furpalmatecontaining cream in comparison to topical corticosteroid in atopic dermatitis of hands of 40 adult patients. 20th Congress of the European Academy of Dermatology and Venereology Lisbon (Abstracts on cd-rom). Geneva, 2011. chapter Lindelöf 1987 Lindelöf B, Wrangsjö K, Lidén S. A double-blind study of Grenz ray therapy in chronic eczema of the hands. British Journal of Dermatology 1987;117:77–80. 2 Lodén 2012 Lodén M, Wirén K, Smerud KT, Meland N, Hønnås H, Mørk C, et al. The effect of a corticosteroid cream and a barrier-strengthening moisturizer in hand eczema. A double-blind, randomized, prospective, parallel group clinical trial. Journal of the European Academy of Dermatology & Venereology 2012;26:597–601. Möller 1983 Möller H, Svartholm H, Dahl G. Intermittent maintenance therapy in chronic hand eczema with clobetasol propionate and flupredniden acetate.Current Medical Research & Opinion 1983;8:640–4. Odia 1996 Odia S, Vocks E, Rakoski J, Ring J. Successful treatment of dyshidrotic hand eczema using tap water iontophoresis with pulsed direct current. Acta Dermato-Venereologica 1996;76:472–4. Pacor 2006 PacorML, Di Lorenzo G,Martinelli N,Mansueto P, Friso S, Pellitteri ME, et al. Tacrolimus ointment in nickel sulphate induced steroid-resistant allergic contact dermatitis. Allergy & Asthma Proceedings 2006;27:527–31. Pigatto 1990 Pigatto PD, Gibelli E, Fumagalli M, Bigardi A, Morelli M, Altomare GF. Disodiumcromglycate versus diet in the treatment and prevention of nickel-positive pompholyx. Contact Dermatitis 1990;22:27–31. Polderman 2003 Polderman MC, Govaert JC, le Cessie S, Pavel S. A double blind placebo- controlled trial of UVA-1 in the treatment of dyshidrotic eczema. Clinical & Experimental Dermatology 2003;28:584–7. Schnopp 2002 Schnopp C, Remling R, Möhrenschlager M, Weigl L, Ring J, Abeck D. Topical tacrolimus (FK506) and mometasone furoate in treatment of dyshidrotic palmar eczema: a randomized, observer-blinded trial. Journal of the American Academy of Dermatology 2002;46:73–7. Sezer 2007 Sezer E, Etikan I. Local narrowband UVB phototherapy vs. local PUVA in the treatment of chronic hand eczema. Photodermatology, Photoimmunology & Photomedicine 2007; 23:10–4. Sharma 2006 Sharma AD. Disulfiram and low nickel diet in the management of hand eczema: a clinical study. Indian Journal of Dermatology, Venereology & Leprology 2006;72:113–8. Sheehan-Dare 1989 Sheehan-Dare RA, Goodfield MJ, Rowell NR. Topical psoralen photochemotherapy (PUVA) and superficial radiotherapy in the treatment of chronic hand eczema.British Journal of Dermatology 1989;121:65–9. Sjövall 1987 Sjövall P, Christensen OB. Local and systemic effect of UVB irradiation in patients with chronic hand eczema. Acta Dermato-Venereologica 1987;67:538–41.

99 Tzaneva 2009 Tzaneva S, Kittler H, Thallinger C, Hönigsmann H, Tanew A. Oral vs. bath PUVA using 8-methoxypsoralen for chronic palmoplantar eczema. Photodermatolology, Photoimmunology & Photomedicine 2009;25:101–5. Uggeldahl 1986 Uggeldahl PE, Kero M, Ulshagen K. Comparative effects of desonide cream 0.1% chapter and 0.05% in patients with hand eczema. Current Therapeutic Research, Clinical & Experimental 1986;40:969–73. 2 van Coevorden 2004 van Coevorden AM, Kamphof WG, van Sonderen E, Bruynzeel DP, Coenraads PJ. Comparison of oral psoralen- UV-A with a portable tanning unit at home vs hospital administered bath psoralen-UV-A in patients with chronic hand eczema: an open-label randomized controlled trial of efficacy.Archives of Dermatology 2004;140:1463–6. Veien 1995 Veien NK, Kaaber K, Larsen PO, Nielsen AO, Thestrup-Pedersen K. Ranitidine treatment of hand eczema in patients with atopic dermatitis: a double blind, placebo-controlled trial. Journal of the American Academy of Dermatology 1995;32:1056–7. Veien 1999 Veien NK, Olholm Larsen P, Thestrup-Pedersen K, Schou G. Long-term, intermittent treatment of chronic hand eczema with mometasone furoate. British Journal of Dermatology 1999;140:882–6. Whitaker 1996 Whitaker DK, Cilliers J, De Beer C. Evening primrose oil (Epogam) in the treatment of chronic hand dermatitis: disappointing therapeutic results. Dermatology 1996;193:115–20. Yousef 2012 Yousefi M, Barikbin B, Kamalinejad M, Abolhasani E, Ebadi A, Younespour S,et al. Comparison of therapeutic effect of topical Nigella with Betamethasone and Eucerin in hand eczema. Journal of the European Academy of Dermatology & Venereology 2013;27:1498-504.

Additional references 1. Thyssen J, Johansen J, Linneberg A, Menné T. The epidemiology of hand eczema in the general population-prevalence and main findings.Contact dermatitis 2010:62:75-82. 2. Meding B, Jarvholm B. Incidence of hand eczema-a population-based retrospective study. J Invest Dermatol 2004:122:873-877. 3. Yngveson M, Svensson A, Johannisson A, Isacsson A. Hand dermatosis in upper secondary school pupils: 2-year comparison and follow-up. Br J Dermatol 2000:142:485-489. 4. Nilsson E, Mikaelsson B, Andersson S. Atopy, occupation and domestic work as risk factors for hand eczema in hospital workers. Contact Dermatitis 1985:13:216-223. 5. Diepgen TL. Occupational skin-disease data in Europe. Int Arch Occup Environ Health 2003:76:331-338. 6. Meding B, Jarvholm B. Hand eczema in Swedish adults - changes in prevalence between 1983 and 1996. J Invest Dermatol 2002:118:719-723. 7. Menné B, Maibach HI. Hand Eczema. 7:158 edition:CRC Press,2000.

100 8. Diepgen TL, Andersen KE, Brandao FM, Bruze M, Bruynzeel DP, Frosch P, et al. Hand eczema classification: a cross-sectional, multicentre study of the aetiology and morphology of hand eczema. Br J Dermatol 2009:160:353-358. 9. Coenraads PJ. Hand eczema. N Engl J Med 2012:367:1829-1837. 10. Holness DL, Nethercott JR. Work outcome in workers with occupational skin disease. Am J Ind chapter Med 1995:27:807-815. 11. Agner T, Andersen KE, Brandao FM, Bruynzeel DP, Bruze M, Frosch P, et al. Hand eczema severity 2 and quality of life: a cross-sectional, multicentre study of hand eczema patients. Contact Dermatitis 2008:59:43-47. 12. Moberg C, Alderling M, Meding B. Hand eczema and quality of life: a population-based study. Br J Dermatol 2009:161:397-403. 13. Wallenhammar LM, Nyfjall M, Lindberg M, Meding B. Health-related quality of life and hand eczema--a comparison of two instruments, including factor analysis. J Invest Dermatol 2004:122:1381-1389. 14. van Coevorden AM, van Sonderen E, Bouma J, Coenraads PJ. Assessment of severity of hand eczema: discrepancies between patient- and physician-rated scores. Br J Dermatol 2006:155:1217-1222. 15. Meding B, Wrangsjo K, Jarvholm B. Fifteen-year follow-up of hand eczema: persistence and consequences. Br J Dermatol 2005:152:975-980. 16. Veien NK, Hattel T, Laurberg G. Hand eczema: causes, course, and prognosis II. Contact Dermatitis 2008:58:335-339. 17. Hald M, Agner T, Blands J, Veien NK, Laurberg G, Avnstorp C, et al. Clinical severity and prognosis of hand eczema. Br J Dermatol 2009:160:1229-1236. 18. Diepgen TL, Agner T, Aberer W, Berth-Jones J, Cambazard F, Elsner P, et al. Management of chronic hand eczema. Contact Dermatitis 2007:57:203-210. 19. Van Coevorden AM, Coenraads PJ, Svensson A, Bavinck JN, Diepgen TL, Naldi L, et al. Overview of studies of treatments for hand eczema-the EDEN hand eczema survey. Br J Dermatol 2004:151:446-451. 20. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions 5.1.0 edition: www.cochrane-handbook.org.,The Cochrane Collaboration,2011. 21. Held E, Skoet R, Johansen JD, Agner T. The hand eczema severity index (HECSI): a scoring system for clinical assessment of hand eczema. A study of inter- and intraobserver reliability. Br J Dermatol 2005:152:302-307. 22. Odia S, Vocks E, Rakoski J, Ring J. Successful treatment of dyshidrotic hand eczema using tap water iontophoresis with pulsed direct current. Acta Derm Venereol 1996:76:472-474. 23. van der Valk PG, van Gils RF, Boot CR, Evers AW, Donders R, Alkemade HA, et al. A simple tool with which to study the course of chronic hand eczema in clinical practice: a reduced-item score. Contact Dermatitis 2013:69:112-117.

101 24. Coenraads PJ, Van Der Walle H, Thestrup-Pedersen K, Ruzicka T, Dreno B, De La Loge C, et al. Construction and validation of a photographic guide for assessing severity of chronic hand dermatitis. Br J Dermatol 2005:152:296-301. 25. Weistenhofer W, Baumeister T, Drexler H, Kutting B. An overview of skin scores used for chapter quantifying hand eczema: a critical update according to the criteria of evidence-based medicine. Br J Dermatol 2010:162:239-250. 2 26. Schmitt J, Williams H, HOME Development Group. Harmonising Outcome Measures for Eczema (HOME). Report from the First International Consensus Meeting (HOME 1), 24 July 2010, Munich, Germany. Br J Dermatol 2010:163:1166-1168. 27. English J, Aldridge R, Gawkrodger DJ, Kownacki S, Statham B, White JM, et al. Consensus statement on the management of chronic hand eczema. Clin Exp Dermatol 2009:34:761-769. 28. Lynde C, Guenther L, Diepgen TL, Sasseville D, Poulin Y, Gulliver W, et al. Canadian hand dermatitis management guidelines. J Cutan Med Surg 2010:14:267-284. 29. Menné T, Johansen JD, Sommerlund M, Veien NK, Danish Contact Dermatitis Group. Hand eczema guidelines based on the Danish guidelines for the diagnosis and treatment of hand eczema. Contact Dermatitis 2011:65:3-12. 30. Veien NK, Menne T. Treatment of hand eczema. Skin Therapy Lett 2003:8:4-7. 31. Diepgen TL, Elsner P, Schliemann S, Fartasch M, Kollner A, Skudlik C, et al. Guideline on the management of hand eczema ICD-10 Code: L20. L23. L24. L25. L30. J Dtsch Dermatol Ges 2009:7 Suppl 3:S1-16. 32. Drake LA, Dinehart SM, Farmer ER, Goltz RW, Graham GF, Hordinsky MK, et al. Guidelines of care for the use of topical glucocorticosteroids. American Academy of Dermatology. J Am Acad Dermatol 1996:35:615-619. 33. Halpern SM, Anstey AV, Dawe RS, Diffey BL, Farr PM, Ferguson J, et al. Guidelines for topical PUVA: a report of a workshop of the British photodermatology group. Br J Dermatol 2000:142:22-31. 34. Charman CR, English JS. Getting to grips with hand eczema: measuring skin disease severity objectively. Br J Dermatol 2005:152:199-201. 35. Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 Statement: updated guidelines for reporting parallel group randomised trials. Trials 2010:11:32-6215-11-32.

102 chapter 2

103 Chapter3

104 Evidence based dermatology – chapter Hand eczema 3 Evidence Based Dermatology, 3rd edition, august 2014, chapter 22, 117-126

Wietske Andrea Christoffers1, Marie-Louise Anna Schuttelaar1, Pieter-Jan Coenraads1

1Department of Dermatology, University of Groningen, University Medical Center Groningen, Groningen, Netherlands

Key words: evidence-based dermatology, hand eczema, treatment

105 Key points There is insufficient evidence on which to base a choice between short bursts of potent topical corticosteroids compared with continuous application of mild corticosteroids. • There is insufficient evidence for oral immunosuppressants as maintenance therapy. • There is little evidence of a steroid-sparing effect of emollients, although these are widely prescribed. • PUVA and UVB are effective, but there is no evidence of a clinical advantage of one modality over the other. chapter • Oral retinoids appear to be effective and well tolerated in hand eczema, especially in 3 hyperkeratotic hand eczema. • There is insufficient evidence of an additive effect of iontophoresis. • There is insufficient evidence for a low-nickel diet or chelating agents in hand eczema accompanied by nickel allergy. • There is insufficient evidence of an additive effect of topical antibacterial agents. • There is insufficient evidence of the superiority of topical calcineurin inhibitors to topical corticosteroids. • Education for secondary and tertiary prevention seems to be effective; however, long-term data are needed.

106 Introduction Definition The term “hand eczema” implies an inflammation of the skin (dermatitis) that is generally confined to the hands, but sometimes the feet are involved as well. Clinically, the condition is characterized by signs of redness, vesicles (tiny blisters), papules, edema, scaling, fissures (cracks), erosions, and hyperkeratosis (callus-like thickening), all of which may be present at different points in time (Fig. 1). Itch, sometimes severe, is a common feature. Fissures and blisters might be painful and impair manual work. Microscopically, the disease is character- chapter ized by spongiosis with varying degrees of acanthosis, and a superficial perivascular infiltrate of lymphocytes and histiocytes. 3 Incidence and prevalence Hand eczema is considered a common condition, with a point prevalence of 1–5% among adults in the general population, and a 1-year prevalence of up to 10%, depending on whether the disease definition includes more pronounced or mild cases.1 The prevalence may be higher in some countries. Hand eczema is twice as common in women as in men, with the highest prevalence in young women. The reasons for this sex difference are unknown, although greater exposure of women to wet work is probably contributory. Reliable data on the incidence are scarce, and are mainly confined to estimates in particular occupational groups. The incidence of notified work-related cases is between 0.7 and 1.5 cases per 1,000 workers per year with much higher annual incidences among high-risk occupations, such as bakers and hairdressers.2

Etiology The etiology is multifactorial. Contact irritants are the commonest external causes. Hand eczema caused by such irritants, or mild toxic agents, is known as irritant contact dermatitis. Water is a contact irritant, and thus an external causal or contributing factor. Causal factors that are less common than irritants are contact allergens. Hand eczema caused by skin contact with allergens is called allergic contact dermatitis. Ingested allergens (for example, nickel) may also provoke hand eczema. Being atopic (to produce immunoglobulin E (IgE) antibodies in response to ordinary exposures to allergens, usually proteins and, as a consequence, the tendency to develop asthma, rhinoconjunctivitis, or eczema) is a major predisposing factor responsible for hand eczema. A combination of the above-mentioned factors appears to play a role in many patients. Hand eczema can be classified by etiology: irritant contact dermatitis, allergic contact dermatitis, atopic hand eczema, protein contact dermatitis, or a combination of these types. A morphological classification of hand eczema is: vesicular hand eczema (dyshidrotic,

107 pompholyx), hyperkeratotic hand eczema, fingertip dermatitis (pulpitis), nummular hand eczema, interdigital eczema, and dry fissured hand eczema.

Prognosis When there is a single, easily avoidable contact allergic factor, the prognosis is good. Several studies, however, have suggested that hand eczema tends to run a long-lasting and chronic relapsing course, probably because of the multifactorial origin. chapter Diagnostic tests The diagnosis is mainly based on history and clinical signs; there are no standardized 3 diagnostic criteria. Patients are patch-tested to detect or rule out a contact allergy. In addition, prick tests or determination of specific IgE are performed to detect atopy, and skin scrapings are performed to rule out a mycotic infection. In the majority of cases, no relevant contact allergy can be detected. Specific prick tests of specific IgE are of additional value only in very special cases (contact urticarial reactions that can become eczematous: protein contact dermatitis). Differential diagnostic psoriasis, mycosis, lichen planus, scabies, granuloma annulare, herpes simplex, and artefacts might be considered.

Treatment The treatment of hand eczema is aimed at reducing clinical symptoms (including the disabling itch), preventing relapses, and reducing the burden of disease by allowing the resumption of everyday manual tasks. The outcome of the treatment can be assessed in dif¬ferent ways. Relevant outcome parameters include: • the percentage of patients reporting a good/excellent response; • the percentage of patients with an investigator-reported good/ excellent response; • reduction in severity (patient-rated and physician-rated scoring systems); • dose reduction; • time until relapse.

Objectives In daily practice, we often ask ourselves what treatment would be best for the patient with hand eczema who is sitting in front of us. This usually involves a comparison between different treatment modalities. Against this background, we formulated 14 clinically relevant questions. Because of the tendency of hand eczema to develop a chronic or relapsing course, all of the questions are concerned with chronic hand eczema. In the context of this chapter, chronicity can arbitrarily be defined as a duration of more than 6 months.

108 Methods of search Controlled trials dating back to 1977 were located by searching the Skin Group Specialized Register, Cochrane Central Register of Controlled Trials (CENTRAL), PubMed, and Embase in February 2013 for original articles in English, French, German, or Dutch. Uncontrolled trials were discarded, unless systematic reviews and controlled trials were lacking on a specific subject. Also, papers studying different dermatoses and not specifically stating the results for the patients with hand eczema were ignored. The questions were formulated before the search, and only papers pertaining to these questions were included. chapter 3

Fig. 1. Hand eczema: redness, erosions, and tiny blisters, accompanied by severe itching.

Questions Because of the tendency of hand eczema to develop a chronic or relapsing course, all of the questions considered below deal with chronic hand eczema, arbitrarily defined as a duration of more than 6 months. Prescription of topical corticosteroids is at present the most common treatment; therefore, it is the major comparator in the questions listed below. • In adults with chronic hand eczema, do topical coal-tar preparations lead to a better patient-rated or physician-rated reduction in symptom scores than topical corticosteroids?

109 • In adults with hyperkeratotic hand eczema, does dithranol lead to an improvement in patient-rated and physician-rated symptom scores and longer remission periods after clearance in comparison with topical corticosteroids? • In adults with chronic hand eczema, do short bursts of potent topical corticosteroids (class 3 or 4) lead to better patient-rated or physician-rated symptom scores than continuous mild (class 1 or 2) topical steroids? • In adults with chronic hand eczema, are oral immunosuppressive agents (cyclosporine, methotrexate, mycophenolate mofetil, azathioprine) better in maintaining a long-term chapter (more than 6 months) reduction of patient-rated or physician-rated symptom scores than topical corticosteroids? 3 • Is the treatment of chronic hand eczema with local psoralen– ultraviolet A (PUVA) or ultraviolet B (UVB) irradiation better in reducing patient-rated and physician-rated symptom scores than topical corticosteroids? • In adults with chronic hand eczema, does treatment with PUVA irradiation (oral or topical psoralen) lead to a better reduction in patient-rated and physician-rated symptom scores and remission periods than UVB irradiation? • In adults with chronic hand eczema, is oral treatment with retinoids better in terms of patient-rated and physician-rated symptom scores than topical corticosteroids? • Is the treatment of chronic hand eczema with calcineurin inhibitors better in reducing patient-rated and physician-rated symptom scores than topical corticosteroids? • In adults with dyshidrotic hand eczema, does iontophoresis lead to an improvement in patient-rated and physician-rated symptom scores in comparison with topical steroids or UVB/PUVA irradiation? • In adults with relapsing vesicular hand eczema based on contact allergy to nickel, does dietary intervention or oral therapy with chelating agents lead to an improvement in patient-rated and physician-rated symptom scores in comparison with topical corticosteroids? • Does the daily application of a bland emollient lead to dose reduction and/or frequency reduction of topical corticosteroids in adults with chronic hand eczema? • In adults with chronic clinically active hand eczema, do protective or occlusive gloves or barrier creams lead to better patient-rated or physician-rated symptom scores than topical steroids? • Does the addition of a topical antibacterial agent to topical corticosteroids result in better patient-rated or physician-rated symptom scores than topical corticosteroids alone? • In patients with chronic hand eczema, is an educational intervention program effective for secondary and tertiary prevention on hand eczema compared with limited education as usual?

110 In adults with chronic hand eczema, do topical coal-tar preparations lead to a better patient-rated or physician-rated reduction in symptom scores than topical corticosteroids? One clinical trial was identified on the effect of coal tar. Additional trials may be found in the older literature (pre-1977).

Effcacy No systematic reviews were found. chapter

Coal tar In a self-controlled, randomized study, the efficacy of coal tar paste with zinc oxide paste 3 was compared with betamethasone valerate 0.1% in 19 patients with a within-patient design. Coal tar gave significant improvement compared with baseline, but no significant difference was observed between the treatments.3

Drawbacks Six participants dropped out because they experienced problems with wearing hand gloves (problems were not specified). One participant dropped out due to pompholyx as a result of contact allergy to 5% coal tar.

Comment Small number of participants (n = 9) with relatively high drop-out rate (n = 7). The results are listed as overall mean scores and no exact data are given.

Implications for clinical practice The scientific-based evidence for coal tar treatment is scarce; however, trials may be found in the earlier literature (pre-1977). Contrary to belief, coal tar is not associated with an increased risk of cancer.4

In adults with hyperkeratotic hand eczema, does dithranol lead to an improvement in patient-rated and physician-rated symptom scores and longer remission periods after clearance in comparison with topical corticosteroids? No systematic reviews were found, and no trials (controlled or uncontrolled) of dithranol for any type of hand eczema were identified. Trials may be found in the earlier literature (pre-1977).

111 In adults with chronic hand eczema, do short bursts of potent topical corticosteroids (class 3 or 4) lead to better patient-rated or physician-rated symptom scores than continuous mild to moderate (class 1 or 2) topical steroids? We found no studies comparing the effect of short bursts of strong (class 3 or 4) topical corticosteroids (for example, twice weekly, or at weekends only) with continuous application of milder (class 1 or 2) topical corticosteroids. One randomized controlled trial (RCT) compared thrice-weekly application versus weekend application of the same steroid, with limited evidence that the thrice-weekly application was better. chapter

Effcacy 3 No systematic reviews were found.

Thrice-weekly versus weekend application There is limited evidence of a preferential effect of thrice-weekly application of mometasone in a parallel group RCT with 106 participants in a 30-week maintenance phase (i.e., after induction of remission).5

Once-daily versus twice-daily application In a controlled trial with once-daily halcinonide 0.1% versus twice-daily betamethasone dipropionate 0.05% both showed good efficacy in 53 participants during four weeks.6 In half of the patients, once-daily halcinonide 0.1% was superior.

Two different concentrations One left–right RCT of 2 weeks’ duration comparing different concentrations of the same corticosteroid applied twice daily detected no difference in 46 participants.2

Class 2 versus class 3 corticosteroids One RCT compared short-term (three weeks) application of fluprednidene acetate 0.1% cream (class 2) with betamethasone valerate 0.1% cream (class 3), both in a once-daily regimen in 76 participants.7 There were no differences in the time to onset of effect or clinical efficacy.

Class 2 versus class 4 corticosteroids In a double-blind left–right RCT with 61 participants, more patients remained free of relapses with clobetasol propionate than with fluprednidene acetate8. In addition, the time to relapse was longer with clobetasol propionate. Treatment initially and in case of recur- rence was twice daily; in the maintenance phase, the application was twice weekly. The side effects were comparable between the two groups.

112 Drawbacks Mild skin atrophy was reported in two studies.5,8

Comment With the exception of the study on thrice-weekly versus weekend application, all of the studies were of short duration. The primary outcome was not always clearly stated, and some studies included patients in relapse. None of the studies investigated tachyphylaxis or atrophy. Earlier reports (pre-1977) may provide some insight into this issue. chapter

Implications for clinical practice The appropriate choice of an optimal topical steroid treatment schedule cannot be derived 3 from the current literature on hand eczema trials. Evidence from studies on other eczematous diseases may have to be considered.

In adults with chronic hand eczema, are oral immunosuppressive agents (cyclosporine, methotrexate, mycophenolate mofetil, azathioprine) better in maintaining a long-term (more than 6 months) reduction of patient-rated or physician-rated symptom scores than topical corticosteroids? One RCT was identified, showing that cyclosporine was effective, but not significantly better in terms of clinical signs or quality of life. One RCT concluded that azathioprine is an effective adjunctive treatment option. We identified no RCTs studying other oral immunosuppressive agents.

Effcacy No systematic review was found.

Cyclosporine versus topical betamethasone An RCT compared cyclosporine with betamethasone dipropionate 0.05% twice daily in 41 participants.9 The study had three phases, none of which showed a comparative advantage in terms of clinical signs, global assessment, or cumulative relapse rate. The first treatment phase was six weeks; the second and third amounted to 30 weeks.

Methotrexate We identified no controlled trials. An uncontrolled case study showed low-dose methotrexate to be effective in five patients with recalcitrant palmoplantar pompholyx.10

Mycophenolate mofetil and azathioprine We identified no controlled trials.

113 Azathioprine as adjunctive One RCT compared the additional value of azathioprine to topical clobetasol propionate 0.05% cream with topical clobetasol propionate monotherapy in 108 patients with chronic hand eczema for 24 weeks.11 This study reported a significant better improvement with regard to hand eczema severity index score and itch in the azathioprine group.

Drawbacks Paresthesia (“tingling”), dizziness, facial edema, and increase in serum creatinine were re- chapter ported during cyclosporine use. Surprisingly, no side effects were reported for azathioprine.

3 Comment The comparator in the cyclosporine studies was a relatively strong corticosteroid.

Implications for clinical practice Cyclosporine may be useful for achieving short-term control, but cannot be recommended for maintenance therapy. Long-term side effects such as blood dyscrasia, renal failure, increased blood pressure, and skin malignancies are of major concerns. Choices regarding systemic treatment of hand eczema cannot be derived from the current literature on hand eczema trials.

Is the treatment of chronic hand eczema with local psoralen+ultraviolet A or ultraviolet B irradiation better in reducing patient-rated and physician-rated symptom scores than topical corticosteroids? We identified no trials explicitly comparing PUVA or UVB therapy with topical corticosteroids; only one RCT had ordinary topical treatment with emollients as the comparator. Several controlled trials were identified that compared the efficacy of PUVA, UVB, or UVA1 therapy with a control group or using a left–right design. There is insufficient evidence that PUVA or UVB therapy is more effective than conventional topical corticosteroid therapy.

Effcacy No systematic reviews were found.

Topical psoralen+ultraviolet A In a double-blind randomized within-patient trial of 15 patients with chronically relapsing vesicular hand eczema, topical PUVA and UVA treatment showed improvement of the severity score over the 8-week treatment period, but no statistical difference between the treated hands at any stage.12 In a controlled clinical trial (CCT) with a left–right design, topical cream PUVA was compared with UVA1.13 The study comprised 27 patients with

114 bilateral dyshidrotic hand eczema. Almost all patients showed a good response to both treatments, with a reduction of physician-rated scores of 50%. There were no statistically significant differences between the left and right hands. In an observer-blinded, left–right design, there was little difference between topical 8-methoxypsoralen (8-MOP) bath PUVA and topical 8-MOP lotion PUVA therapy in 24 patients with chronic hand or foot eczema; there was greater than 80% clearing with both modalities.14 After one month, the most successful treatment was continued on both sides until the lesions cleared; there were no differences in the length of the relapse-free period. An open-label RCT with 158 participants showed that oral PUVA at home was equally effective as topical bath PUVA in the hospital.15 chapter In addition, it appeared to result in lower costs and less time off work. In a within-participant RCT with 15 participants, the effectiveness of middle-dose UVA1 irradiation was compared 3 with topical cream PUVA therapy.16 Treatment was given thrice weekly during a period of five weeks. Both groups improved, but there was no significant difference between the groups. An observer-blinded RCT in 29 participants showed no significant difference between topical and oral psoralen, although oral PUVA might be preferred in hyperkeratotic hand eczema.17

Ultraviolet A1 In a double-blind RCT with 28 patients with dyshidrotic hand eczema, five-times weekly irradiation with UVA1 was compared with placebo.18 After 1 week of treatment, a significant difference was seen between the two groups, with greater efficacy for UVA1.

Ultraviolet B Eighteen patients with chronic hand eczema resistant to potent topical corticosteroids were randomly divided into three treatment groups: UVB of the hands only, placebo irradiation, and whole-body UVB irradiation.19 Local UVB irradiation of the hands was significantly better than placebo; whole-body UVB irradiation with additional irradiation of the hands was significantly better than continuing the local treatment alone (not specified), according to a simple clinical grading (cleared, improved, unchanged/worse). A 3-month follow-up period demonstrated fast relapse of the hand eczema. In an RCT with 48 patients with occupational hand eczema, UVB at home 5 days a week for 8 weeks was compared with nonspecific topical treatment.20 Physician-rated scores and transepidermal water loss improved in both groups, although the improvement did not reach statistical significance for most parameters.

Drawbacks Ultraviolet therapy can cause side effects such as burning episodes, subacute eczema, patchy hypopigmentation, and acute exacerbation of eczema. It may also induce skin cancer as a long-term effect.

115 Comment In some studies, patients continued their topical medication or emollients. There are no studies comparing PUVA, UVA1, or UVB therapy with the conventional topical corticosteroid therapy. There is also no evidence that ultraviolet therapy is the most effective for hand eczema (see the next question).

Implications for clinical practice PUVA and UVB are effective; UVA1 also appears to be effective. The choice of these treat- chapter ment options is guided by considerations other than proven clinical superiority over other 3 modalities. In adults with chronic hand eczema, does treatment with psoralen+ultraviolet A irradiation (oral or topical psoralen) lead to a better reduction in patient- rated and physician-rated symptom scores and remission periods than ultraviolet B irradiation? We identified two RCTs on oral PUVA and two CCTs on oral/topical PUVA. The controlled trial on topical bath PUVA demonstrated no comparative advantage, whereas the RCT on oral PUVA showed an effect in favor of PUVA.

Effcacy No systematic review was found.

Topical bath psoralen+ultraviolet A versus ultraviolet B A 6-week left–right design CCT including 13 patients showed that, although effective, topical bath PUVA was not better than UVB.21

Oral psoralen+ultraviolet A versus ultraviolet B An RCT showed an effect in favor of oral PUVA in a 3-month study of 35 patients. In this study, only one hand was treated, but in most patients the untreated hand also improved.22 In a 9-week RCT with left–right design, 15 patients were treated with local narrowband UVB or PUVA thrice a week. Both groups showed significant improvement compared with baseline and a local narrowband UVB phototherapy regimen is as effective as PUVA therapy in patients with chronic hand eczema of dry and dyshidrotic types.23 A CCT comparing UVB used at home with PUVA at the clinic in 26 patients during approximately 10 weeks, showed no compara¬tive advantage.24

Drawbacks Nausea caused by the oral psoralen was reported. Pain, burning, itching, and redness were

116 reported with both therapies, but slightly more from PUVA irradiation. In both PUVA and UVB, mild xerosis was reported, which responded to emollients.

Comment Long-term adverse effects could not be assessed. Improvement of the untreated hand may be the result of compliance with topical emollients. More than 17 uncontrolled studies were identified, claiming a beneficial effect of ultraviolet treatment (PUVA or UVB), but there was no comparator in any of the studies. chapter

Implications for clinical practice PUVA or UVB is effective in treating hand eczema. The question of which modality is better 3 remains unsolved.

In adults with chronic hand eczema, is oral treatment with retinoids better in terms of patient-rated and physician-rated symptom scores than topical corticosteroids? We identified several trials on the use of retinoids in hand eczema; there were no trials comparing oral retinoids with corticosteroids. Both topical and oral treatment with retinoids appeared to be effective.

Effcacy No systematic reviews were found.

Topical retinoid versus topical corticosteroids In a symmetrical double-blind nonrandomized study, the efficacy of triamcinolone acetonide 0.1% cream was compared with the same cream containing, in addition, 0.25% retinoic acid.25 The study included 18 patients with different types of eczema (12 with atopic dermatitis, four with allergic contact dermatitis, one with nummular eczema, and one with dyshidrosis); the palms and soles were involved in only five patients. The duration of treatment was planned for 2 weeks, with the option of extending the treatment to 3 weeks. No statistically significant differences were observed between the treatments. An open-label RCT with 55 patients compared bexarotene gel monotherapy (ligand for retinoid X receptors) with the same gel in combination with either mometasone furoate 0.1% ointment or with hydrocortisone acetonide 1% ointment.26 The steroids were applied twice daily, whereas bexarotene gel was applied in an increasing regimen, starting at once every other day up to three times daily, unless adjustment was needed because of irritation. All groups showed a meaningful decrease in physician-rated scores, without significant differences between the groups.

117 Oral retinoids An RCT including 29 patients with hyperkeratotic hand eczema compared once-daily 30 mg acitretin with placebo.27 A significant improvement in comparison with the placebo group was seen in relation to hyperkeratosis, fissures, and scaling, but not in relation to itch, redness, or vesicles. The improvement occurred in the first four weeks, with no additional effect seen in the following four weeks. A multicenter, double-blind RCT assessed the efficacy of three different dosages of 9-cis-retinoic acid (alitretinoin) and placebo.28 The 319 patients were equally randomized over four groups: oral alitretinoin 10 mg/day, 20 mg/day, 40 mg/ chapter day, and placebo. Alitretinoin led to a significant and dose-dependent improvement in the physician-rated score. In a large randomized trial alitretinoin (10 or 30 mg per day for up 3 to 24 weeks) was superior to placebo in 1,032 patients with chronic severe hand eczema.29 Of the patients treated with 30 mg alitretinoin, 40% rated their hand eczema as “clear” or “almost clear” at the end of therapy. In the 10 mg group this was 24% and in the placebo group 15%. In an extended open-label trial, 243 patients received 30 mg alitretinoin and the drugs remained well tolerated for overall treatment durations of up to 48 weeks.30 In addition, the beneficial effects of alitretinoin over placebo were confirmed in a retreatment trial among a subgroup of 117 patients who had relapsed.31

Drawbacks Topical use of retinoic acid plus corticosteroids is reported to cause significantly more subjective irritation than topical corticosteroids without retinoic acid.25,26 Side effects of oral acitretin are common and almost all patients experience dry skin (especially the lips). The side effects were dose dependent; the most frequently reported were headache (14%, but this was also reported in the placebo group), mucocutaneous signs as dry lips (5%), and flushing (3%). Clinically insignificant increases in serum triglyceride, cholesterol, and creatinine kinase were reported in several trials on alitretinoin. Central hypothyroidism, with no clinical expression, was observed more rarely. Several papers studied the safety of oral alitretinoin, reporting comparable side effects.32,33

Comment Oral 9-cis-retinioic acid appears to be a promising treatment option, but it remains to be demonstrated that this drug is more effective than conventional topical corticosteroids or UVB/PUVA therapy. In addition, evidence of the efficacy and safety of alitretinoin beyond 48 weeks and of the efficacy in vesicular hand eczema is lacking. A pregnancy prevention program is mandatory for women of fertile age because of the teratogenicity of oral retinoids.

118 Implications for clinical practice Oral retinoids appear to be effective in hand eczema, especially in hyperkeratotic hand eczema. However, as there is no comparison with conventional therapy, it is unclear whether it should be a therapy of first choice.

Is the treatment of chronic hand eczema with calcineurin inhibitors better in reducing patient-rated and physician-rated symptom scores than topical corticosteroids? Two RCTs were found that compared the efficacy of topical tacrolimus with mometasone chapter furoate, which appeared to be equivalent and in two RCTs tacrolimus was significantly more effective than vehicle. Two RCTs compared topical pimecrolimus with vehicle cream. 3

Effcacy No systematic reviews were found.

Tacrolimus Sixteen patients were included in a left–right RCT comparing topical tacrolimus 0.1% ointment with mometasone furoate 0.1% ointment (a class III corticosteroid).34 The treatment period was four weeks, with a follow-up period up to 8 weeks. Both treatments led to a statistically significant decrease in clinical severity, with no significant differences between the groups. In another RCT, topical tacrolimus 0.1% was compared with mometasone furoate in 30 patients for 90 days.35 Both treatments showed similar therapeutic results. A randomized pilot study in 32 patients with moderate to severe hand eczema suggested that tacrolimus might prolong the time until relapse compared with vehicle cream during 14 weeks.36 In an RCT, 28 participants with moderate to severe nickel sulfate-induced allergic hand eczema were treated with 0.1% tacrolimus ointment twice daily versus twice daily vehicle ointment during 14 days. The symptom scores were significantly lower in the tacrolimus group compared with the vehicle group during treatment and seven days afterwards.37

Pimecrolimus In a multicenter RCT, 294 patients with hand eczema were allocated to up to 3 weeks’ treatment with pimecrolimus 1% cream or to vehicle.38 Twice-daily application of the study creams (evening application under occlusion) was continued until clearance or completion of three weeks’ treatment. The efficacy of pimecrolimus 1% cream increased over time, while that of the vehicle stagnated after the second week. There were no statistically significant differences between the two groups, unless stratification for palmar involvement was applied; pimecrolimus 1% cream was superior in these patients. In a multicenter, double-blind, RCT 652 patients were treated for 6 weeks with pimecrolimus 1% or vehicle cream twice

119 daily with overnight occlusion, followed by a 6-week open-label pimecrolimus treatment.39 There were no significant differences with regard to disease signs; however, pruritus relief was significantly better in the pimecrolimus group.

Drawbacks Tacrolimus 0.1% ointment produced stinging upon application; with pimecrolimus 1% cream, however, this was uncommon (0.7% vs 2.1% in the vehicle group) or comparable to the vehicle cream.39 chapter

Comment 3 The effect of pimecrolimus 1% cream in comparison with the vehicle might have reached significance if the follow-up period had been longer, as the efficacy of pimecrolimus 1% cream increased over time. Pruritus relief was greater in the pimecrolimus group, but the other studies did not include patient-rated scores.

Implications for clinical practice Topical tacrolimus is more effective than vehicles, but is at best equally effective as topical corticosteroids. Pimecrolimus is as effective as vehicles. With the present evidence, calcineurin inhibitors may be used for rotational therapy with topical corticosteroids, with potent corticosteroids for (severe) exacerbations and topical calcineurin inhibitors in the maintenance phase and for mild exacerbations.

In adults with dyshidrotic hand eczema, does iontophoresis lead to an improvement in patient-rated and physician-rated symptom scores in comparison with topical steroids or ultraviolet B/psoralen+ultraviolet A irradiation? We identified only one RCT using iontophoresis in patients with dyshidrotic hand eczema, showing a significant improvement on the iontophoresis-treated side in comparison with the untreated side. We found no trials comparing iontophoresis with topical corticosteroids or UVB/PUVA therapy.

Effcacy No systematic reviews were found.

Iontophoresis versus no treatment In a randomized left–right comparison, the effects of tap-water iontophoresis in addition to steroid-free topical therapy were investigated in 20 patients with dyshidrotic hand eczema.40 After three weeks (20 iontophoresis applications), the parameters “itching” and “vesicle for-

120 mations” scored significantly better on the iontophoresis-treated side than on the untreated side, but redness and desquamation did not differ significantly.

Drawbacks Tap-water iontophoresis was always associated with subjective sensations such as stinging and discrete paresthesia (“tingling”). No severe side effects or possible harmful effects were reported.

Comment chapter We found insufficient evidence for the benefit of additional iontophoresis therapy in comparison with conventional corticosteroid or UVB/PUVA therapy. 3

Implications for clinical practice Iontophoresis appears to be harmless, but has not been proved to be effective.

In adults with relapsing vesicular hand eczema based on contact allergy to nickel, does dietary intervention or oral therapy with chelating agents lead to an improvement in patient-rated and physician-rated symptom scores in comparison with topical corticosteroids? We identified three small RCTs and one CCT. None of the studies compared the intervention with topical corticosteroids. An RCT on triethylenetetramine found no significant improvement of hand eczema. Another RCT, on tetraethylthiuram disulfide (disulfiram), found only very limited evidence in favor of this treatment. One controlled trial found no evidence that a low-nickel diet improves dyshidrotic hand eczema.

Effcacy No systematic reviews were found.

Oral therapy with a nickel-chelating compound In a multicenter, randomized, double-blind, crossover study, oral treatment with triethylenetetramine 300 mg daily for a 6-week period or placebo were given to 23 nickel-positive patients with chronic hand eczema.41 No significant improvement occurred in hand eczema on the basis of either the patients’ or the doctors’ evaluations. The study was terminated prematurely because of literature reports on teratogenicity in rats. In a double-blind, placebo-controlled RCT, tetraethylthiuram disulfide at a gradually increased dosage was given for at least six weeks after the full dosage of 200 mg had been reached.42 During the treatment period, the hand eczema healed in five of the 11 patients treated with tetraethylthiuram disulfide in comparison with two of 13 in the placebo group (not significant).

121 Using a semi quantitative scoring system, the results in favor of tetraethylthiuram disulfide were statistically significant for scaling and frequency of flares, but not for the sum of the parameters.

Low-nickel diet In a nonrandomized trial including 24 patients with dyshidrotic hand eczema caused by nickel, the effects of a low-nickel diet for three months (eight patients) were compared with oral disodium cromoglycate for three months (nine patients) and with seven patients who did not chapter give consent to the study and did not receive any treatment.43 All 24 patients were evaluated blindly for itching and number of vesicles. The low-nickel diet did not lead to improvement 3 in these patients, but those treated with disodium cromoglycate improved significantly and had fewer vesicles than the controls and the patients treated by diet.

Combination A randomized placebo-controlled trial including 21 patients with chronic vesicular hand eczema with nickel sensitivity stated that the combination of low-nickel diet and short course of oral disulfiram therapy reduced severity of hand eczema statistically after four weeks.44

Drawbacks One patient treated with disulfiram had toxic hepatitis after 8 weeks of treatment and two of 30 patients showed signs of hepatic toxicity.42 Another RCT only noted slight, transient elevation of liver enzymes after four weeks.44 The study on triethylenetetramine was ended prematurely due to a literature report on teratogenicity in rats.41 No studies using a low- nickel diet assessed possible harmful effects.

Comment None of the trials showed sufficient evidence for the benefit of either a low-nickel diet or a nickel-chelating compound. Only four CTs with small numbers of patients were carried out. On the basis of the harm and possible side effects, oral treatment with a nickel-chelating compound cannot be recommended. None of the trials compared treatments with conventional topical medication (for example, corticosteroids).

Implications for clinical practice Given the side effects and the lack of efficacy, oral therapy with a nickel-chelating compound cannot be recommended. There is no evidence that a low-nickel diet improves pompholyx- type hand eczema.

122 Does the daily application of a bland emollient lead to dose reduction and/ or frequency reduction of topical corticosteroids in adults with chronic hand eczema? One RCT and two uncontrolled studies compared emollients with topical corticosteroids, showing better clinical assessments for emollients, albeit not significant.

Effcacy No systematic review was found. chapter

Emollient versus topical corticosteroids A double-blind RCT of two weeks’ duration compared twice-daily application of betametha- 3 sone valerate 0.1% with once daily application of betamethasone valerate 0.1% in addition to maintenance therapy with a 5% urea-containing moisturizer in 44 participants. Once-daily treatment combined with emollients showed a better clinical assessment (albeit not statistically significant) compared with twice-daily treatment, especially in the group of patients with a moderate eczema at inclusion.45

Versus each other In one left–right RCT with 30 participants, using patient preference as the outcome parameter, there was limited evidence in favor of Aquacare-HP over Calmurid, both of which contain 10% urea.46 An RCT in 32 participants confirmed that the frequent application of emollients resulted in better hand eczema scores.47 However, a superior effect of emollient with ceramides versus a regular petrolatum-based emollient was not demonstrated. This RCT showed that an emollient with ceramides was able to reduce the use of topical corticosteroids. The beneficial effect of emollients on hand eczema was also seen in a CCT comparing two bland emollients.48 There was a decrease in transepidermal water loss, as well as an improvement in physician-rated and patient-rated severity scores. Uncontrolled studies noted a reduction in steroid use in patients treated with a moisturizing cream and in patients treated with a protective foam in 31 and 37 patients.45,49

Drawbacks No major side effects were reported. Burning and worsening of the preexisting hand eczema were reported.46 Patients were concerned about greasiness of their hands and with staining of objects they handled.

Comments Several poor-quality uncontrolled studies were also identified, none of which had steroid dose reduction as the outcome parameter.

123 Implications for clinical practice Despite the widespread use of emollients, there is only little evidence of any steroid-sparing or additive effect in the treatment of hand eczema. In general, there seems to be no harm either, apart from occasional contact allergy against an ingredient.

In adults with chronic clinically active hand eczema, do protective or occlusive gloves or barrier creams lead to better patient-rated or physician-rated symptom scores than topical steroids? chapter Information on avoidance of allergens or irritants on a case-by-case basis can be found in the major textbooks on contact dermatitis. The effect of emollients was covered in the 3 previous question. One controlled trial on protective creams was found. We found a few uncontrolled, rather descriptive studies indicating some benefit of gloves and/or barrier creams, and one study used a within-patient left–right design.50,51

Effcacy A number of issues in connection with this question are dealt with in a Cochrane systematic review on irritant hand eczema.52 They concluded that industrial barrier creams may have a similar role as emollients in the prevention of occupational contact dermatitis; however, there is insufficient evidence that it has a long-term protective effect.

Barrier creams In a randomized open trial in 53 patients a barrier-strengthening moisturizer (5% urea) seemed to prolong the disease-free interval in patients with controlled hand eczema compared with no treatment at all (20 vs 2 days).53

Comments Protective gloves offer protection when manual (wet) tasks are performed; however, prolonged occlusion may be a risk factor for hand eczema.54 Based on practical experience, supported by an experimental study on the healthy skin of volunteers, a cotton lining or inner glove is recommended.55

Implications for clinical practice Insufficient evidence for long-term protective effect.

Does the addition of a topical antibacterial agent to topical corticosteroids result in better patient-rated or physician-rated symptom scores than topical corticosteroids alone? No trials compared the additional effect of topical antibacterial agents with topical corticos-

124 teroids alone. Only one RCT comparing betamethasone cream with the addition of either fusidic acid or clioquinol was found, showing a similar effect on clinical severity.

Effcacy No systematic reviews were found.

Addition of fusidic acid or clioquinol to betamethasone In a multicenter open-label RCT with 120 patients, four weeks’ twice-daily application of betamethasone 0.1% clioquinol 3% cream was compared with betamethasone 0.1% fusidic chapter acid 2% cream.56 The two preparations were equally effective in reducing the observer-rated severity score. However, the combination of betamethasone cream and fusidic acid pro- 3 duced a better bacteriological response.

Drawbacks Betamethasone 0.1% fusidic acid 2% cream was considered more cosmetically acceptable than the preparation with clioquinol. The difference was statistically highly significant. Staining of the skin and clothing were the major problems.

Comment Staphylococcal superantigens in infected areas elsewhere on the body, although the study protocol allowed them to be treated, might have had an effect on the hands.

Implications for clinical practice There were no comparisons of a corticosteroid with a combination of corticosteroid and antibacterial agents. Evidence of an additional effect of antibacterial agents in patients with hand eczema is still lacking.

In patients with chronic hand eczema, is an educational intervention program effective for secondary and tertiary prevention on hand eczema compared with limited education as usual? Several RCTs have been conducted on primary prevention, and a systematic review concluded that there is moderate evidence for the effectiveness of primary prevention programs. However, there is low evidence for the effect on improving clinical and self- reported outcomes.57 We identified two RCTs, one CCT, and one controlled trial on secondary or tertiary prevention.

125 Effcacy A Cochrane review concluded that, although the findings of the review were generally positive, there is insufficient evidence, at present, for the effectiveness of most of the treatments identified for primary prevention of occupation-induced hand eczema in the workplace.52 No systematic reviews were identified on secondary or tertiary prevention.

Education programs An RCT in 255 hospital workers with symptoms of hand eczema showed a beneficial effect of chapter skin care education and individual counseling compared with care as usual (only topical steroids, emollients, and care by general practitioner) after six months.58,59 In a CCT, 209 geri- 3 atric nurses with hand eczema received either care as usual by a dermatologist or a person- alized secondary intervention program with education concerning barrier cream and gloves with lectures and hands-on training during four visits. After three months the hand eczema significantly improved in the intervention group. More nurses were able to continue their job because of the invention program compared with those without the program (96 vs 86%).60 The Tertiary Individual Prevention study is a multicenter trial with an interdisciplinary, integrated inpatient rehabilitation measure with three weeks of inpatient treatment. The 1-year data from 1,617 individuals revealed a significant reduction in the severity of occupational skin diseases, the use of topical corticosteroids, and days of sick leave: 87.4% were able to return to work and to remain in the workforce.61

Integrated care A multicenter RCT with 196 patients received usual care by a dermatologist or integrated care. The integrated care included allergo-dermatological evaluation by a dermatologist, occupational intervention by a clinical occupational physician, and counseling by a specialized nurse on optimizing topical treatment and skin care. After 26 weeks the intervention group scored significantly better on clinical effectiveness, but not on quality of life and days of sick leave.62

Drawbacks Participation in the TIP study was voluntary, but patients were obliged to cooperate with the respective insurance organizations, which makes it difficult to compare this study with other countries.

Comment In several studies, no comparator was used. It is difficult to compare different educational programs and to implement in daily practice. Three weeks of inpatient treatment might not be feasible in every setting.

126 Implications for clinical practice Education and counseling seems to improve the effectiveness of hand eczema care and prevent relapse; however, long-term efficacy is still under investigation.

chapter 3

127 References 1. Thyssen J, Johansen J, Linneberg A, Menné T. The epidemiology of handeczema in the general population--prevalence and main findings.Contact dermatitis 2010:62:75-82. 2. Uggeldahl PE, Kero M, Ulshagen K, Solberg VM. Comparative effects of desonide cream 0.1% and 0.05% in patients with hand eczema. Curr Therap Res 1986:40:969-73. 3. Kemper MFP, Van der Valk PGM, Arnold WP. Effectivity of coal tar paste versus betamethasone valerate ointment 0.1% in hand eczema. Nederlands tijdschrift voor Dermatologie en Venereologie 1998:8:289-291. chapter 4. Roelofzen JH, Aben KK, Oldenhof UT, Coenraads PJ, Alkemade HA, van de Kerkhof PC, et al. 3 No increased risk of cancer after coal tar treatment in patients with psoriasis or eczema. J Invest Dermatol 2010:130:953-961. 5. Veien NK, Olholm Larsen P, Thestrup-Pedersen K, Schou G. Long-term, intermittent treatment of chronic hand eczema with mometasone furoate. Br J Dermatol 1999:140:882-886. 6. Levy A. Comparison of 0.1% halcinonide with 0.05% betamethasone dipropionate in the treatment of acute and chronic dermatoses. Curr Med Res Opin 1977:5:328-332. 7. Bleeker J, Anagrius C, Iversen N, Stenberg B, Cullberg Valentin K. Double-blind comparative study of Corticoderm cream + Unguentum Merck and Betnovate cream + Unguentum Merck in hand dermatitis. J Dermatol Treat 1989:1:87-90. 8. Möller H, Svartholm H, Dahl G. Intermittent maintenance therapy in chronic hand eczema with clobetasol propionate and flupredniden acetate.Curr Med Res Opin 1983:8:640-644. 9. Granlund H, Erkko P, Eriksson E, Reitamo S. Comparison of cyclosporine and topical betamethasone-17,21-dipropionate in the treatment of severe chronic hand eczema. Acta Derm Venereol 1996:76:371-376. 10. Egan CA, Rallis TM, Meadows KP, Krueger GG. Low-dose oral methotrexate treatment for recalcitrant palmoplantar pompholyx. J Am Acad Dermatol 1999:40:612-614. 11. Agarwal US, Besarwal RK. Topical clobetasol propionate 0.05% cream alone and in combination with azathioprine in patients with chronic hand eczema: an observer blinded randomized comparative trial. Indian J Dermatol Venereol Leprol 2013:79:101-103. 12. Grattn CE, Carmichael AJ, Shuttleworth GJ, Foulds IS. Comparison of topical PUVA with UVA for chronic vesicular hand eczema. Acta Derm Venereol 1991:71:118-122. 13. Petering H, Breuer C, Herbst R, Kapp A, Werfel T. Comparison of localized high-dose UVA1 irradiation versus topical cream psoralen-UVA for treatment of chronic vesicular dyshidrotic eczema. J Am Acad Dermatol 2004:50:68-72. 14. Shephard SE, Schregenberger N, Dummer R, Panizzon RG. Comparison of 8-MOP aqueous bath and 8-MOP ethanolic lotion (Meladinine) in local PUVA therapy. Dermatology 1998:197:25-30.

128 15. van Coevorden AM, Kamphof WG, van Sonderen E, Bruynzeel DP, Coenraads PJ. Comparison of oral psoralen-UV-A with a portable tanning unit at home vs hospital-administered bath psoralen-UV-A in patients with chronic hand eczema: an open-label randomized controlled trial of efficacy.Arch Dermatol 2004:140:1463-1466. 16. Adams S, Bayerl C. Medium dose UVA-1 irradiation and topical PUVA therapy in chronic dyshidrotic hand dermatitis - A prospective study. Aktuelle Dermatologie 2007:33:142-5. 17. Tzaneva S, Kittler H, Thallinger C, Honigsmann H, Tanew A. Oral vs. bath PUVA using 8-methoxypsoralen for chronic palmoplantar eczema. Photodermatol Photoimmunol Photomed 2009:25:101-105. chapter 18. Polderman MC, Govaert JC, le Cessie S, Pavel S. A double-blind placebo-controlled trial of UVA-1 in the treatment of dyshidrotic eczema. Clin Exp Dermatol 2003:28:584-587. 3 19. Sjövall P, Christensen OB. Local and systemic effect of UVB irradiation in patients with chronic hand eczema. Acta Derm Venereol 1987:67:538-541. 20. Bayerl C, Garbea A, Peiler D, Rzany B, Allgäuer T, Kleesz P, et al. Pilotstudie zur Therapie des beruflich bedingten Handekzems mit einer neuen tragbaren UVB Bestrahlungseinheit. Akt Dermatol 1999:25:302-5. 21. Simons JR, Bohnen IJ, van der Valk PG. A left-right comparison of UVB phototherapy and topical photochemotherapy in bilateral chronic hand dermatitis after 6 weeks’ treatment. Clin Exp Dermatol 1997:22:7-10. 22. Rosén K, Mobacken H, Swanbeck G. Chronic eczematous dermatitis of the hands: a comparison of PUVA and UVB treatment. Acta Derm Venereol 1987:67:48-54. 23. Sezer E, Etikan I. Local narrowband UVB phototherapy vs. local PUVA in the treatment of chronic hand eczema. Photodermatol Photoimmunol Photomed 2007:23:10-14. 24. Sjövall P, Christensen OB. Treatment of chronic hand eczema with UV-B Handylux in the clinic and at home. Contact Dermatitis 1994:31:5-8. 25. Schmied C, Piletta PA, Saurat JH. Treatment of eczema with a mixture of triamcinolone acetonide and retinoic acid: a double-blind study. Dermatology 1993:187:263-267. 26. Hanifin JM, Stevens V, Sheth P, Breneman D. Novel treatment of chronic severe hand dermatitis with bexarotene gel. Br J Dermatol 2004:150:545-553. 27. Thestrup-Pedersen K, Andersen KE, Menné T, Veien NK. Treatment of hyperkeratotic dermatitis of the palms (eczema keratoticum) with oral acitretin. A single-blind placebo-controlled study. Acta Derm Venereol 2001:81:353-355. 28. Ruzicka T, Larsen FG, Galewicz D, Horvath A, Coenraads PJ, Thestrup-Pedersen K, et al. Oral alitretinoin (9-cis-retinoic acid) therapy for chronic hand dermatitis in patients refractory to standard therapy: results of a randomized, double-blind, placebo-controlled, multicenter trial. Arch Dermatol 2004:140:1453-1459.

129 29. Ruzicka T, Lynde CW, Jemec GB, Diepgen T, Berth-Jones J, Coenraads PJ, et al. Efficacy and safety of oral alitretinoin (9-cis retinoic acid) in patients with severe chronic hand eczema refractory to topical corticosteroids: results of a randomized, double-blind, placebo-controlled, multicentre trial. Br J Dermatol 2008:158:808-817. 30. Lynde C, Cambazard F, Ruzicka T, Sebastian M, Brown TC, Maares J. Extended treatment with oral alitretinoin for patients with chronic hand eczema not fully responding to initial treatment. Clin Exp Dermatol 2012:37:712-717. 31. Bissonnette R, Worm M, Gerlach B, Guenther L, Cambazard F, Ruzicka T, et al. Successful chapter retreatment with alitretinoin in patients with relapsed chronic hand eczema. Br J Dermatol 2010:162:420-426. 3 32. Schmitt-Hoffmann AH, Roos B, Sauer J, Spickermann J, Stoeckel K, Edwards D, et al. Pharmacokinetics, efficacy and safety of alitretinoin in moderate or severe chronic hand eczema. Clin Exp Dermatol 2011:36 Suppl 2:29-34. 33. Dirschka T, Reich K, Bissonnette R, Maares J, Brown T, Diepgen TL. An open-label study assessing the safety and efficacy of alitretinoin in patients with severe chronic hand eczema unresponsive to topical corticosteroids. Clin Exp Dermatol 2011:36:149-154. 34. Schnopp C, Remling R, Möhrenschlager M, Weigl L, Ring J, Abeck D. Topical tacrolimus (FK506) and mometasone furoate in treatment of dyshidrotic palmar eczema: a randomized, observer- blinded trial. J Am Acad Dermatol 2002:46:73-77. 35. Katsarou A, Makris M, Papagiannaki K, Lagogianni E, Tagka A, Kalogeromitros D. Tacrolimus 0.1% vs mometasone furoate topical treatment in allergic contact hand eczema: a prospective randomized clinical study. Eur J Dermatol 2012:22:192-196. 36. Krejci-Manwaring J, McCarty MA, Camacho F, Manuel J, Hartle J, Fleischer A,Jr, et al. Topical tacrolimus 0.1% improves symptoms of hand dermatitis in patients treated with a prednisone taper. J Drugs Dermatol 2008:7:643-646. 37. Pacor ML, Di Lorenzo G, Martinelli N, Mansueto P, Friso S, Pellitteri ME, et al. Tacrolimus ointment in nickel sulphate-induced steroid-resistant allergic contact dermatitis. Allergy Asthma Proc 2006:27:527-531. 38. Belsito DV, Fowler JF,Jr, Marks JG,Jr, Pariser DM, Hanifin J, Duarte IA, et al. Pimecrolimus cream 1%: a potential new treatment for chronic hand dermatitis. Cutis 2004:73:31-38. 39. Hordinsky M, Fleischer A, Rivers JK, Poulin Y, Belsito D, Hultsch T. Efficacy and safety of pimecrolimus cream 1% in mild-to-moderate chronic hand dermatitis: a randomized, double-blind trial. Dermatology 2010:221:71-77. 40. Odia S, Vocks E, Rakoski J, Ring J. Successful treatment of dyshidrotic hand eczema using tap water iontophoresis with pulsed direct current. Acta Derm Venereol 1996:76:472-474. 41. Burrows D, Rogers S, Beck M, Kellett J, McMaster D, Merrett D, et al. Treatment of nickel dermatitis with Trientine. Contact Dermatitis 1986:15:55-57.

130 42. Kaaber K, Menné T, Veien N, Hougaard P. Treatment of nickel dermatitis with Antabuse; a double blind study. Contact Dermatitis 1983:9:297-299. 43. Pigatto PD, Gibelli E, Fumagalli M, Bigardi A, Morelli M, Altomare GF. Disodium cromoglycate versus diet in the treatment and prevention of nickel-positive pompholyx. Contact Dermatitis 1990:22:27-31. 44. Sharma AD. Disulfiram and low nickel diet in the management of hand eczema: a clinical study. Indian J Dermatol Venereol Leprol 2006:72:113-118. 45. Lodén M, Wirén K, Smerud KT, Meland N, Hønnås H, Mørk G, et al. The effect of a corticosteroid cream and a barrier-strengthening moisturizer in hand eczema. A double-blind, randomized, chapter prospective, parallel group clinical trial. J Eur Acad Dermatol Venereol 2012:26:597-601. 46. Fredriksson T, Gip L. Urea creams in the treatment of dry skin and hand dermatitis. Int J Dermatol 3 1975:14:442-444. 47. Kucharekova M, Van De Kerkhof PC, Van Der Valk PG. A randomized comparison of an emollient containing skin-related lipids with a petrolatum-based emollient as adjunct in the treatment of chronic hand dermatitis. Contact Dermatitis 2003:48:293-299. 48. Bock M, Wulfhorst B, Gabard B, Schwanits HJ. Effektivität von Hautschutzcremes zur Behandlung irritativer Kontaktekzeme bei Friseurauszubildenden. Dermatol Beruf Umwelt 2001:49:73-6. 49. Fowler JF,Jr. A skin moisturizing cream containing Quaternium-18-Bentonite effectively improves chronic hand dermatitis. J Cutan Med Surg 2001:5:201-205. 50. Baack BR, Holguin TA, Holmes HS, Prawer SE, Scheman AJ. Use of a semipermeable glove during treatment of hand dermatitis. Cutis 1996:58:423-424. 51. Schleicher SM, Milstein HJ, Ilowite R, Meyer P. Response of hand dermatitis to a new skin barrier- protectant cream. Cutis 1998:61:233-234. 52. Bauer A, Schmitt J, Bennett C, Coenraads PJ, Elsner P, English J, et al. Interventions for preventing occupational irritant hand dermatitis. Cochrane Database Syst Rev. 2010:16:CD004414. 53. Lodén M, Wirén K, Smerud K, Meland N, Hønnås H, Mørk G, et al. Treatment with a barrier- strengthening moisturizer prevents relapse of hand-eczema. An open, randomized, prospective, parallel group study. Acta Derm Venereol 2010:90:602-606. 54. Kwon S, Campbell LS, Zirwas MJ. Role of protective gloves in the causation and treatment of occupational irritant contact dermatitis. J Am Acad Dermatol 2006:55:891-896. 55. Ramsing DW, Agner T. Effect of glove occlusion on human skin (II). Long-term experimental exposure. Contact Dermatitis 1996:34:258-262. 56. Hill VA, Wong E, Corbett MF, Menday AP. Comparative efficacy of betamethasone/clioquinol (Betnovate-C) cream and betamethasone/fusidic acid (Fucibet) cream in the treatment of infected hand eczema. J Dermatol Treat 1998:9:15-9. 57. van Gils RF, Boot CR, van Gils PF, Bruynzeel D, Coenraads PJ, van Mechelen W, et al. Effectiveness of prevention programmes for hand dermatitis: a systematic review of the literature. Contact Dermatitis 2011:64:63-72.

131 58. Ibler KS, Agner T, Hansen JL, Gluud C. The Hand Eczema Trial (HET): Design of a randomised clinical trial of the effect of classification and individual counselling versus no intervention among health-care workers with hand eczema. BMC Dermatol 2010:10:8. 59. Ibler KS, Jemec GB, Diepgen TL, Gluud C, Lindschou Hansen J, Winkel P, et al. Skin care education and individual counselling versus treatment as usual in healthcare workers with hand eczema: randomised clinical trial. BMJ 2012:345:e7822. 60. Schürer NY, Klippel U, Schwanitz HJ. Secondary individual prevention of hand dermatitis in geriatric nurses. Int Arch Occup Environ Health 2005:78:149-157. chapter 61. Weisshaar E, Skudlik C, Scheidt R, Matterne U, Wulfhorst B, Schonfeld M, et al. Multicentre Study ‘Rehabilitation of Occupational Skin Diseases – Optimisation and Quality Assurance of Inpatient 3 Management (ROQ)’ – results from 12-month follow-up. Contact dermatitis 2013:68:169-174. 62. van Gils RF, Boot CR, Knol DL, Rustemeyer T, van Mechelen W, van der Valk PG, et al. The effectiveness of integrated care for patients with hand eczema: results of a randomized, controlled trial. Contact Dermatitis 2012:66:197-204.

132 chapter 3

133 Chapter4

134 Drug survival of cyclosporine in the treatment of hand eczema: a multicenter, daily use study chapter Manuscript submitted for publication to JEADV 4

Wietske Andrea Christoffers1, Klaziena Politiek1, Pieter-Jan Coenraads1, Jorien van der Schaft2, Marjolein de Bruin-Weller2, Marie-Louise Anna Schuttelaar1

1 Department of Dermatology, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands 2 Department of Dermatology & Allergology, University Medical Center Utrecht, Utrecht, the Netherlands.

Key words: hand eczema, cyclosporine, drug survival, vesicular eczema

135 Abstract Background: Hand eczema is a common condition; it is often chronic and can be difficult to treat. Cyclosporine is used off-label to treat severe hand eczema; however the evidence for this treatment is scarce. Objective: To examine the drug survival of cyclosporine in a daily practice cohort of patients with chronic hand eczema. Methods: This daily use study included hand eczema patients who were treated with cyclosporine between 01-06-1999 and 01-06-2014 in two Dutch university hospitals. Patient and treatment characteristics were retrospectively collected from medical charts. First treatment episodes were analysed by means of Kaplan-Meier drug survival curves and possible determinants of drug survival were analysed by Cox-regression models. chapter Results: A total of 102 patients was treated with cyclosporine. The median survival rate was 0.86 years. The overall drug survival rates after 6 months, 1, 2 and 3 years were 61.7%, 4 45.2%, 18.6% and 13.9%, respectively. Main reasons for discontinuation were adverse events, especially early in treatment, and ineffectiveness. After three months a good response to treatment was recorded in 59.8% of the patients. Conclusion: Cyclosporine is a valuable treatment option for patients with hand eczema, especially in patients with recurrent vesicular hand eczema. Moreover, cyclosporine can be used for longer periods of time without the occurrence of serious adverse events.

136 Introduction Hand eczema is a common condition affecting approximately 10% of the population.1 Topical corticosteroids are the mainstay of treatment.2 However, an estimated 3-5% of hand eczema patients is unresponsive to topical treatment. Various guidelines recommend either phototherapy and/ or systemic treatment for these patients.3-5 Although, except for one type of retinoid6-8, the evidence for systemic therapy is limited and head to head trials are lacking.9-11 Therefore clinicians base their choice of treatment often on clinical experience. Cyclosporine is a cyclic polypeptide with a potent immunosuppressive effect. It specifically inhibits the proliferation of T-lymphocytes, without influencing phagocyting cells and without suppressing hematopoiesis. Cyclosporine is the only registered systemic treatment for severe atopic dermatitis, but is not registered for the treatment of hand eczema.

In this retrospective daily use study the drug survival of cyclosporine was investigated. chapter Drug survival is the period of time during which a patient remains on a specific treatment.12 This depends on a combination of factors such as drug effectiveness, occurrence of adverse 4 events, patient satisfaction and availability of alternative treatments.12,13 If a treatment is effective, the treatment will be continued, or the patient will discontinue over time due to disease control. Therefore, drug survival is a useful outcome parameter for treatment success which can be used retrospectively. In addition, a retrospective clustered Physician’s Global Assessment (PGA) was used. The aim of this study was to examine the effectiveness of cyclosporine by means of drug survival in a daily practice cohort of patients with chronic hand eczema.

Material & Methods Subjects & treatment This retrospective study was conducted at the Dermatology Departments of the University Medical Center Groningen (UMCG) and the University Medical Center Utrecht (UMCU). To be included, patients had to fulfill the following criteria: a clinical diagnosis of chronic hand eczema and treatment with cyclosporine between 01-06-1999 and 01-06-2014 (data-lock). In the UMCG patients were identified based on the clinical diagnosis and the hematological monitoring of creatinine, since this is mandatory during cyclosporine treatment. In the UMCU patients were identified based on the clinical diagnosis combined with prescription records. Patients with doubtful hand eczema or psoriasis on the remaining of the body were excluded. Patients with predominantly atopic dermatitis in which the hands were also involved, were excluded as well and will be discussed in a separate paper. Two different dosages schedules were used: step-up and step-down regime. In the step-up regime the dermatologist starts with a low-intermediate dosage of cyclosporine (≤ 3.5 mg/

137 kg/day) and gradually increases the dosages in the following weeks to months, which was more often used in the UMCG in the past. In the step-down regime the start dosages was higher (>3.5 mg/kg), and slowly tapered over time, which was more often prescribed in the UMCU and recently in the UMCG. During treatment with cyclosporine, patients used topical emollients and/or topical corticosteroids. Whenever the hand eczema was controlled, the dosage of cyclosporine was gradually tapered.

Data collection Medical records were retrospectively analysed with regard to patient characteristics, type of medication, start and maintenance dosages (<3.5 mg/kg/day or >3.5 mg/kg/day), start and stop dates, adverse events, laboratory parameters, co-medication, patch tests and contact chapter sensitizations. Non-naive patients were treated with another systemic agent in the past, such 4 as acitretin or azathioprine. Hand eczema was classified according to clinical diagnoses of the Danish Contact Dermatitis Group: chronic fissured hand eczema, recurrent vesicular hand eczema, hyperkeratotic palmar eczema, pulpitis, interdigital eczema and nummular hand eczema.3 Retrospective classification was conducted based on the flow chart of Boonstraet al. in addition to the doctors diagnosis.14 The effectiveness of treatment was retrospectively assessed after three months of treatment according to the clustered physician’s global assessment (PGA) score of Hijnen et al.15 This clustered PGA ranges 1 to 3: • PGA 1: good effect; more than 50% improvement, • PGA 2: moderate effect; patients with less than 50% improvement, • PGA 3: failure of treatment; no improvement or worse.

Drug survival and statistical analyses Data analyses were performed with IBM SPSS Statistics (SPSS, Chicago, IL, USA) for Windows (version 22.0). Chi squared test and Mann-Whitney U test were used whenever appropriate. Duration of drug survival was defined as the number of days which an individual continued the first treatment episode. Treatment interruptions up to 14 days were allowed and considered as one continuous treatment episode. Reasons for drug discontinuation were registered. For each subject only the first treatment episode was analysed. Drug survival was descriptively analysed using Kaplan-Meier survival curves. Graphpad prism version 5.04 was used to create Kaplan Meier survival curves. Separate analyses were conducted for discontinuation in general (a), discontinuation due to adverse events (b) and due to ineffectiveness (c). In the overall analysis (a), subjects were censored when still active

138 at the moment of data lock or when lost to follow-up. In sub analyses (b, c) subjects were censored for all other reasons than the reason of interest. Subjects that discontinued because of both adverse events and ineffectiveness were considered to have an event in both sub analyses (b, c). In addition, the influence of the following determinants was analysed by univariate Cox- regression models: gender, age, prior immunosuppressive drugs, dosage, co-medication and clinical type of hand eczema. Determinants with p-values <0.2 were entered in a multivariate Cox-regression model. By backward selection, a full model was built to identify independent determinants which influence the overall drug survival for cyclosporine. Missing values were excluded from analyses. Determinants with a p-value <0.05 were considered statistically significant.

chapter Results Patient characteristics 4 A first episode of cyclosporine was prescribed to 102 hand eczema patients, of whom 71 patients were treated in the UMCG and 31 patients in the UMCU. Cyclosporine was often prescribed to patients with recurrent vesicular hand eczema (64.7%) and 42.2% had a physician-diagnosed history of atopic dermatitis (Table 1).

n (%) Male gender 57 (55.9) Age over 40 years 62 (60.8) Mean age (SD) 44.7 (±13.7) Prior oral immunosuppressive drugs 21 (20.6) At least 1 positive patch test reaction 64 (62.7) (12 missing) A history of atopic dermatitis 43 (42.2) (1 missing) Concomitant potent topical corticosteroids 78 (76.5) Start dosages <3.5 mg/kg/day 50 (49.0) (17 missing) Morphology • Recurrent vesicular hand eczema 66 (64.7) • Hyperkeratotic palmar eczema 15 (14.7) • Chronic fissured dry hand eczema 12 (11.8) • Nummular hand eczema 1 (1.0) • Pulpitis 1 (1.0) • Non classifiable 7 (6.9)

Table 1 Demographic characteristics of included patients

139 Overall drug survival rates of cyclosporine Fig. 1 shows the drug survival for cyclosporine in patients with hand eczema. The overall drug survival rates after 6 months, 1, 2 and 3 years were 61.7%, 45.2%, 18.6% and 13.9%, respectively. The median survival duration for cyclosporine in hand eczema was 0.86 year (10.3 months). The treatment duration ranged from 6 days to 13.7 years.

chapter 4

Fig. 1 Cyclosporine drug survival

Determinants of drug survival Table 2 shows results from the univariate Cox-regression analyses. Potential determinants associated with the overall drug survival for cyclosporine were: male sex, concomitant use of potent topical corticosteroids, recurrent vesicular hand eczema, prior oral immunosuppressive medication and at least one positive patch test. The multivariate Cox regression analysis demonstrated that only male sex was an independent determinant associated with a longer overall drug survival for cyclosporine (HR 0.536, 95%CI 0.334-0.861). Recurrent vesicular hand eczema was borderline significant.

Effectiveness based on PGA The majority of the patients (59.8%) showed an improvement in their hand eczema of at least 50% after three months of treatment (PGA1). Gender, age, being naïve, clinical type of hand eczema and co-medication did not influence PGA score of cyclosporine after three months (data not shown). In the group of patients with recurrent vesicular hand eczema 66.7% reached a PGA1 score.

140 With regards to drug survival, ineffectiveness was a reason to discontinue cyclosporine treatment in 11.1%, 18.0%, 40.9% and 52.2% of the patients after 6 months, 1, 2 and 3 years, respectively (Fig. 1c). Twenty three of 102 patients discontinued cyclosporine because of ineffectiveness (Table 3).

Overall drug-survival Event = discontinuation Hazard ratio [CI95%] Increasing age every 5 years 1.015 [0.927-1.113] Male sex 0.641 [0.412-0.996]* Atopic dermatitis 0.929 [0.597-1.446] (1 missing) Positive patch test 1.457 [0.852-2.492] † (12 missing) Recurrent vesicular hand eczema 0.722 [0.459-1.136] † Prior oral immunosuppressive drugs 1.453 [0.866-2.436] † chapter Concomitant potent topical corticosteroids 0.596 [0.354-1.003] † Start dosage >3.5 mg/kg/day 0.877 [0.535-1.438] (17 missing) 4

Table 2 Determinants of overall drug survival determined by univariate Cox regression analyses † p<0.2 * p< 0.05

n (%) Status at the moment of data lock • Active 13 (12.8) • Stop due to ineffectiveness 18 (17.6) • Stop due to adverse events 44 (43.1) • Stop due to adverse events & ineffectiveness 5 (4.9) • Stop due to disease control 11 (10.8) • Other stop reason (pregnancy wish, non-adherence, other compound) 3 (2.9) • Lost to follow up 8 (7.6) Effectiveness after 3 months15 • PGA 1 61 (59.8) • PGA 2 20 (19.6) • PGA 3 16 (15.7) • Missing PGA 5 (4.9)

Table 3 Treatment characteristics

Adverse events One or more adverse events occurred in 71 patients (69.6%). In 24 patients (23.5%) hypertension was registered, which was a reason to discontinue therapy in 21 patients (Table 4).

141 Eight patients (7.8%) encountered an increase in serum creatinine levels of at least 30% and six patients discontinued therapy because of this. Gastrointestinal complaints (18.6%) and headache (18.6%) were the most reported subjective adverse events. In total 49 (48%) patients discontinued their treatment due to adverse events (Table 3 & Table 4). With regards to drug survival, adverse events were a reason to discontinue treatment with cyclosporine in 28.6%, 40.8%, 59.1% and 62.2% of the patients after 6 months, 1, 2 and 3 years respectively (Fig. 1b). The patient who used cyclosporine for 13.7 years does not suffer from any adverse events, though the hand eczema flares whenever the dosages is reduced and patients refused to discontinue treatment.

Discussion chapter Cyclosporine is a valuable treatment option for patients with hand eczema. Despite the fact that cyclosporine is not registered for the treatment of hand eczema extensive clinical 4 experience exists. This is the first study on drug survival of cyclosporine in patients with hand eczema

Drug survival as outcome parameter As expected, a substantial number of patients discontinued cyclosporine treatment in the first months. The majority of these patients discontinued treatment because of insur- mountable adverse events early in treatment, as represented by the steep curve in Fig. 1b. Systemic treatments require some time before their effectiveness is established, though overall cyclosporine is rapid-acting.16 Therefore, a substantial number of drop outs due to ineffectiveness was expected in the first months of therapy, but not in the first days/weeks: this is shown by the small plateau, followed by a steep drop in Fig. 1c. After one year of treatment, few patients discontinued treatment due to adverse events, and when the drug survival was plotted further then three years (data not shown), none of the patients dropped out because of adverse events. When patients ‘survived’ the first year of treatment, the chances of dropping out diminished and drug survival rates up to 13.7 years were even registered. After three years subjects did stop because of ineffectiveness or disease control, but over time other reasons such as pregnancy wish or non-specific request of the patient, were more often reasons for discontinuation.

Patients who used cyclosporine for several months were expected to eventually discontinue treatment because of disease control. However, even after two years of treatment patients discontinued treatment due to ineffectiveness, though treatment had been effective before. This might be a result of treatment ineffectiveness after dose reduction. If patients use cyclosporine for a substantial period of the time, physicians are inclined to taper the

142 dosages or patients request a lower dosage out of fear for adverse events. This tapering may result in a flare of the hand eczema. In addition the chronic and relapsing course of hand eczema can result in flares without a change is dosages. In the drug survival analyses, some patients were censored early in treatment. The majority of these patients was referred back to their local hospital after successful initiation of treatment. Moreover, active users were censored. Including these censored patients would have resulted in even longer drug survival.

Daily-life versus clinical trial Difference in drop-out rates between clinical trials and daily practice is a well-known phenomenon.17 Clinical studies investigate the efficacy of a drug under ‘ideal’ study conditions, while daily practice studies investigate the effectiveness of a drug under ‘usual’ chapter conditions. Trial patients are overall ‘perfect’ patients who fulfill strict in- and exclusion criteria and follow a strict trail protocol towards a fixed end goal. Moreover, trial patients 4 may be more motivated to continue treatment. Because of this, the internal validity is higher in clinical trials, but the external validity may be lower compared to daily use studies. Cyclosporine was first described to treat recalcitrant vesicular hand eczema in 1992.18 In a small case series by Reitamo et al.11, 6 out of the 7 patients responded moderately to good to cyclosporine during 2-16 weeks of treatment. Granlund et al. compared cyclosporine A to topical 0.05% betamethasone-17,21-dipropionate cream in 41 patients with chronic hand eczema for six weeks: both gave significant improvement, without significant differences between the groups.10,19 Granlund et al. found a decrease in severity of hand eczema of at least 50% in half of the patients using cyclosporine.10 These results are in line with our results: more than half of the patients (59.8%) in our cohort achieved an improvement of at least 50% of their eczema after three months. The same is true for atopic dermatitis: in a meta-analysis which included 15 studies on atopic dermatitis, Schmitt et al. found a relative effectiveness of 55% (95%CI 48-62%) for cyclosporine.16

The study of Granlund and the case series of Reitamo were of short duration and only one follow-up study was conducted, although the subjects in this follow-up study did not receive active cyclosporine treatment.20 For cyclosporine the tendency exists to give the drug for short crisis interventions and to taper as soon as possible. However, the current study demonstrated a median survival for cyclosporine of 10.3 months. Since long term use data of cyclosporine in hand eczema do not exist, it is not possible to compare our results to other studies in hand eczema patients. Cyclosporine is extensively studied in patients with atopic dermatitis and Schmitt et al. conducted an extensive meta-analysis of 602 atopic dermatitis patients in clinical trials of cyclosporine.16 Schmitt concluded that the use of cyclosporine was limited by the occurrence

143 of adverse events. They calculated the risk of adverse events in patient months (the months that patients in a study are followed divided by the event of interest) to compare 15 different studies. In 11% of the patient months serum creatinine levels increased significantly, hypertension occurred in 6% of patient months and gastrointestinal symptoms in 40% of patient months. Our study found a lower number of increased creatinine levels, hypertension and gastrointestinal symptoms, but in our cohort adverse events were the major reason for discontinuation of treatment as well.

Alitretinoin Recently, the retinoid alitretinoin was registered for the European market. In large, well- conducted, trials 30 mg alitretinoin a day resulted in clear or almost clear hand eczema in 48% of the participants.6-8 However, of the participants with vesicular hand eczema only one chapter third (33%) achieved clearance or almost clearance, compared to 16% in the placebo group 4 (based on Physician Global Assessment on a 5-point scale). In our study, cyclosporine gave a PGA1 score in 59.8% of all patients and in the subgroup of 66 patients with vesicular hand eczema 66.7% of them achieved a PGA1 score. The clustered PGA score in our retrospective study is less reliable than the prospective PGA used by Ruzicka et al. However, based on these results, a head-to-head trial comparing cyclosporine to alitretinoin is highly recommended. Schmitt et al. registered a clinical trial comparing cyclosporine to alitretinoin in all kinds of hand eczema, but the trial was ended prematurely (clinicaltrials.gov NCT01231854). Alitretinoin was not included in this daily life study, since it was registered in the Netherlands only recently (September 2013). Moreover the maximum recommended treatment duration for alitretinoin is 24 weeks: therefore drug-survival will not be a suitable outcome.

Limitations of the study The retrospective nature of this study is one of the major limitations. However, in almost all medical records start- and stop dates and reasons for discontinuation were registered in detail and drug survival is reliable even in a retrospective design. With regard to the effectiveness after three months, a clustered PGA score was used, which had been used for retrospective analyses more often.15 Moreover, missing information was excluded from analyses. Selection bias may have occurred: physicians may have been more prone to prescribe for example acitretin in hyperkeratotic hand eczema, and cyclosporine for vesicular hand eczema based on their own clinical experience. One should take this potential skewed distribution into consideration when interpreting these results. Moreover, this study included various tertiary referral patients, who tend to suffer of more recalcitrant eczema. This could limit the generalizability of this study. Only the first episode of cyclosporine was analysed in order to prevent further selection bias and to prevent distortion, because dermatologists may be inclined to re-introduce

144 cyclosporine if the first treatment episode was successful Over the years the attitude of dermatologists towards systemic drugs has changed. In the past many physicians started with a low dosage cyclosporine for short intervals. Nowadays, a higher start dosage is more common. Subgroups analyses regarding the start dosages did not favor a step-up or step-down regime, though these subgroups were relative small and further studies are warranted.

Conclusion This study demonstrated that cyclosporine is a valuable treatment option for patients with severe hand eczema with a long median drug survival of 10.3 months. Cyclosporine is especially effective in patients with recurrent vesicular hand eczema and should be considered in this subgroup as one of the first treatment options. chapter

Acknowledgement 4 We would like to thank Raisa Thybaut and Anna Brandsma for their help with data collection, Wietske Kievit, Elke de Jong and Juul van den Reek for their support with the statistical analyses and finally Tanja Vogel for her input in the manuscript.

145 References 1. Meding B, Jarvholm B. Incidence of hand eczema-a population-based retrospective study. J Invest Dermatol 2004:122:873-877. 2. Soost S, Abdollahnia M, Kostev K, Worm M. Topical therapy of hand eczema - analysis of the prescription profile from dermatologists in private practice.J Dtsch Dermatol Ges 2012:10:180-184. 3. Menné T, Johansen JD, Sommerlund M, Veien NK, Danish Contact Dermatitis Group. Hand eczema guidelines based on the Danish guidelines for the diagnosis and treatment of hand eczema. Contact Dermatitis 2011:65:3-12. 4. Diepgen TL, Elsner P, Schliemann S, Fartasch M, Kollner A, Skudlik C, et al. Guideline on the management of hand eczema ICD-10 Code: L20. L23. L24. L25. L30. J Dtsch Dermatol Ges 2009:7 Suppl 3:S1-16. chapter 5. Lynde C, Guenther L, Diepgen TL, Sasseville D, Poulin Y, Gulliver W, et al. Canadian hand dermatitis management guidelines. J Cutan Med Surg 2010:14:267-284. 4 6. Ruzicka T, Larsen FG, Galewicz D, Horvath A, Coenraads PJ, Thestrup-Pedersen K, et al. Oral alitretinoin (9-cis-retinoic acid) therapy for chronic hand dermatitis in patients refractory to standard therapy: results of a randomized, double-blind, placebo-controlled, multicenter trial. Arch Dermatol 2004:140:1453-1459. 7. Ruzicka T, Lynde CW, Jemec GB, Diepgen T, Berth-Jones J, Coenraads PJ, et al. Efficacy and safety of oral alitretinoin (9-cis retinoic acid) in patients with severe chronic hand eczema refractory to topical corticosteroids: results of a randomized, double-blind, placebo-controlled, multicentre trial. Br J Dermatol 2008:158:808-817. 8. Bissonnette R, Worm M, Gerlach B, Guenther L, Cambazard F, Ruzicka T, et al. Successful retreatment with alitretinoin in patients with relapsed chronic hand eczema. Br J Dermatol 2010:162:420-426. 9. Thestrup-Pedersen K, Andersen KE, Menné T, Veien NK. Treatment of hyperkeratotic dermatitis of the palms (eczema keratoticum) with oral acitretin. A single-blind placebo-controlled study. Acta Derm Venereol 2001:81:353-355. 10. Granlund H, Erkko P, Eriksson E, Reitamo S. Comparison of cyclosporine and topical betamethasone-17,21-dipropionate in the treatment of severe chronic hand eczema. Acta Derm Venereol 1996:76:371-376. 11. Reitamo S, Granlund H. Cyclosporin A in the treatment of chronic dermatitis of the hands. Br J Dermatol 1994:130:75-78. 12. Burden AD. Drug survival rates for tumour necrosis factor-alpha antagonists in psoriasis. Br J Dermatol 2011:164:940-941. 13. Gniadecki R, Kragballe K, Dam TN, Skov L. Comparison of drug survival rates for adalimumab, etanercept and infliximab in patients with psoriasis vulgaris. Br J Dermatol 2011:164:1091-1096.

146 14. Boonstra MB, Christoffers WA, Coenraads PJ, Schuttelaar MLA. Patch test results of hand eczema patients – relation to clinical types. Journal of the European Academy of Dermatology & Venereology ahead of printing:. 15. Hijnen DJ, ten Berge O, Timmer-de Mik L, Bruijnzeel-Koomen CA, de Bruin-Weller MS. Efficacy and safety of long-term treatment with cyclosporin A for atopic dermatitis. J Eur Acad Dermatol Venereol 2007:21:85-89. 16. Schmitt J, Schmitt N, Meurer M. Cyclosporin in the treatment of patients with atopic eczema - a systematic review and meta-analysis. J Eur Acad Dermatol Venereol 2007:21:606-619. 17. Martin K, Begaud B, Latry P, Miremont-Salame G, Fourrier A, Moore N. Differences between clinical trials and postmarketing use. Br J Clin Pharmacol 2004:57:86-92. 18. Petersen CS, Menné T. Cyclosporin A responsive chronic severe vesicular hand eczema. Acta Derm Venereol 1992:72:436-437. chapter 19. Granlund H, Erkko P, Reitamo S. Comparison of the influence of cyclosporine and topical betamethasone-17,21-dipropionate treatment on quality of life in chronic hand eczema. 4 Acta Derm Venereol 1997:77:54-58. 20. Granlund H, Erkko P, Reitamo S. Long-term follow-up of eczema patients treated with cyclosporine. Acta Derm Venereol 1998:78:40-43.

147 Chapter

148 Patch test results of hand eczema patients: relation to clinical types JEADV 2014 Sep 15. doi: 10.1111/jdv.12735 [Epub ahead of print]

Marrit Baukje Boonstra1*, Wietske Andrea Christoffers1*, Pieter-Jan Coenraads1, chapter Marie-Louise Anna Schuttelaar1 5 1 Department of Dermatology, University Medical Center Groningen, University of Groningen, Groningen, Netherlands. * Both authors contributed equally

Key Words: allergic contact dermatitis, patch test, hand eczema, hyperkeratotic, dyshidrotic eczema

149 Abstract Background: Allergic contact dermatitis is a well-known cause of hand eczema, although the influence of contact allergens on different clinical types of hand eczema remains still unclear. Objective: To identify most common positive tested allergens among hand eczema patients and to define the relation between specific contact allergies and clinical types of hand eczema according to the guidelines of the Danish Contact Dermatitis Group (DCDG). Methods: We included 1,571 hand eczema subjects who were patch tested from 1 January 2002 to 31 December 2013. They were retrospectively classified according to the guidelines of the DCDG into six clinical types: recurrent vesicular hand eczema, chronic fissured hand eczema, hyperkeratotic palmar eczema, pulpitis, interdigital eczema and nummular hand eczema according to a newly developed flow chart. The prevalence of sensitizations and association with clinical type, atopic dermatitis, age and gender were studied. Results: A total of 1395 subjects were classified into one of the six clinical types. The most frequently found clinical types were recurrent vesicular hand eczema (39.7%) and chronic chapter fissured hand eczema (35.5%). Subjects with recurrent vesicular hand eczema were significantly more likely to have a contact allergy (OR 1.55), whereas subjects with hyperkeratotic 5 palmar eczema and pulpitis were less likely to be sensitized (OR 0.51; OR 0.44). Overall, metals (nickel sulphate, cobalt chloride), fragrances and preservatives (methylchloroisothiazoline/ methylisothiazoline, methyldibromoglutaronitrile) were the most frequent sensitizers in patients with hand eczema. This did not deviate in the different clinical types, although subjects with recurrent vesicular hand eczema were significantly more frequently sensitized to nickel sulphate and other allergens compared to other clinical types of hand eczema. Conclusion: In the diagnostic work up of hand eczema subjects with recurrent vesicular hand eczema should be patch tested, especially women of older age, although the need for patch testing in males with hyperkeratotic palmar eczema might be less imperative.

150 Introduction Hand eczema is a common condition, with a 1-year prevalence up to 10%, depending on the definition used.1 The pathogenesis of hand eczema is multifactorial, with contribution of internal factors such as an atopic status and external factors, such as allergens and irritants. The precise impact of allergic contact dermatitis is unclear, although patients often suspect an allergy and recently published guidelines stress the importance of patch testing.2,3 Up till now, few studies have been published on the role of contact allergies in different clinical subgroups of hand eczema.4-7 However, since consensus regarding the (clinical) classification of hand eczema is lacking, these results are difficult to compare. Recently, the Danish Contact Dermatitis Group (DCDG) published a clear clinical classification system for hand eczema.2 Subsequently, Johansen et al. conducted a prospective study according to this system in which they classified 508 patients in six clinical types and etiological types.4 In this study, less than 40% had at least one positive patch test reaction and in 24.4% of the patients the final diagnosis was allergic contact dermatitis. Johansenet al. did not differentiate between specific contact allergies, such as nickel or fragrance mixes. To assess the spectrum of allergens related to hand eczema, we conducted a retro-spective chapter analysis of patch test data in hand eczema patients. We hypothesize that specific groups of contact allergies are associated with different clinical types. The aim of this study was 5 to identify most common positive tested allergens among all hand eczema patients and to define the relation between specific contact allergies and clinical types of hand eczema according to the guidelines of the Danish Contact Dermatitis Group in a large retrospective cohort of hand eczema patients.2

Materials and methods Patients In this retrospective study, 1,835 unique subjects who were patch tested at the Dermatology department of the University Medical Center Groningen between 01 January 2002 until 31 December 2013 were extracted from our patch test database. Only subjects in which the hands were affected were selected. Subjects who were not tested with the European Baseline series (n = 51) or subjects with other missing data (n = 129) were excluded. Finally, subjects with other dermatoses on the hands (n = 84) such as dermatomycosis (n = 22), psoriasis (n = 14), urticaria (n = 8) and drug reactions (n = 8) were excluded. After excluding these subjects, a total of 1,571 patients with hand eczema were included in this study.

Patch testing Patch tests were performed with the TRUE Test (Thin-layer Rapid Use Epicutaneous Test, Mekos Laboratories, Hillerød, Denmark) and additional allergens patch tested in van der

151 Bend square chambers® (Van der Bend BV, Brielle, The Netherlands). Patch tests were applied to the upper back for 48 hours occlusion. All subjects were patch tested with the European Baseline series and a baseline extension series. Based on the patient’s history, additional series such as a cosmetic series or a (meth)acrylate series were also patch tested. Results were read on day 2 and day 3 from January 2002 till November 2008, in which the reading at day 3 was conclusive. In November 2008 the day 2 reading was replaced by a day 7 reading to pre-empt late reactions. A reaction was regarded as ‘positive’ if erythema and infiltration developed and covered the whole patch test area8, which was later adopted by the guidelines of the International Contact Dermatitis Group. The patch test results and demographic characteristics of subjects were prospectively registered in the European Surveillance System of Contact allergies (ESSCA).9 For all subjects MOAHLFA characteristics were recorded, which represents the percentages of males, occupational factors, a history of atopic dermatitis, primary site hand, leg or face, and an age over 40 years.10 Polysensitization was considered as positive reactions to three or more contact allergens in the European Baseline series.11 chapter Classification of hand eczema 5 The diagnosis hand eczema was retrospectively classified based on medical records and photographic documentation according to the criteria of the DCDG. Based on these criteria hand eczema was classified in six different clinical types: chronic fissured hand eczema, recurrent vesicular hand eczema, hyperkeratotic palmar eczema, pulpitis, interdigital eczema and nummular hand eczema, alongside a non-classifiable category. To facilitate the classification of hand eczema in different clinical types, a flow chart was developed based on the criteria of the DCDG (Fig. 1). In this flow chart the primary location on the hand was combined with the morphology, which results in a final clinical type of hand eczema. The flow chart was developed to retrieve the data from the medical records in a structured way.

Statistics Data analyses were performed with IBM SPSS Statistics (SPSS, Chicago, IL, USA) for Windows (version 22.0). Data are presented as absolute numbers and in percentages for categorical variables and as median values for continuous variables. The chi-squared test was used for comparison between different clinical types. By means of logistic regression the influence of clinical subgroups among other factors on the presence of a contact allergy was studied and odds ratios (OR) were calculated.

152 Pulpitis Interdigital Nummular Diagnosis Hyperkeratotic Non-classifiable Non-classifiable Non-classifiable Non-classifiable Non-classifiable Non-classifiable Non-classifiable Chronic fissured Chronic fissured Recurrent vesicular no no no no yes yes yes yes yes yes yes unknown unknown unknown no no ll demarcated palmar site? ll-circumscribed, Morphology Hyperkeratosis Dry and scaling Dry and scaling and/or vesicles? and/or fissures? We hyperkeratosis on and examination? We areas and fissures? areas and fissures? Eruptions of vesicles and/or hyperkeratotic and/or hyperkeratotic Erythema and scaling? coin shaped lesions with and/or vesicles in history erythema and/or keratosis chapter no yes yes yes 5 Vesicles on fingers Vesicles and/or examination? Predominantly pulpa? and/or palms in history Exclusively webspace(s)? no no Primary location Fingertips Interdigital Entire hand Dorsal hand Palms and/or fingers

Fig. 1. Flowchart clinical type. To classify the clinical type of hand eczema first select your primary location and subsequently assess the morphology. This will result in a clinical type.

153 Results General description A total of 1,571 subjects were included. About half of the included population was male (45.2%) and the mean age was 38.1 years (range 9.3 to 80.4). About one-third of the subjects had a personal history of atopic dermatitis. Occupational factors were involved in 52.7% of the cases. In 84.1% the hands were the primary site of dermatitis, in 0.6% the legs and in 2.7% the face were the primary site (Table 1).

The majority of the cases (n = 623) were classified as recurrent vesicular hand eczema. Chronic fissured hand eczema was diagnosed in 557 subjects, hyperkeratotic palmar eczema in 119, pulpitis in 63, nummular hand eczema in 21 and interdigital eczema in 12 subjects. In 176 subjects, it was not possible to classify the hand eczema into a clinical subtype. In the hyperkeratotic and pulpitis group, the mean age was statistically significantly higher with a mean age of 50.1 and 46.1 years. Although more than half of the total group was female, in the hyperkeratotic palmar eczema group males were significantly overrepresented (69.7% vs. chapter 30.3%).

5 Contact sensitizations A total of 1,709 positive patch test reactions were found in the European Baseline series. In addition, 879 positive reactions were found in additional series. Half of the subjects with hand eczema (50.7%) had a contact sensitization to one or more of the tested allergens in the European Baseline series and 15% was polysensitized. Males were less often sensitized to one or more allergens in the European Baseline series (37.1% men, OR 0.48, p<0.001) and sensitization was associated with age over 40 years (OR 1.58, p<0.001). Polysensitization to allergens in the European Baseline series was significantly less frequently found in males and more often in subjects with higher age (OR 0.45, p<0.001; OR 2.32, p<0.001, respectively). No relationship between a history of atopic dermatitis and the presence of at least one contact sensitization was found, not for the whole group of hand eczema patients nor for any clinical subtype of hand eczema. However, subjects with a history of atopic dermatitis were more likely to be polysensitized to allergens in the European Baseline series (OR 1.43, p = 0.019) when adjusted for the same characteristics. Subjects with recurrent vesicular hand eczema were more likely to be sensitized to one or more allergens in the European Baseline series compared to sensitization in all other subjects with hand eczema, even after adjustment for gender, age, atopic dermatitis and occupational factors (OR 1.41, p = 0.001). This difference remained for sensitizations in additional test series included in the analyses (OR 1.55, p<0.001).

154 (11.2) (33.0) (56.3) (40.3) (75.6) (2.3) (33.5) (54.0) (15.9) (59.1) (26.1) (%) (8.0) able f

176 58* 99 71 133* 4* 14* 59* 95 28 104 46 n Non- classi (3.1) (57.1) (52.4) (19.0) (81.0) (4.8) (4.8) (23.8) (42.9) (9.5) (47.6) (14.3) (%) 21 12 11 4 17 1* 1 5 9 2 10 3 n Nummular eczema (0.8) (33.3) (58.3) (33.3) (58.3) (0.0) (16.7) (16.7) (50.0) (8.3) (66.7) (33.3) (%) 12 4 7 4 7* 0 2* 2 6 1 8 4 n Interdigital Interdigital eczema (4.0) (50.8) (44.4) (17.5) (98.4) (0.0) (0.0) (66.7) (33.3) (6.3) (39.7) (12.7) (%) 63 32 28 11* 62* 0 0 2* 21* 4 25* 8* n Pulpitis

chapter - (7.6) (69.7) (36.1) (16.8) (84.9) (0.0) (0.8) (79.8) (35.3) (6.7) (37.8) (9.2) (%) 5 119 83* 43* 20* 101 0 1 95* 2* 8* 45* 11* n Hyperk eratotic eczema (39.7) (44.0) (46.2) (37.4) (83.0) (0.3) (1.0) (41.3) (55.2) (17.0) (61.8) (25.4) (%) 623 274 288* 233 517 2 6* 257 344* 106 385* 158 n Recurrent Recurrent vesicular (35.5) (44.3) (63.2) (40.9) (86.9) (0.5) (3.4) (38.1) (50.3) (15.4) (55.8) (24.1) (%) ssured 557 247 352* 228* 484* 3 19 212* 280 86 311 134 Chronic Chronic n f (100) (45.2) (52.7) (36.3) (84.1) (0.6) (2.7) (42.8) (50.7) (15.0) (56.5) (23.2) (%)

Overall hand eczema n 1571 710 828 571 1321 10 43 672 797 235 888 364 Demographic characteristics of included subjects and the characteristics in the different clinical types. the different Demographic characteristics of included subjects and the in Total Men Occupational Atopic dermatitis Primary site hand Primary site leg Primary site face Age over 40 in EBS ≥1 positive reactions in EBS ≥3 positive reactions at all ≥1 positive reactions at all ≥3 positive reactions Table 1. Table Baseline Series and a baseline extension series EBS = European test all other clinical types by Chi-squared from * p<0.05: statistically significantly different

155 Subjects with hyperkeratotic palmar eczema and pulpitis were less likely to have positive reactions to one or more allergens in the European Baseline series after adjustment for the same characteristics (OR 0.51, p<0.001; OR 0.44, p = 0.003) or in additional series (OR 0.46, p<0.001; OR 0.46 p = 0.005).

Frequent sensitizers Table 2 displays most common positive tested allergens in all patients with hand eczema; allergens with at least 1% positive reactions are shown. The most frequent sensitizer was nickel sulphate (18.8%). Subsequently, methylchloroisothiazolinone/ methylisothiazolinone (MCI/MI), cobalt chloride and fragrance mix I caused a substantial amount of positive reactions. Overall, the most frequent sensitizers did not deviate between the different clinical types (Table 3). Although subjects with recurrent vesicular hand eczema were more likely to react to nickel sulphate (OR 1.50, p = 0.002), potassium dichromate (OR 1.78, p = 0.012), cobalt chloride (OR 1.81, p = 0.003), compositae mix (OR 2.97, p = 0.004) and mercapto mix (OR 2.10, p = 0.037), compared to the other clinical types. Subjects with chapter pulpitis reacted more often to acrylates.

5 Discussion Subjects with recurrent vesicular hand eczema were significantly more likely to have a contact allergy, whereas subjects with hyperkeratotic palmar eczema and pulpitis were less likely to be sensitized. Subjects with chronic fissured and non-classifiable hand eczema did not show more frequent contact allergies compared to all other types of hand eczema. Present findings are in line with Johansenet al. as they encountered contact allergies more often in recurrent vesicular hand eczema.4 They also reported that the role of contact allergies was less pronounced in pulpitis and hyperkeratotic palmar eczema, although the subgroups were small (12 and 41 patients respectively). Thus, the current relatively large retrospective study supports the results of Johansen et al. with regards to sensitizations in different clinical types of hand eczema. Diepgen et al. investigated the percentage of patients with contact sensitization in different subtypes of hand eczema including vesicular and hyperkeratotic hand eczema among etiological subtypes.5 In more than 60% of the subjects with vesicular hand eczema and less than 40% with hyperkeratotic hand eczema, contact sensitizations to one or more allergens in the European Baseline series and additional series were found. These results are in agreement with the current results of recurrent vesicular hand eczema and hyperkeratotic palmar eczema (i.e. 61.8% and 37.8%). However, these results are difficult to compare since Diepgenet al. gave no clear definitions of vesicular and hyperkeratotic hand eczema. Moreover they included etiological diagnoses. Since the etiology of hand eczema is a complex combined action of major causes and other

156 factors, where the degree of influence of each factor is virtually impossible to determine and where this influence may even vary over time, it is challenging to classify hand eczema into subgroups based on etiological types. The addition of two clinical types makes this classification system perhaps even more difficult to apply in daily practice. Therefore, we prefer a classification system based on clinical types. According to our data, a history of atopic dermatitis does not seem to be a risk factor for sensitization in patients with hand eczema. However, we did show that a history of atopic dermatitis was a risk factor for polysensitization in hand eczema patients to allergens in the European Baseline series when corrected for age and gender (OR 1.43, p = 0.019). Agner et al. found a borderline decreased frequency of contact sensitization in hand eczema patients with a history of atopic dermatitis.12 Handa investigated the association between sensiti-zation and an atopic status in hand eczema patients and showed that being atopic was not a risk factor for contact allergies in patients with hand eczema.13 Study results on the relationship between contact sensitization and a history of atopic dermatitis are also conflicting.12,14,15 Clemmensen found similar frequencies of positive patch test reactions in patients 14 with and without atopic dermatitis. Thyssen et al. found a significant association between chapter atopic dermatitis and contact sensitization to at least one allergen in the general population of Copenhagen, with the exception of nickel and thiomersal.15 Carlsen et al. found an OR 5 of 1.43 for atopic eczema in polysensitized patients.11 These conflicting results illustrate the ongoing debate on the relation between atopic dermatitis and contact sensitizations. The female gender and an older age were risk factors for the development of contact allergies. These findings are supported by various studies.16,17

Contact sensitizations in clinical types of hand eczema Subjects with recurrent vesicular hand eczema were significantly more frequently sensitized to nickel sulphate, potassium dichromate, cobalt chloride, compositae mix and mercapto mix compared to the other clinical types of hand eczema. Nickel is a well-known sensitizer and various efforts have been made to reduce the number of sensitizations in the general population. The role of nickel sensitizations in vesicular hand eczema is controversial and some authors suggested that oral challenge with nickel sulphate in nickel-sensitive patients exacerbates vesicular hand eczema.18 Although skin contact with nickel does not seem to result in an increased number of vesicles, as was demonstrated by Bryld et al.19 Bryld found no association between nickel sensitization and recurrent vesicular attacks (relative risk 0.45, confidence interval 0.06 – 3.36). An explanation for the overrepresentation of nickel sensitizations in the recurrent vesicular type might be the female preponderance in this type, as women are more often exposed to nickel in, for example, jewellery.

157 Overall hand eczema n = 1571

Rank Allergen n (%) 1 Nickel sulphate 296 (18.8) 2 MCI/MI 144 (9.2) 3 Cobalt chloride 107 (6.8)

4 Fragrance mix I 97 (6.2) 5 MDBGN 95 (6.0) 6 Potassium dichromate 79 (5.0) 7 Fragrance mix II 77 (4.9) 8 Carba mix 75 (4.8) 9 Colophony 72 (4.6) 10 Thiuram mix 69 (4.4) 11 CAPB 66 (4.2) 12 PPD 53 (3.4) 13 Quaternium 44 (2.8) 14 Balsam of Peru 43 (2.7) chapter 15 Epoxy resin 42 (2.7) 5 16 Formaldehyde 42 (2.7) 17 Lyral 39 (2.5) 18 Mercapto mix 33 (2.1) 19 Compositae mix 32 (2.0) 20 MBT 31 (2.0) 21 Ammonium persulphate 30 (1.9) 22 PTBFR 26 (1.7) 23 4-Aminoazobenzene 26 (1.7) 24 Wool alcohols 25 (1.6) 25 Disperse orange 24 (1.5) 26 Thiomersal 23 (1.5) 27 TETD 23 (1.5) 28 SL mix 23 (1.5) 29 Isoeugenol 22 (1.4) 30 PTD 20 (1.3) 31 Hydroxycitronellal 19 (1.2) 32 Oakmoss 19 (1.2)

Table 2. The most frequent sensitizers in hand eczema patients with their prevalence. CAPB cocamidopropyl betaine, DPG 1,3-diphenyl guanidine, MBT 2-mercaptobenzothiazole, MCI/MI methylchloroisothiazolinone Methylisothiazolinone, MDBGN methyldibromo glutaronitrile, PPD p-phenylene diamine, PTBFR p-tert-butylphenol formaldehyde resin, PTD p-toluene diamine, SL mix Sesquiterpene lactonemix, TETD Tetra-ethylthi- uram disulfide.Table slightly adapted from the original article.

158 However, even after correction for age and gender, nickel sensitization was significantly more present in patients with vesicular hand eczema compared to the group of other types of hand eczema in our study. The metals nickel sulphate, potassium dichromate and cobalt chloride are known to give concomitant positive patch test results.20 Compositae mix is a combination of chemicals from different plants. Paulsen et al. described a subgroup of patients with compositae dermatitis21, which suffered from vesicular hand eczema. Mercapto mix is a mixture of three curing additives used as accelerators in the manufacture of, for example, rubber gloves. A typical feature of allergic contact dermatitis to rubber additives in hand eczema is localization of eczema on the dorsal side of the hand and fingers and on the flexor or extensor surfaces of the forearms, not extending to the area outside glove contact. In this study, we found an association between sensitization to mercapto mix and recurrent vesicular hand eczema; however, vesicular eczema was defined as recurrent vesicles on the palms and/or sides/palmar aspects of the fingers. Other sources than rubber gloves such as glues, cutting oils, household products and sports equipment made with natural rubber should be considered. A clinical relevant association between contact chapter allergy to mercapto mix and recurrent vesicular hand eczema is unclear. Agner et al. found no significant association between hand eczema subgroups, i.e. vesicular and hyperkeratotic 5 hand eczema and specific allergens or allergen groups including nickel sulphate, potassium dichromate, cobalt chloride, compositae mix and mercapto mix.12 Again, no definitions for vesicular and hyperkeratotic hand eczema were given which makes these results hard to compare. In our study, subjects with pulpitis were more often sensitized for (meth)acrylates compared to subjects without pulpitis, although the additional (meth)acrylate series was also more often patch tested in subjects with pulpitis and this group was relatively small (n = 63).

Frequent sensitizers in all hand eczema patients Metals, fragrances and preservatives are important sensitizers in the general patch test population.22,23 As shown in Table 2, the frequent sensitizers in our cohort of hand eczema patients are comparable to the general patch test population. Nickel sulphate is a well- known sensitizer in the European Baseline series with a prevalence between 11.9% and 27.4% in European countries.22 We found a similar sensitization rate of 18.8% in subjects with hand eczema. The prevalence of the second most frequent found allergen, the preservative MCI/MI, is much higher in our hand eczema cohort than in the most recent publication on the general European patch test population (1-4% vs. 9.2%). Alhough this low rate of MCI/MI sensitizations dates from 2007 to 2008, while methylisothiazoline (MI) is an upcoming allergen which underwent an alarming increase over the last years.24

159 * p<0.05: statistically significantly different from all other clinical types by Chi-squared test MDBGN methyldibromo glutaronitrile, PPD CAPB cocamidopropyl betaine, FM fragrance mix, MCI/MI methylchloroisothiazolinone/ methylisothiazolinone, Table 3. 16 15 14 13 Rank 12 11 10 9 8 7 6 5 4 3 2 1 Most frequent sensitizers in the different clinical types of hand eczema patients. Allergen n = 557 Peru balsam Ammonium persulphate Quaternium Formaldehyde Chronic PPD Colophony Potassium dichromate CAPB MDBGN Carba mix Thiuram mix FMII FMI Cobalt chloride MCI/MI Nickel sulphate f ssured hand eczema n 15 15 16 20 21 23 24 27 30 31 31 33 34 35 58 92 p (%) (2.7) (2.7) (2.9) (3.6) (3.8) (4.1) (4.3) (4.8) (5.4) (5.6) (5.6) (5.9) (6.1) (6.3) (10.4) (16.5) -phenylene diamine, PTBFR

Lyral Mercapto mix PPD Compositae mix Allergen n = 623 Recurrent vesicular hand eczema Thiuram mix Quaternium Carba mix CAPB FMII Colophony MDBGN Potassium dichromate FMI Cobalt chloride MCI/MI Nickel sulphate

chapter p - tert 19 19* 20 21* n 22 23 29 29 32 36 42 42* 42 57* 61 141* 5 -butylphenol formaldehyde resin. (3.0) (3.0) (3.2) (3.4) (%) (3.5) (3.7) (4.7) (4.7) (5.1) (5.8) (6.7) (6.7) (6.7) (9.1) (9.8) (22.6) Epoxy resin 2-methyl-4-isothiazolinone FMII Neomycin PPD Allergen n = 119 Hyperkeratotic palmar eczema Thiuram mix Peru balsam PTBFR CAPB FMI Carba mix Cobalt chloride Potassium dichromate MDBGN MCI/MI Nickel sulphate

2 2 2 2* 2 n 2 3 3 3 4 4 5 6 7 8 15 (1.7) (1.7) (1.7) (1.7) (1.7) (%) (1.7) (2.5) (2.5) (2.5) (3.4) (3.4) (4.2) (5.0) (5.9) (6.7) (12.6)

160 Our results support the fact that the use of MCI and MI should be restricted further to stop this epidemic. A few studies primarily focused on specific contact allergens in hand eczema patients.6,12,25,26 Overall all the results were comparable with our findings with nickel sulphate, cobalt chloride and potassium dichromate as major allergens in the metal groups. Fragrances are important sensitizers, with fragrance mix II as upcoming allergen mix. Since fragrance mix II was introduced in our standard series in 2005, it is remarkable that the fragrance mix II gained the seventh position in the overall list of frequent sensitizers.

Limitations of the study The major limitation of this study is its retrospective nature. Although we included a large population and were able to classify 1,395 of the 1,571 subjects in one of the six clinical types of hand eczema, despite this retrospective approach. This may be attributed to the detailed medical records in our hospital and the newly developed flow chart, which facili- tated the retrospective classification and the reproducibility of our results. Due to the retrospective nature of the study, it was for each sensitization not possible to retrieve the chapter clinical relevance to the hand eczema. One of the difficulties we encountered was the classification of chronic fissured hand 5 eczema. When closely studying the definition of this subtype, the group can be very divers. Although the DCDG classification is very clear, we would like to recommend fine tuning of the chronic fissured type. The subjects in our study were drawn from a tertiary hospital population. This might have resulted in more severe cases of hand eczema, in which a higher prevalence of sensitizations may be present. In our centre, various additional series are available such as a hairdresser’s series or a bakery series. Since this might result in more positive reactions and more cases of polysensitization, we therefore mainly focused on the European Baseline series. Some type of selection bias might have occurred, since physicians might have been more prone to patch test vesicular types of hand eczema. Subjects with hyperkeratotic palmar eczema might have been less frequently patch tested, which might explain the predominance of recurrent vesicular hand eczema (n = 623) in our study population and the relatively small number of hyperkeratotic palmar eczemas. However, the prospective study of Johansen et al. found a similar ratio between recurrent vesicular and hyperkeratotic palmar eczema.4 Overall, our findings are in line with the study of Johansen, but add additional insight into specific contact allergens.

161 Conclusion Subjects with recurrent vesicular hand eczema are more likely to have positive patch test reactions to the European Baseline series compared to other types of hand eczema. Therefore, all subjects with recurrent vesicular hand eczema should be patch tested, especially women of older age, in the diagnostic work up of hand eczema. Moreover, we opine that a diagnostic patch test in males with hyperkeratotic palmar eczema might be less imperative. The newly developed flow chart contributes to the classification of clinical types and -in creases the reproducibility of the results. However, there is an urgent need for consensus regarding a single classification system of hand eczema in which the guidelines of the DCDG could be the foundation.

Acknowledgement We would like to acknowledge drs. T.A. Vogel for her thorough reviewing of the manuscript and her useful comments. chapter 5

162 References 1. Thyssen J, Johansen J, Linneberg A, Menné T. The epidemiology of handeczema in the general population--prevalence and main findings.Contact dermatitis 2010:62:75-82. 2. Menné T, Johansen JD, Sommerlund M, Veien NK, Danish Contact Dermatitis Group. Hand eczema guidelines based on the Danish guidelines for the diagnosis and treatment of hand eczema. Contact Dermatitis 2011:65:3-12. 3. Diepgen TL, Elsner P, Schliemann S, Fartasch M, Kollner A, Skudlik C, et al. Guideline on the management of hand eczema ICD-10 Code: L20. L23. L24. L25. L30. J Dtsch Dermatol Ges 2009:7 Suppl 3:S1-16. 4. Johansen JD, Hald M, Andersen BL, Laurberg G, Danielsen A, Avnstorp C, et al. Classification of hand eczema: clinical and aetiological types. Based on the guideline of the Danish Contact Dermatitis Group. Contact Dermatitis 2011:65:13-21. 5. Diepgen TL, Andersen KE, Brandao FM, Bruze M, Bruynzeel DP, Frosch P, et al. Hand eczema classification: a cross-sectional, multicentre study of the aetiology and morphology of hand eczema. Br J Dermatol 2009:160:353-358. 6. Magina S, Barros MA, Ferreira JA, Mesquita-Guimaraes J. Atopy, nickel sensitivity, occupation, and chapter clinical patterns in different types of hand dermatitis. Am J Contact Dermat 2003:14:63-68. 7. Cronin E. Clinical patterns of hand eczema in women. Contact Dermatitis 1985:13:153-161. 5 8. Magnusson B, Blohm SG, Fregert S, Hjorth N, Hovding G, Pirila V, et al. Routine patch testing. II. Proposed basic series of test substances for Scandinavian countries and general remarks on testing technique. Acta Derm Venereol 1966:46:153-158. 9. Uter W, Hegewald J, Aberer W, Ayala F, Bircher AJ, Brasch J, et al. The European standard series in 9 European countries, 2002/2003 -- first results of the European Surveillance System on Contact Allergies. Contact Dermatitis 2005:53:136-145. 10. Schnuch A, Geier J, Uter W, Frosch PJ, Lehmacher W, Aberer W, et al. National rates and regional differences in sensitization to allergens of the standard series. Population-adjusted frequencies of sensitization (PAFS) in 40,000 patients from a multicenter study (IVDK). Contact Dermatitis 1997:37:200-209. 11. Carlsen BC, Andersen KE, Menné T, Johansen JD. Patients with multiple contact allergies: a review. Contact Dermatitis 2008:58:1-8. 12. Agner T, Andersen KE, Brandao FM, Bruynzeel DP, Bruze M, Frosch P, et al. Contact sensitisation in hand eczema patients-relation to subdiagnosis, severity and quality of life: a multi-centre study. Contact Dermatitis 2009:61:291-296. 13. Handa S, Kaur I, Gupta T, Jindal R. Hand eczema: correlation of morphologic patterns, atopy, contact sensitization and disease severity. Indian J Dermatol Venereol Leprol 2012:78:153-158. 14. Clemmensen KK, Thomsen SF, Jemec GB, Agner T. Pattern of contact sensitization in patients with and without atopic dermatitis in a hospital-based clinical database. Contact Dermatitis 2014:71:75-81.

163 15. Thyssen JP, Linneberg A, Engkilde K, Menné T, Johansen JD. Contact sensitization to common haptens is associated with atopic dermatitis: new insight. Br J Dermatol 2012:166:1255-1261. 16. Schnuch A, Geier J, Lessmann H, Arnold R, Uter W. Surveillance of contact allergies: methods and results of the Information Network of Departments of Dermatology (IVDK). Allergy 2012:67:847-857. 17. Agner T, Andersen KE, Brandao FM, Bruynzeel DP, Bruze M, Frosch P, et al. Hand eczema severity and quality of life: a cross-sectional, multicentre study of hand eczema patients. Contact Dermatitis 2008:59:43-47. 18. Nielsen GD, Jepsen LV, Jorgensen PJ, Grandjean P, Brandrup F. Nickel-sensitive patients with vesicular hand eczema: oral challenge with a diet naturally high in nickel. Br J Dermatol 1990:122:299-308. 19. Bryld LE, Agner T, Menné T. Relation between vesicular eruptions on the hands and tinea pedis, atopic dermatitis and nickel allergy. Acta Derm Venereol 2003:83:186-188. 20. Hegewald J, Uter W, Pfahlberg A, Geier J, Schnuch A, IVDK. A multifactorial analysis of concurrent patch-test reactions to nickel, cobalt, and chromate. Allergy 2005:60:372-378. 21. Paulsen E, Andersen KE. Compositae dermatitis in a Danish dermatology department in 1 year (II). chapter Clinical features in patients with Compositae contact allergy. Contact Dermatitis 1993:29:195-201. 5 22. Uter W, Aberer W, Armario-Hita JC, Fernandez-Vozmediano JM, Ayala F, Balato A, et al. Current patch test results with the European baseline series and extensions to it from the ‘European Surveillance System on Contact Allergy’ network, 2007-2008. Contact Dermatitis 2012:67:9-19. 23. Schnuch A, Uter W, Lessmann H, Geier J. Clinical epidemiology and prevention of contact allergies. The Information Network of Departments of Dermatology (IVDK) as a register and surveillance system. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2012:55:329-337. 24. Aerts O, Baeck M, Constandt L, Dezfoulian B, Jacobs MC, Kerre S, et al. The dramatic increase in the rate of methylisothiazolinone contact allergy in Belgium: a multicentre study. Contact Dermatitis 2014:71:41-48. 25. Warshaw EM, Ahmed RL, Belsito DV, DeLeo VA, Fowler JF,Jr, Maibach HI, et al. Contact dermatitis of the hands: cross-sectional analyses of North American Contact Dermatitis Group Data, 1994-2004. J Am Acad Dermatol 2007:57:301-314. 26. Ni C, Dou X, Chen J, Zhu X, Liu L. Contact sensitization in Chinese patients with hand eczema. Dermatitis 2011:22:211-215.

164 chapter 5

165 Chapter

166 Severe bullous allergic contact dermatitis caused by glycidyl methacrylate and other acrylates Contact Dermatitis 2014;71:247-249

Tatiana Alexandra Vogel1, Wietske Andrea Christoffers1, Malin Engfeldt2, Magnus Bruze2, Pieter-Jan Coenraads1, Marie-Louise Schuttelaar1

1 Department of Dermatology, University Medical Center Groningen, chapter University of Groningen, Groningen, Netherlands. 2 Occupational and Environmental Dermatology, Skåne University Hospital, Malmö, Sweden 6

Key Words: acrylate; allergic contact dermatitis; bullous contact allergy; occupational; 1,6-hexanediol diacrylate, glycidyl methacrylate.

167 chapter 6

168 Cases of contact allergy to acrylates are described quite frequently in, for example, dentists, prosthesists, and printers. In this case report, we describe a rarer case of a severe allergic contact dermatitis caused by acrylates in an occupational setting.

Case Report A 50-year-old healthy man was seen at the outpatient clinic after he had suffered two episodes of skin lesions on his left knee, right wrist, and left ring finger (Fig. 1ab). A few hours after he had finished his work, these lesions appeared as burning, indurated redness, developing into tense blisters within 24 hr. The lesions healed without scar formation within 10 days. An occupational dermatitis was suspected, and history taking revealed that, for three years, the patients had been working as a process operator in the production of semi- finished products. He controlled production processes and manually added chemicals, ammonium persulfate and different liquid acrylates, in a semiclosed process. Material safety data sheets provided by the employer showed that these liquid acrylates included 1,6-hexanediole diacrylate (HDDA) and glycidyl methacrylate (GMA). As protection, the patient wore chemical-resistant, but liquid-permeable, clothing. In addition, he wore protective gloves: cotton gloves with a rubber coating on the palmar sites of the gloves for use during dry activities, and thick nitrile gloves with a cotton lining when handling liquids. He changed his clothes a few times a week and his gloves once or twice a day, depending on his activities. chapter Patch testing with the European baseline series (TRUE Test ®; Mekos Laboratories AS, Hillerød, Denmark) and our additional series was performed, together with a (meth)acrylate 6 series (Chemotechnique®, Vellinge, Sweden), ammonium persulfate 2.5% pet., and products from the patient’s workplace, namely the industrial acrylates GMA and HDDA, in a dilution series (obtained from the Department of Occupational and Environ-mental Dermatology, Skåne University Hospital, Malmö, Sweden), under 48 hr of occlusion. The test materials were prepared shortly before application, in order to prevent evaporation of the acrylates. Reactions to the tests were read according to the ICDRG criteria at D2, D3 and D7 after application. The patient reacted neither to the baseline series, nor to ammonium persulfate. However, multiple strong positive reactions were seen to the (meth)acrylates series and the industrial acrylates GMA and HDDA (Fig. 1c–h, Table 1). Besides AQ5 the severe reactions to the patch test materials on his back, he showed a serous crust on his forehead (Fig. 1i). This was interpreted as a recall reaction resulting from local skin memory. The patient was advised to take strict protective measures against all traces of acrylates, including unnoticed spatter and spill of products containing acrylates. If handling of acrylate- containing products was inevitable, he should wear impermeable, three-layered laminated gloves made of polyethylene/ethylene vinyl. Nitrile gloves were unsuitable for this purpose, because of their short breakthrough time.

169 chapter

Fig. 1. (a,b) Tense blisters on the left knee and right wrist 24 hr after occupational exposure to the acrylates glycidyl 6 methacrylate (GMA) and 1,6-hexanediole diacrylate (HDDA). (c–e) Patch test results for GMA. From left to right: pure GMA 1% pet., industrial GMA 0.3% pet., and industrial GMA 0.1% pet. (f–h) Patch test results for HDDA. From left to right: pure HDDA 0.1% pet., industrial HDDA 0.3% pet., and industrial HDDA 0.1% pet. (i) Serous crust on the patients’ forehead, resulting from local skin memory. Discussion We have described the first case of an occupational bullous allergic contact dermatitis caused by GMA and HDDA. HDDA (CAS 13048-33-4) belongs to the group of multifunctional acrylates. These acrylates are used in dentistry, and are important components of printing inks and waterproof coatings. They are all strong irritants and potent sensitizers. Contact allergies are almost exclusively seen in an occupational setting.1 GMA (CAS 106-91-2) is a potent sensitizer; it is an ester of methacrylate acid and epichlo- rohydrin forming a methacrylate monomer. Apart from a moderately reactive methacrylate group, it has an extremely reactive epoxy group. GMA has a wide range of applications, including the production of semi-finished products, as in the case of our patient’s factory. Moreover, GMA is also used to impregnate clothing and paper, for the production of waterproof sealings, and in dentistry techniques and orthopaedics.2

170 (Meth)acrylate series D2 D3 D7 Ethyl acrylate (EA) 0.1% pet. ++ ++ + Butyl acrylate (BA) 0.1% pet. ++ ++ + 2-Ethylexyl acrylate (2-EHA) 0.1% pet. + ? - Hydroxyethyl acrylate (2-HEA) 0.1% pet. ++ ++ + 2-Hydroxypropyl acrylate (2-HPA) 0.1% pet. ++ ++ + 2-Hydroxyethyl methacrylate (HEMA) 2% pet. - ? ? 2-Hydroxypropyl methacrylate (HPMA) 2% pet. ? + + Ethyleneglycol dimethacrylate (EGDMA) 2% pet. - ? - Triethylene Glycol Dimethacrylate (TEGDMA) 2% pet. + + ? 1,4-Butanediol dimethacrylate (BUDMA) 2% pet. ++ ++ + 1,4-Butanediol diacrylate (BUDA) 0.1% pet. +++ +++ + 1,6-Hexanediol diacrylate (HDDA) 0.1% pet. +++ +++ + Diethyleneglycol diacrylate (DEGDA) 0.1% pet. +++ +++ + Tripropylene glycol diacrylate (TPGDA) 0.1% pet. ? + ? Oligotriacrylate (OTA480) 0.1% pet. - + - Epoxy acrylate (bis-GA) 0.5% pet. ? + ? Urethane diacrylate, aliphatic (al-UDA) 0.1% pet. + ++ ++ Glycidylmethacrylate (GMA) 1% pet. +++ +++ +++ Dilution series industrial HDDA and GMA chapter HDDA 0.01% pet. - - - HDDA 0.03% pet. ? + - HDDA 0.1% pet. + + + 6 HDDA 0.3% pet. +++ +++ +++ GMA 0.01% pet. - - - GMA 0.03% pet. ? ? - GMA 0.1% pet. + ++ + GMA 0.3% pet. +++ +++ +++ GMA 1% pet. +++ +++ +++

Table 1. Positive patch test results for the (meth)acrylates series and the dilution series of the patient’s own industrial acrylates. GMA, glycidyl methacrylate; HDDA, 1,6-hexanediole diacrylate.

Furthermore, contact allergies to GMA have often been described in nail stylists. A severe irritant bullous reaction after occupational contact with GMA was described by Shimizu et al. in 2008.3 The multiple reactivity to various acrylates in this patient can be explained by two different phenomena: first, the phenomenon of cross-reactivity, owing to the almost identical molecular structures of different acrylates or transformation to the same molecular structure; and second, contamination of acrylates with and subsequent exposure to other acrylates

171 used in an industrial setting, as most acrylates contain up to 20% contamination.4,5 Our analysis of GMA and HDDA showed that both substances had been contaminated by other acrylates (Table 2).

GMA Pure GMA 97% HDDA 0.5% 2 probable methacrylates, not identified 1.7% Dichloropropanol 0.2% HDDA Pure HDDA 93% 5 acrylates, not identified 6.5%

Table 2. Results of analysis for confirmation of purity of glycidyl methacrylate (GMA) and 1,6-hexanediole diacrylate (HDDA) in the industrial products by gas chromatography and mass spectrometry.

Making the distinction between a severe irritant reaction and a severe contact allergic reaction, as in the case described, is often difficult. A very low concentration of any acrylate can initiate a very severe reaction. For this reason, a dilution series of the product was tested to determine the origin of the reaction.6 chapter The recall reaction on the patient’s forehead might have been attributable to local skin memory. This memory function of the skin was first described by Scheperet al.7 In this 6 case, the memory recall might have been caused by the severe patch test reactions on the patient’s back combined with earlier contact with one of these acrylates on his forehead. Furthermore, the absence of any reaction on his left knee, right wrist and left ring finger still leaves some questions. This case report stresses the importance of patch testing with a dilution series of acrylates to distinguish between a contact allergy and an irritant reaction. Moreover, it emphasizes the importance of investigating the products in a patient’s workplace and testing the purity of these products.

172 References 1. Botella-Estrada R, Mora E, de la Cuadra J. Hexanediol diacrylate sensitization after accidental occupational exposure. Contact Dermatitis 1992:26:50-51. 2. Sanchez-Perez J, Gonzalez-Arriba A, Goiriz R, Garcia-Diez A. Occupational allergic contact dermatitis to acrylates and methacrylates. Contact Dermatitis 2008:58:252-254. 3. Shimizu A, Kamada N, Kambe N, Matsue H. Chemical burn caused by glycidyl methacrylate. Contact Dermatitis 2008:59:316-317. 4. Kanerva L. Cross-reactions of multifunctional methacrylates and acrylates. Acta Odontol Scand 2001:59:320-329. 5. Rustemeyer T, de Groot J, von Blomberg BM, Frosch PJ, Scheper RJ. Induction of tolerance and cross-tolerance to methacrylate contact sensitizers. Toxicol Appl Pharmacol 2001:176:195-202. 6. Emmett EA, Kominsky JR. Allergic contact dermatitis from ultraviolet cured inks. Allergic contact sensitization to acrylates. J Occup Med 1977:19:113-115. 7. Rustemeyer T, de Groot J, von Blomberg BM, Bruynzeel DP, Frosch PJ, Scheper RJ. Assessment of contact allergen cross-reactivity by retesting. Exp Dermatol 2002:11:257-265.

chapter 6

173 Chapter

174 Two decades of occupational (meth) acrylate patch test results and focus on isobornyl acrylate Contact Dermatitis 2013;69:86-92

Wietske Andrea Christoffers, Pieter-Jan Coenraads, Marie-Louise Anna Schuttelaar

Department of Dermatology, University of Groningen, University Medical Center Groningen, Groningen, Netherlands

chapter 7

Key words: acrylate; allergic contact dermatitis; isobornyl acrylate; occupational.

175 Abstract Background: Acrylates constitute an important cause of occupational contact dermatitis. Isobornyl acrylate sensitization has been reported in only two cases. We encountered an industrial process operator with occupational contact dermatitis caused by isobornyl acrylate. Objectives: (i) To investigate whether it is relevant to add isobornyl acrylate to the (meth) acrylate test series. (ii) To report patients with (meth)acrylate contact allergy at an occupational dermatology clinic. Patients/materials/methods: Our patch test database was screened for positive reactions to (meth)acrylates between 1993 and 2012. A selected group of 14 patients was tested with an isobornyl acrylate dilution series: 0.3%, 0.1%, 0.033%, and 0.01%. Readings were performed on D2, D3, and D7. Results: One hundred and fifty-one patients were tested with our (meth)acrylate series; 24 had positive reactions. Most positive reactions were to 2-hydroxypropyl acrylate, 2-hydroxyethyl acrylate, 2-hydroxypropyl methacrylate, and diethyleneglycol diacrylate. Hypothetical screening with 2-hydroxypropyl acrylate, ethyleneglycol dimethacrylate, ethoxylated bisphenol A glycol dimethacrylate and trimethylolpropane triacrylate identified 91.7% of the 24 patients. No positive reactions were observed in 14 acrylate positive patients tested with the isobornyl acrylate dilution series. The 0.3% isobornyl acrylate concentration induced irritant reactions in three patients. Conclusions: We report a rare case of allergic contact dermatitis caused by isobornyl acrylate. However, this study provides insufficient support for isobornyl acrylate to be added to a (meth)acrylate series. chapter 7

176 Acrylates and methacrylates are important causes of occupational contact dermatitis. Sensitization may be induced by adhesives, dental products, ultraviolet (UV)-cured inks, and coatings. Nail stylists, dental personnel, printers and industrial assembly line workers are particularly at risk, because of daily occupational exposure. Today, screening for (meth) acrylate contact allergy is a topic of interest, and different screening series have been developed.1-3 Isobornyl acrylate (Fig. 1; CAS 5888-33-5) is a photopolymerizable monomer that is used in various industrial products such as UV-cured ink and UV-cured adhesives. Contact allergy to isobornyl acrylate in two young women was reported in 1995.4 Both women developed eczema and abscesses around the injection site of their insulin pump. They showed allergic reactions to the medical adhesives and the plastic of the infusion sets upon patch testing with components of the glues and the plastic scrapings from their infusion set, including positive reactions to phenoxypoly (ethyleneoxy) ethylacrylate 0.1%, β-carboxyethyl acrylate 0.1%, 1-benzoylcyclohexanol 1%, and isobornyl acrylate 0.1%. The latter turned out to be one of the culprit allergens that had been used in the UV-cured adhesive and had diffused into the plastic of the infusion set. To the best of our knowledge, this is the only report of isobornyl acrylate contact allergy. Other studies failed to demonstrate isobornyl acrylate sensitization, even after prolonged exposure at high doses.3,5,6

H3C CH3

CH3 CH2 O

chapter O 7 Fig. 1. Chemical structure of isobornyl acrylate.

Case Report A 47-year-old atopic man was referred to our centre because of therapy-resistant hand eczema. He had been working as a process operator in a factory producing glass fibres for over 20 years. His work involved painting glass fibres with UV-cured paint, printing the glass fibres, covering them with an acrylate coating, and cleaning the machines. His skin problems cleared during holidays, and relapsed when he returned to work. The patient was patch tested with the European baseline series and 12 department-specific additions, the cosmetic series, and our (meth)acrylate series containing 29 commercially available (meth)acrylates (Chemotechnique Diagnostics, Vellinge, Sweden; Table 1). The patch tests were applied on the upper back for 48 hr under occlusion with van der Bend®

177 square chambers (Van der Bend BV, Brielle, The Netherlands) and Fixomull® Stretch (BSN Medical, Hamburg, Germany). The tests were read on D3 and D7 according to the guidelines of the International Contact Dermatitis Research Group (ICDRG). There were no positive reactions to the extended European baseline series and the cosmetic series. The patient showed 1+ and 2+ positive patch test reactions to 11 different (meth)acrylates (Table 1). However, all of the acrylates turned out to be currently clinically irrelevant, because they were not present in the substances that the patient worked with at that moment. Nonethe- less, an acrylate was suspected to be the causative allergen, given the nature of his present occupation. A workplace visit showed that isobornyl acrylate was a component of the glass fibre coatings [Desolite™ (DSM Desotech, Heerlen, The Netherlands) and Bufferlite™ DU- 2002 (DSM Desotech)] and UV-cured ink (Herkula-Ultracoat™ OF 813; Krefeld, Germany) with which he came into contact during the production process. Thus, isobornyl acrylate was suspected as a relevant allergen. A patch test with isobornyl acrylate 0.1% pet. (Sigma Al- drich, Zwijndrecht, The Netherlands; in-house preparation) under 48 hr of occlusion resulted in a 2+ positive reaction on D3 and D7 (Fig. 2). This case of isobornyl acrylate contact allergy and the concomitant various positive reactions to other clinically irrelevant acrylates led to the investigation into potentially missed isobornyl acrylate sensitizations in other patients. The question is whether contact allergy to isobornyl acrylate is rare or just underdiagnosed, and whether isobornyl acrylate should be added to (meth)acrylate patch test series. There- fore, we screened our database for patients with allergic reactions to (meth)acrylates, and patch tested a selected group with a dilution series of isobornyl acrylate, in order to detect any missed sensitization and potential cross-reactivity. In addition, the results of almost 20 years of patch testing with the (meth)acrylate patch test series are presented. chapter 7

Fig. 2. Strong positive reaction to isobornyl acrylate on D3.

178 Materials and Methods Database study The (meth)acrylate series with 29 different (meth) acrylates has been used for patch testing at our dermatology department since 1993. The patch test database was screened for the patients tested with the (meth)acrylate series since 1993. Our (meth)acrylate series consists of 29 different (meth)acrylates (Table 1) supplied by Chemotechnique Diagnostics. The types of acrylate in the series have not been changed over the last 19 years. However, the concentrations of butyl acrylate, 2-hydroxyethyl acrylate (2-HEA) and 2-hydroxypropyl acrylate (2-HPA) were reduced from 0.5% to 0.1% in 2001. The patch tests were read on D2 and D3, according to the guidelines of the ICDRG, between January 1993 and October 2008. They were also read on D3 and D7 between November 2008 and July 2012, in order to register late reactions. Besides the patch test results and concomitant reactions, patient characteristics were registered according to the MOAHLFA index.7 The clinical relevance and the possible exposure to acrylates were assessed by means of a detailed history, including occupational history and potential exposure to acrylates. When contact with an acrylate containing product was suspected, Material Safety Data Sheets (MSDS) were studied to confirm a causative relationship. In some cases, a workplace visit was also undertaken. The clinical relevance was labeled as ‘certain’ when the patient developed an itching dermatitis after exposure to at least one product containing acrylates (according to the product label or MSDS). The clinical relevance was labeled as ‘probable’ when a patient had developed an itching dermatitis after exposure to one or more products probably containing acrylates, but a specific product had not been identified as the cause of the clinical reaction. The clinical chapter relevance was labeled as ‘possible’ when a patient developed an itching dermatitis after the use of various products with or without acrylates, and materials other than acrylates may 7 have been the cause of dermatitis in the patient. Finally, the clinical relevance was labeled as ‘unlikely’ when a patient had no contact at all with products containing acrylates as far as was established.8,9

Isobornyl acrylate sub-study Patients with a previously demonstrated acrylate sensitization were selected from our patch test database and approached by letter. The inclusion criterion was: at least one positive reaction to a (meth)acrylate diagnosed between January 2000 and July 2012. Patients using oral prednisone and pregnant women were excluded. The included subjects were patch tested with an isobornyl acrylate dilution series of 0.3%, 0.1%, 0.033% and 0.01% pet. The patch test consisted of van der Bend® square chambers (Van der Bend BV, Brielle, The Netherlands) filled with 20 mg of isobornyl acrylate (CAS no. 5888-33-5; purity 91.8%) in pet., and fixated with Fixomull® Stretch (BSN Medical, Hamburg, Germany).

179 Acrylate, patient no urethane diacrylate (aliphatic) 0.1% alUDA tripropylene glycol diacrylate 0.1% TPGDA BIS-MA 2% ethyleneglycol dimethacrylate 2% EGDMA methyl methacrylate 2% MMA ethyl acrylate 0.1% EA urehane diacrylate (aromatic) 0.05% arUDA trimethylolpropane triacrylate 0.1% TMPTA BIS-GMA 2% triethylene glycol dimethacrylate 2% TEGDMA ethyl methacrylate 2% EMA butyl acrylate 0.1% BA triethylene glycol diacrylate 0.1% TREGDA pentaerthritol triacrylate 0.1% PETA 1,4-butanediol diacrylate 0.1% BUDA 1,4-butanediol dimethacrylate 2% BUDMA N-butyl methacrylate 2% BMA 2-ethylexyl acrylate 0.1% 2-EHA N,N,-methylene bisacrylamid 1% MBAA oligotriacrylate 0.1% 0TA 1.6-hexanediol diacrylate 0.1% HDDA urethane dimethacrylate 2% UEDMA 2-hydroxyethyl methacrylate 2% 2-HEMA 2-hydroxyethyl acrylate 0.1% 2-HEA Total epoxy acrylate 0.5% diethyleneglycol diacrylate 0.1% DEGDA BIS-EMA 1% 2-hydroxypropyl methacrylate 2% 2-HPMA 2-hydroxypropyl acrylate 0.1% 2-HPA - 1a - - - ++ ? - - - - + + ------+ ++ + 7 - - - ++ + 2b - - - - - + - - - - + + - - + ------+ 8 - + - + - - 3a - - - - + ------1 - - - + + 4a - - - ? - + ------+ ------+ + 7 - + - - - + 5b - - - - - ? ------+ 2 - - - - + 6b ------++ ------? - - - - 3 - ++ - - - 7b ------+ ------1 - - - - + - 8a - - - + IR - - - - - IR ------+ + 4 - - - - + 9a ------+ - - + - - - - - + - - - 4 - + - + + 10b - - - + + + - - - + - - + - + + - - - - + - - + 12 ------11b ------+ + - - - - - + - ? ------3 - ++ - ++ ++ 12a - - - + ------+ ++ - ++ ------++ 8 - + - + + 13a - - - + + + - - - + + - - - - + + - - - - + + + 13 chapter 5 - - - + ? 14b - - - + - + ------+ - - - - + - 11 - + - + + 15c - - - + - + - - - + - + ++ ------+ - 7 + ++ 8 - ++ IR - ++ 16a ------+ + - + + - - - - + - - + 2 + - + - - 17b ------8 - ++ - - - 18a - + - - - - - + - + - - ++ - + + - - ++ IR - - - - 3 - - - - - 19a - - - + + - - - - - + - - - - - IR - - IR - - - - 17 - + - + + 20a - - + ++ + + - - + + + + + - + + IR - - - ? + + + 13 + + - - + 21a - ++ - - - + - + - - - + + - + + - - - + + + - ? 3 + - + - - 22b ------+ ------1 - - + - - 23a ------? ------4 - - - + - 24a - - - + - ? ------+ + 148 3/151 11/151 3/151 11/151 12/151 Total 0/151 2/151 1/151 10/151 5/151 8/151 1/151 3/151 2/151 6/151 6/151 7/151 9/151 1/151 8/151 6/151 1/151 1/151 1/151 1/151 5/151 4/151 8/151 12/151 3 4 1 Rank 5 7 10 6 8 8 1

180 Table 1. Patch test results between January 1993 and July 2012 according to the International Contact Dermatitis Research Group (ICDRG) guidelines Bis-EMA, ethoxylated bisphenol A glycol dimethacrylate; Bis-GMA, bisphenol A glycidyl methacrylate; Bis-MA, 2,2-bis[4-(methacryloxy)phenyl]propane. * The concentrations of butyl acrylate, 2-hydroxyethyl acrylate and 2-hydroxypropyl acrylate were 0.5% in patients 1–3 and were reduced to 0.1% in patients 4–24. A subject included in the isobornyl acrylate sub-study. B subject not included in the isobornyl acrylate sub-study. C index patient with a positive patch test reaction to isobornyl acrylate. Patch test results of the (meth)acrylate series containing 29 different acrylates in pet. are given. Tests were read on D2 and D3 for patients 1–12, and on D3 and D7 for patients 13–24. Results were scored according to the ICDRG guidelines. The strongest patch test results on either day are listed.

The isobornyl acrylate was supplied by Sigma Aldrich (Amsterdam, The Netherlands), and dilutions were prepared in-house by our pharmacy. The patch test syringes were stored in a refrigerator (7°C) and freshly prepared every four weeks. The patch tests were prepared immediately prior to application, and applied on the upper back for 48 hr under occlusion. The results were read on D2, D3 and D7 according to the ICDRG guidelines. The study was approved by the Medical Ethics Committee of the University Medical Center Groningen.

Results Database results A total of 151 patients were tested with the (meth)acrylate series at the dermatology department between January 1993 and July 2012. Twenty-four (15.9%) of these 151 patients had chapter a positive reaction to at least one acrylate (Table 1). Most of the positive reactions were caused by 2-HPA 0.1% pet. (12/148), 2-HEA 0.1% pet. (12/148), 2-hydroxypropyl methacrylate 7 (2-HPMA) 2% pet. (11/148), and diethyleneglycol diacrylate (DEGDA) 0.1% pet. (11/148), whereas no positive reactions to urethane diacrylate (aliphatic) 0.1% pet. were encountered. We noted seven irritant reactions to various acrylates, and only ten doubtful reactions were registered. A minority (n = 8, 33.3%) of the patients were males, and half of them (n = 13, 54.2%) were older than 40 years, with a mean age of 42.7 years (range 25–65 years). Only 4 (16.7%) patients suffered from atopic dermatitis. The MOAHLFA index was M33.3,O87.5, A54.2, H87.5, L0.0, F12.5, A16.7. The duration of their dermatitis ranged from 1 month to 2.5 years (mean 6 months). The duration of exposure to acrylates varied from 2 days to 20 years, but was unknown in several cases. The final diagnosis was ‘allergic contact dermatitis’ in almost all of the patients (95.8%); in one patient, the diagnosis was hyperkeratotic hand eczema, possibly occupationally related. Twenty-one (87.5%) of the 24 sensitized patients had a risk

181 of occupational exposure, and 14 cases (66.7%) were labeled as certainly occupation-related. Of the remaining occupational cases, six cases (28.6%) were labeled as probably occupation- related, and 1 case (4.8%) was labeled as possibly occupation-related. Among the occupationally exposed patients were nail stylists (n = 8), assembly line workers (n = 4), printers (n = 3), laboratory technicians (n = 2), dental technicians (n = 2), a painter (n = 1), and a dairy farmer (n = 1) who used a two component hoof glue. All of the occupational cases presented with fingertip dermatitis or hand eczema, except for one patient who developed facial eczema in addition to the existing atopic dermatitis after starting training as a nail stylist. In two cases, there was secondary spread, and one patient suffered from angioedema. Three (12.5%) of our 24 patients with (meth)acrylate contact allergy were considered to have been sensitized in a non-occupational manner. In one case, the patient suffered from stomatitis caused by new dentures (relevance: labeled as ‘certain’), whereas another case might have been caused by the use of artificial nails (relevance: labeled as ‘possible’). The third case developed erythema and oedema on her face after using an acrylate coating for her boat (relevance: labeled as ‘certain’). The symptoms resolved in all three cases after avoidance of the acrylate containing products.

Isobornyl acrylate sub-study Fourteen of the 24 potential subjects were included in the isobornyl acrylate sub-study (Fig. 3). There were various reasons for non-inclusion: three subjects were excluded because they used immunosuppressive drugs, two subjects declined because they had previously developed intense itching after patch testing, and five subjects either did not respond or were untraceable. The MOAHLFA index of the 14 included subjects was M: 28.6, O: 85.7, A: chapter 50.0, H: 85.7, L: 0, F: 14.3, A: 14.2. The majority of the included subjects (n = 12, 85.7%) had a known occupational risk. Among the occupationally exposed subjects were nail stylists 7 (n = 6), printers (n = 2), assembly line workers (n = 1), laboratory technicians (n = 1), dental technicians (n = 1), and a dairy farmer (n = 1) who used a two-component hoof glue. Two (14.3%) of the 14 cases were sensitized in a non-occupational manner. The 14 included subjects were patch tested with the dilution series of isobornyl acrylate: 0.3%, 0.1%, 0.033% and 0.01% (Table 2). None of them developed a positive patch test reaction to the isobornyl acrylate. However, isobornyl acrylate 0.3% provoked an irritant reaction in 3 (21.4%) of the subjects. Five healthy controls were also patch tested with the isobornyl acrylate dilution series. None of them showed any irritant reactions or signs of sensitization.

182 (meth)acrylate patch test at the UMCG Dermatology department between 01-01-2000 and 01-07-2012 n = 151

Positive reaction (at least +) for at least one (meth) acrylate n = 24

Excluded because of prednisone n = 3 Unwilling to participate n = 2 Untraceable or not responding n = 5

Included and patch tested with isobornyl acrylate n = 14

Fig. 3. Flow chart of subjects included in the isobornyl acrylate sub-study; UMCG, University Medical Center Groningen.

Subject 1 2 3 4 - 14 Day D2 D3 D7 D2 D3 D7 D2 D3 D7 D2 D3 D7 IBA 0.3% IR IR - - IR - IR IR IR - - - chapter IBA 0.1% ------IBA 0.033% ------7 IBA 0.01% ------

Table 2. Patch test results of isobornyl acrylate according to the International Contact Dermatitis Research Group guidelines -, negative reaction; IBA, isobornyl acrylate (% dilution in pet.); IR, irritant reaction. Three subjects developed an irritant reaction to isobornyl acrylate 0.3%; none developed a positive reaction.

Discussion Acrylate contact allergy Acrylates are considered to constitute an uncommon, although important, cause of contact allergy in general dermatology practices. The prevalence, based on screening series added to the baseline, varies between 1.0% and 1.6%.10,11 In our study, 24 of the 151 (15.9%) patients patch tested with the (meth)acrylate series in the last 19 years reacted positively to at least

183 one acrylate. This difference may be explained by the selected population. At our department, (meth)acrylates are not routinely tested, but are only tested if the physician has a strong suspicion of acrylate contact allergy.12 Moreover, a substantial amount of data was derived from our Expert Centre for Eczematous and Occupational Dermatoses, whereby most of the patients were occupationally exposed. This is in line with a previous study by Kanerva et al. who reported positive reactions in 48 out of 275 (17.5%) patients with a history of occupational exposure to (meth)acrylates.13 Between January 1993 and January 2002, we tested 32 patients with the (meth)acrylate series, and only 3 (9.4%) of them developed a positive reaction. Between 2002 and 2011, a total of 104 patients were tested, and 15 (14.4%) of them had positive reactions. The (meth) acrylate series was patch tested in 16 patients during the last year, and 6 (37.5%) of them have had positive reactions so far. This increase may be explained by the more frequent use of (meth)acrylates.14 Another plausible explanation is the selection of our patients who are eligible for patch testing. In addition, the patch test procedure has been improved over the years. Handling of the acrylate allergens was improved by storing them in a refrigerator and preparing them only minutes before application, to prevent evaporation and thus optimize the patch test concentration. Another improvement was the reading of the results on D7, to pre-empt late reactions. Eight of the 88 positive reactions were exclusively identified on D7; however, all of these patients had reacted to other acrylates on D3. However, for an accurate and complete diagnosis, an additional D7 reading is recommended. Most positive reactions in our (meth)acrylate series were elicited by 2-HPA, 2-HEA, 2-HPMA, and DEGDA. In a comprehensive overview, Kanerva et al. found a similar ranking, with 2-HEA (16/132), 2-HPMA (29/242) and 2-HPA (14/132) in the top four.13 DEGDA (13/243) was chapter ranked 10th, but this acrylate caused more positive reactions (23/66) in the more recent overview by Aalto-Korte et al. who tested 541 patients and found that 75 of them reacted to 7 at least one acrylate.3 We did not observe positive reactions to urethane diacrylate (aliphatic) (alUDA), which is known to be a rare sensitizer. Even so, positive reactions to alUDA have been reported, especially in association with artificial nails2,15,16, warranting inclusion of this allergen in the (meth)acrylate series.

Screening for (meth)acrylates is of real interest. Inclusion of a few acrylates in the baseline series as a screening tool for acrylate contact allergy has been proposed. A hypothetical series of four acrylates [2-HPA, ethyleneglycol dimethacrylate (EGDMA), ethoxylated bisphenol A glycol dimethacrylate and trimethylolpropane triacrylate] would have identified 92% of our patients. Goon et al. composed two different acrylate mixes, one with triethylene glycol diacrylate (TREGDA) 0.1%, 2-hydroxyethyl methacrylate (2-HEMA) 1.0%, and EGDMA 1.0%, and another with TREGDA 0.1% and 2-HEMA 2.0%, to screen for acrylate sensitization in the baseline series. 17 Screening with the individual components of these mixes

184 would have identified fewer than half of our patients, respectively 11 (45.8%) and 10 (41.7%), although we did not investigate 2-HEMA at 2.0%, but only at 1.0%. Goon et al. tested 1,632 patients, and 48 had positive results to one or more (meth) acrylates. 1,2 They composed a hypothetical screening series containing 2-HEMA, EGDMA, TREGDA, 2-HPMA, bisphenol A glycidyl methacrylate, and 1,4-butanediol diacrylate (BUDA) or 1.6-hexanediol diacrylate (HDDA), which would have identified all of their past patients (dental, industrial, and nail) with suspected (meth)acrylate contact allergy. However, in our cohort, 17 (70.8%) of our 24 patients would have been identified with this series, irrespective of whether BUDA or HDDA was used. Aalto-Korte et al. produced a hypothetical screening series with four allergens (EGDMA, DEGDA, 2-HPMA, and pentaerythritol triacrylate) that identified 92.4% (61/66) of their occupationally exposed patients3, whereas it identified only 17 (70.8%) of our 24 patients. Owing to this discrepancy and the fact that a substantial number of our patients would have been missed, we prefer to use a supplementary series containing 29 (meth)acrylate allergens in addition to the baseline series in patients.

Isobornyl acrylate contact allergy An additional case of contact allergy caused by isobornyl acrylate is reported in this study. In this occupational case, the sensitizers were uncured UV ink and acrylate coating. With avoidance of these products, the skin problems of the industrial process operator resolved. To date, one year later, he is working at another department in the same factory without any complaints. Sensitization to isobornyl acrylate is uncommon. Our study did not show any new cases of isobornyl acrylate sensitization besides the index patient. Together with the study by Busschots et al., this means that only three cases of isobornyl acrylate sensitization have been identified to date.4 Unfortunately, our index patient was only patch tested with chapter the 0.1% concentration, and declined further patch testing. Otherwise, he would have been tested with the dilution series to ascertain the lowest concentration able to elicit a positive 7 reaction to isobornyl acrylate.18 A concentration of 0.1% is the preferred patch test concentration for acrylates, because of the irritant potency and the risk of active sensitization.19,20 However, acrylates are known for their volatility, and recent studies have shown a poor correlation between the measured (lower) and the stated (higher) concentrations of acrylate patch test concentrations.21,22 In addition, a woman suspected of having occupationally related contact allergy provoked by an adhesive containing 61% isobornyl acrylate did not react to an isobornyl acrylate patch test.5 Moreover, various workers exposed to acrylates showed no reaction to isobornyl acrylate 0.1%.3,6,23 This, together with the lack of evidence of sensitization to 0.1% in other studies, compelled us to include a higher concentration of 0.3%. However, the concentration of 0.3% turned out to be too high, and caused irritant reactions in three patients. These reactions consisted of ‘shiny skin’ in two patients and irritant papules on the corners of the patch test chambers in the remaining subject. Although the

185 index patient was only patch tested with 0.1% isobornyl acrylate, he did develop a 2+ positive reaction, which is in line with the patients reported by Busschots et al., who developed 2+ and 3+ positive patch test reactions at a concentration of 0.1%, making it a legitimate concentration for showing sensitization.4 Acrylates, which are chemically more similar to isobornyl acrylate, were expected to cross- react with isobornyl acrylate. Given the fact that none of the included subjects showed any reaction to isobornyl acrylate, no valid conclusions can be drawn regarding the cross reactivity of isobornyl acrylate with other (meth)acrylates in vivo. Furthermore, cross-reactivity of (meth)acrylates with isobornyl acrylate seems to be less plausible, because cross-reaction was not observed in this study and has not been reported in the literature.

Conclusion We have described an additional rare case of allergic contact dermatitis caused by isobornyl acrylate. Cross-reactivity between isobornyl acrylate and other acrylates could not be shown in a selected group of previously sensitized patients. The ideal patch test concentration for isobornyl acrylate seems to be 0.1%. This study and the current literature provide insufficient support for isobornyl acrylate to be routinely used in the (meth)acrylate patch test series at our department. Isobornyl acrylate contact allergy seems to be rare. However, this allergen should be considered as a potential sensitizer in individual cases.

chapter 7

186 References 1. Goon AT, Isaksson M, Zimerson E, Goh CL, Bruze M. Contact allergy to (meth)acrylates in the dental series in southern Sweden: simultaneous positive patch test reaction patterns and possible screening allergens. Contact Dermatitis 2006:55:219-226. 2. Goon AT, Bruze M, Zimerson E, Goh CL, Isaksson M. Contact allergy to acrylates/methacrylates in the acrylate and nail acrylics series in southern Sweden: simultaneous positive patch test reaction patterns and possible screening allergens. Contact Dermatitis 2007:57:21-27. 3. Aalto-Korte K, Henriks-Eckerman ML, Kuuliala O, Jolanki R. Occupational methacrylate and acrylate allergy--cross-reactions and possible screening allergens. Contact Dermatitis 2010:63:301-312. 4. Busschots AM, Meuleman V, Poesen N, Dooms-Goossens A. Contact allergy to components of glue in insulin pump infusion sets. Contact Dermatitis 1995:33:205-206. 5. Kanerva L, Jolanki R, Leino T, Estlander T. Occupational allergic contact dermatitis from 2-hydroxyethyl methacrylate and ethylene glycol dimethacrylate in a modified acrylic structural adhesive. Contact Dermatitis 1995:33:84-89. 6. Kiec-Swierczynska M, Krecisz B, Swierczynska-Machura D, Zaremba J. An epidemic of occupational contact dermatitis from an acrylic glue. Contact Dermatitis 2005:52:121-125. 7. Schnuch A, Geier J, Uter W, Frosch PJ, Lehmacher W, Aberer W, et al. National rates and regional differences in sensitization to allergens of the standard series. Population-adjusted frequencies of sensitization (PAFS) in 40,000 patients from a multicenter study (IVDK). Contact Dermatitis 1997:37:200-209. 8. Lachapelle JM. A proposed relevance scoring system for positive allergic patch test reactions: practical implications and limitations. Contact Dermatitis 1997:36:39-43. chapter 9. Ale S, Maibach HI. Clinical relevance in allergic contact dermatitis: an algorithmic approach. Derm Beruf Umwelt 1995:43:119-121. 7 10. Drucker AM, Pratt MD. Acrylate contact allergy: patient characteristics and evaluation of screening allergens. Dermatitis 2011:22:98-101. 11. Goon AT, Bruze M, Zimerson E, Goh CL, Soo-Quee Koh D, Isaksson M. Screening for acrylate/ methacrylate allergy in the baseline series: our experience in Sweden and Singapore. Contact Dermatitis 2008:59:307-313. 12. van der Valk PG, Devos SA, Coenraads PJ. Evidence-based diagnosis in patch testing. Contact Dermatitis 2003:48:121-125. 13. Kanerva L, Jolanki R, Estlander T. 10 Years of Patch Testing with the (Meth)acrylate Series. Contact Dermatitis 1997:37:255-258. 14. Geukens S, Goossens A. Occupational contact allergy to (meth)acrylates. Contact Dermatitis 2001:44:153-159.

187 15. Kanerva L, Lauerma A, Estlander T, Alanko K, Henriks-Eckerman ML, Jolanki R. Occupational allergic contact dermatitis caused by photobonded sculptured nails and a review of (meth) acrylates in nail cosmetics. Am J Contact Dermat 1996:7:109-115. 16. Koppula SV, Fellman JH, Storrs FJ. Screening for acrylate dermatitis associated with artificial nails. Am J Contact Dermat 1995:6:78-85. 17. Goon AT, Bruze M, Zimerson E, Goossens A, Goh CL, Isaksson M. High frequency of false- positive reactions in attempted patch testing with acrylate/methacrylate mixes. Contact Dermatitis 2012:67:157-61. 18. Fregert S. Publication of allergens. Contact Dermatitis 1985:12:123-124. 19. Kanerva L, Estlander T, Jolanki R. Sensitization to patch test acrylates. Contact Dermatitis 1988:18:10-15. 20. Cronin E editor. Contact Dermatitis. Edingburgh. London, New York,Churchill-Livingstone,1980. 21. Goon AT, Bruze M, Zimerson E, Sörensen Ö, Goh CL, Koh DS, et al. Correlation between stated and measured concentrations of acrylate and methacrylate allergens in patch-test preparations. Dermatitis 2011:22:27-32. 22. Goon AT, Bruze M, Zimerson E, Sörensen Ö, Goh CL, Koh DS, et al. Variation in allergen content over time of acrylates/methacrylates in patch test preparations. Br J Dermatol 2011:164:116-124. 23. Aalto-Korte K, Alanko K, Kuuliala O, Jolanki R. Occupational methacrylate and acrylate allergy from glues. Contact Dermatitis 2008:58:340-346.

chapter 7

188 chapter 7

189 Chapter

190 Co-sensitization to ascaridole and tea tree oil Contact Dermatitis 2013;69:187-189

Wietske Andrea Christoffers1, Brunhilde Blömeke2, Pieter-Jan Coenraads1, Marie-Louise Anna Schuttelaar1

1 Department of Dermatology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands 2 Department of Environmental Toxicology, University of Trier, Trier, Germany

chapter 8

Key words: allergic contact dermatitis; ascaridole; Melaleuca alternifolia; patch test; tea tree oil.

191 chapter 8

192 Tea tree oil (Melaleuca alternifolia, CAS no. 68647-73-4) is an oil distilled from the leaves of Melaleuca alternifolia. Indigenous populations in Australia used it as an antiseptic and as an alternative herbal medicine for centuries, but tea tree oil is now an increasingly popular ingredient in household and cosmetic products. Various reports of contact allergy to tea tree oil exist; however, the sensitizing component is still an issue of debate.

Case 1 A 50-year-old woman presented at our dermatology department with periorbital dermatitis, which had been present for years. Patch tests were performed with the local extended Euro- pean baseline series and a cosmetics series, applied with van der Bend® square chambers (Van der Bend BV, Brielle, The Netherlands), fixed with Fixomull® Stretch (BSN Medical, Hamburg, Germany), and read according to the guidelines of the International Contact Dermatitis Research Group. No reactions were seen. However, since our previous study, we have added ascaridole (provided by the Institute of Pharmacology, University of Bonn, Germany; purity 99.8%; in-house preparation) and oxidized tea tree oil 5% (Chemotechnique Diagnostics, Vellinge, Sweden) to our cosmetics series1, and this resulted in the following reactions:

D3 D7 Oxidized tea tree oil 5% pet ? ? Ascaridole 1% pet. + + Ascaridole 2% pet. + + Ascaridole 5% pet. IR (Fig. 1a) IR

Table 1. Patch test results of case I.

The patient denied using pure tea tree oil, so she was instructed to study the household and chapter cosmetic products that she used for ‘tea tree oil’ as well as for ‘melaleuca alternifolia’. Tea tree oil turned out to be a component of the soap that she used (Etos® cream soap; Etos, 8 Beverwijk, The Netherlands), but also of Eczefleur® cream (Bloem natuurproducten, Winschoten, The Netherlands), an herbal remedy which the patient used to treat her dermatitis. The patient stopped using Eczefleur® cream, and pimecrolimus (Elidel®) was prescribed, which resolved the periorbital dermatitis.

Case 2 A 26-year-old man suffered from periorbital dermatitis and persistent folliculitis barbae. Patch testing with the local extended European baseline series and our cosmetics series, with van der Bend® square chambers, resulted in the following reactions:

193 D3 D7 Oxidized tea tree oil 5% pet + + Ascaridole 1% pet. + + Ascaridole 2% pet. + + Ascaridole 5% pet. + (Fig. 1b) +

Table 2. Patch test results of case II

Tea tree oil turned out to be a component of the patient’s shaving oil: Williams® shaving oil (The David Somerset Skincare Company, UK). The patient switched to another shaving product, and tacrolimus ointment (Protopic®) was prescribed, which resolved his skin problems.

a b

Fig. 1. (a) Case I: irritant reaction to ascaridole 5% on D3 with sharp demarcated erythema and yellow crustae. (b) Case II: positive reaction to ascaridole 5% on D3 with erythema covering the whole area, infiltration, and a few papules. chapter Discussion Ascaridole (CAS no. 512-85-6) is one of the proposed sensitizers in tea tree oil.2,3 Ascaridole 8 is a terpene endoperoxide that may develop in tea tree oil. Several reports of sensitization to ascaridole have been published; however, the patch test concentration of ascaridole varied from 5% to 10%, and most reactions were strong positive.2,4,5 A previous study by our group investigated ascaridole at concentrations of 1% pet. and 5% pet., recommended testing with 2% pet.1 The patch test results described in the current report support this proposal, because case I developed an obvious irritant reaction to ascaridole 5% (Fig. 1a), whereas she showed positive reactions to ascaridole 1% and 2%. Further studies into the patch test concentration of ascaridole as a sensitizing component of tea tree oil are needed. In our previous study, we were unable to explain the (clinical) relevance of a positive reaction

194 in the majority of patients.1 For ascaridole 1%, the source was identified in 1 (0.16%) of the 602 tested patients, and in only 2 (1.39%) of the 144 tested patients investigated with ascaridole 5%. On the basis of the literature, we composed a checklist of a wide range of products that have been reported to contain tea tree oil. This checklist included pure tea tree oil, and well-known substances such as shampoo, day cream, and aromatherapy products, but also less obvious products, such as shoe cream and deworming products for pets.3,6 With this list, patients can inspect their kitchen and bathroom cabinets to identify products containing tea tree oil or Melaleuca alternifolia. This article reports the first possible relevant reactions to a relatively low concentration of ascaridole as a potentially sensitizing component of tea tree oil.

chapter 8

195 References 1. Bakker CV, Blömeke B, Coenraads PJ, Schuttelaar ML. Ascaridole, a sensitizing component of tea tree oil, patch tested at 1% and 5% in two series of patients. Contact Dermatitis 2011:65:240-241. 2. Hausen BM, Reichling J, Harkenthal M. Degradation products of monoterpenes are the sensitizing agents in tea tree oil. Am J Contact Dermat 1999:10:68-77. 3. Pirker C, Hausen BM, Uter W, Hillen U, Brasch J, Bayerl C, et al. Sensitization to tea tree oil in Germany and Austria. A multicenter study of the German Contact Dermatitis Group. J Dtsch Dermatol Ges 2003:1:629-634. 4. Hausen BM. Evaluation of the main contact allergens in oxidized tea tree oil. Dermatitis 2004:15:213-214. 5. Hausen BM. Kontaktallergie auf Teebaumöl und Ascaridol. Akt Dermatol 1998:24:60-62. 6. Rutherford T, Nixon R, Tam M, Tate B. Allergy to tea tree oil: retrospective review of 41 cases with positive patch tests over 4.5 years. Australas J Dermatol 2007:48:83-87.

chapter 8

196 chapter 8

197 Chapter

198 The optimal patch test concentration for ascaridole as a sensitizing component of tea tree oil Contact Dermatitis 2014;71:129-137

Wietske Andrea Christoffers1, Brunhilde Blömeke2, Pieter-Jan Coenraads1, Marie-Louise Anna Schuttelaar1

1 Department of Dermatology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands 2 Department of Environmental Toxicology, University of Trier, Trier, Germany

chapter 9

Key words: allergic contact dermatitis; ascaridole; Melaleuca alternifolia; tea tree oil.

199 Abstract Background: Tea tree oil is used as a natural remedy, but is also a popular ingredient in household and cosmetic products. Oxidation of tea tree oil results in degradation products, such as ascaridole, which may cause allergic contact dermatitis. Objectives: To identify the optimal patch test concentration for ascaridole, and to investigate the relationship between a positive reaction to ascaridole and a positive reaction to oxidized tea tree oil. Patients/materials/methods: Three hundred and nineteen patients with eczema were patch tested with ascaridole 1%, 2%, and 5%, and 250 patients were patch tested with oxidized tea tree oil 5%. Readings were performed on D3 and D7 according to a patch test calibration protocol. Results: With an increasing ascaridole test concentration, the frequency of positive reactions increased: ascaridole 1%, 1.4%; ascaridole 2%, 5.5%; and ascaridole 5%, 7.2%. However, the frequencies of irritant and doubtful reactions also increased, especially for ascaridole 5%. A positive reaction to ascaridole was related to a positive reaction to tea tree oil. Conclusions: This study is in support of ascaridole being a sensitizer. We recommend patch testing with ascaridole at 2%. The finding that every positive reaction to oxidized tea tree oil is accompanied by a positive reaction to ascaridole suggests that ascaridole might be a contact allergen in oxidized tea tree oil.

chapter 9

200 Introduction Tea tree oil (CAS no. 68647-73-4) is an oil originally distilled from the leaves of Melaleuca alternifolia, which is cultivated in the states of New South Wales and Queensland in Austra- lia. The indigenous Australian population used the oil as an herbal medicine for centuries. Nowadays, tea tree oil is used as a natural remedy for various conditions, such as acne and warts, and the oil is a popular ingredient in household and cosmetic products. Although consumers might assume that natural products are safer than synthetic products, tea tree oil is known to cause allergic contact dermatitis. In 1991, the first cases of allergic contact dermatitis caused by tea tree oil were reported in patients who had used tea tree oil as a natural remedy and had become sensitized.1,2 In these early reports, patch tests were performed with undiluted tea tree oil, sometimes even causing extreme bullous reactions. Over time, tea tree oil diluted 1:3 in alcohol, 1:5 in olive oil or 1% in alcohol was recommended, on the basis of a few case reports.3,4 This was supported by a study of Coutts et al. in 2002 in the United Kingdom, who patch tested 550 dermatitis patients with pure, oxidized tea tree oil.5 In total, 13 patients (2.4%) developed a positive reaction; however, 209 patients (38%) developed an irritant reaction to the pure patch test preparation. In 1999, a standardized variant of oxidized tea tree oil 5% was added to the North American Contact Dermatitis Group screening panel. After an initial rise in prevalence6, the rate of sensitization to oxidized tea tree oil 5% in petrolatum was 1.0% in 4299 dermatitis patients in 2009–2010 in the United States.7 This is comparable with the results of other studies in Europe and Australia that reported prevalence rates of sensitization to oxidized tea tree oil that varied between 0.5% and 1.8% of the patch tested population.8-10 Besides the search for the optimal patch test concentration of tea tree oil, the search for the sensitizing component was started, given the fact that tea tree oil contains >100 different components. In 1999, Hausen postulated that the degradation products of monoterpenes are probably the sensitizers in tea tree oil.11 These monoterpenes develop through photo- oxidation of fresh tea tree oil within days to months. The most important sensitizers are probably the sesquiterpene α-phellandrene and the monoterpenes terpinolene, ascaridole, α-terpinene, 1,2,4-trihydroxy menthane, α-phellandren, and (+)-limonene.8,11-13 Ascaridole (1,4-epodioxy-p-menth-2-ene; CAS no. 512-85-6) (Fig. 1) is a potential sensitizer in oxidized tea tree oil.8,11 It is a monoterpene, derived from oxidized α-terpene. In a previous study chapter we patch tested 602 patients with ascaridole 1%, and found 9 (1.5%) positive reactions.14 Only one positive reaction to ascaridole 1% was judged to be clinically relevant, because 9 the patient recalled the use of pure tea tree oil followed by the development of dermatitis. Furthermore, ascaridole 1% caused 15 (2.5%) doubtful reactions, which led to the assump- tion that the concentration of ascaridole 1% was too low to produce positive reactions. We therefore patch tested ascaridole 5% in 144 patients in the same study; this caused 21

201 positive reactions, but also 14 (9.7%) doubtful and 5 (3.5%) irritant reactions. Only in 1.4% of these patients was the reaction clinically relevant. One patient recalled the use of pure tea tree oil, and the other patient had used a shaving soap containing tea tree oil.15 Interestingly, positive and irritant reactions caused by ascaridole 5% were remarkably similar, especially with regard to bullous reactions, making it difficult to differentiate between positive and irritant reactions. The substantial number of irritant reactions warranted a decrease in the patch test concentration. Thereupon, patch testing with ascaridole 2% was initiated, and the results of this are described in this article. In addition, the clinical relevance of the positive reactions was assessed more thoroughly. Finally, we added tea tree oil to our cosmetic patch test series, to investigate the relationship between a positive reaction to ascaridole and a positive reaction to oxidized tea tree oil.

Fig. 1. Chemical structure of ascaridole (1,4-epodioxy-pmenth-2-ene).

Materials and Methods Chemicals Ascaridole (99.8% purity) was provided by the Institute of Pharmacology, University of Bonn, Germany. It was diluted in pet. in-house by the pharmacy of the University Medical Center Groningen, at concentrations of 1%, 2%, and 5%. The preparations were used for a maximum of three months, and were stored in a refrigerator (7°C) during this period. Oxidized tea tree oil 5% pet. was purchased from Chemotechnique Diagnostics, Vellinge, Sweden. To oxidize the tea tree oil, pure tea tree oil was mixed with air for 5 min daily in an chapter Erlenmeyer flask. The flask was exposed to a daylight fluorescent lamp for 12 hr daily for 9 three days (Chemotechnique Diagnostics, pers. comm.2013). Patch test procedure Patch tests were applied to the upper back for 48 hr of occlusion with van der Bend® square chambers (Van der Bend BV, Brielle, The Netherlands), and fixed with Fixomull® Stretch (BSN Medical, Hamburg, Germany). Approximately 25 μg of the test preparation was applied to each chamber. Results were read on D3 and D7.

202 In order to improve the quality of the readings and to better distinguish the morphology of irritant and allergic reactions, a patch test calibration protocol was used.16,17 Exclusion criteria were pregnancy, an angry back in the past, and the use of oral immunosuppressive drugs in the two weeks prior to the patch test procedure.

Outpatients From November 2011 to March 2013, ascaridole 1%, 2% and 5% was added to the cosmetic patch test series at the Dermatology Department of the University Medical Center Groningen. A total of 290 consecutive outpatients with eczema were patch tested with the three different concentrations of ascaridole. In February 2012, oxidized tea tree oil 5% was added, and was patch tested in 221 of the 290 patients.

Re-patch tested patients Twenty-nine patients who were patch tested with only ascaridole 1% or 5% between March 2008 and October 2011 and had developed a positive, doubtful or irritant reaction in the past were re-patch tested with the complete series of ascaridole 1%, 2% and 5% and oxidized tea tree oil between November 2012 and February 2013.14 This study was approved by the Medical Ethics Committee of the University Medical Center Groningen, and informed consent was obtained from all participants. Combining the 290 outpatients with the 29 re-patch tested patients gives 319 patients who were patch tested with the three concen-trations of ascaridole. In total, 250 of these 319 patients were tested with oxidized tea tree oil in addition to the three different ascaridole concentrations.

Clinical relevance The clinical relevance of a positive patch test result was evaluated in relation to the history of exposure and the subsequent dermatitis pattern. A positive patch test reaction was judged to be clinically relevant when a patient had been exposed to a product containing tea tree oil, and the presence of tea tree oil was confirmed by the product label. In addition, the patient had to develop dermatitis after the exposure, and a causative relationship had to be suspected. A repeated open application test was not conducted to establish clinical chapter relevance. The potential source of exposure and the clinical relevance were assessed for patients with a 9 positive reaction to ascaridole or tea tree oil by means of a questionnaire. Specific questions were asked about the current and past use of pure tea tree oil and natural products, and visits to beauticians, in relation to dermatitis to investigate potential sources of exposure to tea tree oil or other ‘natural’ products. In addition, patients were provided with a list of products

203 that might contain tea tree oil, such as pure tea tree oil, and well-known applications such as body and facial creams, shampoos, shower gels, massage oils, and aromatherapy materials. After a literature study, less obvious products, such as shoe cream and anti-parasite products for pets, were also included in this list.8,10,12 Supported by a list of synonyms provided by us, patients were requested to inspect their kitchen and bathroom cabinets to identify products containing, among others, ‘tea tree oil’ or ‘melaleuca alternifolia’.

Statistics Analyses were performed with IBM spss™ version 20. The chi-squared test and Mann– Whitney U-test were used to compare differences between groups. Fisher’s exact test was used to investigate the relationship between a positive reaction to ascaridole and a positive reaction to oxidized tea tree oil.

Results Patch test results Of the 290 consecutive patch tested outpatients, 89 (30.7%) were male and 143 (49.3%) were aged over 40 years. In 68 (23.4%) patients, occupational factors were suspected to play a role in general. The primary location of the dermatitis was the face in 137 (47.2%) patients, the hands in 90 (31.0%) patients, and the legs in five (1.7%) patients. In 122 (42.1%) patients, a history of atopic dermatitis was reported by the patient. The MOAHLFA index for this population was: M30.7, O23.4, A42.1, H31.0, L1.7, F47.2, and A49.3. The results of the patch testing in terms of total numbers of tested patients, and the frequency of positive, doubtful and irritant reactions, are shown in Table 1. An increase in the number of positive reactions with increasing ascaridole test concentration was recorded: ascaridole 1% caused positive reactions in 1.4% (4/290); ascaridole 2% caused positive reactions in 5.5% (16/290); and ascaridole 5% caused positive reactions in 7.2% (21/290). However, the higher frequencies of positive reactions with increasing concentrations of ascaridole were accompanied by higher frequencies of irritant reactions; ascaridole 1% caused irritant reactions in 0.3% (1/290); ascaridole 2% caused irritant reactions in 1.7% (5/290); and ascaridole 5% caused irritant reactions in 9.0% (26/290). Also, the number of chapter doubtful reactions increased with concentration: ascaridole 1% caused doubtful reactions in 3.1% (9/290); ascaridole 2% caused doubtful reactions in 4.1% (12/290); and ascaridole 5% 9 caused doubtful reactions in 10.3% (30/290). Of the nine patients with doubtful reactions to ascaridole 1%, five patients showed positive reactions and four patients showed negative reactions to ascaridole 2%. Of the 12 patients with doubtful reactions to ascaridole 2%, only one patient showed a positive reaction to ascaridole 5%. Three reactions remained doubtful for ascaridole 5%. Three of 12 patients with doubtful reactions to ascaridole 2% showed irritant reactions to ascaridole 5% and in five patients with doubtful reactions to ascaridole

204 2%, no reaction was caused by ascaridole 5%. All of the patients with positive reactions to ascaridole 1% also showed reactions to ascaridole 2%. In the unselected 221 consecutive outpatients we found 2 (0.9%) positive reactions to oxidized tea tree oil and 1 (0.5%) doubtful reaction; no irritant reactions were registered (Table 1).

No. + (n,%) ++ (n,%) +++ IR (n, %) ? (n,%) Total Relevant tested (n,%) +(n,%) (n,%) Ascaridole 1% 290 4 (1.4) - - 1 (0.3) 9 (3.1) 4 (1.4) 1 (0.3) Ascaridole 2% 290 15 (5.2) 1 (0.3) - 5 (1.7) 12 (4.1) 16 (5.5) 5 (1.7) Ascaridole 5% 290 20 (6.9) 1 (0.3) - 26 (9.0) 30 (10.3) 21 (7.2) 4 (1.4) Tea tree oil 5% 221 2 (0.9) - - - 1 (0.5) -

Table 1. Patch test results for ascaridole at 1%, 2% and 5% and oxidized tea tree oil, patch tested from February 2012 to March 2013 in consecutive dermatitis patients of the Department of Dermatology. The strongest result on either D3 or D7 is registered. Two hundred and twenty-one of the 290 patients were patch tested with ascaridole 1%, ascaridole 2%, ascaridole 5%, and oxidized tea tree oil 5%. MOAHLFA index: M30.7, O23.4, A42.1, H31.0, L1.7, F47.2, and A49.

Co-sensitization to ascaridole and tea tree oil Combining the 221 outpatients consecutively tested with ascaridole 1%, 2% and 5%and tea tree oil 5% with the 29 re-patch tested patients gives a total of 250 patients who were patch tested with this entire series (Table 2). Six patients developed a positive reaction upon patch testing with oxidized tea tree oil. These six patients also showed a positive reaction to ascaridole (Tables 2 and 3). The association between a positive reaction to oxidized tea tree oil and a positive reaction to ascaridole was statistically significant for all concentrations (ascaridole 1%, p<0.001; ascaridole 2%, p<0.001; ascaridole 5%, p<0.001) and ascaridole and tea tree oil overall (p<0.001). An increased number of patients reacting to other fragrances, such as linalool, eugenol, or isoeugenol, was not observed, and neither did the number of sensitizations to other oils such as peppermint oil or eucalyptus oil deviate from that in the general patch tested population (Table 3).

chapter Tea tree oil positive Tea tree oil negative Total Any ascaridole concentration positive 6 24 30 All ascaridole concentrations negative 0 220 220 9 Total 6 244 250

Table 2. Co-sensitization to oxidized tea tree oil 5% and any ascaridole concentration. The table includes the patch test results of 250 patients who were all patch tested with ascaridole 1%, ascaridole 2%, ascaridole 5%, and oxidized tea tree oil. Two hundred and twenty-one of the 250 were outpatients patch tested between February 2012 and March 2013, and 29 were re-patch tested patients. MOAHLFA index: M31.0, O22.6, A48.9, H30.7, L1.6, F47.3, and A50.5

205 49 52 38 58 23 55 28 24 60 58 59 43 57 Age F F F M F F M F F M M F Gender F ? - + - + - + - + - + ? TTO5% + + + - + ? + + + + IR + ? Asc1% + + + + + + + + + + IR ++ Asc2% + IR IR ++ + + + + + IR ++ + + Asc5% + IR ibital Perior Hand Hand Hand Hand Facial bital Perior Hand Hand Facial Hand Facial Facial tion Loca - - - cream* Eczefleur Etos* Baby soap - - - - - oil* shaving Williams - - - - PhD wax* - source Exposure current never never never current never never current past in the never never past in the never complaints Beautician, no no yes yes yes current never current current never never past in the past in the never never never never current complaints products, Natural yes yes no yes no no never never never never never never never never never never never never never complaints pathic, Homeo-

never never never never never never never never never never never never never therapy Aroma- Balsam of Peru, wool alcohol MDBGN, CAPB Co, colophony, Cr, Ni, PTBFR, MCI/MI Isoeugenol CAPB - bamate CAPB, iodoproponynyl butalcar mix, Ni FMI, MDBGN, MBT, mercapto - lactone mix, sorbic acid Parthenolide, sesquiterpene moss abs, PTBFR, thiolactic acid methyl heptene carbonate, oak monothioglycolate, MDBGN, CAPB, Captan, FMI, Glycerol CAPB, Ni Concomittant reactions - chapter 9 -

206 Co, HMBA, Ni, thiomersal Balsam of Peru, clove oil, FMI, FMII, lemon grass oil, wool alcohol Co, Colophony, Cr, FMII, PTBFR, Cr, Co, Colophony, thiomersal, Anisyl alcohol FMI, MDBGN, Wool alcohol FMI, MDBGN, Wool - Budesonide, FMII, MCI/MI - Sesquiterpene lactone mix, Laurel Sesquiterpene lactone mix, Laurel oil Carba mix, MDBGN, oak moss Balsam of Peru, Isoeugenol - never never never never never never never never never never never yes no yes current current never never never never never current never never never yes no yes no no current current never never current in the past in the past never never never never yes yes no no current current never never never never never in the past current never never - - - - Pure tea Pure oil for tree onychomy cosis - Babor cos metics* - - de Tuinen de Tuinen Tree Tea oil facial mask* Obagi cos metics* - Pure tea Pure oil as tree disinfec tants - - - - - Perior ibital Trunk Feet Hand Hand Facial Facial Hand Gener alized Hand Hand + + ++ ? + IR + + + + + + + + + + + + + + + +

chapter + - + - ? - ? ? - - - 9 ------F F F M F F M F F M M 69 55 66 22 34 65 49 68 58 28 20

207

52 59 49 66 51 24 40 24 39 57 50 24 65 F F F F F F F M F M M F F NT NT NT - - NT - NT - NT - NT - - - + ------+ + + ------? - + IR + + ? + + + + + + + + IR + Facial Facial Facial Facial Foot Facial Hand Hand Hand Trunk alized Gener Facial Facial - - - ing Data miss ------current current - never never current never past in the never never past in the past in the never no yes no no no no never never - past in the current never never current never past in the past in the never never no no no no no never never - past in the never never never never never never never never current no yes never never - never never never never never never never never never never - CAPB, formaldehyde, MCI/MI PTBFR, Caine mix, Ni, phenyl salicylate, Balsam of Peru, Benzyl benzoate, Ni thiuram mix disperse blue, FM2, Ni, PPD, Black rubber mix, caine MCI/MI CAPB oil Co, FMI, Ni, peppermint oil, clove - MCI/MI MDBGN, FMII ethylamine, PPD urea, Ni, oleamidodipropyldim Diazolidnyl urea, imidiazolidnyl amyl cynnamic aldehyde aldehyde, lillial, linalool, alpha alcohol, farnesol, hexyl cinnamic BHA, benzyl salicylate, cinnamic chapter 9 -

208 Table 3. Overview of patch test results from all patients reacting to tea tree oil and/or ascaridole. The strongest result on D3 or D7 is shown. Exposure was established by the patients themselves by means of a questionnaire and a checklist. Asc, ascaridole; BHA, butylatedhydroxyanisole; CAPB,cocamidopropylbetaine; Co, cobalt chloride; Cr, potassium dichromate; F, female;FM, fragrance mix; HMBA, 2-hydroxy 4-methoxy benzoic acid; IR, irritant reaction; M, male; MBT, mercaptobenzothiazole; MCI/MI, methylchloroisothiazolinone/methylisothiazolinone; MDBGN, methyldibromo glutaronitrile; MHC, methyl heptine carbonate; Ni, nickel sulfate; NT, not tested; PPD, p-phenylenediamine; PTBFR, p-tertbutylphenolformaldehyde resin; TTO, oxidized tea tree oil. *Contains tea tree oil according to the product information.

Aspect of irritant reactions The difference between an irritant and allergic reaction was scored according to a patch test calibration protocol.16 The irritant reactions caused by ascaridole 1% had a silk-like (or cigarette paper-like) appearance. For ascaridole 2%, the morphology of the irritant reactions was also silk-like inmost cases. One patient had a positive reaction to ascaridole 1%, but papulo-pustular irritant reactions to ascaridole 2% and ascaridole 5% (Fig. 2d). The irritant reactions caused by ascaridole 5% showed a wide range of different morphologies: silk-like skin, sharp demarcated redness, pustules, bullae, and erosions (Fig. 2). Some irritant reactions caused by ascaridole 5% closely resembled very strong positive reactions, especially with regard to the erosions and potential remnants of bullae.

Positive reactions in the total study population In our total study population (n = 319), we identified 37 positive reactions to either tea tree oil or ascaridole (Table 3). The mean age of patients with a positive reaction to ascaridole was 47 years, ranging from 20 to 69 years (Table 3). A minority of patients were male (n = 11, 29.7%), and 14 (37.8%) patients had atopic dermatitis diagnosed by a physician. Sixteen (43.2%) patients suffered from facial dermatitis, and three of these had mainly periorbital dermatitis. Fifteen (40.5%) patients suffered from hand dermatitis, and none of them had leg dermatitis. We were unable to show a relationship between sex, location of dermatitis and atopic eczema and a positive reaction to ascaridole or tea tree oil. Ascaridole 5% caused significantly more irritant reactions in patients who reported a history of atopic dermatitis (p = 0.016).

Clinical relevance and potential exposure chapter The 37 patients with a positive reaction to tea tree oil and/or ascaridole were provided with a questionnaire to analyse potential exposure to tea tree oil and clinical relevance. One 9 patient did not return the questionnaire. Five of 36 patients admitted preferring natural treatments, and two of these patients recalled the use of pure tea tree oil to treat either onychomycosis or wound infections. None of the patients recalled using aromatherapy. Fifteen patients, all women, regularly visited a beautician. Six of them had ever developed a rash afterwards, and for two of them we were able

209 to identify tea tree oil in the products used by the beauticians (PhD safewax® and Babor™ cosmetics). These reactions were judged to be clinically relevant. Sixteen of the 36 patients admitted preferring natural products, and in six patients a natural cosmetic product turned out to contain tea tree oil and caused subsequent dermatitis. In total, known exposure to a product containing ascaridole was shown in 7 of 36 patients (Table 3); two of these were exposed to multiple sources of tea tree oil. After exposure to tea tree oil, all of the patients developed a dermatitis pattern consistent with the use of the product, and all of these reactions were judged to be clinically relevant. Considering only the 290 consecutively tested outpatients, ascaridole 1% caused a clinically relevant positive reaction in 0.3% (1/290), ascaridole 2% in 1.7% (5/290), and ascaridole 5% in 1.4% (4/290). We were unable to show a clinically relevant reaction caused by oxidized tea tree oil 5% in 221 outpatients. A relevant reaction to oxidized tea tree oil was found in one of the 29 re-patch tested patients (Table 3).

a b c

d e f chapter

9 Fig. 2. A wide variety of irritant reactions caused by ascaridole 5%. (a, b, d–f) D3 readings. (c) D7 reading.

210 Discussion Ascaridole is considered to be a moderate skin sensitizer, as it has the ability to induce antigen-specific cell proliferation in a murine local lymph node assay.18 In vitro investigations looking for activation of human monocyte-derived dendritic cells andmonocytic cell lines (THP-1) as a model for dendritic cell activation showed augmentation of CD86 by ascaridole.19 The current study supports the idea that ascaridole is a sensitizer, although it seems to express more irritant features at higher patch test concentrations. In the past, ascaridole 10% has been patch tested, and caused very strong positive, bullous reactions.20 These might have been irritant reactions.

Prevalence In our unselected 221 outpatients, we found 2 (0.9%) positive reactions to oxidized tea tree oil 5% (Table 1). This is comparable with the number of sensitizations to oxidized tea tree oil of the North American Contact Dermatitis Group, which found a prevalence of 1.0% in 2009–2010.7 This frequency is also comparable to those in studies from Germany, Austria, Denmark, and Australia, where the prevalence varied between 0.5% and 1.8% of the patch tested population.8-10 In our population, the frequencies of sensitization to ascaridole were 1.4% for ascaridole 1%, 5.5% for ascaridole 2%, and 7.2% for ascaridole 5%. However, as ascaridole is not routinely tested, as far as we know, we are unable to compare our results with those obtained in other populations.

Patch test results Ascaridole 2% seems to be the most favourable patch test concentration, on the basis of the ratio of positive (5.5%), irritant (1.7%) and doubtful (4.1%) reactions. Ascaridole 5% caused more irritant reactions (9.0%) and doubtful reactions (10.3%) than positive reactions (7.2%). Despite the patch test calibration protocol, it was difficult to differentiate between the irritant and positive reactions for ascaridole 5%. Therefore, we do not recommend testing ascaridole at 5%. Ascaridole 1% caused more doubtful than positive reactions (1.4% versus 3.1%). Although larger cohorts of patients need to be investigated to calculate a reliable reaction index, on the basis of these preliminary results we recommend a patch test concentration of 2%. chapter In this study, we found a total of nine doubtful reactions to ascaridole 1%. In five of these, the reaction was positive upon patch testing with ascaridole 2%, whereas four doubtful reac- 9 tions could not be reproduced (were negative) upon concomitant testing with ascaridole 2%. For ascaridole 2%, in only one patient was the doubtful reaction positive upon patch testing with ascaridole 5%; the remaining 11 doubtful reactions to ascaridole 2% were negative, remained doubtful or became irritant reactions upon patch testing with ascaridole 5%.

211 A doubtful reaction becoming positive, becoming irritant or even disappearing upon patch testing with higher concentrations is also seen for other allergens, such as formaldehyde.21,22

Co-sensitization and clinical relevance All six patients with positive reactions to oxidized tea tree oil showed positive reactions to ascaridole (Tables 2 and 3). This finding suggests that ascaridole is a contact allergen in oxidized tea tree oil, although we should not neglect the potential presence of other sensitizers in oxidized tea tree oil, such as α-phellandrene, terpinolene, α-terpinene, 1,2,4-trihydroxy menthane, α-phellandren, and (+)-limonene.8,11-13 Twenty-four patients had a positive reaction to ascaridole without a positive reaction to oxidized tea tree oil. Recently, Sciarrone et al. investigated the components of tea tree oil, and found ascaridole at a concentration of only 0.2%.23 Rudbäck et al. investigated the auto-oxidation of α-terpinene in four bottles of tea tree oil, and in all of the bottles α-terpinene and oxidation products were found, but no ascaridole.18 The exact composition of the oxidized tea tree oil used in our study was unknown; it is possible that ascaridole was present at a concentration too low to elicit a reaction in all those who are sensitized to ascaridole, or, although unlikely, was not present at all. Among the six patients with positive reactions to oxidized tea tree oil and ascaridole, there was only one patient in whom clinical relevance could be shown. A possible explanation is that patients are unaware of being exposed to tea tree oil, as different chemical names are used on product labels. Although we supplied the patients with a list of synonyms,15 in the future we can improve the identification by instructing patients to bring all of their suspected products. It was remarkable that, in six patients with a positive reaction to ascaridole but without a positive reaction to tea tree oil, clinical relevance was shown, because there was exposure to products containing tea tree oil and the subsequent appearance of eczema after use of these products. This suggests that patch tests with ascaridole may provide clinically relevant positive reactions in patients that would be missed if they were only tested with oxidized tea tree oil. An explanation for positive reactions to ascaridole unrelated to tea tree oil might be the presence of allergenic degradation products of ascaridole or other substances in the as- chapter caridole preparation that are not present in the oxidized tea tree oil. We used 99.8% pure ascaridole for patch test preparations, although we did not verify whether this concentration 9 changed over time. Perhaps we have patch tested with auto-oxidation products, although these should, in theory, be present in the oxidized tea tree oil as well. In addition, the discrepancy between sensitization to ascaridole and the lack of exposure to tea tree oil could be explained by other, independent sources of ascaridole. The best known source of ascaridole is Chenopodium ambrosioides24. The essential oil of C. ambrosioides contains 40–70% ascaridole, and is used is an antihelminthic. Because of its toxicity, this oil is no longer used

212 in humans.25 The leaves of the plant are sometimes used, for example in South American cuisine or in order to produce medicinal tea. Another potential source, Boldo oil (Peumus boldus oil)26, is used as a herbal remedy for various conditions. However, Boldo oil is banned from use in cosmetics as well27, and its use as a herbal remedy is discouraged, in view of the potential risks associated with the toxicity of ascaridole.28 Therefore, these other sources of ascaridole are unlikely to be sensitizers in this population.

We collected a cohort of subjects with reactions to oxidized tea tree oil and/or ascaridole. For future studies, we suggest conducting thin layer chromatography on oxidized tea tree oil and performing patch tests with the chromatograms in this selected population, to gain further insights into other potential allergens in oxidized tea tree oil and their relationship with ascaridole. In conclusion, the current study supports the idea that ascaridole might be a sensitizer. We recommend patch testing with ascaridole at a concentration of 2%. The finding that every positive reaction to oxidized tea tree oil is accompanied by a positive reaction to ascaridole suggests that ascaridole might be a contact allergen in oxidized tea tree oil.

chapter 9

213 References 1. Apted JH. Contact dermatitis associated with the use of tea-tree oil. Australas J Dermatol 1991:32:177. 2. de Groot AC, Weyland JW. Systemic contact dermatitis from tea tree oil. Contact Dermatitis 1992:27:279-280. 3. van der Valk PG, de Groot AC, Bruynzeel DP, Coenraads PJ, Weijland JW. Allergic contact eczema due to ‘tea tree’ oil. Ned Tijdschr Geneeskd 1994:138:823-825. 4. Knight TE, Hausen BM. Melaleuca oil (tea tree oil) dermatitis. J Am Acad Dermatol 1994:30:423-427. 5. Coutts I, Shaw S, Orton D. Patch testing with pure tea tree oil – 12 months experience. Br J Dermatol 2002:147:62-74. 6. Zug KA, Warshaw EM, Fowler JF,Jr, Maibach HI, Belsito DL, Pratt MD, et al. Patch-test results of the North American Contact Dermatitis Group 2005-2006. Dermatitis 2009:20:149-160. 7. Warshaw EM, Belsito DV, Taylor JS, Sasseville D, DeKoven JG, Zirwas MJ, et al. North American Contact Dermatitis Group patch test results: 2009 to 2010. Dermatitis 2013:24:50-59. 8. Pirker C, Hausen BM, Uter W, Hillen U, Brasch J, Bayerl C, et al. Sensitization to tea tree oil in Germany and Austria. A multicenter study of the German Contact Dermatitis Group. J Dtsch Dermatol Ges 2003:1:629-634. 9. Veien NK, Rosner K, Skovgaard GL. Is tea tree oil an important contact allergen? Contact Dermatitis 2004:50:378-379. 10. Rutherford T, Nixon R, Tam M, Tate B. Allergy to tea tree oil: retrospective review of 41 cases with positive patch tests over 4.5 years. Australas J Dermatol 2007:48:83-87. 11. Hausen BM, Reichling J, Harkenthal M. Degradation products of monoterpenes are the sensitizing agents in tea tree oil. Am J Contact Dermat 1999:10:68-77. 12. Hausen BM. Evaluation of the main contact allergens in oxidized tea tree oil. Dermatitis 2004:15:213-214. 13. Rubel DM, Freeman S, Southwell IA. Tea tree oil allergy: what is the offending agent? Report of three cases of tea tree oil allergy and review of the literature. Australas J Dermatol 1998:39:244-247. 14. Bakker CV, Blömeke B, Coenraads PJ, Schuttelaar ML. Ascaridole, a sensitizing component of tea chapter tree oil, patch tested at 1% and 5% in two series of patients. Contact Dermatitis 2011:65:240-241. 15. Christoffers WA, Blömeke B, Coenraads PJ, Schuttelaar MLA. Co-sensitization to ascaridole and 9 tea tree oil. Contact Dermatitis 2013:69:187-189. 16. Svedman C, Isaksson M, Björk J, Mowitz M, Bruze M. ‘Calibration’ of our patch test reading technique is necessary. Contact Dermatitis 2012:66:180-187. 17. Bruze M, Svedman C, Andersen KE, Bruynzeel D, Goossens A, Johansen JD, et al. Patch test concentrations (doses in mg/cm2 ) for the 12 non-mix fragrance substances regulated by European legislation. Contact Dermatitis 2012:66:131-136.

214 18. Rudbäck J, Bergström MA, Börje A, Nilsson U, Karlberg AT. alpha-Terpinene, an antioxidant in tea tree oil, autoxidizes rapidly to skin allergens on air exposure. Chem Res Toxicol 2012:25:713-721. 19. Tietze C, Blomeke B. Sensitization assays: monocyte-derived dendritic cells versus a monocytic cell line (THP-1). J Toxicol Environ Health A 2008:71:965-968. 20. Hausen BM. Kontaktallergie auf Teebaumöl und Ascaridol. Akt Dermatol 1998:24:60-62. 21. Trattner A, Johansen JD, Menné T. Formaldehyde concentration in diagnostic patch testing: c omparison of 1% with 2%. Contact Dermatitis 1998:38:9-13. 22. Pontén A, Aalto-Korte K, Agner T, Andersen KE, Giménez-Arnau AM, Gonçalo M, et al. Patch testing with 2.0% (0.60 mg/cm 2) formaldehyde instead of 1.0% (0.30 mg/cm 2) detects significantly more contact allergy. Contact Dermatitis 2013:68:50-53. 23. Sciarrone D, Ragonese C, Carnovale C, Piperno A, Dugo P, Dugo G, et al. Evaluation of tea tree oil quality and ascaridole: a deep study by means of chiral and multi heart-cuts multidimensional gas chromatography system coupled to mass spectrometry detection. J Chromatogr A 2010:1217:6422-6427. 24. Dembitsky V, Shkrob I, Hanus LO. Ascaridole and related peroxides from the genus Chenopodium. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub 2008:152:209-215. 25. International Fragrance Association. IRFA Standard Chenopodium Oil. October 14, 2009. 26. Johnson MA, Croteau R. Biosynthesis of ascaridole: iodide peroxidase-catalyzed synthesis of a monoterpene endoperoxide in soluble extracts of Chenopodium ambrosioides fruit. Arch Biochem Biophys 1984:235:254-266. 27. International Fragrance Association. IRFA Standard Boldo Oil. October 14, 2009. 28. European Medicines Agency: Committee on Herbal Medicinal Products (HMPC). Assessment Report on Peumus Boldus Molina, Folium EMEA/HMPC/591648/2007. 2009.

chapter 9

215 Chapter10

216 Discussion How to treat hand eczema in daily practice?

Wietske Andrea Christoffers

Department of Dermatology, University of Groningen, University Medical Center Groningen, Groningen, Netherlands

chapter 10

217 Overview of main findings & general discussion This thesis discusses the various treatment options for patients with hand eczema. In chapter 2 we reviewed extensively, according to the Cochrane protocol, all the randomized controlled trials (RCT) on interventions for hand eczema. The slightly disappointing conclusion was that, with a few exceptions, the overall quality of the studies is poor and head-to-head trials are lacking. A systematic review published in 2004 came to the same conclusion.1 In spite of well-designed studies regarding alitretinoin and topical calcineurin inhibitors, the evidence for the efficacy of topical corticosteroids or cyclosporine and acitretin, often used in daily practice, has remained scarce. Clinicians often wonder what treatment would be best for their specific patient with hand eczema. The choice for an optimal treatment should, ideally, be based on a comparison between different treatment modalities, such as topical corticosteroids versus topical calcineurin inhibitors. In chapter 3 we discussed questions from daily practice. Here the overall conclusion was in line with that of chapter 2: in general, there is insufficient high-quality evidence to recommend one treatment over another. For example, PUVA and UVB are both effective in the treatment of hand eczema, but there is no evidence of a clinical advantage of one over the other. The same is true of topical calcineurin inhibitors; although they are more effective than placebo, their superiority over topical corticosteroids in cases of hand eczema has not been demonstrated. Oral retinoids (alitretinoin), however, do appear to be effective and well tolerated in hand eczema, especially in hyperkeratotic hand eczema. In chapter 2 we included a potentially important study by Schmitt et al. which compared the effectiveness of cyclosporine to alitretinoin. Unfortunately, this investigator- initiated study was ended prematurely due both to an insufficient number of naïve participants and to limited resources. These problems are characteristic for intervention studies with patents: once a drug has been established in daily practice or has received market authorization, there is no financial motivation for pharmaceutical companies to conduct a comparative study. This could explain the scarcity of head-to-head trials.

Besides randomized controlled trials, daily life studies like chapter 4 provide useful information. One advantage is the applicability of their results to daily clinical practice; therefore the external validity is higher in daily life studies than in clinical trials, though the internal validity is lower. Limitations in external validity make it difficult to apply results of clinical trials in daily practice. Trials recruit only ‘perfect patients’ who fulfill strict in- and exclusion criteria, are relatively healthy, motivated and adherent to treatment. In daily practice patients chapter are unselected: only patients with true contraindications are excluded from treatment.2 In general the number of drop-outs due to adverse events is higher in daily practice than in 10 clinical trials. Differences in duration of active treatment may also contribute to the higher

218 discontinuation rates. Moreover patients who choose to participate in clinical trials are often very motivated to fulfill the trial period despite adverse events.3 Finally, artificial endpoints and a fixed duration of treatment might influence the external validity of RCTs. The end points might not be relevant for a patient at all: for example, a reduction in erythema after 14 days of using tacrolimus might not be very relevant to a patient with a chronic condition like hand eczema.4 For this reason, in chapter 4 we conducted a daily practice study regarding the drug survival of cyclosporine. This study concluded that cyclosporine is a safe treatment for patients with recurrent vesicular hand eczema and patients used it for an average of 10.3 months, which is substantially longer than the average trial.

One of the first steps after diagnosing hand eczema is to provide education regarding the use of emollients, gloves and the avoidance of allergens and irritants. Contact allergens such as thiuram, mercaptams and acrylates are well known for their role in hand eczema. How- ever, the role of contact allergies is an issue of debate in patients with chronic hand eczema. In chapter 5 we studied the most frequent sensitizers in subjects with hand eczema. We demonstrated that 50.3% of patients with clinical subtype chronic fissured hand eczema were sensitized to at least one allergen in an extended European baseline series. This number was substantially higher than the sensitization rate in the general population, where the median prevalence of contact allergy to at least one allergen is 21.2% (range 12.5–40.6%).5 In subjects with recurrent vesicular hand eczema this number was even higher: 55.2% of the patients was sensitized to at least one allergen in the extended European Baseline series and 17.0% of this clinical subtype was polysensitized. In patients with pulpitis, acrylates were frequent sensitizers. Acrylates are in general known to be potent sensitizers. In chapters 6 and 7 we describe two cases of hand eczema due to acrylate contact allergy. In chapter 6 a process operator developed bullae on the hands and knee due to contact with various acrylates, of which 1,6-hexanediol diacrylate and glycidyl methacrylate were, among others, the main sensitizers. Subsequently, in chapter 7 we went on to describe therapy-resistant hand eczema in a printer who worked with acrylates. A rare contact allergy to isobornyl acrylate turned out to be the explanation for his persistent hand eczema. Triggered by these cases we further studied our patients who were sensitized to (meth)acrylates; 87.5% turned out to have a contact allergy of the hands. These findings highlight the importance of (meth)acrylates in (occupational-related) hand eczema.

When conventional therapies do not give desired results, patients tend to search for alterna- chapter tives. One marketed alternative for the treatment of eczema is tea tree oil (melaleuca alternifolia), a natural oil derived from the Australian Melaleuca Alternifolia. This oil is claimed 10 to have antibacterial, anti-inflammatory properties and to reduce histamine-induced skin

219 inflammation.6-8 Because it is a natural product consumers assume it to be safe. However, when tree tea oil oxidizes, several sensitizing degradation products are formed. Chapters 8 and 9 discussed one of these products, ascaridole; more than one third of our patients who were sensitized to ascaridole suffered from hand eczema.

How to treat hand eczema in daily practice? In spite of our focus on different interventions for hand eczema, we have not been able to give a one-off recommendation for the treatment of patients with this disease. Although, a Dutch guideline for the treatment of hand eczema does not exist, various well-respected international groups have proposed guidelines, including Canadian9, Danish10 and German11 guidelines and the British consensus.12 These all acknowledge the importance of (primary, secondary and tertiary) prevention, avoidance strategies and education and propose that treatment be tailor made to the needs of the specific patient. All guidelines go on to recommend moderate to super potent topical corticosteroids. If the hand eczema is unresponsive to topical corticosteroids, different guidelines give different recommendations, but the potential options are often listed in alphabetic order without clear stratification. In this final chapter of the thesis we aim to provide a clear, stratified, practical tool to treat hand eczema based on current literature, existing guidelines and the findings of this thesis.

In chapter 1 we discussed the steps for diagnosing hand eczema and the differential diagnosis. We later in chapter 5 presented a flow chart to classify the clinical subtypes of hand eczema according the criteria of Danish Contact Dermatitis Group. Once the diagnosis of hand eczema and the clinical subtype have been established, the treatment of hand eczema consists of several important steps as displayed in Fig. 1.

Immediate treatment is recommended and watchful waiting should be discouraged, since various studies have demonstrated that the prognosis of chronic hand eczema is poor. Hald et al. found in a 6-month follow up study that fissures and scaling, signs of chronic hand eczema, had a poorer prognosis than, for example, vesicles.13 Meding demonstrated that the extent of the hand eczema also influences the prognosis negatively.14 Since patients experience a median patient delay of three months, and 11.2% of the patients wait longer than one year before consulting a physician15, physicians should be prompt with proactive treatment.

chapter Basic skin care & protection The first step in the treatment of hand eczema is education and basic skin care advice. One 10 must emphasize that no quick and easy solution for hand eczema exists. As proper skin care

220 will not cure the eczema completely patients may fail to see its value, but skin care is vital and patients should continue with it for the rest of their lives. Patients need to know that frequent contact with irritants such as water, detergents and cutting fluids can impair the skin barrier function. This impaired barrier function can evoke or maintain hand eczema. Insight into potential triggers and the mechanism behind the disease makes the patient self-reliant and increases self-management.16 Whenever possible, causative factors should be removed. Chapter 5 demonstrated that about half of hand eczema patients are sensitized to at least one allergen. The clinician and patient should try to establish whether this sensitization is relevant for the hand eczema, though this could be a complex quest. Contact allergens that are clinically relevant for the hand eczema can be present in cosmetic, household and occupational products; these should be checked, and where necessary, avoided. Websites such as www.huidarts.com or www.huidziekten.nl contain extensive lists of allergens with clear information as to potential sources of allergens. The physician ought to provide the patient with specific written information, since it is difficult to remember complex information like names of contact allergens. Unfortunately, the complete elimination of a contact allergen does not guarantee complete healing of the hand eczema, since hand eczema is typically a multifactorial disease.

A silent causative factor in hand eczema is frequent washing of the hands (>20 times a day). If the hands are not visibly dirty but must be cleaned to prevent infection, as in the case of health care workers, the use of alcohol-based disinfectants instead of traditional full hand washing should be encouraged.17,18 If hands are visibly dirty, traditional hand washing is necessary.19 Therefore lukewarm water and preferably a fragrance-free soap without well-known sensitizers such as MCI/MI should be used. The hands should be dried well, preferably with disposable paper towels, which have higher drying efficiency and are more hygienic than (hot) air dryers or common-use towels.20 Subsequently, an emollient should be applied immediately. One should avoid rings, especially when washing the hands, as dirt and soap can get trapped under a ring, causing irritation; the skin underneath a ring is also difficult to dry.21

Wet work in general is an important stressor for the skin and plays a prominent role in the majority of irritant hand eczema cases. Whenever possible, wet work should be avoided, for example by using a washing machine and a dishwasher. Protective (household) gloves can also be worn, but gloves be worn in the correct way.

Gloves are essential in the avoidance of contact with allergens and irritants. Different gloves chapter are suitable for different purposes. It is important to select a specific glove for a specific job, since gloves might provide good protection against one substance but be easily penetrated 10 by another.22 Moreover, gloves may themselves contain sensitizers which should be avoided.

221 • Nitrile gloves provide good protection against solvents, oils, greases, selected acids and bases. These gloves can contain common sensitizing rubber accelerators such as carbamates, thiurams, and mercaptams. • Vinyl gloves protect against acids, bases, oils, greases, peroxides, and amines. These gloves might contain the rare sensitizers phthalates and bisphenol A and are not very elastic. • Latex gloves provide proper protection against biological and water based materials. These gloves can contain common sensitizing rubber accelerators such as carbamates, thiurams, and mercaptams. Moreover, latex gloves can cause a protein contact dermatitis or an immediate type I allergic reaction. • Polychloroprene gloves protect against acids, bases, alcohols, fuels, peroxides, hy- drocarbons, oils, greases, and phenols. Polychloroprene gloves might contain the rare sensitizer thioureas. • Polyvinyl alcohol gloves protect against aromatic and chlorinated solvents, ketones, esters and methacrylate. These gloves are quite expensive. • Butyl gloves protect against ketones, aldehydes and esters. These gloves are also expensive. • Multilayer laminated gloves (laminated gloves of ethylene- vinyl-alcohol-polyethylene, 4H®) protect against practically everything, but their fit is poor.22 As discussed in chapters 5, 6 and 7, acrylates are well known sensitizers. Multilayer laminated gloves provide the best protection against acrylates;23 however these gloves have a poor fit and are not practical for fine manual work like that of nail stylists and dental technicians. In such cases thicker nitrile gloves are more effective than latex or vinyl gloves because of their longer breakthrough time for various acrylates. Their elasticity is also superior to that of multilayer laminated gloves.23-26 Not only must patients select the proper gloves, but they must wear gloves only for a short period of time and not re-use them. The barrier function of a glove is impaired after the first contact with an allergen, whether or not the glove has been worn. One should immediately discard gloves with holes and regularly replace all gloves. Moreover, when putting gloves on or off, a subject can come into contact with the, often contaminated, outer lining of the glove and transfer the allergen from the outside to the inside of the glove. Selection of proper gloves remains challenging, since besides exposure to allergens and irritants, one must also consider the risk of latex allergy and sensitization to rubber chemicals. Additional

chapter information regarding gloves can be found on http://www.ansell.nl/, http://www.kcl.de/ and http://www.marigoldindustrial.com/. We also recommend contacting the occupational physi- 10 cian. Another point to consider is that although gloves protect the skin from contact with allergens and irritants (especially wet work), the occlusion of the skin caused by the glove

222 itself is a risk factor for hand eczema.22,27 Therefore, cotton linings or inner gloves underneath occlusive gloves are recommended when gloves are worn for longer than 10 minutes.28 In conclusion, gloves must be worn as long as necessary, but as short as possible.

Patients should be instructed to use emollients frequently. Lipid-rich emollients (ointments) are preferred over creams: because these are more lipid and contain less water, thus there is less evaporation. Moreover, creams more often contain sensitizing preservatives and mildly irritating emulsifiers. The Cochrane review chapter( 2) demonstrated that a petrolatum- based emollient (Vaseline lanette) did not differ from a ceramide-containing emollient (Locobase® Repair). However, one must consider not only the efficacy of an emollient but, even more important, the patients’ preference. Patient-rated factors such as cosmetic acceptability, odor, greasiness and amount of staining should therefore be taken into account. Providing various alter- natives during the first visit allows the patient to select the emollient of his personal preference, which could contribute to his adherence. Patients must be instructed to use emollients several times a day, especially after contact with water. Multiple samples should be available in several places in the house, as next to every sink, or in a handbag. The patient must understand the use of emollients and that emollients must be applied to the entire hands, including web spaces, finger tips and the back of the hands.21

Topical treatment Topical corticosteroids are the mainstream treatment for hand eczema. The rationale for topical corticosteroids is based largely on pharmacological reasoning and clinical experience; actual evidence is limited. Nine randomized controlled trials with topical corticosteroids were carried out with hand eczema patients. These studies had different designs; different corticosteroids and different vehicles, dosages or application frequencies were used. It is thus difficult to compare and recommend an evidence-based dosing regime. To prevent chronicity of hand eczema, we recommend as first treatment step a potent topical corticosteroid, applied once or twice daily. To treat one entire hand, one fingertip unit (0.5 gram) of topical corticosteroids should be applied.29 Patients should be instructed not to apply corticosteroids at the same time as their emollients, since this might diminish the total dosage.

If after two to four weeks a potent topical corticosteroid seems to be insufficient, one should evaluate the patient’s adherence to treatment. This should be discussed openly, since treat- chapter ment adherence to topical therapy is even poorer than to oral therapy.30,31 In a small study, 10 patients with moderate to severe atopic dermatitis were instructed to apply fluocinonide 10 0.1% cream twice daily for 5 days.32 The median treatment adherence was only 40% (range

223 0-100). For treatments with longer durations, the adherence might diminish even further. To improve treatment adherence, the dosage schedule should be easy to comprehend and follow. Therefore a once daily application is preferable to a twice daily application, and complex tapering schedules should be avoided. Although little studied, the correct way of tapering is an issue of debate. Given the reservoir function of the epidermis, application every other day might be preferred over application on three or four successive days. During the tapering phase, the patient should be reminded to continue with frequent use of emollients. Lack of information and (irrational) fear of adverse events can diminish treatment adherence. Patients may suffer from corticophobia due to fear of skin atrophy and systemic effects.33 However, a systematic review of atopic dermatitis reported that randomized controlled trials had shown no evidence for the development of skin atrophy provided that the corticosteroids were used appropriately.34 Full body application of very potent corticosteroids such as clobetasol propionate does result in inhibited cortisol production in patients with severe atopic dermatitis or bullous pemphigoid; however these effects disappeared with dosages of less than 50 mg a week and with less potent corticosteroids.35,36 Thus the risk of systemic effects in the topical treatment of hand eczema was found to be negligible. This, combined with the poor prognosis of chronic hand eczema, should encourage physicians to prescribe potent topical corticosteroids in an early phase. If topical corticosteroids are necessary over a longer time period, for example during the maintenance phase, one can consider rotation with less potent topical corticosteroids or topical calcineurin inhibitors (tacrolimus and pimecrolimus). Other potential explanations for ineffectiveness of topical treatment are contact allergies to lanolin or corticosteroids. This can be the case especially in hand eczema that worsens despites adequate topical treatment, or if the eczema is more active in the borders than in the center of the lesion (after ruling out dermatomycosis). In case of lanolin-sensitizations, one must prescribe lanolin-free emollients (unguentum leniens (koelzalf FNA), paraffin, Vase- line and Vaseline album) and lanolin free corticosteroids (Emovate, Elocon, and Dermovate among others.37-39 Diagnosing a contact allergy to corticosteroids can be difficult, since the patch test with corticosteroids suppresses the delayed type reaction by the anti-inflammatory action of the corticosteroids themselves. When patch testing, one should at least test the screening markers tixocortol pivalate, budesonide and hydrocortisone-17-butyrate, along with the patient’s own corticosteroids.40 A day-7 reading can be introduced to pre-empt late reactions to corticosteroids, but even this cannot exclude false-negative reactions. Corti- costeroids are known to cross-react, based on C16-methylated and non-methylated mol-

chapter ecules.41,42 Recommending a safe topical corticosteroid can be difficult.42,43 However, topical calcineurin inhibitors are a valuable treatment alternative for every patient with a sensitiza- 10 tion to corticosteroids.

224 The Cochrane review demonstrated that topical calcineurin inhibitors are comparable to or more effective than placebos, but they are not more effective than mometasone furoate.44,45 Moreover, topical calcineurin inhibitors are almost seven times as expensive as topical corticosteroids (on average €12 versus €80 per 100 gram). Finally, long term adverse events are not yet known, as long-term studies in atopic dermatitis are still in progress. A further suggestion is to add salicylic acid to the treatment for patients with hyperkeratotic hand eczema and water baths for patients with recurrent vesicular hand eczema, though these recommendations are based purely on experience and not on evidence.

If topical treatment is successful after initial crisis intervention, one should switch to the maintenance phase. Frequent application of topical corticosteroids should be tapered in an easy to follow schedule and the patient should be encouraged to continue the application of emollients. If tapering of corticosteroids results in a flare, one can prescribe short bursts of potent topical corticosteroids or longer term continuous application of less-potent corticosteroids. One can also consider using rotation with topical calcineurin inhibitors instead of maintenance therapy with topical corticosteroids.

Phototherapy When potent corticosteroids have failed to help patients with vesicular hand eczema one should, before starting systemic treatment, consider phototherapy. Clinical experience has also shown phototherapy to be effective in treating chronic fissured hand eczema. The effectiveness of topical PUVA in treating hyperkeratotic palmar eczema might be less, however, probably because of the thick hyperkeratotic plaques and the absence of inflammation. In Chapter 2 we demonstrated that phototherapy was more effective than placebo, but there was little evidence to indicate which method, either UV-B or PUVA was superior. If frequent visits to the hospital are feasible and the travel distance to the hospital is acceptable, one can prescribe phototherapy, although several treatment centers may decide to skip this treatment option. First of all, phototherapy needs to be tailored to specific patients and clinics. Moreover, although it is effective, phototherapy has drawbacks. Studies of psoriasis patients concluded that patients found phototherapy demanding and time consuming and some even doubted the therapeutic value of the time they had invested.46 Moreover, frequent hospital visits resulted in considerable direct and indirect costs and absence from work.47 Phototherapy at home could resolve some of these issues: in patients chapter with chronic hand eczema oral PUVA with a portable tanning unit at home is as effective as hospital-administered bath PUVA and more cost-effective.48 In the Netherlands, the main 10 disadvantage of phototherapy at home is the uncertainty of reimbursement.

225 Fig. 1: Flowchart for the treatment of a patient with moderate to severe hand eczema. The flow chart is based on the current level of evidence and clinical experience. The timeline in the upper part of the flow chart is just an indication of recommended treatment duration.

Systemic interventions for hand eczema Before starting systemic therapy for hand eczema, one must re-evaluate the patient’s treatment adherence and reconsider the diagnosis. The decision to start systemic treatment should be made by both physician and patient, taking into consideration possible adverse events, benefits and costs for the specific patient. chapter Patients should be given verbal and written information regarding the systemic options and 10 offered sufficient time to consider these options.

226 In the Netherlands, alitretinoin is the only registered systemic treatment option for chronic hand eczema. In well-designed pharmaceutical sponsored trials49-51 30 mg alitretinoin a day resulted in clear or almost clear responses in 48% of the participants, compared to 17% with a placebo.50 In the hyperkeratotic subtype 49% reacted, as compared to 12% in the placebo group. However, only one third (33%) of the participants with recurrent vesicular hand eczema (defined in the study as pompholyx) reached clearance or almost clearance, as compared to 16% in the placebo group. Based on these statistics and the mechanism of alitretinoin, and the fact that alitretinoin is the only registered treatment option for hand eczema, we recommend alitretinoin as the chapter first systemic intervention step for all types of chronic hand eczema, except for recurrent vesicular hand eczema (Fig. 1). A dosage of 30 mg/day should be prescribed. If there is no 10 considerable improvement after 12 weeks, alitretinoin should be stopped. Otherwise it can be continued until 24 weeks. After 24 weeks of treatment, according to the manufacturer’s guidelines the patient should stop with alitretinoin. In case of an exacerbation, alitretinoin can be restarted. If alitretinoin is insufficient, acitretin 25 mg/day is a valuable alternative for hyperkeratotic hand eczema. For other clinical subtypes of hand eczema cyclosporine can be considered.

For patients with recurrent vesicular hand eczema a low dose of oral cyclosporine might be preferred over alitretinoin. As demonstrated in chapter 4, the drug survival of cyclosporine was especially long for patients with recurrent vesicular hand eczema, and more than half of the patients achieved an improvement of at least 50%. Meta-analysis in patients with atopic dermatitis demonstrated an efficacy of cyclosporine of 55% after 6 to 8 weeks.52 In patients with atopic dermatitis, starting with a high dose of cyclosporine (>3.5 mg/kg/day) and subsequent tapering of the dose (drop-down) seemed more effective and resulted in a longer drug survival than starting with a low dose of <3.5 mg/kg/day and slowly increasing the dose to reach the optimum treatment effect (step-up).53 Moreover, in chapter 4 we demonstrated that cyclosporine can be used safely for several months to years and only one patient discontinued treatment because of an increase in serum creatinine levels. Unpublished data demonstrate the same; atopic dermatitis patients whose median treatment duration was more than one year experienced no substantial influence on renal function.53 After 6 weeks one should evaluate the effect of cyclosporine; if this is satisfactory, the drug can be continued for 3-6 months as needed. Even longer treatment did not seem to result in more adverse events and can therefore be considered for refractory cases. During treatment with cyclosporine, one must carefully monitor blood pressure and serum creatinine, among others. If cyclosporine is not effective, one can then consider alitretinoin for patients with recurrent vesicular hand eczema. If alitretinoin, acitretin and cyclosporine are ineffective, various other off-label treatment options are available. Methotrexate (10-15 mg/week) with folic acid may be beneficial in hyperkeratotic hand eczema54, while azathioprine (50-150 mg/day) can be effective in recurrent vesicular hand eczema and chronic hand eczema. 55,56 For severe refractory chronic or vesicular hand eczema, unresponsive to conventional treatment, mycophenolate mofetil (0.5-3 gram/day),57 enteric-coated mycophenolate sodium (720 mg twice a day),58 or a low-dose of prednisolon (5-10 mg/day) with osteoporosis pro- phylaxes can be prescribed, though there is no evidence for the effects of these treatment

chapter options in patients with hand eczema. Furthermore, based on experience, a course of the oral antifungal agent itraconazol occasionally seems effective for recurrent vesicular hand 10 eczema. If a subject is prone to superimposed infections of the skin this may aggravate the hand eczema; long-term treatment with oral claritromycin (500 mg/day) can then be

228 considered. For hyperkeratotic palmar eczema dithranol and coal tar can be prescribed. Combinations of different oral treatments can be considered in extreme refractory cases, though this should be done with extreme caution and careful monitoring. In the future, biologicals may provide alternative options, though to date no effective biological for (hand) eczema is known.

The ESCD Guideline: Guidelines for diagnosis, prevention and treatment of hand eczema An expert working group of the European Society of Contact Dermatitis (ESCD) is currently working on a guideline for the diagnosis and treatment of hand eczema59, partly based on the findings of the Cochrane review fromchapter 2. Overall, the guideline of the ESCD does not differ essentially from our flow chart. The skin protection program of the ESCD with regards to gloves, hand washing and emollients overlaps with the advices described in the text above.

The ESCD strongly recommends quick and vigorous treatment to avoid the development of chronic hand eczema. The use of topical corticosteroids is the first line treatment in the management of hand eczema (evidence of high quality, strong consensus-based recommendation), but states a maximum duration of six weeks. For patients with chronic hand eczema refractory to topical corticosteroids, the ESCD strongly recommends alitretinoin (high quality evidence, consensus-based recommendation) or photo-therapy (UVB, PUVA) (evidence of moderate quality, strong consensus-based recommendation). We have positioned phototherapy at the same level as the ESCD. The ESCD emphasizes that other treatment options are off-label. The off-label use of cyclosporine is recommended as alternative option by the ESCD (evidence of moderate quality, strong consensus-based recommendation). Other options such as azathioprine, acitretin or methotrexate can be considered once other therapeutic options have failed. A substantial difference between the ESCD guideline and our treatment plan is the inclusion of specific subtypes of hand eczema in our treatment plan. The ESCD emphasizes that in research and clinical trials some kind of classification needs to be applied, though current consensus is lacking. Thereupon the ESCD proposes to use the classification system of Diepgen et al.60 This classification, as discussed in chapter 5, includes a mixture of etiological and clinical types. In the guideline of the ESCD the classification of subtypes is not further incorporated to recommend specific treatment options to specific subgroups of hand eczema. Our flow chapter chart distinguishes clinical subtypes of hand eczema according to classification of the Danish contact Dermatitis Group.10 Because of this classification, we slightly differ from the 10 guidelines of the ESCD. Alitretinoin is a highly appreciated treatment option for the most

229 subtypes of hand eczema. However, in our opinion, alitretinoin should not be preferred for recurrent vesicular hand eczema, but we recommend the use of cyclosporine. This can be considered as a controversial statement given the level of evidence for alitretinoin.49-51,61 In addition, cyclosporine is not registered for the treatment of hand eczema, only for the treatment of atopic dermatitis. However, based on clinical experience, the findings in chapter 4 and the limited efficacy of alitretinoin in clinical trials in this subtype, we consider this justified. Though, further studies are needed.

Future perspectives This thesis demonstrates various hiatuses in our knowledge regarding hand eczema. First, there is no consensus regarding the definition of hand eczema and the classification of its subtypes. The textbook definition of hand eczema is “noninfectious skin inflammation of the hands”. For chronic hand eczema Diepgen’s definition is widely accepted: “hand eczema which exists for more than three months or which returns twice or more within 12 months despite adequate dermatological therapy and patient compliance.”11 In chapter 2 we encountered various time slots for the definition of ‘chronic’, such as six weeks or six months. International consensus regarding this last definition is important in order to improve unifor- mity in clinical trials. However, the differences in the definition of chronic hand eczema seem relatively insignificant compared to the lack of consensus regarding classification of subtypes of hand eczema. Several contradicting articles have been published. 60,62,63 Some groups use etiological subtypes, others clinical, and still others a combination of etiological and clinical subtypes. We would like to recommend the clinical classification as described by the Danish Contact Dermatitis Group,10,62 though slightly adjusted. We would include a more detailed description of the clinical subtype ‘chronic fissured dry hand eczema’ since, as reported in chapter 5, a wide range of different morphological types can be included under this subtype. The tool in chapter 5 includes the localization on the hand and the morphology, which results in a clinical diagnosis. This tool should be expanded with the factor ‘time’ to distinguish chronic from acute hand eczema and to further determine the nature of recurrent vesicular hand eczema. This could be an important step toward consensus regarding these subtypes. In the Cochran review we encountered first of all a wide range of systems for scoring severity of the disease, and newer and improved scoring systems are still being developed. Weistenhöfer et al. demonstrated a maze of different, often unvalidated, scoring systems: 45 different methods for quantifying hand eczema were identified in 69 articles.64 The different chapter scoring systems included either the extent of the hand eczema, the morphological characteristics, subjective criteria, or a combination of these. Moreover, for the majority of the 10 scoring systems, inter and intra observer reliability were not studied. To reach a consensus regarding the scoring of severity we would advocate the approach

230 of OMERACT (Outcome Measures in Rheumatology) 65 and HOME (Harmonizing Outcome Measures for Eczema).66 The HOME initiative aims to facilitate an international, multidisci- plinary consensus on core outcome measures to be included in all eczema trials and clinical record keeping: it also aims to identify other relevant issues for eczema outcome research. The use of standardized endpoints in randomized controlled trials and in observational studies is vital in that it facilitates comparisons of outcomes across studies in different study populations, which as we learned in chapter 2 is frankly impossible in current studies. Moreover, a minimal set of outcome measures would improve the quality of studies overall; some studies included in chapter 2 used outcomes which were barely relevant in daily practice. Consensus regarding the definition of hand eczema and its subtypes, and the development or recommendation of a single severity scoring system are essential for evidence based medicine.

A second important observation of the Cochrane review was the lack of well-designed randomized controlled trials comparing major treatment groups. We therefore present an urgent appeal for such studies. At the time of writing, alitretinoin is, in the Netherlands, the only registered treatment for chronic hand eczema. Other treatment options such as cyclosporine or acitretin are used off-label. In clinical trials alitretinoin was effective in the majority of hand eczema patients, though it is effective only in 33% of patients with vesicular hand eczema. Based on the findings of chapter 4, cyclosporine might be more effective in patients with recurrent vesicular hand eczema. Therefore we would like to propose a clinical trial which compares the efficacy of alitretinoin and cyclosporine. Schmitt designed such a trial and registered it at clincialtrials.gov. This trial aimed to compare the efficacy and safety of cyclosporine and alitretinoin in the treatment of severe atopic hand dermatitis. Although unfortunately this study was ended prematurely because of the inability to include a sufficient number of patients, the need for such research is obvious.

The perfect study would compare the efficacy of cyclosporine and alitretinoin in different clinical types of chronic hand eczema, starting with recurrent vesicular hand eczema, since a relatively large number of patients suffer from recurrent vesicular hand eczema and this group seems to benefit the least from alitretinoin. The study should include proper outcome measures. The primary outcome would preferably be an objective observer-rated severity score, such as the HECSI. Patient-rated outcomes chapter should be included prominently in the secondary outcomes. Quality of life is an important subjective outcome parameter. The DLQI seemed to correlate with the HECSI67, though a 10 scoring system specific for hand eczema patients would be preferred. Recently, an interna-

231 tional group of experts developed the German Quality of Life in Hand Eczema Question- naire (QOLHEQ).68 The QOLHEQ scores the health related quality of life for hand eczema patients in different domains (symptoms, emotions, functioning, and treatment) and is currently being translated into Dutch. This instrument will be a valuable addition to the existing dermatological quality of life scoring systems such as the SKINDEX and the DLQI. Further, other patient-rated outcomes such as itching, loss of sleep and productivity loss should be included. The number and severity of adverse events should be registered throughout the study. Finally, we highly recommend a cost-effectiveness analysis, including health care costs and patient expenses (absence from work, out-of-pocket expenses).69 The study would require a minimum duration of 24 weeks, since hand eczema is a chronic relapsing condition. During the study, all participants would be advised to continue the use of emollients.

One of the main difficulties Schmitt encountered in his study was the lack of naïve patients (patients who had never before received alitretinoin or cyclosporine). However, as demonstrated by the multivariate analysis in chapter 4, being naive does not significantly influence the outcome of the systemic therapy. Moreover, the majority of our patients received multiple episodes of systemic therapy, and the effectiveness during a second or third episode is not necessarily better or worse in the following episodes. The requirement to include naïve patients can therefore be negated, or the study could be conducted in a multicenter setting to achieve a substantial number of patients. Including a sufficient number of patients is important in order to draw valid conclusions with sufficient statistical power. The majority of the studies in the Cochrane review did not, however, include a sample size rationale. The sample size calculation is usually based on the power (ß = 80 or 90%), the level of significance α( = 0.05) and the expected difference in the primary outcome of the study. To do so, consensus regarding the primary outcome and the clinical relevant difference is necessary. Reports on the study should adhere to the updated CONSORT (Consolidated Standards of Reporting Trials) statement. 70,71 The CONSORT statement provides a checklist of 25 items which should be reported for each randomized controlled trial. Such reporting facilitates complete and transparent reporting of trials and aids in critical appraisal and interpretation. Observational studies should be reported according to the STROBE (STrengthening the Reporting of OBservational studies in Epidemiology) statement.72 Several questions remain unanswered, including: a comparison between alitretinoin and

chapter cyclosporine; a comparison of various potencies of topical corticosteroids and the taper regime; a comparison between phototherapy and potent topical therapy; and so on. We are 10 also still waiting for long-term results of studies regarding counseling and education. With regard to intervention studies, much remains to be explored.

232 Not only interventions, but also the causality of hand eczema needs more attention in the research. Studies of its genetic makeup should be included. Thyssen et al. took the first steps with their research into the association between xerosis, atopic dermatitis, hand eczema and contact sensitization, and filaggrin gene mutations. 73-75 Filaggrin mutations are known for their role in atopic dermatitis and skin barrier dysfunction, and seemed to be associated with the chronic dry fissured type of hand eczema. Moreover, hand eczema in patients with atopic dermatitis and filaggrin null mutations had an earlier onset and was more persistent (OR 2.98; 95% confidence interval 1.27-7.01) than in controls.76 We currently lack studies regarding the distinct groups of hyperkeratotic palmar eczema and recurrent vesicular hand eczema; however, since large population based studies are being conducted and gene mapping is a developing area, in the next few years the genes responsible for these specific subtypes may be revealed. Several times in this thesis we have emphasized the need for better definitions of subtypes; one of these subtypes is hyperkeratotic palmar eczema. Hyperkeratotic eczema could be caused by repeated trauma or a dysfunction in keratinisation, but the exact mechanism is unknown.77 Further studies into the histology and pathophysiology of this eczema might provide insight into this mechanism and whether hyperkeratotic palmar eczema is truly a form of hand eczema or something entirely different.

In this thesis I would like to advocate setting up well-designed randomized controlled trials of at least 24 weeks in order to compare different treatment groups. The quality of such studies would greatly benefit from more precise definitions of the types of hand eczema and their clinical subtypes. For this purpose international consensus is indispensable.

chapter 10

233 References 1. Van Coevorden AM, Coenraads PJ, Svensson A, Bavinck JN, Diepgen TL, Naldi L, et al. Overview of studies of treatments for hand eczema-the EDEN hand eczema survey. Br J Dermatol 2004:151:446-451. 2. Rothwell PM. External validity of randomised controlled trials: “to whom do the results of this trial apply?” Lancet 2005:365:82-93. 3. Andrade SE, Walker AM, Gottlieb LK, Hollenberg NK, Testa MA, Saperia GM, et al. Discontinuation of antihyperlipidemic drugs--do rates reported in clinical trials reflect rates in primary care settings? N Engl J Med 1995:332:1125-1131. 4. Pacor ML, Di Lorenzo G, Martinelli N, Mansueto P, Friso S, Pellitteri ME, et al. Tacrolimus ointment in nickel sulphate-induced steroid-resistant allergic contact dermatitis. Allergy Asthma Proc 2006:27:527-531. 5. Thyssen JP, Linneberg A, Menné T, Johansen JD. The epidemiology of contact allergy in the general population--prevalence and main findings.Contact Dermatitis 2007:57:287-299. 6. Koh KJ, Pearce AL, Marshman G, Finlay-Jones JJ, Hart PH. Tea tree oil reduces histamine-induced skin inflammation.Br J Dermatol 2002:147:1212-1217. 7. Wallengren J. Tea tree oil attenuates experimental contact dermatitis. Arch Dermatol Res 2011:303:333-338. 8. Carson CF, Riley TV. Safety, efficacy and provenance of tea tree (Melaleuca alternifolia) oil. Contact Dermatitis 2001:45:65-67. 9. Lynde C, Guenther L, Diepgen TL, Sasseville D, Poulin Y, Gulliver W, et al. Canadian hand dermatitis management guidelines. J Cutan Med Surg 2010:14:267-284. 10. Menné T, Johansen JD, Sommerlund M, Veien NK, Danish Contact Dermatitis Group. Hand eczema guidelines based on the Danish guidelines for the diagnosis and treatment of hand eczema. Contact Dermatitis 2011:65:3-12. 11. Diepgen TL, Elsner P, Schliemann S, Fartasch M, Kollner A, Skudlik C, et al. Guideline on the management of hand eczema ICD-10 Code: L20. L23. L24. L25. L30. J Dtsch Dermatol Ges 2009:7 Suppl 3:S1-16. 12. English J, Aldridge R, Gawkrodger DJ, Kownacki S, Statham B, White JM, et al. Consensus statement on the management of chronic hand eczema. Clin Exp Dermatol 2009:34:761-769. 13. Hald M, Agner T, Blands J, Veien NK, Laurberg G, Avnstorp C, et al. Clinical severity and prognosis of hand eczema. Br J Dermatol 2009:160:1229-1236. 14. Meding B, Wrangsjo K, Jarvholm B. Hand eczema extent and morphology--association and influence on long-term prognosis.J Invest Dermatol 2007:127:2147-2151. chapter 15. Hald M, Agner T, Blands J, Johansen JD, Danish Contact Dermatitis Group. Delay in medical 10 attention to hand eczema: a follow-up study. Br J Dermatol 2009:161:1294-1300.

234 16. Mollerup A, Johansen JD, Thing LF. Knowledge, attitudes and behaviour in everyday life with chronic hand eczema: a qualitative study. Br J Dermatol 2013:169:1056-1065. 17. Boyce JM, Kelliher S, Vallande N. Skin irritation and dryness associated with two hand-hygiene regimens: soap-and-water hand washing versus hand antisepsis with an alcoholic hand gel. Infect Control Hosp Epidemiol 2000:21:442-448. 18. Slotosch CM, Kampf G, Loffler H. Effects of disinfectants and detergents on skin irritation. Contact Dermatitis 2007:57:235-241. 19. Pittet D, Allegranzi B, Boyce J, World Health Organization World Alliance for Patient Safety First Global Patient Safety Challenge Core Group of Experts. The World Health Organization Guidelines on Hand Hygiene in Health Care and their consensus recommendations. Infect Control Hosp Epidemiol 2009:30:611-622. 20. Huang C, Ma W, Stack S. The hygienic efficacy of different hand-drying methods: a review of the evidence. Mayo Clin Proc 2012:87:791-798. 21. Agner T, Held E. Skin protection programmes. Contact Dermatitis 2002:47:253-256. 22. Kwon S, Campbell LS, Zirwas MJ. Role of protective gloves in the causation and treatment of occupational irritant contact dermatitis. J Am Acad Dermatol 2006:55:891-896. 23. Andersson T, Bruze M, Björkner B. In vivo testing of the protection of gloves against acrylates in dentin-bonding systems on patients with known contact allergy to acrylates. Contact Dermatitis 1999:41:254-259. 24. Rietschel RL, Huggins R, Levy N, Pruitt PM. In vivo and in vitro testing of gloves for protection against UV-curable acrylate resin systems. Contact Dermatitis 1984:11:279-282. 25. Munksgaard EC. Permeability of protective gloves to (di)methacrylates in resinous dental materials. Scand J Dent Res 1992:100:189-192. 26. Andreasson H, Boman A, Johnsson S, Karlsson S, Barregard L. On permeability of methyl methacrylate, 2-hydroxyethyl methacrylate and triethyleneglycol dimethacrylate through protective gloves in dentistry. Eur J Oral Sci 2003:111:529-535. 27. Fartasch M, Taeger D, Broding HC, Schoneweis S, Gellert B, Pohrt U, et al. Evidence of increased skin irritation after wet work: impact of water exposure and occlusion. Contact Dermatitis 2012:67:217-228. 28. Ramsing DW, Agner T. Effect of glove occlusion on human skin (II). Long-term experimental exposure. Contact Dermatitis 1996:34:258-262. 29. Long CC, Finlay AY. The finger-tip unit- a new practical measure.Clin Exp Dermatol 1991:16:444-447. 30. Gupta G, Mallefet P, Kress DW, Sergeant A. Adherence to topical dermatological therapy: lessons

from oral drug treatment. Br J Dermatol 2009:161:221-227. chapter 31. Krejci-Manwaring J, McCarty MA, Camacho F, Carroll CL, Johnson K, Manuel J, et al. Adherence with topical treatment is poor compared with adherence with oral agents: implications for effective 10 clinical use of topical agents. J Am Acad Dermatol 2006:54:S235-6.

235 32. Hix E, Gustafson CJ, O’Neill JL, Huang K, Sandoval LF, Harrison J, et al. Adherence to a five day treatment course of topical fluocinonide 0.1% cream in atopic dermatitis. Dermatol Online J 2013:19:20029. 33. Charman CR, Morris AD, Williams HC. Topical corticosteroid phobia in patients with atopic eczema. Br J Dermatol 2000:142:931-936. 34. Hoare C, Li Wan Po A, Williams H. Systematic review of treatments for atopic eczema. Health Technol Assess 2000:4:1-191. 35. van Velsen SG, Haeck IM, Bruijnzeel-Koomen CA. Percutaneous absorption of potent topical corticosteroids in patients with severe atopic dermatitis. J Am Acad Dermatol 2010:63:911-913. 36. Terra JB, Potze WJ, Jonkman MF. Whole body application of a potent topical corticosteroid for bullous pemphigoid. J Eur Acad Dermatol Venereol 2014:28:712-718. 37. Miest RY, Yiannias JA, Chang YH, Singh N. Diagnosis and prevalence of lanolin allergy. Dermatitis 2013:24:119-123. 38. Coloe J, Zirwas MJ. Allergens in corticosteroid vehicles. Dermatitis 2008:19:38-42. 39. Green CM, Holden CR, Gawkrodger DJ. Contact allergy to topical medicaments becomes more common with advancing age: an age-stratified study. Contact Dermatitis 2007:56:229-231. 40. Baeck M, Chemelle JA, Terreux R, Drieghe J, Goossens A. Delayed hypersensitivity to corticosteroids in a series of 315 patients: clinical data and patch test results. Contact Dermatitis 2009:61:163-175. 41. Baeck M, Chemelle JA, Goossens A, Nicolas JF, Terreux R. Corticosteroid cross-reactivity: clinical and molecular modelling tools. Allergy 2011:66:1367-1374. 42. Baeck M, Goossens A. Immediate and delayed allergic hypersensitivity to corticosteroids: practical guidelines. Contact Dermatitis 2012:66:38-45. 43. Baeck M, Goossens A. New insights about delayed allergic hypersensitivity to corticosteroids. G Ital Dermatol Venereol 2012:147:65-69. 44. Katsarou A, Makris M, Papagiannaki K, Lagogianni E, Tagka A, Kalogeromitros D. Tacrolimus 0.1% vs mometasone furoate topical treatment in allergic contact hand eczema: a prospective randomized clinical study. Eur J Dermatol 2012:22:192-196. 45. Schnopp C, Remling R, Möhrenschlager M, Weigl L, Ring J, Abeck D. Topical tacrolimus (FK506) and mometasone furoate in treatment of dyshidrotic palmar eczema: a randomized, observer-blinded trial. J Am Acad Dermatol 2002:46:73-77. 46. Waernulf L, Moberg C, Henriksson EW, Evengard B, Nyberg F. Patients’ views on care and treatment after phototherapy for psoriasis and atopic eczema including a gender perspective. J Dermatolog Treat 2008:19:233-240.

chapter 47. van Coevorden AM, Coenraads PJ, Stant AD, Duizendstraal A. Thuisbehandeling van handeczeem met PUVA (THP). 2002:VAZ/CVZ 99252. 10

236 48. van Coevorden AM, Kamphof WG, van Sonderen E, Bruynzeel DP, Coenraads PJ. Comparison of oral psoralen-UV-A with a portable tanning unit at home vs hospital-administered bath psoralen-UV-A in patients with chronic hand eczema: an open-label randomized controlled trial of efficacy.Arch Dermatol 2004:140:1463-1466. 49. Ruzicka T, Larsen FG, Galewicz D, Horvath A, Coenraads PJ, Thestrup-Pedersen K, et al. Oral alitretinoin (9-cis-retinoic acid) therapy for chronic hand dermatitis in patients refractory to standard therapy: results of a randomized, double-blind, placebo-controlled, multicenter trial. Arch Dermatol 2004:140:1453-1459. 50. Ruzicka T, Lynde CW, Jemec GB, Diepgen T, Berth-Jones J, Coenraads PJ, et al. Efficacy and safety of oral alitretinoin (9-cis retinoic acid) in patients with severe chronic hand eczema refractory to topical corticosteroids: results of a randomized, double-blind, placebo-controlled, multicentre trial. Br J Dermatol 2008:158:808-817. 51. Bissonnette R, Worm M, Gerlach B, Guenther L, Cambazard F, Ruzicka T, et al. Successful retreatment with alitretinoin in patients with relapsed chronic hand eczema. Br J Dermatol 2010:162:420-426. 52. Schmitt J, Schmitt N, Meurer M. Cyclosporin in the treatment of patients with atopic eczema - a systematic review and meta-analysis. J Eur Acad Dermatol Venereol 2007:21:606-619. 53. van der Schaft J, Politiek K, Schuttelaar MLA, de Bruin-Weller MS. Drug survival for cyclosporin A in a long-term daily practice cohort of adult patients with atopic dermatitis. manuscript in preparation. 54. Egan CA, Rallis TM, Meadows KP, Krueger GG. Low-dose oral methotrexate treatment for recalcitrant palmoplantar pompholyx. J Am Acad Dermatol 1999:40:612-614. 55. Agarwal US, Besarwal RK. Topical clobetasol propionate 0.05% cream alone and in combination with azathioprine in patients with chronic hand eczema: an observer blinded randomized comparative trial. Indian J Dermatol Venereol Leprol 2013:79:101-103. 56. Scerri L. Azathioprine in dermatological practice. An overview with special emphasis on its use in non-bullous inflammatory dermatoses.Adv Exp Med Biol 1999:455:343-348. 57. Neuber K, Schwartz I, Itschert G, Dieck AT. Treatment of atopic eczema with oral mycophenolate mofetil. Br J Dermatol 2000:143:385-391. 58. Haeck IM, Knol MJ, Ten Berge O, van Velsen SG, de Bruin-Weller MS, Bruijnzeel-Koomen CA. Enteric-coated mycophenolate sodium versus cyclosporin A as long-term treatment in adult patients with severe atopic dermatitis: a randomized controlled trial. J Am Acad Dermatol 2011:64:1074-1084. 59. Diepgen TL, Andersen KE, Chosidow O, Coenraads PJ, Elsner P, English J, et al. Guidelines for

diagnosis, prevention and treatment of hand eczema. J Deutsche Dermatol Gesellschaft - ahead of chapter printing. 10

237 60. Diepgen TL, Andersen KE, Brandao FM, Bruze M, Bruynzeel DP, Frosch P, et al. Hand eczema classification: a cross-sectional, multicentre study of the aetiology and morphology of hand eczema. Br J Dermatol 2009:160:353-358. 61. Lynde C, Cambazard F, Ruzicka T, Sebastian M, Brown TC, Maares J. Extended treatment with oral alitretinoin for patients with chronic hand eczema not fully responding to initial treatment. Clin Exp Dermatol 2012:37:712-717. 62. Johansen JD, Hald M, Andersen BL, Laurberg G, Danielsen A, Avnstorp C, et al. Classification of hand eczema: clinical and aetiological types. Based on the guideline of the Danish Contact Dermatitis Group. Contact Dermatitis 2011:65:13-21. 63. Molin S, Diepgen TL, Ruzicka T, Prinz JC. Diagnosing chronic hand eczema by an algorithm: a tool for classification in clinical practice.Clin Exp Dermatol 2011:36:595-601. 64. Weistenhofer W, Baumeister T, Drexler H, Kutting B. An overview of skin scores used for quantifying hand eczema: a critical update according to the criteria of evidence-based medicine. Br J Dermatol 2010:162:239-250. 65. Tugwell P, Boers M, Brooks P, Simon L, Strand V, Idzerda L. OMERACT: an international initiative to improve outcome measurement in rheumatology. Trials 2007:8:38. 66. Schmitt J, Williams H, HOME Development Group. Harmonising Outcome Measures for Eczema (HOME). Report from the First International Consensus Meeting (HOME 1), 24 July 2010, Munich, Germany. Br J Dermatol 2010:163:1166-1168. 67. Agner T, Jungersted JM, Coenraads PJ, Diepgen T. Comparison of four methods for assessment of severity of hand eczema. Contact Dermatitis 2013:69:107-111. 68. Ofenloch RF, Weisshaar E, Dumke AK, Molin S, Diepgen TL, Apfelbacher C. The Quality of Life in Hand Eczema Questionnaire (QOLHEQ): validation of the German version of a new disease- specific measure of quality of life for patients with hand eczema.Br J Dermatol 2014:17:304-312. 69. Schuttelaar ML, Vermeulen KM, Coenraads PJ. Costs and cost-effectiveness analysis of treatment in children with eczema by nurse practitioner vs. dermatologist: results of a randomized, controlled trial and a review of international costs. Br J Dermatol 2011:165:600-611. 70. Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 Statement: updated guidelines for reporting parallel group randomised trials. Trials 2010:11:32-6215-11-32. 71. Williams HC. Cars, CONSORT 2010, and clinical practice. Trials 2010:11:33-6215-11-33. 72. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007:370:1453-1457. 73. Thyssen JP, Ross-Hansen K, Johansen JD, Zachariae C, Carlsen BC, Linneberg A, et al. Filaggrin

chapter loss-of-function mutation R501X and 2282del4 carrier status is associated with fissured skin on the hands: results from a cross-sectional population study. Br J Dermatol 2012:166:46-53. 10

238 74. Thyssen JP, Johansen JD, Zachariae C, Menné T, Linneberg A. Xerosis is associated with atopic dermatitis, hand eczema and contact sensitization independent of filaggrin gene mutations. Acta Derm Venereol 2013:93:406-410. 75. Molin S, Vollmer S, Weiss EH, Ruzicka T, Prinz JC. Filaggrin mutations may confer susceptibility to chronic hand eczema characterized by combined allergic and irritant contact dermatitis. Br J Dermatol 2009:161:801-807. 76. Thyssen JP, Carlsen BC, Menné T, Linneberg A, Nielsen NH, Meldgaard M, et al. Filaggrin null mutations increase the risk and persistence of hand eczema in subjects with atopic dermatitis: results from a general population study. Br J Dermatol 2010:163:115-120. 77. Hersle K, Mobacken H. Hyperkeratotic dermatitis of the palms. Br J Dermatol 1982:107:195-201.

chapter 10

239 240 Summary Hand eczema: interventions & contact allergies

Wietske Andrea Christoffers

Department of Dermatology, University of Groningen, University Medical Center Groningen, Groningen, Netherlands

241 Summary This thesis contains a series of review articles and clinical studies on hand eczema and allergic contact dermatitis. A general introduction of hand eczema is presented in chapter 1, with additional focus on allergic contact dermatitis. Hand eczema is a common skin condition with prevalence up to 10%. The etiology of hand eczema is often multifactorial, although atopic dermatitis and wet work are well-known risk factors. Chapter 1 describes the diagnostic steps to establish hand eczema in daily practice and discusses the risk factors, etiology and consequences of hand eczema. Moreover the objectives of this thesis are presented. Primary objective of this thesis is to provide physicians more insight in the treatment of hand eczema.

In chapter 2 we present a part of the Cochrane review on different interventions for hand eczema. The original Cochrane review included 55 randomized controlled trials with a total of 4,619 participants. The review covered a wide variety of different interventions: emollients, topical corticosteroids, phototherapy, topical calcineurin inhibitors, systemic interventions and some rather unconventional treatments. Most studies used a placebo and only five studies compared two different classes of interventions. The studies had substantial clinical heterogeneity in interventions, outcome measures and duration. Statistical pooling was not possible for the majority of the studies, because they were considered too heterogeneous. For each study the risk of bias was weighted according to the Cochrane Hand Book Guidelines. Many studies were at unclear risk of bias in one or more components of trial design, however the quality of studies and reporting improved over time. The results of this review cannot be used to inform clinical practice with regard to the best way of managing hand eczema due to the diversity and quality of the included studies and the lack of comparing studies. Topical corticosteroids and phototherapy appear to be the major standard treatment modalities. However, there is little evidence of a comparative advantage of one specific treatment within these categories, and also little evidence for a comparative advantage over other treatment modalities: further studies are needed. In this thesis only the conventional systemic interventions are presented. We included seven studies with a total of 1,661 participants investigating different interventions: alitretinoin (n = 3), acitretin (n = 1), cyclosporine (n = 1), azathioprine (n = 1) and one study comparing cyclosporine to alitretinoin. In general, the studies on systemic interventions were well conducted, with few biases in trial design, except from sponsoring by pharmaceutical companies. Alitretinoin was effective in the treatment of chronic hand eczema in well-conducted randomized controlled trials with large numbers of patients (n = 1,468). However, since the only study that compared alitretinoin to another treatment modality (cyclosporine) was ended prematurely and no other comparisons were conducted, the superiority of alitretinoin over other treatment modalities remains unclear.

242 In chapter 3 we discussed questions physicians encounter during daily-practice from an evidence-based perspective. Overall, insufficient evidence was found to base a choice between short bursts of potent topical corticosteroids compared with continuous application of mild corticosteroids, systemic therapy as maintenance therapy, iontophoresis, additive effect of topical antibacterial agents and topical calcineurin inhibitors. Little evidence was found for the steroid-sparing effect of emollients, although these are widely prescribed. With regards to phototherapy both PUVA and UVB are effective, although there is no evidence of a clinical advantage of one modality over the other. Oral retinoids, especially alitretinoin, are effective and well tolerated in hand eczema, especially in hyperkeratotic hand eczema. With regards to education for secondary and tertiary prevention: these seemed to be effective; however, long-term data are needed.

Cyclosporine is used off-label to treat hand eczema. In chapter 4 we retrospectively studied the effectiveness of cyclosporine in a daily practice cohort of 102 patients with chronic hand eczema. Drug survival and a clustered Physician Global Assessment (PGA) were used to study the effectiveness of treatment with cyclosporine in patients at two Dutch Dermatology Departments. The median drug survival rate was 10.3 months. The overall drug survival rates after 6 months, 1, 2 and 3 years were 61.7%, 45.2%, 18.6% and 13.9%, respectively. Main reasons for discontinuation were adverse events, mainly early in treatment, and ineffectiveness. After three months a good response to treatment was recorded in 59.8% of the patients. In the subgroup of patients with recurrent vesicular hand eczema cyclosporine was especially effective: 66.7% achieved a PGA1 (more than 50% improvement compared to the baseline) score after three months of treatment and the overall drug survival was borderline significant longer than in other groups. Based on this study we would like to recommend cyclosporine as valuable treatment option for patients with hand eczema, especially in patients with recurrent vesicular hand eczema. Moreover, cyclosporine might be used for longer periods of time instead of short pulses of three months. The etiology of hand eczema is multifactorial and sometimes unclear, but allergic contact dermatitis is a well known cause. The most common sensitizations among 1,571 hand eczema patients are discussed in chapter 5. Moreover the relation between specific contact allergens and clinical subtypes of hand eczema according to the guidelines of the Danish Contact Dermatitis Group (DCDG) was studied. The most frequently found clinical subtypes were recurrent vesicular hand eczema (39.7%), chronic fissured hand eczema (35.5%), hyperkeratotic palmar eczema (7.6%) and pulpitis (4.0%). Subjects with recurrent vesicular hand eczema were significantly more likely to be sensitized (OR 1.55, p<0.001), whereas subjects with hyperkeratotic palmar eczema and pulpitis were less likely to be sensitized (OR 0.51, p<0.001; OR 0.44, p = 0.003). In addition females and patients over the age of 40 were more likely to be sensitized. Overall, metals (nickel sulphate, cobalt chloride), fragrances and

243 preservatives (methylchloroisothiazoline/ methylisothiazoline, methyldibromoglutaronitrile) were the most frequent sensitizers in patients with hand eczema. This did not deviate in the different clinical subtypes, although subjects with recurrent vesicular hand eczema were significantly more frequently sensitized to nickel sulphate (OR 1.50, p = 0.002) and other allergens compared to other clinical subtypes of hand eczema. Therefore, we concluded that in the diagnostic work up of hand eczema subjects with recurrent vesicular hand eczema should be patch tested, especially women of older age. However, the need for patch testing males with hyperkeratotic palmar eczema is less imperative. Allergic contact dermatitis can be very severe and may result in absence from work or change of occupation, as was demonstrated in chapter 6. Herein we discussed a process operator that suffered from a severe occupational-related bullous allergic contact dermatitis due to various acrylates. In this case report 1,6-hexanediole diacrylate (HDDA) and glycidyl methacrylate (GMA) were the main culprits, alongside various other acrylates. The importance of patch testing with a dilution series of acrylates to distinguish between a contact allergy and an irritant reaction was emphasized. Moreover, the importance of investigating the products used at work by a patient and testing the purity of these products was highlighted.

The need to consider allergens which are not included in a standard patch test series was also emphasized in chapter 7. Here we studied a process operator which was sensitized to isobornyl acrylate. We tried to establish whether isobornyl acrylate was an underdiagnosed allergen. However, besides the process operator and an incidental case report, no other cases of isobornyl acrylate contact allergy were recorded: neither in literature, nor in our cohort of acrylate-sensitized patients which were re-patch tested with isobornyl acrylate. In addition, the results of our (meth)acrylate patch test series over the last two decades were studied. A total of 151 patients was patch tested; 24 had positive reactions. The most frequent sensitizers were: 2-hydroxypropyl acrylate, 2-hydroxyethyl acrylate, 2-hydroxypropyl methacrylate, and diethyleneglycol diacrylate. Because of the risk of active sensitization, the (meth) acrylate patch test series should only be patch tested on indication. To minimize the exposure to acrylates, we composed a hypothetical screening series. Screening with 2-hydroxypropyl acrylate, ethyleneglycol dimethacrylate, ethoxylated bisphenol A glycol dimethacrylate and trimethylolpropane triacrylate identified 91.7% of our patients that were sensitized to (meth)acrylates. Since hand eczema is a chronic, often relapsing condition, patients might be inclined to search alternative treatments as was demonstrated in chapter 2. Another example of alternative treatments is tea tree oil, an oil distilled from the Melaleuca alternifolia. Aside from a natural remedy, tree tea oil is also a popular ingredient in household and cosmetic products. Oxidation of tea tree oil results in degradation products, such as the endoper-

244 oxidase ascaridole, which can cause allergic contact dermatitis. In chapter 8 we presented two cases of allergic contact dermatitis due to oxidized tea tree oil and ascaridole. In chapter 9 we aimed to identify the optimal patch test concentration for ascaridole, and to investigate the relationship between a positive reaction to ascaridole and a positive reaction to oxidized tea tree oil in 319 (hand) eczema patients. With an increasing ascaridole test concentration, the frequency of positive reactions increased: ascaridole 1% caused positive reactions in 1.4% of our patients; ascaridole 2% in 5.5%; and ascaridole 5% in 7.2%. However, the frequencies of irritant and doubtful reactions also increased with an increase in concentration. Therefore we recommended patch testing ascaridole at 2% in petrolatum. Every positive reaction to oxidized tea tree oil was accompanied by a positive reaction to ascaridole, but not every positive reaction to ascaridole was accompanied by a positive reaction to tree tea oil. Some clinical relevant reactions were missed by patch testing tea tree oil only. Thus patch testing distinct components such as ascaridole, should be preferred over patch testing complex mixtures, such as oxidized tea tree oil. In chapter 10 we gave a general discussion on this thesis and the main findings. All these findings were combined into a flowchart which provides the physician insight in the treatment of different clinical subtypes of hand eczema. Guidance, education and emollients are the basic of hand eczema management. Topical corticosteroids are the mainstream treatment. In patients with chronic hand eczema, unresponsive to topical corticosteroids, alitretinoin is highly recommended. However, in patients with the clinical subtype recurrent vesicular hand eczema, cyclosporine can be considered as first line systemic treatment. To further confirm our treatment algorithm, we provided suggestions for future studies.

245 246 Nederlandse samenvatting Handeczeem: interventies & contactallergieën

Wietske Andrea Christoffers

Department of Dermatology, University of Groningen, University Medical Center Groningen, Groningen, Netherlands

247 Nederlandse samenvatting Deze thesis bevat verscheidene reviewartikelen en klinische studies over handeczeem en contact allergie. In hoofdstuk 1 werd een algemene inleiding gegeven over handeczeem met extra aandacht voor allergisch contacteczeem. Handeczeem is een veelvoorkomende huidaandoening met prevalenties tot 10%. De etiologie is multifactorieel en soms on- bekend, maar atopisch eczeem en beroepen met veel nat werk zijn bekende risicofactoren voor het ontwikkelen van handeczeem. Een contact allergie kan zowel een oorzakelijke als een bijdragende rol spelen. Hoofdstuk 1 bevat een diagnostisch stappenplan om handeczeem vast te stellen in de dagelijkse praktijk. Daarnaast zijn de risicofactoren, oorzaken en gevolgen van handeczeem besproken. Ten slotte zijn de doelstellingen van deze thesis belicht. Het belangrijkste doel van deze thesis is om artsen meer inzicht te geven in de behandeling van handeczeem.

In hoofdstuk 2 zijn de verschillende behandelopties voor handeczeem geëvalueerd middels een Cochrane review. De oorspronkelijke Cochrane review bevatte 55 randomized controlled trials (RCT) met een totaal van 4,619 deelnemers, waarin verschillende behandelingopties werden bestudeerd. Deze verschillende opties waren: emollientia, topicale corticosteroïden, topicale calcineurine remmers, lichttherapie, systemische therapie en enkele onconventionele behandelingen zoals kruidentherapie. De meeste studies bevatten een placebo en slechts vijf studies vergeleken verschillende behandelgroepen. De studies waren erg divers wat betreft design, uitkomstmaten en duur van interventie en follow-up. Hierdoor was het niet mogelijk om de studies onderling goed te vergelijken en was het slechts mogelijk om de data van twee studies statistisch te poolen. Voor elke studie werden de risico’s voor bias (risk of bias) beoordeeld volgens de richtlijnen van het Cochrane Handboek. De meeste studies vertoonden een of meerdere vormen van bias in verschillende onderdelen van de studie, zoals randomisatie, blindering of sponsoring. De kwaliteit van de studies was sterk wisselend, maar in de loop der tijd verbeterde de kwaliteit van studies en het niveau van rapporteren aanzienlijk. Topicale corticosteroïden en lichttherapie zijn de meest gangbare behandelopties. Echter, er is weinig bewijs om een van deze behandelopties aan te bevelen. Door de wisselende kwaliteit van studies, de verschillende designs en het gebrek aan vergelijkende studies, konden we gebaseerd op de Cochrane review geen uniforme aanbeveling doen voor de behandeling van handeczeem. In deze thesis is slechts een deel van de Cochrane review geïncludeerd: alleen de conventionele systemische therapieën zijn besproken. Zeven studies met een totaal van 1,661 deelnemers die verschillende systemische middelen onderzochten werden geïncludeerd: alitretinoïne (n = 3), acitretine (n = 1), ciclosporine (n = 1), azathioprine (n = 1) en een vergeli- jking tussen ciclosporine en alitretinoïne. Over het algemeen waren de studies betreffende

248 systemische therapie goed opgezet met weinig tekortkomingen in het design. Voor alitretinoïne konden de resultaten van twee studies geclusterd worden en dit middel was effectief voor de behandeling van chronisch handeczeem in RCT’s met grote aantallen patiënten (n = 1,468). Het is niet mogelijk om uitspraken te doen over de superioriteit van alitretinoïne ten opzichte van andere behandelopties, omdat de enige studie die alitretinoïne met een ander middel (ciclosporine) vergeleek prematuur werd beëindigd en er verder geen andere vergelijkende studies zijn uitgevoerd.

In hoofdstuk 3 hebben we op een evidence-based methode praktische vragen beantwoord die artsen in hun dagelijkse praktijk tegen kunnen komen. Er was onvoldoende bewijs om aanbevelingen te doen omtrent de duur en potentie van topicale corticosteroïden, systemische therapie als onderhoudsbehandeling, iontoforese, de toegevoegde waarde van topicale antibiotica en topicale calcineurine remmers. Beperkt bewijs werd gevonden voor het steroïdsparende effect van emollientia, alhoewel deze wel wijdverbreid worden voorgeschreven. Met betrekking tot lichttherapie bleken zowel PUVA als UVB effectief in de behandeling van handeczeem, echter er was geen bewijs om een van beide opties te prefereren. Orale retinoïden en in het bijzonder alitretinoïne, zijn effectief voor de behandeling van handeczeem, met name bij patiënten met hyperkeratotisch handeczeem. Secundaire en tertiaire preventieprogramma’s leken effectief, hoewel lange termijn data nog ontbreken.

Ciclosporine wordt off-label toegepast om ernstig handeczeem te behandelen. In hoofdstuk 4 hebben we retrospectief de doelmatigheid van ciclosporine in de dagelijkse praktijk geanalyseerd met behulp van drug survival en een geclusterde physician global assessment (PGA). In totaal werden 102 handeczeempatiënten uit twee Nederlandse centra geïncludeerd. De mediane drug survival bedroeg ruim 10 maanden. De algehele drug survival na 6 maanden, 1, 2 en 3 jaar was 67,7%, 45,2%, 18,6% en 13,9%, respectievelijk. De belangrijkste redenen om de behandeling te staken waren bijwerkingen, voornamelijk gedurende de eerste maanden, en ineffectiviteit. De meerderheid (59,8%) van de patiënten reageerde goed op ciclosporine na drie maanden (PGA1 score). Ciclosporine was bijzonder werkzaam in het klinische subtype van patiënten met terugkerend vesiculair handeczeem; 66,7% behaalde een PGA1 score na drie maanden en de algehele drug survival leek langer. Gebaseerd op deze studie concluderen wij dat ciclosporine een waardevolle behandeloptie is, met name voor vesiculair handeczeem. Bovendien kan ciclosporine gedurende langere periodes worden gebruikt.

De etiologie van handeczeem is multifactorieel en allergisch contact eczeem is een bekende oorzaak. De meest voorkomende contactallergenen in 1,571 handeczeempatiënten werden besproken in hoofdstuk 5. Daarnaast werd de relatie tussen specifieke contactallergieën en

249 klinische subtypes van handeczeem volgens de richtlijnen van de Deense Contact Dermatitis Groep (DCDG) bestudeerd. De meest voorkomende klinische subtypen waren terugkerend vesiculair handeczeem (39,7%), chronisch gefissureerd handeczeem (35,5%), hyperkeratotisch palmair eczeem (7,6%) en pulpitis (4,0%). Patiënten met vesiculair handeczeem waren significant vaker gesensibiliseerd (OR 1.55, p<0.001) dan andere subgroepen, terwijl sensibilisatie minder waarschijnlijk was in patiënten met hyperkeratotisch palmair eczeem (OR 0.51, p<0.001) en pulpitis (OR 0.44, p = 0.003). Geslacht en leeftijd correleerden ook met het risico op sensibilisatie: vrouwen en mensen met een hogere leeftijd liepen meer risico gesensibiliseerd te zijn voor een contactallergeen. De meest voorkomende contactallergenen waren metalen (nikkelsulfaat, kobaltchloride), geurstoffen en conserveermiddelen (methylchloorisothiazoline/ methylisothiazoline, methyldibromoglutaronitriel). Dezelfde contactallergenen werden gevonden in de verschillende klinische subtypen, alhoewel patiënten met terugkerend vesiculair handeczeem significant vaker gesensibiliseerd waren voor nikkelsulfaat (OR 1.50, p = 0.002) en andere allergenen vergeleken met andere subtypes handeczeem. Samenvattend is in de diagnostiek van vesiculair handeczeem een epicutane allergietest essentieel, met name bij vrouwen van middelbare leeftijd. Echter, bij mannen met hyperkeratotisch palmair eczeem is een epicutane allergietest minder van belang. Allergisch contacteczeem kan zeer ernstig zijn en leiden tot ziekteverzuim of het veranderen van baan, zoals besproken in hoofdstuk 6. In dit hoofdstuk werd een productiemedewerker gepresenteerd met een ernstig werk-gerelateerd bulleus contacteczeem veroorzaakt door verscheidene acrylaten waaronder 1,6-hexanedioldiacrylaat (HDDA) en glycidyl methacrylaat (GMA). Het belang van patch testen met een verdunningsreeks van acrylaten om het onderscheid te maken tussen een contact allergische en een irritatieve reactie werd benadrukt door deze casus. Bovendien werd de noodzaak om producten van de werkplaats van de patiënt te testen in hoofdstuk 6 onderstreept. Dat het belangrijk is om contactallergenen te overwegen die niet in een standaard patch test serie voorkomen werd ook benadrukt in hoofdstuk 7. In dit hoofdstuk bespraken we de casus van een drukkerijmedewerker die gesensibiliseerd was voor isobornyl acrylaat, een acrylaat dat niet in onze (meth) acrylaten reeks voorkomt. De hypothese of isobornyl acrylaat een ondergediagnosticeerd allergeen is, werd onderzocht. Echter, op de drukker en een enkel case report na, konden we geen andere gevallen van isobornyl acrylaat sensibilisatie vinden, noch in de literatuur, noch onder onze eigen patiënten. Een deel van onze acrylaat- gesensibiliseerde patiënten werd gepatchtest met een verdunningsserie van isobornyl acrylaat, maar dit leverde geen nieuwe gevallen op. Daarnaast hebben we de resultaten van onze (meth)acrylaat patch test serie van de laatste twee decennia geëvalueerd. In totaal zijn 151 patiënten gepatchtest en 24 patiënten hadden minstens een positieve reactie. De meest frequente allergenen waren: 2-hydroxypropyl acrylaat, 2-hydroxyethyl acrylaat, 2-hydroxypropyl methacrylaat en diethyleenglycol diacrylaat. Gezien het risico van actieve sensibilisatie,

250 wordt aanbevolen de (meth)acrylaten serie alleen te testen bij een hoge a priori verdenking. Om de blootstelling aan acrylaten te beperken, hebben we een hypothetische screeningsreeks samengesteld. Screenen met 2-hydroxypropyl acrylaat, ethyleenglycol dimethacrylaat, ethoxylated bisphenol A glycol dimethacrylaat en trimethylolpropaan triacrylaat identificeerde 91,7% van onze gesensibiliseerde patiënten. Echter, omdat het vaak belangrijk is het exacte contactallergeen te achterhalen, zoals al vaker benoemd is in deze thesis, bevelen wij deze screeningsreeks niet aan bij een sterke a priori verdenking.

Aangezien handeczeem een chronische, terugkerende aandoening is, kunnen patiënten geneigd zijn naar alternatieve behandelingen te zoeken. Zoals vermeld in hoofdstuk 2, zijn natuurlijke producten een behandelalternatief voor conventionele therapie. Theeboomolie is een olie die gedistilleerd wordt van de Melaleuca alternifolia. Theeboomolie wordt als natuurlijke remedie gebruikt voor verscheidene (huid)aandoeningen, maar is ook een populair ingrediënt in cosmetica en huishoudproducten. Als theeboomolie oxideert, ontstaan er degradatieproducten waaronder de endoperoxidase ascaridol. Ascaridol kan een allergisch contact eczeem veroorzaken. In hoofdstuk 8 werden twee casus van allergisch contacteczeem door theeboomolie en/of ascaridol besproken. In hoofdstuk 9 hebben we de optimale patch testconcentratie voor ascaridol geïden- tificeerd en de relatie tussen positieve patch test reacties op ascaridol en theeboomolie bestudeerd bij 319 (hand)eczeempatiënten. Met het toenemen van de ascaridol testconcentratie, nam het aantal positieve reacties ook toe: ascaridol 1% veroorzaakte positieve reacties in 1,4% van de onderzochte patiënten, ascaridol 2% in 5,5% en ascaridol 5% in 7,2%. Echter, het aantal twijfelachtige en irritatieve reacties nam ook toe met een oplopende concentratie. Daarom raden we aan om ascaridol te patch testen in een concentratie van 2% in petrolatum. Iedere positieve reactie op geoxideerde theeboomolie ging samen met een positieve reactie op ascaridol, maar niet iedere positieve reactie op ascaridol werd vergezeld door een positieve reactie op theeboomolie. Door alleen theeboomolie te onderzoeken werden een aantal klinisch relevante reacties gemist. Het patch testen van de specifieke componenten (bijvoorbeeld ascaridol) is daarom te prefereren boven het onderzoeken van complexe mixen zoals geoxideerde theeboomolie.

In hoofdstuk 10 is een algemene samenvatting van de resultaten van deze thesis gegeven. De bevindingen van deze thesis zijn vervolgens gecombineerd in een behandelalgoritme. Deze bestaat uit een flowchart waarin we de arts praktische handvatten bieden om patiënten met handeczeem te behandelen met extra aandacht is voor de specifieke subtypes van handeczeem. Voorlichting en het gebruik van emollientia zijn de basis van een goed behandelplan, gevolgd door het gebruik van topicale corticosteroïden. Voor een

251 chronisch handeczeem wat onvoldoende reageert op topicale therapie wordt systemische behandeling met alitretinoïne aanbevolen. Echter, voor het subtype terugkerend vesiculair handeczeem kan ciclosporine als eerste therapeutische optie worden overwogen. Daarnaast bevat de flowchart alternatieve behandelopties voor therapieresistente gevallen. Om het behandelalgoritme verder te versterken, geven wij tot slot aanbevelingen voor toekomstig onderzoek waarbij vooral een vergelijkende studie tussen alitretinoïne en ciclosporine geïndiceerd lijkt.

252 253 254 Dankwoord

255 256 Dankwoord Het schrijven van je dankwoord is misschien nog wel het moeilijkste gedeelte van je proefschrift, omdat je weet dat iedereen dit leest. Dus mocht ik je nou vergeten zijn, dan is deze zin special voor jou. Bedankt voor al je steun tijdens mijn promotietraject!

Ten eerste wil ik mijn promotor, prof. dr. P.J. Coenraads en copromotor, dr. M.L.A. Schuttelaar hartelijk bedanken. Pieter Jan, bedankt voor je wijsheid, je rust, je connecties, het inzicht en de kansen die je mij gegeven hebt. We hebben samen een hoop losse eindjes weggewerkt en METc’s doorstaan en ik heb veel over de leuke en minder leuke kanten van wetenschap van je geleerd. Be- dankt hiervoor. Marie-Louise, zonder jou had ik niet gestaan waar ik nu sta. Bedankt voor je vertrouwen in mij, je motivatie en je kennis. Ik hoop dat het speerpunt eczeem binnen de afdeling verder geconsolideerd wordt mede door deze promotie.

I would also like to thank the members of the reading committee; prof. dr. T. Agner, prof. W. Uter and prof. dr. T. Rustemeyer. Thank you for reading and judging my extensive manuscript.

Mijn paranimfen, Jorinde Dies-Talman en Tanja Vogel wil ik bedanken voor al hun steun: niet alleen tijdens deze geweldige promotiedag, maar ook in het algemeen. Lieve Jorinde, jij bent mijn rots in de branding, die mij door en door kent. We weten allebei wat we aan elkaar hebben en hoewel onze levens compleet anders lopen, kan ik me geen leven zonder jou voorstellen. Bedankt dat je er altijd voor mij bent. Lieve Tanja, ik vind het bijna zonde om de kliniek in te gaan en niet meer iedere dag van je aanwezigheid te kunnen genieten (of profiteren). Samen met jou sparren, of het nou over wetenschappelijke onderwerpen, klussen, watersnoodrampen of mannen is, samen zijn wij een super team. Je bent naast een geweldige collega en kamergenoot inmiddels ook een dierbare vriendin geworden.

Lieve Laura, wij hebben samen jarenlang een kamer gedeeld. Jij hebt mij als studentje alles geleerd wat je wist en me geholpen tijdens mijn eerste stappen in de wetenschap. Dat ik deze mooie dag samen met jou mag delen, vind ik een hele eer.

Alle proefpersonen die ik in de afgelopen tijd aan een plakproef, vragenlijst of een andere vorm van marteling heb mogen onderwerpen, wil ik hartelijk danken voor hun medewerking, hun geduld en hun vertrouwen.

257 Een proefschrift schrijven, kun je niet alleen. Ik wil het complete (en dynamische) eczeemteam bedanken voor hun steun in alle mogelijke vormen: Annemarie Bosch, Nelleke Drukker, Ramona van Houten, Vivian Loonstra, Carolien Scholten en Jessica Wilbrink.

Ik wil alle AIOS dermatologie bedanken voor hun steun en geduld tijdens mijn promotietraject en voor het feit dat deel mag zijn van jullie hechte groep. Bedankt.

Daarnaast wil ik alle stafleden dermatologie bedanken en het dagelijks bestuur van de afdeling dermatologie. In het bijzonder wil ik prof. dr. Jonkman en dr. Horvath bedanken voor de mogelijkheid om mijn promotietraject tijdens de opleiding af te ronden.

Aan dit proefschrift hebben ook veel (voormalig) studenten meegewerkt en ik wil hen allemaal bedanken voor hun inzet. In het bijzonder wil ik Klaziena Politiek en Marrit Boonstra bedanken. Het is geweldig om je enthousiasme voor een onderwerp te kunnen delen en over te dragen aan ‘je student’. Ik kijk uit naar jullie boekjes!

Ik wil iedereen bedanken die met ons heeft samengewerkt aan een van de projecten in dit boekje of daarbuiten. Alle perifere dermatologen, bedankt voor het verwijzen van patiënten en proefpersonen. Marjolein de Bruin-Weller en Jorien van der Schaft bedankt voor de ener- verende samenwerking binnen het eczeemconsortium. Moreover, I would like to thank our international collaborators: Frau Blömeke from Trier, Germany; Magnus Bruze, Marléne Isaksson and Malin Engfeldt from Malmö, Sweden. I would also like to thank all of the people who worked on the Cochrane review: Cathy Bennett, Janine Dickinson-Blok, Thomas Diepgen, Uwe Matterne, Åke Svensson and Hywell Williams.

Verder bedank ik natuurlijk alle verpleegkundigen, verpleegkundig specialisten, doktersassistenten, administratief medewerkers, secretaresses, laboranten, Buurman en Piet Toonder van de afdeling Dermatologie van het UMCG.

Familie 43, Lisanne, Olaf, Paul, Rakita en Yvette, jullie zijn mijn uitlaatklep. Zonder jullie en zwarte sambucca was het nooit goed gekomen met me (of juist wel).

Musketiers, lieve Lonneke, Marchien, Nienke en Tjitske, maar natuurlijk ook de mannen en kleine Elyne, bedankt voor jullie vriendschap. All our dreams can come true if we have the courage to pursue them. Lieve Nienke, jou wil ik extra bedanken (weer een voor op je scorelijstje): bedankt dat jij de perfecte date en reispartner bent, ookal ben ik dat niet altijd.

258 Mariëtte, je bent mijn oudste vriendin. Soms zien we elkaar maanden niet, maar het voelt altijd goed en vertrouwd. Ik hoop dat jouw promotietraject ook snel vorm mag krijgen.

Suriname-vrienden, ik heb met jullie een geweldige tijd in SU mogen beleven en nu nog steeds. Bedankt voor het vleugje no-spang wat tijdens mijn een promotietraject af en toe erg hard nodig was.

Gerlien, lieve grote zus, jij bent in heel veel dingen in het leven mijn voorbeeld. At jij geen spruitjes, dan ik ook niet, ging jij balletten, dan moest ik ook. Inmiddels hebben we elk ons eigen leven, maar jouw mening blijft belangrijk voor mij. Roel, lieve, kleine, grote broer, jij helpt mij om nuchter in het leven te blijven staan en je bent mijn redder in nood als het gaat om Excel of Ikea meubels. Lieve Arjan, bedankt dat je z’n goede man voor mijn zus bent en zo geduldig met al mijn “feedback” over dit prachtige boekje omging. Lieve Annelies, ik ben blij dat jij bij onze familie hoort en op mijn broertje past.

Als laatste, wil ik mijn ouders bedanken. Jullie hebben mij altijd op alle mogelijke manieren gesteund. Bij iedere beslissing die ik nam in mijn leven, waren jullie er voor mij. Ook wanneer het nodig was om de stukken weer bijeen te lijmen... Bedankt. Geweldige ouders, nu hebben jullie pas echt een doctór gemaakt.

259 260 Bibliografie

261 Publications Christoffers WA, Coenraads PJ, Schuttelaar M-LA. Two decades of occupational (meth)acrylate patch test results and focus on isobornyl acrylate. Contact Dermatitis. 2013 Aug;69:86-92.

Christoffers WA, Blömeke B, Coenraads PJ, Schuttelaar M. Co-sensitization to ascaridole and tea tree oil. Contact Dermatitis. 2013 Sep 1;69:187-189.

Schuttelaar MLA, Christoffers WA, Blömeke B, Coenraads PJ. Is ascaridol het contactallergeen in tea tree oil? Nederlands Tijdschrift voor Dermatologie en Venereologie. 2013 Sep 1;23:431-436.

Christoffers WA, Coenraads PJ, Schuttelaar M-L. Bullous allergic reaction caused by colophonium in medical adhesives. Contact Dermatitis. 2014 Apr;70:256-7.

Christoffers WA, Schuttelaar M, Coenraads PJ. Hand eczema. In Evidence Based Dermatology. Wiley-Blackwell. 2014. p.117.

Christoffers WA, Blömeke B, Coenraads PJ, Schuttelaar M-L. The optimal patch test concentration for ascaridole as a sensitizing component of tea tree oil. Contact Dermatitis. 2014 Sep;71:129-37.

Boonstra MB, Christoffers WA, Coenraads PJ, Schuttelaar MLA. Patch test results of hand eczema patients: relation to clinical types. Journal of the European Academy of Dermatology and Venereology : JEADV. 2014 Sep 15.

Vogel TA, Christoffers WA, Engfeldt M, Bruze M, Coenraads PJ, Schuttelaar ML. Severe bullous allergic contact dermatitis caused by glycidyl methacrylate and other acrylates. Contact Dermatitis. 2014 Oct;71:247-9.

262 Abstracts (oral presentations, posters) Christoffers WA, Coenraads PJ, Schuttelaar MLA. Isobornyl acrylate contact allergy: Rare or underdiagnosed? Contact Dermatitis. 2012 Jun 1;66:46.

Christoffers WA, Schuttelaar MLA, Coenraads PJ. Trends and developments in two decades of patch testing. Allergy: European Journal of Allergy and Clinical Immunology. 2013 Sep 1;68:598-599.

Brakel TM, De Flokstra-Blok BMJ, Elberink JNG, Schuttelaar MLA, Christoffers WA, Roerdink EM et al. Assessing allergy management information needs in primary care based on requests for sIgE determinations. Allergy: European Journal of Allergy and Clinical Immunology. 2013 Sep 1;68:280.

Flokstra-De Blok BM, Roerdink EM, Brakel TM, Oude Elberink JN, Oei R, Schuttelaar M-LA et al. Developing an allergy management support system (AMSS) for primary care: Agreement between the AMSS and the allergy specialist. Allergy: European Journal of Allergy and Clinical Immunology. 2014 Sep 1;69:602.

Schuttelaar ML, Christoffers WA, Boonstra M, Coenraads PJ. Treatment of hand eczema, any help from classification?Contact Dermatitis. 2014 Jun 1;70:20-21.

Christoffers, W. A., Blömeke, B., Coenraads, PJ, Schuttelaar, Is ascaridole a sensitizing degradation product in tea tree oil? Contact Dermatitis 2014 Jun 1;70:60.

263 264 Curriculum Vitae

265 266 Curriculum Vitae

Wietske Andrea (Wianda) Christoffers werd geboren op 29 april 1987 te Assen, waar ze ook opgroeide. In 2005 behaalde zij haar Gymnasium diploma aan het Vincent van Gogh College te Assen. Aansluitend volgde zij een studie Geneeskunde aan de Rijksuniversiteit Groningen. Tijdens haar studie werkte ze mee aan de EDEN Fragrance Study (prof. dr. P.J. Coenraads) en werkte ze als 24-uurs verzorgende bij een vrouw met multiple sclerose. Wianda doorliep haar coschappen achtereenvolgens in het Universitair Medisch Centrum Groningen, het Wilhelmina Ziekenhuis Assen, Diaconessenhuis te Meppel en de Isala Klinieken Zwolle. Haar semi-arts stage liep zij bij de afdeling Dermatologie van het Universitair Medisch Centrum Groningen en in het Academisch Ziekenhuis Paramaribo, Suriname. In 2012 sloot ze haar geneeskundeopleiding af met haar wetenschappelijke stage bij de afdeling Dermatologie van het UMCG onder begeleiding van dr. M.L.A. Schuttelaar. Deze stage resulteerde in haar eerste publicatie (hoofdstuk 7). Vervolgens startte Wianda met haar promotietraject onder begeleiding van prof. dr. P.J. Coenraads en dr. M.L.A. Schuttelaar. In februari 2014 is Wianda begonnen met haar opleiding tot dermatoloog welke zij in 2019 hoopt af te ronden.

267 268