Oral Immunosuppressive Drugs in the Treatment of Atopic Dermatitis

Home , ATC code D02, ATC code D05, Dihydrofolic acid

Oral immunosuppressive drugs in the treatment of atopic dermatitis

IMPROVING PERFORMANCE AND SAFETY

FLOOR M. GARRITSEN Colofon Oral immunosuppressive drugs in the treatment of atopic dermatitis Improving performance and safety

Thesis with a summary in Dutch, Utrecht University, the Netherlands

No part of this thesis may be reproduced, stored or transmitted, in any form or by any means, without prior permission of the author.

ISBN: 978-90-393-6931-9 Cover design and layout: Die Jongens Printed and published by: Proefschriftmaken.nl Oral immunosuppressive drugs in the treatment of atopic dermatitis: improving performance and safety

Orale immunosuppressiva bij de behandeling van constitutioneel eczeem: het verbeteren van effectiviteit en veiligheid

(met een samenvatting in het Nederlands)

Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof. dr. G.J. van der Zwaan, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op

donderdag 8 februari 2018 des middags te 2.30 uur

door

Floralie Maria Garritsen geboren op 22 oktober 1986 te Beek en Donk Promotor: Prof. dr. C.A.F.M. Bruijnzeel-Koomen

Copromotoren: Dr. M.S. de Bruin-Weller Dr. M.P.H. van den Broek Table of contents

Chapter 1 General introduction 7

Part I – Qualitative and quantitative exploration of oral immunosuppressive drug use in the Netherlands

Chapter 2 Ten years experience with oral immunosuppressive treatment 23 in adult patients with atopic dermatitis in two academic centres

Chapter 3 Use of oral immunosuppressive drugs in the treatment of atopic 39 dermatitis in the Netherlands

Part II - Safety of long-term treatment with oral immunosuppressive drugs

Chapter 4 Lymphopenia in atopic dermatitis patients treated 57 with oral immunosuppressive drugs

Chapter 5 Risk of non-melanoma skin cancer in patients with 71 atopic dermatitis treated with oral immunosuppressive drugs

Chapter 6 Is there an increased risk of cervical neoplasia in atopic 89 dermatitis patients treated with oral immunosuppressive drugs?

Chapter 7 Pregnancy and fetal outcomes after paternal exposure to azathioprine, 99 methotrexate or mycophenolic acid: a critically appraised topic

Part III - Optimizing the therapeutic potential of thiopurines

Chapter 8 Thiopurine metabolite levels in patients with atopic dermatitis 123 and/or chronic hand/foot eczema treated with azathioprine

Chapter 9 Improving outcome of azathioprine treatment in chronic eczema 141 by allopurinol co-prescription

Chapter 10 First experience with thioguanine in the treatment 153 of patients with atopic dermatitis

Chapter 11 General discussion 165

Chapter 12 Summary / Samenvatting 189

Appendices List of co-authors 204

List of publications 206

Acknowledgement 208

Curriculum vitae 212

General introduction 8 CHAPTER 1

Background Atopic dermatitis (AD, atopic eczema, eczema) is a chronic relapsing inflammatory skin disease, characterized by remissions and exacerbations. The prevalence varies between 2 and 5% in young adults and 20% in children, making it one of the most common skin diseases.1 It often occurs in families with other atopic diseases (bronchial asthma, allergic rhino conjunctivitis, food allergy).1,2 It is characterized by intense pruritus, excoriations and erythematous, xerotic, lichenified skin and has a characteristic age-dependent distribution 1,3 pattern. AD and the increased risk of skin infections may have an effect on quality of life, sleep, self-esteem and systemic comorbidities, that may influence social contacts and activities, interpersonal relationships, participations in leisure and sports, and attendance or performance at school or work.4,5

Treatment of atopic dermatitis with oral immunosuppressive drugs A combination of emollients, topical corticosteroids, topical calcineurin inhibitors, oral antihistamines and antibiotics, is often used in the management of AD.2 Different forms of phototherapy may also be effective.6-8 However, a subgroup of severe and difficult to treat patients remains. Patients in who AD is not controlled by topical treatment or UV-light treatment are candidates for oral immunosuppressive treatment. In addition, in patients who need daily treatment with potent topical corticosteroids of large body areas to control their AD, oral immunosuppressive drugs may be used to taper these corticosteroids to safe maintenance levels.1,2,9-12 Besides, sometimes side effects of the (intensive) treatment with topical corticosteroids, like striae and skin atrophy, are an indication for the start of oral immunosuppressive drugs. In some cases of difficult to treat AD patients the involved body surface may be limited, but the intensity of the eczema may still have huge impact on the quality of life. Examples are eyelid dermatitis and hand/foot eczema. Symptoms in these patients may be so severe and obstructive, that oral immunosuppressive treatment is indicated, although the eczema affects only a small part of the body service.

All the above mentioned patients may be candidates for the treatment with oral immunosuppressive drugs. In addition, recently, the expert panel of the international eczema council concluded that the decision to start systemic medication should include severity and quality of life assessments over time while considering the individual’s general health status, psycho- logical needs and personal attitudes towards systemic therapies.13

Oral immunosuppressive drugs that are used in the treatment of AD are systemic corticosteroids, cyclosporine A, azathioprine, methotrexate, mycophenolate mofetil and enteric-coated mycophenolate sodium. Based on the mode of action (table 1), clinical efficacy during treatment with azathioprine, methotrexate and mycophenolate mofetil / enteric-coated mycophenolate sodium can only be reached after 12-16 weeks; therefore these drugs are not suitable for treatment of acute exacerbations.1 GENERAL INTRODUCTION 9

Table 1 Mode of action of oral immunosuppressive drugs Drug Mode of action 1 Cyclosporine A11,14,15 - Acts primarily on T cells by inhibiting calcineurin and thus signal trans- duction mediated by T-cell receptor activation - Affects interleukin-2 production, B cells and dendritic cells - Can also suppress some growth-related pathways in keratinocytes Azathioprine11,16 - Purine synthesis inhibitor - Inhibits DNA synthesis and therefore affects active replicating cells, such as B cells and T cells Methotrexate11,16 - Antifolate metabolite - Affects the synthesis of DNA, RNA and purines - Negatively affects T-cell function, targeting several key T-cell functions Myfophenolate mofetil - Contains mycophenolic acid / enteric-coated myco- - Inhibits synthesis of DNA and RNA in B- and T-cell development by phenolate sodium11,14 inhibition of inosine monophosphate dehydrogenase and therefore prevents immune cell proliferation - Selectively affects B cells and T cells

Clinical trials on oral immunosuppressive drugs Clinical efficacy and safety have been proven in clinical trials for most oral immunosuppressive drugs, including some head-to-head trials.17 Schmitt et al investigated the comparative efficacy of prednisolone and cyclosporine A for adults with severe eczema in a double-blind randomized multicenter trial.18 Patients were treated with prednisolone (initial dose 0.5-0.8mg/ kg daily) for two weeks followed by placebo for 4 weeks or cyclosporine A (2.7-4.0 mg/kg daily) for 6 weeks. They concluded that cyclosporine A was more efficacious than prednisolone, but the study was terminated early because of unexpected high numbers of withdrawals in the prednisone arm, mainly due to exacerbations of eczema. In the study of Haeck et al enteric- coated mycophenolate sodium and cyclosporine A were compared in an observer-blinded randomized controlled trial (RCT).14 Fifty-five patients with AD were treated with cyclosporine A (5mg/kg) for 6 weeks. Thereafter, patients received cyclosporine A (3mg/kg, n=26) or enteric-coated mycophenolate sodium (1440mg, n=24). Enteric-coated mycophenolate sodium was as effective as cyclosporine A as maintenance therapy in patients with AD, but clinical improvement with enteric-coated mycophenolate sodium was delayed compared to cyclosporine A. In addition, prednisone courses were often needed in the patients treated with enteric-coated mycophenolate sodium. In the study of Schram et al the efficacy and safety of methotrexate versus azathioprine was compared.16 In this RCT, 42 patients with severe AD were randomly assigned to receive methotrexate (10-22.5 mg/week) or azathioprine (1.5-2.5 mg/kg/day) for 12 weeks. Both treatments achieved clinically relevant improvement and were safe, but two (10%) patients in the methotrexate group and 4 (18%) patients in the azathioprine group required a prednisone course. This study clearly demonstrates the delayed mechanism of action of both methotrexate and azathioprine. 10 CHAPTER 1

Based on a systematic review performed by Roekevisch et al, including the above mentioned head-to-head trials and 31 other RCTs (12 different systemic treatments, total of 1653 patients), cyclosporine A is recommended as first-line treatment for short-term 17 use. Azathioprine is recommended as second-line treatment option and methotrexate can be considered as a third- line treatment option.

Oral immunosuppressive drugs in atopic dermatitis in daily practice The clinical trials with oral immunosuppressive drugs show encouraging results. However, a clinical trial setting is different from a real life daily practice situation and results from clinical trials are not always generalizable to daily practice. Treatment duration in clinical trials is limited and patients participating in clinical trials often differ from patients in daily practice due to the strict in- and exclusion criteria. In addition, patients participating in clinical trials are generally more compliant, probably resulting in a better clinical efficacy of the drug.

Real life data from patients with AD, treated with oral immunosuppressive drugs in daily practice, provide important and essential information in addition to the existing clinical trial data. In the past years, a notable increased interest in daily practice data was seen, also in the field of AD. Several retrospective cohort studies evaluating the use of oral immunosuppressive drugs have been published in the past years.19-23 One of the newer developments in the evaluation of long-term treatment is ‘drug survival’. Drug survival is the time a patient continues to take a particular drug and is a reflection of daily practice.24 Drug survival is a measure which encom- passes factors such as effectiveness, safety, and patients’ and doctors’ preferences. It reflects the real life situation in daily practice in a large unselected patient group without intervening of a fixed research protocol. The first drug survival studies on the treatment of patients with AD were performed by van der Schaft et al and Politiek et al.25-27 These studies clearly demonstrate that oral immunosuppressive drugs are regularly used for longer treatment periods in daily practice, but that a considerable number of patients discontinues treatment due to side effects (ranging from 14% in enteric-coated mycophenolate sodium to 36% in azathioprine) or ineffectiveness (ranging from 15% in methotrexate to 38% in enteric-coated mycophenolate sodium).

Open questions concerning the percentage of patients with difficult to treat atopic dermatitis in need for systemic immunosuppressive treatment Although AD is a very common disease, data on the percentage of patients with really difficult to treat AD are scarce. From socio-economic perspective it is very important to have more insight in the number of patients in need of systemic immunosuppressive treatment, as new very effective, but expensive, biological treatment options will be available in the near future. Since patients with severe and difficult to treat AD are commonly treated with oral immunosuppressive drugs, analyzing the number of AD patients using these drugs can give an impression of the percentage of difficult to treat patients in the total AD population. GENERAL INTRODUCTION 11

In 2009, the study of Schmitt et al made an effort to investigate the number of difficult to treat AD patients. They found that about 10% of the AD patients in Germany was treated with systemic 1 corticosteroids and that these patients were supposed to have moderate to severe eczema.28 Because in this study also short courses of prednisone were included, this percentage is not representative for the percentage really difficult to treat AD patients. Data on maintenance treatment (> 3 months) with different oral immunosuppressive drugs will give a better indication of the number of difficult to treat severe AD patients, but these data are lacking in literature.

Question 1: What proportion of atopic dermatitis patients is treated with oral immunosuppressive drugs in the Netherlands and can be defined as difficult to treat?

Question 2: Which oral immunosuppressive drugs are used in the Netherlands in the management of atopic dermatitis and in what order?

Open questions concerning safety of treatment with oral immunosuppressive drugs in atopic dermatitis The immunosuppressive drugs used in the treatment of AD can cause a broad spectrum of objective and subjective side effects that often require withdrawal of therapy. Hypertension and increased serum creatinine levels are often seen during treatment with cyclosporine A, while liver function abnormalities are more commonly seen during treatment with methotrexate and azathioprine. These side effects are known and screening protocols include monitoring of the blood pressure, liver- and kidney function. Less is known about the susceptibility of infections during systemic immunosuppressive drugs in AD patients. Monitoring of lymphocyte counts may be valuable for selecting AD patients at risk for (severe) infections, as lymphopenia is associated with an increased infection risk. Mild lymphopenia might be a reflection of a sufficient immunosuppressive effect of medication, but it is known that persistent lymphopenia, regardless of its cause, is associated with an increased risk of opportunistic infections, as seen in HIV-positive patients, primary immunodeficiency and idiopathic CD4+ lymphopenia.29-31 Drug-induced lymphopenia is a known phenomenon in patients with autoimmune- and rheumatic diseases, inflammatory bowel disease and in organ transplant recipients.32-35 All systemic immunosuppressive treatment options used in AD (prednisone, cyclosporine A, azathioprine, methotrexate and mycophenolic acid) can induce lymphopenia. However, the clinical implications are currently not fully understood and data on the occurrence of persistent lymphopenia in patients with AD are lacking. 12 CHAPTER 1

Question 3: Is oral immunosuppressive treatment in atopic dermatitis (regularly) associated with lymphopenia? If so, what is the clinical relevance?

Due to the chronic course of AD, long-term treatment with oral immunosuppressive drugs is often necessary to maintain adequate disease control. However, for many dermatologists, an important barrier for maintenance treatment with oral immunosuppressive drugs in AD patients is the possible increased risk of non-reversible side effects and the development of malignancies. Especially the risk of the development of human papilloma virus (HPV) associated tumors, like non-melanoma skin cancer (including basal cell carcinoma and squamous cell carcinoma), is often stated as reason for avoiding long-term treatment with oral immunosuppressive drugs in the treatment of patients with AD. Indeed, immunosuppressive agents may increase the risk of these malignancies by causing DNA damage and affecting DNA repair mechanisms. Besides, it is thought that immunocompromised patients are more at risk for developing HPV infections.36 Clinical data on the risk of developing non-melanoma skin cancer is derived from studies in organ transplant recipients. Since these transplant patients often use more than one oral immunosuppressive drug at the same time and in higher doses, it is not possible to directly project these data on patients with AD.

Another concern is the possible increased risk of cervical carcinoma, that is also related to HPV infections. Oral immunosuppressive drugs are frequently prescribed in young women with AD, also in women who are not yet included in the national screening programme for cervical carcinoma (women are screened from the age of 30 years in the Netherlands). However, most literature on this topic is again derived from studies on organ transplant recipients. So the question arises whether the risk of cervical carcinoma is increased in AD patients and whether more intensive screening or early screening in these women should be advised.

Question 4: Is there an increased risk of non-melanoma skin cancer in patients with atopic dermatitis treated with oral immunosuppressive drugs?

Question 5: Is there an increased risk of cervical neoplasia in patients with atopic dermatitis treated with oral immunosuppressive drugs?

Finally, safety of oral immunosuppressive drugs during pregnancy is an important issue, as oral immunosuppressive drugs are frequently used in relatively young patients with AD. This also includes pregnancies fathered by men using oral immunosuppressive drugs. Studies on the risk of maternal use of oral immunosuppressive drugs during pregnancy have been evaluated and summarized before: cyclosporine A should be preferable avoided, methotrexate and mycophenolate mofetil are contraindicated and in azathioprine the dose should be reduced when lymphopenia was detected in the third trimester.37 Several studies on paternal use of oral immuno- GENERAL INTRODUCTION 13

suppressive drugs at the time of conception have also been published before. The opinions regarding the use of cyclosporine A in men who wish to conceive are fairly straightforward (it 1 can be used), but there are different perceptions about the use of the other oral immunosuppressive drugs.38 A systematic overview of the available studies was missing in literature.

Question 6: What is the risk of adverse pregnancy and fetal outcomes in pregnancies fathered by men exposed to azathioprine, methotrexate and mycophenolic acid?

Improving performance of oral immunosuppressive drugs AD is a very heterogeneous disease. However treatment protocols are still universal (‘one size fits all’ approach). Preferable, treatment options should be adapted to the individual patient, to prevent discontinuation due to side effects or ineffectiveness. Nowadays, patients undergo a ‘trial and error’ approach, delaying time to treatment response and negatively impacting quality of life. In order to increase the number of patients treated successfully with oral immunosuppressive drugs, the performance of the different drugs has to be improved. An optimal drug performance is defined as an adequate efficacy with limited and/or acceptable side effects. Recently, an increased interest for the so called ‘personalized treatment’ was noticed, with different foci and interests.

Demographic factors, such as age, may influence the choice for best fitting treatment. A long- term drug survival analysis of cyclosporine A in patients with AD showed that treatment with cyclosporine A was less successful in the older patients.25 Side effects, especially serum creatinine rise, was also more often observed in eldery.39 One of the other options to individualize AD treatment, is a better understanding of the different genetic and immunologic mechanisms underlying the wide spectrum of disease; this will able the development of a roadmap for individualized treatments.40 The measurement of biomarker levels in AD patients may enable the dermatologist to provide a more tailored management.40,41 A recent publication identified four clusters of patients with AD that could represent endotypes with unique biological mechanisms.41 These endotypes seem to be promising in the field of personalized care, especially in future treatment with different biologics.

Finally, the clinical effect of oral immunosuppressive drugs depends on the drug metabolism and the biological availability of the drug is influenced by different enzymes. Understanding these mechanisms in individual patients may improve the performance of the oral immunosuppressive drugs. Therapeutic drug monitoring and pharmacogenomics may further improve the performance of oral immunosuppressive drugs.42,43 14 CHAPTER 1

Improving performance of thiopurines Azathioprine is frequently used off-label in the management of AD but its therapeutic potential may have not been fully utilized. The azathioprine metabolism is very complex due to involvement of various enzymes (figure 1). Azathioprine is a thiopurine pro-drug and has no immunosuppressive activity itself. In the liver, it is converted into 6-mercaptopurin (6-MP). Subsequently, 6-MP undergoes metabolic transformations, via a complex enzymatic pathway, resulting in a variety of pharmacological active metabolites.44 The most important metabolites are 6-thioguanine nucleotide (6-TGN) and methylated 6-methylmercaptopurine (6-MMP). The immunosuppressive effect of azathioprine is caused by the 6-TGN metabolites.

6-MTITP 6-MTGTP

kinase kinase

6-MTIDP 6-MTGDP

kinase kinase

HHGGPPRRTT 6-MMP 6-MTIMP 6-MTGMP 6-MTG

TPMT TPMT kinase TPMT

GDC HHGGPPRRTT IIMMPPDDHH GGMMPPSS HHGGPPRRTT AZA 6-MP 6-TIMP 6-TXMP 6-TGMP 6-TG

XO kinase kinase XO

6-TUA 6-TIDP 6-TGDP 6-TUA ITPase NUDT15

kinase kinase

6-TITP 6-GTP

Figure 1 The azathioprine metabolism

AZA azathioprine; 6-MP 6-mercaptopurine; 6-TG 6-thioguanine; 6-MMP 6-methylmercaptopurine; 6-TUA 6-thiouric acid; 6-TIMP 6-thioinosine monophosphate; 6-TIDP 6-thioinosine diphosphate; 6-TITP 6-thioinosine triphosphate; 6-MTIMP 6-methylthioinosine monophosphate; 6-MTIDP 6-methylthioinosine diphosphate; 6-MTITP 6-methylthioinosine triphosphate; 6-TXMP 6-thioxhanthosine monophosphate; 6-TGMP 6-thioguanine monophosphate; 6-TGDP 6-thioguanine diphosphate; 6-TGTP 6-thioguanine triphosphate; 6-MTGMP 6-methylthioguanine monophosphate, 6-MTGDP 6-methylthioguanine diphosphate; 6-MTGTP 6-methylthioguanine triphosphate.

Enzymatic conversions: GDC glutathione-dependent cleavage; TPMT thiopurine S-methyl transferase; XO xanthine oxidase; HGPRT hypoxanthine- guanine phosphoribosyl transferase; IMPDH inosine monophosphate dehydrogenase; GMPS guanosine monophosphate synthetase; ITPase inosine triphosphate pyrophosphohydrolase; NUDT15 nudix hydrolase 15

Green: 6-thioguanine nucleotides (i.e. 6-TGN), Red: 6-methyl-mercapto-ribonucleotides (i.e. 6-MMP) GENERAL INTRODUCTION 15

In an approach to individualize treatment, the azathioprine dose can be based on the levels of 6-TGN and 6-MMP (measured in red blood cells). This approach is already commonly used 1 in patients with inflammatory bowel disease and organ transplant recipients. In these patient groups, several studies tried to define cut-off values of 6-TGN and 6-MMP in order to predict the risk of myelotoxicity and hepatotoxicity. Besides, attempts were made to define the ‘therapeutic range’, by investigating the 6-TGN level that resulted in maximum decrease of disease (table 2).45-47

Table 2 Therapeutic and toxic levels of 6-TGN and 6-MMP, based on studies in inflammatory bowel disease, according to two different measurement methods Lennard Dervieux Therapeutic 6-TGN 230-450 pmol/8x108 RBC 600-1200 pmol/8x108 RBC 6-MMP/ 6-TGN ratio 5-25 2-10 Toxic 6-TGN >450 pmol/8x108 RBC >1200 pmol/8x108 RBC 6-MMP >5700 pmol/8x108 RBC >5700 pmol/8x108 RBC

Question 7: Can measuring thiopurine metabolite levels in atopic dermatitis patients treated with azathioprine improve the performance?

In studies in patients with inflammatory bowel disease, attempts have been made to influence the metabolism of azathioprine by the co-prescription of allopurinol.48-50 In several studies, this resulted in a clinical improvement and a decrease of (subjective) side effects and hepatotoxicity. Besides, studies in inflammatory bowel disease showed promising results of the treatment with thioguanine. The metabolism of thioguanine is less complex than the metabolism of azathioprine and leads more directly towards the intended pharmacologically active product 6-TGN.51 Azathioprine with allopurinol co-prescription and thioguanine may also be interesting therapeutic options in AD patients, but there is no experience with these treatment options in AD.

Question 8: Are azathioprine with allopurinol co-prescription and thioguanine effective and safe treatment options for patients with atopic dermatitis? 16 CHAPTER 1

Outline of this thesis This thesis consists of three parts. In chapter 2 and 3, the prescription behavior of oral immunosuppressive drugs in AD in the Netherlands was evaluated. In chapter 2, the situation in two university medical centres was analyzed; in chapter 3 we investigated the prescribing behavior for oral immunosuppressive drugs, by deriving data from a national pharmacy database. From chapter 4 until 7 the safety of maintenance treatment with oral immunosuppressive drugs is described. Chapter 4 describes the occurrence of persistent lymphopenia in patients on maintenance treatment with oral immunosuppressive drugs. In chapter 5 we evaluated the risk of non-melanoma skin cancer in patients treated with oral immunosuppressive drugs. In chapter 6, the risk of cervical carcinoma was evaluated and recommendations were made for cervical screening in (young) patients using oral immunosuppressive drugs. In the critically appraised topic of chapter 7, a literature search was performed to evaluate the risk of negative pregnancy and fetal outcomes when pregnancies are fathered by men using oral immunosuppressive drugs. In the last part of this thesis, the use of thiopurines in AD was investigated. In chapter 8, we measured thiopurine metabolite levels in patients with AD and/or chronic hand/foot eczema. In chapter 9, effectiveness of azathioprine treatment was optimized by co-prescription of allopurinol. Finally, in chapter 10, we described our first experience with the use of thioguanine.

Table 3 Outline of this thesis and the corresponding chapters PART I: Qualitative and quantitative exploration of oral immunosuppressive drug use in the Netherlands Chapter 2 Ten years experience with oral immunosuppressive treatment in adult patients with atopic dermatitis in two academic centres Chapter 3 Use of oral immunosuppressive drugs in the treatment of atopic dermatitis in the Netherlands PART II: Safety of long-term treatment with oral immunosuppressive drugs Chapter 4 Lymphopenia in atopic dermatitis patients treated with oral immunosuppressive drugs Chapter 5 Risk of non-melanoma skin cancer in patients with atopic dermatitis treated with oral immunosuppressive drugs Chapter 6 Is there an increased risk of cervical neoplasia in atopic dermatitis patients treated with oral immunosuppressive drugs? Chapter 7 Pregnancy and fetal outcomes after paternal exposure to azathioprine, methotrexate or mycophenolic acid: a critically appraised topic PART III: Optimizing the therapeutic potential of thiopurines Chapter 8 Thiopurine metabolite levels in patients with atopic dermatitis and/or chronic hand/ foot eczema treated with azathioprine Chapter 9 Improving outcome of azathioprine treatment in chronic eczema by allopurinol co-prescription Chapter 10 First experience with thioguanine in the treatment of patients with atopic dermatitis GENERAL INTRODUCTION 17

References 1 1 Wollenberg A, Oranje A, Deleuran M et 10 Eichenfield LF, Tom WL, Berger TG et al. al. ETFAD/EADV Eczema task force 2015 Guidelines of care for the management of position paper on diagnosis and treatment atopic dermatitis: section 2. Management of atopic dermatitis in adult and paediatric and treatment of atopic dermatitis with patients. J Eur Acad Dermatol Venereol topical therapies. J Am Acad Dermatol 2016; 30: 729-47. 2014; 71: 116-32.

2 Ring J, Alomar A, Bieber T et al. Guidelines 11 Sidbury R, Davis DM, Cohen DE et al. for treatment of atopic eczema (atopic Guidelines of care for the management of dermatitis) part I. J Eur Acad Dermatol atopic dermatitis: section 3. Management Venereol 2012; 26: 1045-60. and treatment with phototherapy and systemic agents. J Am Acad Dermatol 3 Eichenfield LF, Tom WL, Chamlin SL et al. 2014; 71: 327-49. Guidelines of care for the management of atopic dermatitis: section 1. Diagnosis and 12 Sidbury R, Tom WL, Bergman JN et al. assessment of atopic dermatitis. J Am Acad Guidelines of care for the management of Dermatol 2014; 70: 338-51. atopic dermatitis: Section 4. Prevention of disease flares and use of adjunctive thera- 4 Blauvelt A, de Bruin-Weller M, Gooderham pies and approaches. J Am Acad Dermatol M et al. Long-term management of 2014; 71: 1218-33. moderate-to-severe atopic dermatitis with dupilumab and concomitant topical 13 Simpson EL, Bruin-Weller M, Flohr C et corticosteroids (LIBERTY AD CHRONOS): a al. When does atopic dermatitis warrant 1-year, randomised, double-blinded, place- systemic therapy? Recommendations from bo-controlled, phase 3 trial. Lancet 2017. an expert panel of the International Eczema Council. J Am Acad Dermatol 2017. 5 Sibbald C, Drucker AM. Patient Burden of Atopic Dermatitis. Dermatol Clin 2017; 35: 14 Haeck IM, Knol MJ, Ten Berge O et al. 303-16. Enteric-coated mycophenolate sodium versus cyclosporin A as long-term treat- 6 Meduri NB, Vandergriff T, Rasmussen H ment in adult patients with severe atopic et al. Phototherapy in the management dermatitis: a randomized controlled trial. J of atopic dermatitis: a systematic review. Am Acad Dermatol 2011; 64: 1074-84. Photodermatol Photoimmunol Photomed 2007; 23: 106-12. 15 Khattri S, Shemer A, Rozenblit M et al. Cyclosporine in patients with atopic derma- 7 Gambichler T. Management of atopic titis modulates activated inflammatory path- dermatitis using photo(chemo)therapy. ways and reverses epidermal pathology. J Arch Dermatol Res 2009; 301: 197-203. Allergy Clin Immunol 2014; 133: 1626-34.

8 Garritsen FM, Brouwer MW, Limpens J et al. 16 Schram ME, Roekevisch E, Leeflang MM Photo(chemo)therapy in the management et al. A randomized trial of methotrexate of atopic dermatitis: an updated systematic versus azathioprine for severe atopic review with implications for practice and eczema. J Allergy Clin Immunol 2011; 128: research. Br J Dermatol 2014; 170: 501-13. 353-9.

9 Ring J, Alomar A, Bieber T et al. Guidelines 17 Roekevisch E, Spuls PI, Kuester D et al. for treatment of atopic eczema (atopic Efficacy and safety of systemic treatments dermatitis) Part II. J Eur Acad Dermatol for moderate-to-severe atopic dermatitis: Venereol 2012; 26: 1176-93. a systematic review. J Allergy Clin Immunol 2014; 133: 429-38. 18 CHAPTER 1

18 Schmitt J, Schakel K, Folster-Holst R et al. 26 van der Schaft J, Politiek K, van den Reek Prednisolone vs. ciclosporin for severe JM et al. Drug survival for azathioprine and adult eczema. An investigator-initiated enteric-coated mycophenolate sodium double-blind placebo-controlled multi- in a long-term daily practice cohort of centre trial. Br J Dermatol 2010; 162: 661-8. adult patients with atopic dermatitis. Br J Dermatol 2016; 175: 199-202. 19 Garritsen FM, Roekevisch E, van der Schaft J et al. Ten years experience with oral 27 Politiek K, van der Schaft J, Coenraads PJ et immunosuppressive treatment in adult al. Drug survival for methotrexate in a daily patients with atopic dermatitis in two practice cohort of adult patients with severe academic centres. J Eur Acad Dermatol atopic dermatitis. Br J Dermatol 2016; 174: Venereol 2015; 29: 1905-12. 201-3.

20 Hegazy S, Tauber M, Bulai-Livideanu C et al. 28 Schmitt J, Schmitt NM, Kirch W et al. Systemic treatment of severe adult Atopic Outpatient care and medical treatment of dermatitis in clinical practice: analysis of children and adults with atopic eczema. J prescribing pattern in a cohort of 241 Dtsch Dermatol Ges 2009; 7: 345-51. patients. J Eur Acad Dermatol Venereol 2017. 29 Gluck T, Kiefmann B, Grohmann M et al. Immune status and risk for infection in 21 Vedie AL, Ezzedine K, Amazan E et al. patients receiving chronic immunosup- Long-term Use of Systemic Treatments for pressive therapy. J Rheumatol 2005; 32: Moderate-to-Severe Atopic Dermatitis in 1473-80. Adults: A Monocentric Retrospective Study. Acta Derm Venereol 2016; 96: 802-6. 30 Duncan RA, von Reyn CF, Alliegro GM et al. Idiopathic CD4+ T-lymphocytopenia--four 22 Totri CR, Eichenfield LF, Logan K et al. patients with opportunistic infections and Prescribing practices for systemic agents no evidence of HIV infection. N Engl J Med in the treatment of severe pediatric atopic 1993; 328: 393-8. dermatitis in the US and Canada: The PeDRA TREAT survey. J Am Acad Dermatol 31 Cunningham-Rundles C, Bodian C. 2017; 76: 281-5. Common variable immunodeficiency: clinical and immunological features of 248 23 Schmitt J, Abraham S, Trautmann F et al. patients. Clin Immunol 1999; 92: 34-48. Usage and effectiveness of systemic treatments in adults with severe atopic eczema: 32 Vogelin M, Biedermann L, Frei P et al. First results of the German Atopic Eczema The Impact of Azathioprine-Associated Registry TREATgermany. J Dtsch Dermatol Lymphopenia on the Onset of Ges 2017; 15: 49-59. Opportunistic Infections in Patients with Inflammatory Bowel Disease. PLoS One 24 van den Reek JM, Kievit W, Gniadecki 2016; 11: e0155218. R et al. Drug Survival Studies in Dermatology:Principles, Purposes, and 33 Al Rifai A, Prasad N, Shuttleworth E et al. Pitfalls. J Invest Dermatol 2015; 135: e34. Natural history of azathioprine-associated lymphopenia in inflammatory bowel 25 van der Schaft J, Politiek K, van den Reek disease patients: a prospective observa- JM et al. Drug survival for ciclosporin A tional study. Eur J Gastroenterol Hepatol in a long-term daily practice cohort of 2011; 23: 153-8. adult patients with atopic dermatitis. Br J Dermatol 2015; 172: 1621-7. 34 Hutchinson P, Chadban SJ, Atkins RC et al. Laboratory assessment of immune function in renal transplant patients. Nephrol Dial Transplant 2003; 18: 983-9. GENERAL INTRODUCTION 19

35 Gergely P. Drug-induced lymphopenia: 44 Lennard L. The clinical pharmacology of focus on CD4+ and CD8+ cells. Drug Saf 6-mercaptopurine. Eur J Clin Pharmacol 1999; 21: 91-100. 1992; 43: 329-39. 1

36 Wieland U, Kreuter A, Pfister H. Human 45 Cuffari C, Seidman EG, Latour S et al. papillomavirus and immunosuppression. Quantitation of 6-thioguanine in peripheral Curr Probl Dermatol 2014; 45: 154-65. blood leukocyte DNA in Crohn’s disease patients on maintenance 6-mercaptopurine 37 Murase JE, Heller MM, Butler DC. Safety of therapy. Can J Physiol Pharmacol 1996; 74: dermatologic medications in pregnancy 580-5. and lactation: Part I. Pregnancy. J Am Acad Dermatol 2014; 70: 401 e1-14; quiz 15. 46 Osterman MT, Kundu R, Lichtenstein GR et al. Association of 6-thioguanine nucleotide 38 Millsop JW, Heller MM, Eliason MJ et al. levels and inflammatory bowel disease Dermatological medication effects on male activity: a meta-analysis. Gastroenterology fertility. Dermatol Ther 2013; 26: 337-46. 2006; 130: 1047-53.

39 van der Schaft J, van Zuilen AD, Deinum J 47 Dubinsky MC, Lamothe S, Yang HY et et al. Serum creatinine levels during and al. Pharmacogenomics and metabolite after long-term treatment with cyclosporine measurement for 6-mercaptopurine A in patients with severe atopic dermatitis. therapy in inflammatory bowel disease. Acta Derm Venereol 2015; 95: 963-7. Gastroenterology 2000; 118: 705-13.

40 Bieber T, D’Erme AM, Akdis CA et al. 48 Smith MA, Blaker P, Marinaki AM et al. Clinical phenotypes and endopheno- Optimising outcome on thiopurines in types of atopic dermatitis: Where are we, inflammatory bowel disease by co-prescrip- and where should we go? J Allergy Clin tion of allopurinol. J Crohns Colitis 2012; 6: Immunol 2017; 139: S58-S64. 905-12.

41 Thijs JL, Strickland I, Bruijnzeel-Koomen C 49 Ansari A, Patel N, Sanderson J et al. et al. Moving toward endotypes in atopic Low-dose azathioprine or mercaptopurine dermatitis: Identification of patient clusters in combination with allopurinol can bypass based on serum biomarker analysis. J many adverse drug reactions in patients Allergy Clin Immunol 2017. with inflammatory bowel disease. Aliment Pharmacol Ther 2010; 31: 640-7. 42 Belaiche J, Desager JP, Horsmans Y et al. Therapeutic drug monitoring of azathio- 50 Kiszka-Kanowitz M, Theede K, Mertz- prine and 6-mercaptopurine metabolites in Nielsen A. Randomized clinical trial: a Crohn disease. Scand J Gastroenterol 2001; pilot study comparing efficacy of low-dose 36: 71-6. azathioprine and allopurinol to azathioprine on clinical outcomes in inflammatory 43 Thijs JL, Van Der Geest BAM, Van Der bowel disease. Scand J Gastroenterol 2016; Schaft J et al. Predicting therapy response 51: 1470-5. to mycophenolic acid using UGT1A9 genotyping: towards personalized medicine in 51 Meijer B, Mulder CJ, Peters GJ et al. Efficacy atopic dermatitis. J Dermatolog Treat 2017; of thioguanine treatment in inflammatory 28: 242-5. bowel disease: A systematic review. World J Gastroenterol 2016; 22: 9012-21. Part I Qualitative and quantitative exploration of oral immunosuppressive drug use in the Netherlands

Ten years experience with oral immunosuppressive treatment in adult patients with atopic dermatitis in two academic centres

Floor M. Garritsen, Evelien Roekevisch, Jorien van der Schaft, Joukje Deinum, Phyllis I. Spuls†, Marjolein S. de Bruin-Weller†

† These authors contributed equally to this study

J Eur Acad Dermatol Venereol 2015;29(10):1905-12 24 CHAPTER 2

Abstract Background: There is a lack of information on the use of oral immunosuppressive drugs in atopic dermatitis (AD) in daily practice. Objective: A 10-year overview of the use of oral immunosuppressive drugs in patients with severe AD. Methods: Medical charts of patients with AD, who received oral immunosuppressive drugs at the Academic Medical Center Amsterdam and in the University Medical Center Utrecht between January 2001 and January 2011, were analysed. Particular attention was paid to patient characteristics, prior treatment, prescribed oral immunosuppressive drugs, the order of use, doses, treatment durations and reasons for discontinuation of treatment. Results: 334 patients [53% male, mean age at start of an oral immunosuppressive drug 36.9 years (SD 13.6)] with AD received oral immunosuppressive treatment of which 102 (31%) participated in clinical trials. Cyclosporine A (CsA) was given in 80% of the patients, mycophenolate mofetil or enteric-coated mycophenolate sodium in 31%, azathioprine in 14%, methotrexate in 11%, systemic glucocorticosteroids in 7% and systemic tacrolimus in 5%. In these academic centres, CsA was the first choice oral immunosuppressive drug in 252 patients. Reasons for discontinuation of oral immunosuppressive drugs were controlled AD disease, ineffectiveness and adverse events. Conclusion: Various types of oral immunosuppressive drugs have been used over the past ten years for the treatment of severe AD with a prominent first choice for CsA. Adverse events and ineffectiveness were the most frequent reasons for discontinuation. A prospective database of patients using oral immunosuppressive treatments in daily practice will give more insight in the effectiveness and safety and may help to formulate future recommendations. TEN YEARS EXPERIENCE WITH ORAL IMMUNOSUPPRESSIVE DRUGS 25

Introduction Atopic dermatitis (AD) is a chronic inflammatory skin disease. The prevalence varies between 2-10% in adults and 15-30% in children.1,2 A combination of emollients, topical glucocorticosteroids, calcineurin inhibitors, antihistamines and antibiotics, is often used in the management of AD.3 Different forms of phototherapy may also be effective.4-6 2

In severe AD patients, with insufficient response to topical treatment or needing continuous treatment with mid to high potent corticosteroids, oral immunosuppressive therapy is indicated. Studies show that cyclosporine A (CsA), azathioprine (AZA), methotrexate (MTX), mycophenolate mofetil (MMF) and enteric-coated mycophenolate sodium (EC-MPS) are effective in the treatment of severe AD patients. Long term treatment with oral glucocorticosteroids is not recommended.7-11

Guidelines for the use of oral immunosuppressive drugs have been published and include recommendations to support daily practice in the management of patients with AD.7,11-18 These guidelines are mostly based on results from clinical trials, preferably randomized clinical trials (RCT). However, patients participating in clinical trials often differ from daily practice patients due to the strict in- and exclusion criteria. In addition, patients participating in clinical trials are in general more compliant, probably resulting in better clinical efficacy of the drug. Daily practice data concerning oral immunosuppressive drugs used in AD patients may help to give additional insight in the effectiveness and safety. Therefore a medical chart review was performed of all 334 AD patients treated with oral immunosuppressive drugs at the dermatology department of the Academic Medical Center Amsterdam (AMC) and University Medical Center Utrecht (UMCU) in the Netherlands over a period of ten years (from January 2001 till January 2011). The aim of this study was to give an overview of the use of oral immunosuppressive drugs in patients with severe AD, with respect to patient characteristics, prior treatment, prescribed oral immunosuppressive drugs, the order of use, doses, treatment durations and reasons for discontinuation of treatment. Both daily practice data and trial (follow up) data were evaluated in order to provide a complete overview of oral immunosuppressive drug use in AD patients.

Materials and methods Patient selection In both centres, AD patients were diagnosed by a dermatologist or validated diagnostic criteria.19 All patients using oral immunosuppressive drugs for AD between January 2001 and January 2011 were included. For all patients, follow up ended on January 1st 2011, regardless whether treatment was continued after this date. Patients who started an oral immunosuppressive drug before January 1st 2001 and continued this treatment after this date were included. Only patients, born before January 1st 1995, were investigated, excluding young children. In the AMC all patients diagnosed with AD in the digital registration system using oral immunosuppressive drugs were included. Additionally patients treated in a RCT performed between July 2009 and December 2010, comparing AZA with MTX were included.8 26 CHAPTER 2

In the UMCU, data of all adult patients diagnosed with AD and treated with oral immunosuppressive drugs from 1996 until now were retrospectively registered into a database. Patients who were treated with oral immunosuppressive drugs between January 2001 and January 2011 were selected. In addition, data from a RCT performed on EC-MPS versus CsA from November 2005 to November 2007 were included.9

Excluded were patients who received oral immunosuppressive drugs as a therapy for a disease other than AD. When patients used the same oral immunosuppressive treatment more than once, only the first treatment episode was evaluated. When patients used different oral immunosuppressive drugs simultaneously, the combination of both treatments was evaluated. In case of monotherapy with systemic glucocorticosteroids, only patients who were treated for a continuous period of more than three months were included.

Data-extraction Information about patient characteristics, prior oral immunosuppressive drug use and phototherapy were collected. Information about the order in which oral immunosuppressive drugs were prescribed, doses and durations of treatment and the reasons for discontinuation of treatment were collected. The manner of discontinuation (acute or stepwise) was also investigated.

Study outcome The primary outcome parameter of this study was the reason for discontinuation of treatment and was subdivided in i) controlled disease, ii) adverse events iii) ineffectiveness, iv) end of study, v) miscellaneous. Controlled disease was defined as improvement of AD (in the opinion of the patient and/or physician) which resulted in discontinuation of the oral immunosuppressive drug. Adverse events leading to discontinuation of treatment included both subjective and objective adverse events, such as laboratory abnormalities. Ineffectiveness was registered as reason for discontinuation when the patient and/or physician were not satisfied with the clinical response. Primary (not effective at all) and secondary (initially improvement but a loss of effect after a certain amount of time) ineffectiveness were not specified but the mean number on days on treatment were provided. End of the study was the reason for discontinuation in all patients participating in the RCT performed on EC-MPS versus CsA as part of the study protocol. Patients participating in the RCT comparing AZA and MTX did not have to discontinue treatment after the trial period. These patients continued treatment with AZA or MTX as in daily practice. Any other reasons for discontinuation were summarized under miscellaneous. More than one reason for discontinuation per patient could be recorded. This review followed the STROBE guidelines as much as possible.20

Statistical analysis SPSS version 20 for windows (SPSS , Inc, Chicago, IL, USA) was used to calculate means, medians, standard deviations and interquartile ranges. TEN YEARS EXPERIENCE WITH ORAL IMMUNOSUPPRESSIVE DRUGS 27

Results Patient characteristics A total of 334 patients were included (85 from the AMC and 249 from the UMCU). Forty-two (13%) of them participated in the AMC RCT and 60 patients (18%) participated in the UMCU RCT. 8,9 Patient characteristics are shown in table 1. 2

Table 1 Patient characteristics N (%) (TOTAL = 334) Male sex 177 (53) Mean age 36.89 years (SD 13.64) Presence of asthma* 180 (54) Presence of allergic rhinitis* 211 (63) Presence of food allergy* 146 (44) Serum IgE determined 217 (65) Serum IgE increased (>100 kU/L) 204 (61) Mean number of previous oral immunosuppressive treatment per patient 0.60 (SD 0.86) before referral to academic centres Patients who used more than one type of oral immunosuppressive treat- 114 (34) ment in the ten years period

*=anamnestic

Prior treatment The mean number of oral immunosuppressive therapies per patient, used before referral to our academic centres, is 0.60 (SD 0.86) (table 1). CsA was most frequently used (98 times, 37% of all patients) (table 2), a requirement for starting in the MTX vs AZA RCT.8

Table 2 An overview of prior immunosuppressive or UV interventions Therapy N (%) (TOTAL = 262) Cyclosporine A 98 (37) Systemic glucocorticosteroids* 68 (26) Azathioprine 7 (3) Mycophenolate mofetil / enteric-coated mycophenolate sodium 4 (2) Efalizumab 1 (0) Methotrexate 2 (1) Neotigason 1 (0) UVB** 50 (19) PUVA** 18 (7) UVA** 13 (5)

UVB ultraviolet B; UVA ultraviolet A; PUVA psoralen plus UVA * Data include short additional courses of a few days ** Data were only collected in the AMC 28 Table 3 An overview of the treatment characteristics, reasons for discontinuation and the median duration (days) of treatment until discontinuation. CHAPTER 2

st 2011 (%) st Mode of discontinuation (%) for discontinuation Reasons (days) ofand median duration until discontinuation treatment (IQR) Therapy patients Total Male (%) 1 January before Started 2001 and continuing after this 2001 and continuing after (%) date (SD) in years start at age Mean dose (SD) start Mean maximum dose (SD) Mean in ofMedian duration treatment days (IQR) 1 January after Continuing 2011 (%) January Discontinuing before 1

Cyclosporine A 267 146 (55) 18/267 (7) 35.50 4.23 mg/kg/day 4.27 mg/kg/day 258 (108-434) 33/267 234/267 Acute: 108 Controlled disease: 69/234 (29) 294 (184-486) (12.90) (0.95) (0.91) (12) (88) Stepwise: 94 Adverse events: 57/234 (24) 217 (100-512) Unknown: 32 Not effective: 35/234 (15) 244 (162-512) End of trial: 37/234 (16) 50 (43-267) Miscellaneous: 193 377 (241-814) Mycophenolate 104 63(61) 0 42.10 MMF: 1050 mg/ MMF: 1650 mg/ MMF: 84 (31.25- 22/104 82/104 Acute: 64 Controlled disease: 11/82 (13) mofetile / (12.17) day (437.80) day (474.34) 151.75) and (21) (79) Stepwise: 11 Adverse events: 18/82 (22) 177 (70-222) enteric-coated EC-MPS: EC-MPS: EC-MPS: 212 Unknown: 7 Not effective: 36/82 (44) 114 (80-283) mycophenolate 1371.06 (275.72) 1371.06 (113.75-380.75) End of trial: 23/82 (28) 217 (207-222) sodium (275.72) Miscellaneous: 35 Azathioprine 46 22 (48) 0 40.24 99.77 mg/day 121.56 mg/day 208 20/46 26/46 Acute: 14 Controlled disease: 5/26 (19) 270 (194-366) (14.03) (43.91) (64.09) (27.25-340.25) (43) (57) Stepwise: 2 Adverse events: 10/26 (38) 27 (15-128) Unknown: 10 Not effective: 4/26 (15) 146 (27-214) Miscellaneous: 12

Methotrexate 37 19 (51) 0 43.98 8.75 mg/week 20.90 mg/week 298 (147-384.5) 20/37 17/37 Acute: 11 Controlled disease: 1/17(6) 261 (14.42) (2.57) (18.50) (54) (46) Stepwise: 1 Adverse events: 7/17 (41) 96 (11-204) Unknown: 5 Not effective: 11/17 (65) 175 (11-204) Miscellaneous: 3

Systemic 138 (104-409) 24 14 (58) 3/24 (13) Prednisone: Prednisone: Prednisone: Prednisone: 266 4/24 20/24 Acute: 0 Controlled disease: 4/20 (20) 2930 glucocorti- 44.65 22.50 mg/day 23.00 mg/day (129.25-492) and (17) (83) Stepwise: 16 Adverse events: 1/20 (5) 323 (118-1574) costeroids (18.78); (9.93) (9.51); Celestone: 1363 Unknown: 4 Not effective: 4/20 (20) Celestone: Celestone: 1.83 Celestone: (258.75-2688.5) Miscellaneous: 31 36.40 (6.32) mg/day (0.76) 1.50mg/day (0.5) Tacrolimus 18 13 (72) 1 /18 (6) 38.67 (SD 6.53 mg/day (SD 6.53 mg/day 106.5 (59-556.5) 2/18 16/18 Acute: 15 Controlled disease: 1/16 (6) 223 11.97) 1.81) (SD 1.81) (11) (89) Stepwise: 1 Adverse events: 9/16 (56) 77 (33-586) Unknown: 0 Not effective: 7/16 (44) 91 (77-294) Miscellaneous:9 TEN YEARS EXPERIENCE WITH ORAL IMMUNOSUPPRESSIVE DRUGS 29

Oral immunosuppressive drugs More than one type of oral immunosuppressive drugs was used by 114 patients (34%) between January 2001 and January 2011. Table 3 shows an overview of the treatment characteristics, reasons for discontinuation and median duration (days) of treatment until discontinuation. In table 4 the adverse events that caused discontinuation of the different drugs are described in 2 more detail.

Cyclosporine A In total, 267 patients (80%) were treated with CsA. Thirty-seven patients (14%) received CsA in the UMCU RCT.9 Fifty-seven patients had more than one treatment episode with CsA, of which only the first episode was evaluated. The median duration of treatment till January st1 2011 was 258 days (IQR 108-434). Of the 234 patients that discontinued treatment, 69 patients (29%) stopped because of controlled disease, 57 patients (24%) discontinued treatment because of adverse events. In 35 patients ineffectiveness (15%) was the reason for discontinuation of therapy. Thirty-seven patients (16%) discontinued because of the end of the trial in which they participated. One hundred and ninety-three other miscellaneous reasons caused discontinuation of treatment. Six patients used >3 months of prednisone simultaneously.

Mycophenolate mofetil and enteric-coated mycophenolate sodium In total, 104 (31%) patients were treated with MMF and EC-MPS. Twenty-three patients (22%) received EC-MPS in the RCT that was performed in the UMCU.9 Four patients had more than one treatment episode with EC-MPS, of which only the first episode was evaluated. The median duration of treatment was 84 days (IQR 31.25-151.75) (MMF) and 212 days (IQR 113.75-380.75) (EC-MPS). Among the 82 patients (79%) who discontinued treatment before January 1st 2011, 11 patients (13%) stopped because of controlled disease, 18 patients (22%) because of adverse events. In 36 patients (44%) MMF or EC-MPS was ineffective. Twenty-three patients (28%) discontinued because of the end of the trial in which they participated. Thirty-five other miscellaneous reasons caused discontinuation of treatment. Twenty-eight patients used >3 months of prednisone simultaneously.

Azathioprine In total, 46 patients (14%) were treated with AZA. Twenty-three patients (50%) received AZA as part of the AMC RCT.8 One patient had more than one treatment episode with AZA of which only the first episode was evaluated. In our 10-year observation period, the median duration of treatment was 208 days (IQR 27.25-340.25). Twenty patients (43%) continued treatment with AZA after January 1st 2011. Of the 26 patients who stopped treatment, five patients (19%) stopped because of controlled disease, ten patients (38%) due to adverse events. In 4 patients (15%) AZA was ineffective. Twelve other miscellaneous reasons caused discontinuation of treatment. Twelve patients used >3 months of prednisone simultaneously. 30 CHAPTER 2

Table 4 Adverse events that caused discontinuation of therapy Therapy Adverse events that caused discontinuation of therapy Cyclosporine A Serum creatinine increase 24/234 (10%) Hypertension 17/234 (7%) Neurological symptoms* 14/234 (6%) Fatigue 9/234 (4%) Gastrointestinal symptoms 9/234 (4%) Myalgia 7/234 (3%) Pitting oedema 5/234 (2%) Flu like symptoms 4/234 (2%) Hypertrichosis 2/234 (1%) Physical condition loss 2/234 (1%) Gum hyperplasia 2/234 (1%) Rushed feeling / mood changes 2/234 (1%) Recurrent viral infection with herpes simplex 1/234 (0%) Flushing 1/234 (0%) Petechia 1/234 (0%) Joint problems 1/234 (0%) Anaemia 1/234 (0%) Mycophenolate mofetile/ enteric-coated Neurological symptoms* 8/82 (10%) mycophenolate sodium Gastro-intestinal symptoms 5/82 (6%) Flu like symptoms 3/82 (4%) Shortness of breath 3/82 (4%) Myalgia 2/82 (2%) Fatigue 1/82 (1%) Oedema 1/82 (1%) Hair loss 1/82 (1%) Azathioprine Nausea 7/26 (27%) Neurological symptoms* 2/26 (8%) Liver dysfunction 2/26 (8%) Fatigue 1/26 (4%) Flu like symptoms 1/26 (4%) Anaemia 1/26 (4%) Lymphocytopenia 1/26 (4%) Pancytopenia 1/26 (4%) Methotrexate Neurological symptoms* 3/17 (18%) Nausea 2/17 (12%) Fatigue 2/17 (12%) Transaminase increase 1/17 (6%) Systemic corticosteroids Unknown 1/20 (5%) Tacrolimus Gastro-intestinal symptoms 3/16 (19%) Serum creatinine increase 3/16 (19%) Neurological problems* 2/16 (13%) Hypertension 1/16 (6%) * Including trembling of the hands, visual impairments, tingling of the fingers, jactation, vertiginous, feeling weak and heaviness of the arms TEN YEARS EXPERIENCE WITH ORAL IMMUNOSUPPRESSIVE DRUGS 31

Methotrexate Of the 37 patients (11%) who received MTX, 23 patients (62%) started with this therapy as part the AMC RCT.8 No patients had more than one treatment episode with MTX. In our 10-year observation period, the median duration of treatment was 298 days (IQR 147-384.5). Twenty patients (54%) continued treatment with MTX after January 1st 2011. Of the seventeen patients 2 who stopped treatment, one patient (6%) stopped because of controlled disease. Seven (41%) discontinued because of adverse events. Eleven patients discontinued treatment because of ineffectiveness (65%). Three other miscellaneous reasons caused discontinuation of treatment. Six patients used >3 months of prednisone simultaneously.

Systemic glucocorticosteroids monotherapy Twenty-four patients (7%) were treated with a monotherapy of systemic glucocorticosteroids, of which twenty patients (83%) used prednisone and four patients (17%) used celestone. Four patients had more than one treatment episode with systemic glucocorticosteroids, of which only the first episode was evaluated. All courses lasted longer than 3 months (inclusion crite- rium), with a median duration of treatment with prednisone of 266 days (IQR 129.25-492) and 1363 days (IQR 258.75-2688.5) for celestone. The mean starting dose was 22.50 mg/day (SD 9.93) for prednisone and 1.83 mg/day (SD 0.76) for Celestone. The mean maximum dose was 23.00 mg/day (SD 9.51) for prednisone and 1.50 mg/day (SD 0.5) for celestone. Twenty patients (83%) discontinued treatment, of which four patients (20%) discontinued treatment because of controlled disease. One patient discontinued treatment with celestone because of unknown adverse events. In four patients (20%) systemic glucocorticosteroids were ineffective. Thirty-one other miscellaneous reasons caused discontinuation of treatment.

Tacrolimus Eighteen patients (5%) were treated with tacrolimus, with a median duration of treatment of 107 days (IQR 59.0-556.5). One patient had more than one treatment episode with tacrolimus, of which only the first episode was evaluated. Of the sixteen patients (89%) who discontinued treatment, one patient (6%) discontinued because of controlled disease. Nine patients (56%) discontinued treatment because of adverse events. In seven patients (44%), the treatment was not effective (table 3). Nine other miscellaneous reasons caused discontinuation of treatment.

Order of use The order of use of the different oral immunosuppressive drugs in the included patients is shown in table 5. These data include only oral immunosuppressive drugs that were given to the patients in our centres and do not include previous oral immunosuppressive treatments, prescribed by dermatologists of peripheral hospitals or by general practitioners. First choice treatment was CsA, given in 252 patients. The absolute number of patients receiving AZA, MTX or EC-MPS in first instance were comparable. Most of them participated in the earlier mentioned clinical trials. 32 CHAPTER 2

Table 5 Order of use in oral immunosuppressive treatments in steps Therapy First Second Third Fourth step step step step Total Cyclosporine A 252 13 2 - 267 Systemic corticosteroids 16 4 3 1 24 Mycophenolate mofetil - 6 3 1 10 Azathioprine 22 8 13 3 46 Methotexate 21 9 6 1 37 Enteric-coated mycophenolate sodium 19 66 9 - 94 Tacrolimus 4 9 3 2 18

Discussion The present study provides an overview of oral immunosuppressive drugs used in the treatment of adult AD patients in two academic centres in a time period of 10 years. These data reflect the situation in the Netherlands and may be different from other European countries. CsA was the most commonly used oral immunosuppressive drug, followed by MMF/EC-MPS, AZA, MTX, systemic glucocorticosteroids and tacrolimus. Although long term treatment with oral glucocorticosteroids is not recommended in guidelines, 24 patients were treated with prednisone for more than 3 months. These patients were monitored well and only one patient had to discontinue due to unknown side effects. The mean amount of 0.60 previous oral immunosuppressive treatments per patient indicates that the use of oral immunosuppressive drugs is not limited to academic centres. Although the major part of the patients used oral immunosuppressive drugs in a daily practice setting, about 30% of the patients participated in one of the earlier mentioned trials. These patients may have a different profile due to the inclusion and exclusion criteria of the trials. The aim of the present study was to provide a complete overview of the oral immunosuppressive drugs that were prescribed in AD patients over a period of ten years. It was not the intention to compare the efficacy of the different oral immunosuppressive drugs, as reliable efficacy measures, such as clinical scores were not available in all patients (only in trial patients) and bias could influence the data. Therefore we chose the reason for discontinuation of the oral immunosuppressive drug as primary outcome measure. In our opinion this is the most reliable measure in retrospective studies. Still, this outcome has its shortcomings, because no predefined definitions for controlled disease and inefficacy were used. Discontinuation of treatment may have been influenced by different factors like local - differ ences in treatment strategies and physician’s experience. Guidelines and insights have changed during the 10-year evaluation period. For instance in the first part of the evaluation period only CsA and oral steroids were described for severe AD, while the prescription of second line drugs like MTX, AZA and MMF/EC-MPS started in the last part of the evaluation period. This may partly explain the high number of finished CsA treatment episodes, while relatively more patients on second line drugs are still under treatment at the moment of datalock. TEN YEARS EXPERIENCE WITH ORAL IMMUNOSUPPRESSIVE DRUGS 33

Adverse events were reported often as a reason for discontinuation of treatment. The lowest percentages were seen in patients treated with systemic glucocorticosteroids (5%), MMF/ EC-MPS (22%) and CsA (24%). In CsA-treated patients objective adverse events (serum creatinine rise and hypertension) were the most common reason for discontinuation, while subjective side effects as a reason for discontinuation were more prominent in the other treatment modal- 2 ities. A relatively high percentage of patients discontinued treatment with AZA (38%), MTX (41%) and tacrolimus (56%) due to subjective adverse events, especially gastro-intestinal (AZA and MTX) and neurological (tacrolimus) side effects. Although these side effects are reported in clinical trials, the percentage of patients that discontinued treatment in this daily practice population is higher compared to published clinical trials.8-10 The most logical explanation is the difference in patients selection and probably compliance differs between trials and daily practice.

Although AD is a chronic disease, it is in general not the intention to use oral immunosuppressive drugs for many years, despite good clinical efficacy. This is in contrast to the treatment with biologics in for instance psoriasis patients where continuous treatment for many years in stable disease is more common. Therefore we also focussed on adequate disease control (controlled disease) as a reason for discontinuation of treatment.

Controlled disease was the reason for discontinuation in relatively few patients. The highest number was seen in patients treated with CsA (29%), followed by systemic glucocorticosteroids (20%) and AZA (19%). For MMF/EC-MPS, MTX and tacrolimus the percentages of discontinuation due to controlled disease were 13%, 6% and 6% respectively.

It is important to realize that discontinuation due to controlled disease is not the same as clinical effectiveness, as patients who were still under treatment at the moment of data analysis, probably have a good response to treatment. In the future, these patients may also discontinue their treatment because of controlled AD disease.

In the present study, patients with multiple comorbidities and difficult treatable AD are included. This in contrast to RCTs were guidelines are based on. Due to the strict inclusion criteria, patients participating in RCTs have less comorbidities and are often more motivated and compliant. A recent study of anti-TNF-a in psoriasis patients, showed that patients who received prior anti- TNF-a in a clinical trial setting, had a better drug survival rate afterwards when treated in a daily practice setting compared to patients who had not participated in clinical trials before. Possible explanations could be that former trial patients may be healthier and have fewer comorbidities, or show a better adherence to treatment due to positive experiences in a previous trial.21 Another important difference between clinical trials and daily practice is the duration of treatment: in clinical trials treatments are often limited to 3-6 months, while treatment periods in daily practice are much longer. 34 CHAPTER 2

There is a need for more practical guidelines in the treatment of patients with AD with oral immunosuppressive drugs. A recent published systematic review provided an evidence-based treatment algorithm for patients with moderate-to-severe AD and the recent AAD guideline gives an overview of the existing literature.7,11 In children, the TREAT survey gives an indication on which systemic immunomodulating drugs are used.22 Data from the present study provide practical information on dosing, order of use and reasons for discontinuation of oral immunosuppressive drugs in adult AD patients, which is valuable for future recommendations in guidelines.

Limitations of this study In this retrospective study, the quality of the included data depended on the completeness of medical records. Some patients used prednisolone >3 months simultaneously with other oral immunosuppressive drugs. It was not possible to evaluate these treatments separately. The use of oral steroids was supportive. Therefore the data of these combinations were described in the evaluation of the main compound. In the result section of the compound, long term treatment with prednisone is mentioned for each oral immunosuppressive drug.

In order to provide an extensive complete overview of oral immunosuppressive drugs that were prescribed in the last 10 years, both daily practice and RCT data were included. We were encouraged to do so because the trials were set up as much as possible according to daily standard care. Nevertheless, it should be taken into account that the data of trial patients are influenced by the trial process. Table 5 for instance gives an overview of the order of use of oral immunosuppressive drugs in our centres, but it should be taken into consideration that the order has been influenced by the randomisation process that was applied in the AMC and UMCU RCTs. Finally, we are aware of the fact that these data may not be generalizable for the Netherlands and may show differences with other European academic AD centres.

Recommendations Besides clinical trial data, observational daily practice data should be taken into account in updating existing guidelines and formulating recommendations. A prospective database of adults and children with AD using oral immunosuppressive drugs in daily practice, including validated and standardized diagnostic criteria, outcome domains and measurements, could be of great value for further recommendations and is currently being developed. TEN YEARS EXPERIENCE WITH ORAL IMMUNOSUPPRESSIVE DRUGS 35

References

1 Asher MI, Montefort S, Bjorksten B et al. 10 Schmitt J, Schakel K, Folster-Holst R et al. Worldwide time trends in the prevalence Prednisolone vs. ciclosporin for severe of symptoms of asthma, allergic rhinocon- adult eczema. An investigator-initiated junctivitis, and eczema in childhood: ISAAC double-blind placebo-controlled multi- Phases One and Three repeat multicountry centre trial. Br J Dermatol 2010; 162: 661-8. 2 cross-sectional surveys. Lancet 2006; 368: 733-43. 11 Roekevisch E, Spuls PI, Kuester D et al. Efficacy and safety of systemic treatments 2 Bieber T. Atopic dermatitis. Ann Dermatol for moderate-to-severe atopic dermatitis: 2010; 22: 125-37. a systematic review. J Allergy Clin Immunol 2014; 133: 429-38. 3 Ring J, Alomar A, Bieber T et al. Guidelines for treatment of atopic eczema (atopic 12 Hanifin JM, Cooper KD, Ho VC et al. dermatitis) part I. J Eur Acad Dermatol Guidelines of care for atopic derma- Venereol 2012; 26: 1045-60. titis, developed in accordance with the American Academy of Dermatology (AAD)/ 4 Gambichler T. Management of atopic American Academy of Dermatology dermatitis using photo(chemo)therapy. Association “Administrative Regulations Arch Dermatol Res 2009; 301: 197-203. for Evidence-Based Clinical Practice Guidelines”. J Am Acad Dermatol 2004; 50: 5 Meduri NB, Vandergriff T, Rasmussen H 391-404. et al. Phototherapy in the management of atopic dermatitis: a systematic review. 13 Hoare C, Li Wan Po A, Williams H. Photodermatol Photoimmunol Photomed Systematic review of treatments for atopic 2007; 23: 106-12. eczema. Health Technol Assess 2000; 4: 1-191. 6 Garritsen FM, Brouwer MW, Limpens J et al. Photo(chemo)therapy in the management 14 Schmitt J, Schakel K, Schmitt N et al. of atopic dermatitis: an updated systematic Systemic treatment of severe atopic review with implications for practice and eczema: a systematic review. Acta Derm research. Br J Dermatol 2014; 170: 501-13. Venereol 2007; 87: 100-11.

7 Sidbury R, Davis DM, Cohen DE et al. 15 Saeki H, Furue M, Furukawa F et al. Guidelines of care for the management of Guidelines for management of atopic atopic dermatitis: section 3. Management dermatitis. J Dermatol 2009; 36: 563-77. and treatment with phototherapy and systemic agents. J Am Acad Dermatol 16 Williams HC. Clinical practice. Atopic 2014; 71: 327-49. dermatitis. N Engl J Med 2005; 352: 2314-24. 8 Schram ME, Roekevisch E, Leeflang MM et al. A randomized trial of methotrexate 17 Dutch Guidelines for the treatment versus azathioprine for severe atopic of atopic dermatitis. Dutch Society of eczema. J Allergy Clin Immunol 2011; 128: Dermatology and Venereology. Quality 353-9. institute for healthcare CBO. Utrecht.

9 Haeck IM, Knol MJ, Ten Berge O et al. 18 Ring J, Alomar A, Bieber T et al. Guidelines Enteric-coated mycophenolate sodium for treatment of atopic eczema (atopic versus cyclosporin A as long-term treat- dermatitis) Part II. J Eur Acad Dermatol ment in adult patients with severe atopic Venereol 2012; 26: 1176-93. dermatitis: a randomized controlled trial. J Am Acad Dermatol 2011; 64: 1074-84. 36 CHAPTER 2

19 Brenninkmeijer EE, Schram ME, Leeflang MM et al. Diagnostic criteria for atopic dermatitis: a systematic review. Br J Dermatol 2008; 158: 754-65.

20 von Elm E, Altman DG, Egger M et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol 2008; 61: 344-9.

21 van den Reek JM, van Lumig PP, Driessen RJ et al. Determinants of drug survival for etanercept in a long-term daily practice cohort of patients with psoriasis. Br J Dermatol 2014; 170: 415-24.

22 Proudfoot LE, Powell AM, Ayis S et al. The European TREatment of severe Atopic eczema in children Taskforce (TREAT) survey. Br J Dermatol 2013; 169: 901-9. TEN YEARS EXPERIENCE WITH ORAL IMMUNOSUPPRESSIVE DRUGS 37

Use of oral immunosuppressive drugs in the treatment of atopic dermatitis in the Netherlands

Floor M. Garritsen†, J. Maurik van den Heuvel†, Carla A.F.M. Bruijnzeel-Koomen, Anke H. Maitland-van der Zee, Marcel P.H. van den Broek, Marjolein S. de Bruin-Weller

† These authors contributed equally to this study

Manuscript submitted This manuscript was financially supported by Sanofi 40 CHAPTER 3

Abstract Introduction: Although atopic dermatitis (AD) is a very common skin disease, data on the percentage of patients with really difficult to treat AD are scarce. From socio-economic perspective it is important to have more insight in these numbers, as new very effective, but expensive, treatment options will be available in the near future for difficult to treat AD. Estimating the number of AD patients using oral immunosuppressive drugs can give an impression of the percentage of difficult to treat patients in the total AD population. Aim: To give an overview of the use of oral immunosuppressive drugs in patients with AD in the Netherlands. Methods: Prescription data of oral immunosuppressive drugs in the Netherlands were extracted from a pharmaceutical database (NControl) containing data of 557 million prescriptions and 7.2 million patients. An algorithm, based on the WHO Anatomical Therapeutic Chemical (ATC) codes, was used to identify patients with AD. The prescription of oral immunosuppressive drugs in patients with AD between January 1st 2012 and January 1st 2017 was evaluated. Results: Based on the algorithm, 65 943 patients with AD were selected. 943 AD patients (1.4%) used cyclosporine A, methotrexate, azathioprine or mycophenolic acid. Methotrexate was most commonly used, followed by azathioprine and cyclosporine A. A switch in medication was rarely seen. In the evaluation period a decrease in the prescription of cyclosporine A was seen, together with an increase of the prescription of methotrexate. In 31% of the patients who stopped treatment, the discontinuation took place within the first months of treatment. Conclusion: In this study population, 1.4% of the AD patients used oral immunosuppressive drugs for their eczema in a five year period. Methotrexate was the most commonly used systemic drug in the Netherlands for the treatment of AD. ORAL IMMUNOSUPPRESSIVE DRUG USE IN THE NETHERLANDS 41

Introduction Atopic dermatitis (AD) is a chronic, itchy, inflammatory skin disease. In most patients, treatment consists of emollients in combination with topical glucocorticosteroids and/or calcineurin inhibitors. In patients with moderate to severe AD, photo(chemo)therapy may also be effective. Oral immunosuppressive drugs may be indicated when optimal treatment with topical regimens does not adequately control the signs and symptoms of disease.1

Although AD is a very common skin disease, data on the percentage of patients with difficult to 3 treat AD are scarce. From a socio-economic perspective it is very important to have more insight in the number of difficult to treat AD patients, as new very effective, but expensive, treatment options will be available in the near future for these patients, such as biologics. Since patients with severe and difficult to treat AD are commonly treated with oral immunosuppressive drugs, analyzing the number of AD patients using these drugs can give information on the percentage of these patients in the total AD population.

Schmitt et al made an effort to investigate the number of difficult to treat AD patients. They found that about 10% of the AD patients in Germany was treated with systemic glucocorticosteroids and that these patients were supposed to have moderate to severe eczema.2 Because this study probably also evaluated short treatment courses of prednisone, this percentage may be an overestimation. Data on (maintenance) treatment with different oral immunosuppressive drugs will give a better indication of the number of difficult to treat severe AD patients, but these data are lacking in literature.

Drugs that are commonly used in AD are cyclosporine A (CsA), methotrexate (MTX), mycophenolic acid (MPA) and azathioprine (AZA).3,4 CsA is the only registered drug for AD in the Netherlands and most parts of Europe, but the other drugs are often prescribed off label, in the absence of other therapeutic options. Current guidelines do not provide suggestions for the order of use of the different oral immunosuppressive drugs. Therefore, the prescribing mode will vary between countries and individual physicians, based on preferences, experiences, costs and the drug availability. Nowadays, it remains unclear how often oral immunosuppressive drugs are prescribed for AD patients in the Netherlands and which drugs are the most popular.

In 2014, we evaluated 334 AD patients who had been treated with oral immunosuppressive drugs in two academic centres in the Netherlands with a special focus on severe, difficult to treat AD.5 CsA was prescribed in 80%, MPA in 31%, AZA in 14%, MTX in 11%, systemic glucocorticosteroids in 7% and systemic tacrolimus in 5% of the patients. In these academic centres, CsA was oral immunosuppressive drug of first choice in 252 patients. The results of this Dutch study reflect a selected group of academic patients and therefore may not be applicable to the general population. Besides, preferences for the use of different drugs may have changed, since several studies showing positive effects of other oral immunosuppressive drugs than CsA have been published in the past. 42 CHAPTER 3

Therefore, the aim of this study is to give an up to date overview of the use of oral immunosuppressive drugs in patients with atopic dermatitis throughout the Netherlands, by evaluating data from a nation-wide pharmacy prescription database.

Methods Pharmacy database NControl Data on the prescription of oral immunosuppressive drugs in the Netherlands were extracted from the NControl database. This database includes general practitioner or specialist prescribed healthcare products dispensed by community pharmacies throughout the Netherlands. The database also contains a small number (<10) outpatient pharmacies located in hospitals. Since 2011, the NControl database contains data related to over 557 million prescriptions and 7.2 million patients. Patients in the NControl database cannot be identified, but can be tracked over time across pharmacies. Prescribers are anonymized and cannot be identified nor tracked over time. NControl is allowed to use these prescription data for research purposes. NControl adheres to data protection and privacy regulations, as established in amongst others the Personal Data Protection Act in the Netherlands as well as the Nederlandse Norm (NEN) 7510 standard, related to information protection in healthcare, which is derived from the International Organisation for Standardization (ISO) 27001 and 27002.

The following prescription information was used from the database: the dispensed medication, dispensing date and prescriber type.

Patient selection Patients were selected from the NControl database between January 1st 2012 and January 1st 2017. No information on disease diagnosis is registered in the NControl database. Therefore an algorithm, based on the WHO Anatomical Therapeutic Chemical (ATC) codes, was used to identify patients with AD (table 1). This algorithm included drugs dispensing that were supposed to be used by patients with AD, like topical glucocorticosteroids, topical tacrolimus or pimecrolimus, emollients and oral antihistamines. To select atopic patients, and to exclude patients with other types of inflammatory skin diseases like psoriasis, the presence of a prescription for medication for allergic rhinitis and/or allergic asthma was mandatory. Besides, patients with dispensings of anti-psoriatic drugs like calcipotriol, and patients using DMARDS (suggesting rheumatological related skin diseases) were excluded from the selection.

From the group of AD patients, all patients using CsA, MTX, AZA and/or MPA were included. Because oral tacrolimus is commonly used by organ transplant recipients, patients using oral tacrolimus were excluded from the patient selection. All patients with at least one of the prescriptions made by a rheumatologist were excluded, to limit the inclusion of patients with rheumatic diseases. ORAL IMMUNOSUPPRESSIVE DRUG USE IN THE NETHERLANDS 43

Outcome Prescriptions between January 1st 2012 and January 1st 2017 were evaluated.

Current treatment Current treatment at the end of the evaluation period (December 31st 2016) was investigated and was defined as the most recent known treatment for active patients. An active patient was defined as a patient with a (final) dispensing during the 182 days (half a year) before January 1st 2017. 3

Prescription type From all patients who received oral immunosuppressive drugs after January 1st 2012, the prescription types were analyzed. Prescriptions were divided into ‘new prescription’ (i.e. the first prescription of an oral immunosuppressive drug in a patient), ‘change of treatment’ or ‘repeated prescription’.

Trends in prescribed treatments Trends in prescribing treatments, including both monotherapy and combinations of two oral immunosuppressive drugs, within the study period were evaluated.

Evaluation of treatment discontinuation In all patients, treatment discontinuation was evaluated. Discontinuation was defined as when the final prescription was dispensed more than 182 days before January st1 2017. In all patients that discontinued their medication, the time between the first and final prescription of oral immunosuppressive drugs was analyzed, in order to estimate the duration of treatment before the discontinuation.

Calculation of the total number of AD patients using oral immunosuppressive drugs The number of AD patients identified with the algorithm in this study is a sample of the total number of patients with AD with actual treatment. In order to estimate the total number of AD patients in the Netherlands treated with oral immunosuppressive drugs, data need to be corrected for the total amount of Dutch pharmacies. In January 2017, 1994 public pharmacies were registered in the Netherlands.

Besides, in the used algorithm for defining AD patients, the use of medication for allergic rhinitis and/or allergic asthma was mandatory to ensure the atopic factor. In recent large clinical trials in a comparable group of AD patients around two third of the patients with AD has allergic asthma and/or allergic rhinitis.6,7 This information was used to estimate the actual number of AD patients with actual treatment. 44 CHAPTER 3

Analytical tools For analysis and reporting Microsoft SQL Server Management Studio 2014 and Excel 2013 were used.

Results Patient characteristics Information on 556 pharmacies throughout the Netherlands was complete for the studied period. A total of 65 943 patients was supposed to have AD, based on the patient selection algorithm. From this group, 943 patients (1.4%) used CsA, MTX, AZA or MPA in this five year period (table 1).

Table 1 The patient selection algorithm used to define patients with atopic dermatitis Steps Patients (n) Step 1 373 507 Selecting patients with - at least two dispensings of topical glucocorticosteroid preparations (ATC code D07) in a one year period in combination with at least one dispensing of emollients (ATC code D02) or oral antihistamines (ATC code R06) or - at least one dispensing of topical tacrolimus (ATC code D11AH01) or topical pimecrolimus (ATC code D11AH02), both in combination with at least one dispensing of emollients (ATC code D02) or oral antihistamines (ATC code R06) within a month before or a year after the first D11AHA01 or D11AH02 prescription Step 2 70 336 - Selecting patients with ≥ 3 dispensings of asthma medication (ATC code R03) or ≥ 3 dispensings of nasal corticosteorids (ATC code R01AD) within two years of the first topical prescription of step 1 Step 3 66 146 - Selecting patients without any dispensing of anti-psoriatic drugs (ATC code D05) during the study period Step 4 65 943 - Selecting only patients that did not receive DMARDS: hydroxychloroquine (P01BA02), Sulfasalazine (A07EC01), Leflunomide (L04AA13) and Tumor necrosis factor alpha inhibitors (L04AB)

Step 5 1261 - Selecting patients receiving cyclosporine A (L04AD01), azathioprine (L04AX01), methotrexate (L04AX03) or mycofenolic acid (L04AA06) within the complete study period, but did not receive oral tracrolimus (L04AD02) Step 6 943 - Selecting patients who did not receive an oral immunosuppressive drug prescription made by a rheumatologist

Drug dispensing is coded according to the WHO Anatomical Therapeutic Chemical (ATC) Classification System. ORAL IMMUNOSUPPRESSIVE DRUG USE IN THE NETHERLANDS 45

Current treatment At the end of the evaluation period, on December 31st 2016, 411 patients were defined as active patients on oral immunosuppressive drugs (figure 1). Most of them used MTX (n=158, 38%), followed by AZA (n=156, 38%). Seventy patients (17%) were on CsA treatment and 25 patients (6%) were on MPA treatment. A combination of CsA and AZA and a combination of CsA and MPA were both seen in 1 patient (0%).

MYCOFENOLIC ACID CYCLOSPORINE+ 6% MYCOFENOLIC ACID 0% 3 AZATHIOPRINE+ CYCLOSPORINE 0% METHOTREXATE 38%

AZATHIOPRINE 38%

CYCLOSPORINE 17%

Figure 1 Prevalence of treatments among active patients (n=411)

Types of prescription Types of prescriptions were evaluated from all patients who started oral immunosuppressive drugs after January 1st 2012 (figure 2). Most prescriptions were repeat prescriptions. A change of treatment was occasionally observed.

Change of treatment First prescription 2% 11%

Repeat prescripton 87%

Figure 2 Types of prescriptions 46 CHAPTER 3

Trends in registered treatments Throughout the study period, MTX was the most commonly prescribed oral immunosuppressive drug, followed by AZA and CsA. During the study evaluation period from 2012 until 2017, a decrease in the total part of CsA prescriptions was seen, together with an increase of the total part of MTX prescriptions (figure 3 and 4). Although the exact numbers are low, combinations of oral immunosuppressive drug use were seen in some patients, including AZA+CsA, AZA+MTX, CsA+MTX and CsA+MPA.

120,00%

100,00%

METHOTREXATE

80,00% CYCLOSPORINE

AZATHIOPRINE

60,00% AZATHIOPRINE+CYCLOSPORINE

AZATHIOPRINE+METHOTREXATE 40,00%

CYCLOSPORINE+METHOTREXATE

20,00% CYCLOSPORINE+MYCOFENOLIC ACID

MYCOFENOLIC ACID 0,00% 1-1-2012 1-2-2012 1-3-2012 1-4-2012 1-5-2012 1-6-2012 1-7-2012 1-8-2012 1-9-2012 1-1-2013 1-2-2013 1-3-2013 1-4-2013 1-5-2013 1-6-2013 1-7-2013 1-8-2013 1-9-2013 1-1-2014 1-2-2014 1-3-2014 1-4-2014 1-5-2014 1-6-2014 1-7-2014 1-8-2014 1-9-2014 1-1-2015 1-2-2015 1-3-2015 1-4-2015 1-5-2015 1-6-2015 1-7-2015 1-8-2015 1-9-2015 1-1-2016 1-2-2016 1-3-2016 1-4-2016 1-5-2016 1-6-2016 1-7-2016 1-8-2016 1-9-2016 1-10-2012 1-11-2012 1-12-2012 1-10-2013 1-11-2013 1-12-2013 1-10-2014 1-11-2014 1-12-2014 1-10-2015 1-11-2015 1-12-2015 1-10-2016 1-11-2016 1-12-2016

Figure 3 Trends in treatments (all unique patients, per month)

100%

METHOTREXATE 90% 27,4% 26,9% 32,1% 38,2% 80% CYCLOSPORINE

70% AZATHIOPRINE

28,5% 60% 29,9% 25,5% AZATHIOPRINE+CYCLOSPORINE 21,9% 50%

AZATHIOPRINE+METHOTREXATE 40%

30% CYCLOSPORINE+METHOTREXATE 42,2% 41,3% 38,4% 37,0% 20% CYCLOSPORINE+MYCOFENOLIC ACID

10% MYCOFENOLIC ACID 7,3% 6,6% 7,2% 6,2% 0% 2013 2014 2015 2016

Figure 4 Trends in treatments (only new patients, per year) ORAL IMMUNOSUPPRESSIVE DRUG USE IN THE NETHERLANDS 47

Discontinuation of treatment In the complete study group of 943 patients, a total of 532 patients (56.0%) discontinued treatment: these patients received their last prescription >182 days before January 1st 2017. Figure 5 shows an overview of the duration of treatment of these patients. It is notable that in 165 patients (31.0%) the time between the first and the last prescription was <30 days. Sometimes even no repeat prescription was disposed. It can be concluded that treatment in these patients was discontinued early, within the first months. No information on reasons for discontinuation was available. 3

180 120,00%

160

100,00%

140

120 80,00%

100

60,00% Frequency Frequency 80 Cumulative %

60 40,00%

20,00%

0 0,00%

Days between first and last prescription

Figure 5 Information on duration of treatment of all ‘stoppers’

Calculation of the total number of AD patients using oral immunosuppressive drugs The total of 943 patients identified with the algorithm in this study is considered as a representative sample of the total number of patients with AD using oral immunosuppressive drugs during a period of 5 years. In order to estimate the total number of AD patients in the Netherlands treated with oral immunosuppressive drugs in this 5-years period, data were corrected for the total number of Dutch pharmacies. In addition, also patients with AD without other atopic diseases, such as allergic asthma and allergic rhinitis, needed to be evaluated. 48 CHAPTER 3

Total amount of pharmacies Information on 556 pharmacies throughout the Netherlands was used for the analysis. This is 27.9% of the 1994 pharmacies located in the Netherlands in 2017. Therefore the factor 3.584 was used to calculate the actual number of patients with AD.

Atopy In the used algorithm for defining AD patients, the use of medication for allergic rhinitis and/ or allergic asthma was mandatory to ensure the atopic factor. In recent large clinical trials in a comparable group of AD patients around two third of the patients with AD has allergic asthma and/or allergic rhinitis.6,7 Therefore the factor 1.5 was used to calculate the total number of patients with AD, including those without allergic rhinitis and/or asthma.

Total number of AD patients treated with oral immunosuppressive drugs in the Netherlands (2012-2017) The total number of AD patients in the Netherlands is estimated to be: 65 943 x 3.584 x 1.5 = 354 510

The total number of AD patients treated with oral immunosuppressive drugs in the Netherlands is estimated to be: 943 x 3.584 x 1.5 = 5070

Total number of AD patients on active treatment with oral immunosuppressive drugs on December 31st 2016 The total number of AD patients on active treatment on December 31st 2016 was 411. When corrected for the total number of pharmacies and the atopy factor, the total number of active AD patients in the Netherlands is estimated to be: 411 x 3.584 x 1.5 = 2210 ORAL IMMUNOSUPPRESSIVE DRUG USE IN THE NETHERLANDS 49

Discussion In this retrospective pharmacy database study, the prescription behavior regarding oral immunosuppressive drugs for patients with AD in the Netherlands was evaluated. MTX was most commonly prescribed between January 1st 2012 and January 1st 2017, followed by AZA and CsA. In this study period, a decrease in the prescriptions of CsA was observed, together with an increase of the prescriptions of MTX. Surprisingly, around 30% of the patients discontinued treatment with oral immunosuppressive drugs soon after start.

3 In this study, a remarkable increase in the prescription of MTX was seen over the years. This might have been a result of the fact that several studies prescribing the clinical benefit of MTX in the treatment of AD have been published in the past years, including a randomized controlled trial.8 Besides, in 2014, an update of the Dutch guideline of atopic dermatitis was published, including recent publications and suggestions for dermatologist to use MTX, AZA and MPA for the treatment of AD, in cases of non-effectiveness or side effects to CsA.

The increased interest in the use of other oral immunosuppressive drugs than CsA was also seen some other recently published studies. Hegazy et al evaluated the use of oral immunosuppressive drugs in 139 adult patients. CsA was the first choice in 59 patients and MTX was the first choice in 19 patients.9 MMF was prescribed in three patients: in all situations, it was the third step of treatment. Vedie et al evaluated 54 patients with AD, treated with oral immunosuppressive drugs.10 Twenty-eight patients had been treated with MTX, 17 patients with AZA and 43 patients with CsA. Seven patients received a combination of AZA and MTX (mean duration of treatment 27.9 months). Totri et al surveyed systemic agent prescribing practices for severe childhood AD, by developing an online survey for pediatric dermatologist.11 In total, 133 dermatologists completed the survey and 115 of them used systemic treatment for severe pediatric AD. First-line drugs of choice were CsA (45.2%), methotrexate (29.6%), and mycophenolate mofetil (13.0%). The most commonly used second-line agents were methotrexate (31.3%) and mycophenolate mofetil (30.4%); azathioprine was the most commonly cited third-line agent. Schmitt et al evaluated 78 adult AD patients during a period of 24 months in different centra in Germany.12 CsA was prescribed the most, but AZA, MTX, prednisone, MMF, alitretinoine and leflunomide were also given to some patients. 50 CHAPTER 3

The study in this manuscript is different from the abovementioned publications. Most published studies are monocentric and performed in dedicated hospitals with a focus on AD and oral immunosuppressive drugs. In addition, the study of Totri et al only focused on children.11 The study described in this manuscript is different and gives an unique overview of the prescribing behavior throughout the country, including smaller regional hospitals. Therefore this study is more representative than the aforementioned studies. The fact that CsA was often the most prescribed first choice drug in the aforementioned published studies, shows that specialized (university) hospitals are more tended to follow the guidelines. In this manuscript, also evaluating prescribing behavior of smaller, regional hospitals, MTX was more popular than CsA.

In this study, discontinuation of treatment was seen in 56% of the patients. Reasons for discontinuation were lacking, but our data are in line with previously published drug survival studies, in which also high rates of discontinuation were seen.13-15 In these studies, discontinuation rates ranged from 55% to 66% after 1 year, for CsA, MTX, EC-MPS and AZA. In the study in this manuscript, remarkably, 31% of the patients discontinued treatment within the first months: the time between the first and last prescription in these patients was <30 days. Sometimes even no repeat prescription was disposed. In the studies of van der Schaft et al and Politiek et al, median drug survival was 256 days for CsA, 201 days for AZA, 322 days for enteric coated mycophenolate sodium and 9.8 months for MTX.13-15 It is not clear how many patients in these studies discontinued treatment within 30 days.

It should be noted that these drug survival studies have been performed in expertise centres with a focus on AD. The data described in the current study give a more accurate overview of the current use of oral immunosuppressive drugs in the Netherlands, including mostly non- specialized centres. Because expertise with (long-term) use of oral immunosuppressive drugs is often lacking in these centres, treatments are often discontinued early, in the hope to prevent the development of side effects.

The nation-wide NControl database offers us the opportunity to get an impression of the use of oral immunosuppressive drugs in the Netherlands. The database is large and includes detailed information on prescriptions throughout the Netherlands. Nevertheless, some limitations need to be discussed. ORAL IMMUNOSUPPRESSIVE DRUG USE IN THE NETHERLANDS 51

The NControl database possesses data from community pharmacies. In general, hospital pharmacies (including pharmacies in tertiary academic hospitals specialized in the treatment of AD), are not included. As a result, this study sample could be an underestimation of the real number of AD patients using oral immunosuppressive drugs. However, according to our clinical experience, most patients pick up their medication at their own community pharmacy.

Because no diseases or prescription criteria were documented, an algorithm was designed in order to select patients with AD from the NControl database. This algorithm was created 3 with the highest possible precision and was discussed with various experts, resulting in the best possible selection of AD patients. Patients who were supposed to have psoriasis or rheumatic diseases instead of AD, were excluded by removing patients using anti-psoriatic drugs or DMARDS from the selection. Organ transplant recipients were excluded by removing patients using oral tacrolimus. Still, our selection may include patients with other (skin) diseases who fortuitously used the combination of drugs that was selected for our algorithm.

By using the filter with concomitant use of nasal and/or inhaled steroids, the selection of patients was further refined. So, the analysis was performed in the subset of atopic patients also suffering from asthma and/or allergic rhinitis. From earlier studies we estimate that this includes about two third of the total AD population. We do not expect that the prescription behavior regarding oral immunosuppressive drugs is different in AD patients, not suffering from other atopic diseases. For the estimation of the total number of AD patients in the Netherlands using oral immunosuppressive drugs, we used a multiplication factor of 1.5.

In conclusion, oral immunosuppressive drugs have been prescribed to 1.4% of the AD patients in the Netherlands over a five year period. MTX was most commonly used. These data may be of great value for updating guidelines in future and for the positioning of new treatment options in AD. 52 CHAPTER 3

References

1 Drucker AM. Treating moderate-to-severe 9 Hegazy S, Tauber M, Livideanu CB et al. atopic dermatitis: what do dermatologists Systemic treatment of severe adult Atopic do with limited evidence? Br J Dermatol dermatitis in clinical practice: Analysis 2017; 176: 1441-2. of prescribing pattern in a cohort of 241 patients. J Eur Acad Dermatol Venereol 2 Schmitt J, Schmitt NM, Kirch W et al. 2017. Outpatient care and medical treatment of children and adults with atopic eczema. J 10 Vedie AL, Ezzedine K, Amazan E et al. Dtsch Dermatol Ges 2009; 7: 345-51. Long-term Use of Systemic Treatments for Moderate-to-Severe Atopic Dermatitis in 3 Sidbury R, Davis DM, Cohen DE et al. Adults: A Monocentric Retrospective Study. Guidelines of care for the management of Acta Derm Venereol 2016; 96: 802-6. atopic dermatitis: section 3. Management and treatment with phototherapy and 11 Totri CR, Eichenfield LF, Logan K et al. systemic agents. J Am Acad Dermatol Prescribing practices for systemic agents 2014; 71: 327-49. in the treatment of severe pediatric atopic dermatitis in the US and Canada: The 4 Roekevisch E, Spuls PI, Kuester D et al. PeDRA TREAT survey. J Am Acad Dermatol Efficacy and safety of systemic treatments 2017; 76: 281-5. for moderate-to-severe atopic dermatitis: a systematic review. J Allergy Clin Immunol 12 Schmitt J, Abraham S, Trautmann F et al. 2014; 133: 429-38. Usage and effectiveness of systemic treatments in adults with severe atopic eczema: 5 Garritsen FM, Roekevisch E, van der Schaft First results of the German Atopic Eczema J et al. Ten years experience with oral Registry TREATgermany. J Dtsch Dermatol immunosuppressive treatment in adult Ges 2017; 15: 49-59. patients with atopic dermatitis in two academic centres. J Eur Acad Dermatol 13 van der Schaft J, Politiek K, van den Reek Venereol 2015; 29: 1905-12. JM et al. Drug survival for ciclosporin A in a long-term daily practice cohort of 6 Simpson EL, Bieber T, Guttman-Yassky E et adult patients with atopic dermatitis. Br J al. Two Phase 3 Trials of Dupilumab versus Dermatol 2015; 172: 1621-7. Placebo in Atopic Dermatitis. N Engl J Med 2016; 375: 2335-48. 14 van der Schaft J, Politiek K, van den Reek JM et al. Drug survival for azathioprine and 7 Blauvelt A, de Bruin-Weller M, Gooderham enteric-coated mycophenolate sodium M et al. Long-term management of in a long-term daily practice cohort of moderate-to-severe atopic dermatitis adult patients with atopic dermatitis. Br J with dupilumab and concomitant topical Dermatol 2016; 175: 199-202. corticosteroids (LIBERTY AD CHRONOS): a 1-year, randomised, double-blinded, place- 15 Politiek K, van der Schaft J, Coenraads PJ et bo-controlled, phase 3 trial. Lancet 2017. al. Drug survival for methotrexate in a daily practice cohort of adult patients with severe 8 Schram ME, Roekevisch E, Leeflang MM atopic dermatitis. Br J Dermatol 2016; 174: et al. A randomized trial of methotrexate 201-3. versus azathioprine for severe atopic eczema. J Allergy Clin Immunol 2011; 128: 353-9. ORAL IMMUNOSUPPRESSIVE DRUG USE IN THE NETHERLANDS 53

3 Part II Safety of long-term treatment with oral immunosuppressive drugs

Lymphopenia in atopic dermatitis patients treated with oral immunosuppressive drugs

Daphne S. Bakker, Floor M. Garritsen, Helen L. Leavis, Jorien van der Schaft, Carla A.F.M. Bruijnzeel-Koomen, Marcel P.H. van den Broek, Marjolein S. de Bruin-Weller

Manuscript submitted 58 CHAPTER 4

Abstract Introduction: Oral immunosuppressive drugs are commonly used in the treatment of atopic dermatitis (AD). In patients with autoimmune- and rheumatic diseases, these drugs have been associated with lymphopenia. The presence of lymphopenia is related to an increased risk of opportunistic infections. The incidence of lymphopenia in patients with AD treated with oral immunosuppressive drugs is yet unknown. Objective: To evaluate the occurrence of persistent lymphopenia in patients with AD treated with oral immunosuppressive drugs and to make recommendations for screening in daily practice. Methods: In this retrospective study, medical records of all adult AD patients treated with oral immunosuppressive drugs in the University Medical Center Utrecht, the Netherlands between January 1st, 2005 and September 10th, 2016 were systematically screened. Patients with persistent lymphopenia (i.e. > 5 times lymphocyte counts below 0.8x10^9/L) during treatment with oral immunosuppressive drugs were included for further analysis. Results: A persistent lymphopenia during treatment with oral immunosuppressive drugs was found in 11 patients, who were treated with azathioprine, cyclosporine A, methotrexate, enteric-coated mycophenolate sodium and oral tacrolimus, sometimes in combination with prednisone. In 8/11 patients persistent lymphopenia was observed during combination treatment of one of the oral immunosuppressive drugs with prednisone. Interestingly, total white blood cell counts were normal in all patients at the time of decreased lymphocyte counts. Immunophenotyping was available in seven patients. CD4+ counts were decreased in all of these seven patient. No serious infections were observed. Conclusion: Oral immunosuppressive drug treatment in AD is occasionally associated with persistent lymphopenia. Concomitant treatment with prednisone seems to be a risk factor. Monitoring total white blood cell counts is not sufficient, since a lymphopenia can occur while total white blood cell count is normal. In case of persistent lymphopenia, the indication for immunosuppressive treatment should be reevaluated and referral to an immunologist for further screening should be considered. LYMPHOPENIA AND ORAL IMMUNOSUPPRESSIVE DRUGS 59

Introduction Atopic dermatitis (AD) is a common chronic inflammatory skin disease, with a prevalence of 2-10% in adults.1-3 The majority of AD patients can be adequately controlled with topical treatment and/or photo(chemo)therapy. However, in patients with moderate to severe eczema without sufficient response or the need of continuous treatment with mid to high potent topical corticosteroids, oral immunosuppressive treatment is often required. Drugs that are commonly used in the treatment of AD include cyclosporine A (CsA), methotrexate (MTX), azathioprine (AZA), enteric-coated mycophenolate sodium (EC-MPS), mycophenolate mofetil (MMF) and systemic corticosteroids.4

These immunosuppressive drugs have the potential to reduce disease activity or even achieve remission, but can also cause a broad spectrum of side effects which often require the with- 4 drawal of therapy. One of these side effects is the development of lymphopenia, as an effect of myelosuppression.5 Drug-induced lymphopenia is a known phenomenon in patients with autoimmune- and rheumatic diseases, inflammatory bowel disease (IBD) and in organ transplantation recipients.6-9 Although it is a common finding, the clinical implications are currently not fully understood. Mild lymphopenia might be a reflection of a sufficient immunosuppressive effect of the drugs, but it is known that persistent lymphopenia, regardless of its cause, is associated with an increased risk of opportunistic infections, as seen in HIV-positive patients, primary immunodeficiency and idiopathic CD4+ lymphopenia.5,10,11 Recently, treatment-induced lymphopenia in patients with psoriasis and multiple sclerosis, treated with the immunosuppressive drug dimethyl fumarate, has been associated with the development of progressive multifocal leukoencephalopathy (PML).12-17 Due to the assumption that PML develops when cellular immunity is severely disturbed, it could thereby also be associated with other immunosuppressive drugs.

Remarkably, although oral immunosuppressive drugs are commonly used in severe, difficult to treat AD for long time periods, data on the incidence and the clinical implication of lymphopenia in these patients is lacking. Therefore, the aim of this study is to evaluate the occurrence of lymphopenia in patients with AD treated with oral immunosuppressive drugs and to make recommendations for screening in daily practice. 60 CHAPTER 4

Methods Study population This study was exempted from review by our institutional review board. A retrospective explor- ative study was conducted at the Dermatology and Allergology Department of the University Medical Center Utrecht, the Netherlands. Patients were selected from our immunosuppressive drugs database, which includes all AD patients treated with oral immunosuppressive drugs in our center, including CsA, MTX, AZA, EC-MPS, MMF, oral tacrolimus and prednisolone (>3 months). This database contains about 360 patients. Electronic medical records were manually screened for the appearance of a persistent lymphopenia (defined as five or more lymphocyte counts below 0.8x109/L).

All patients receiving oral immunosuppressive drugs for the treatment of AD between January 1st 2005 and September 10th 2016 and a persistent lymphopenia during treatment were included. AD was diagnosed by a dermatologist, according to the Hanifin and Rajka criteria.18 Patients’ lymphocyte counts were monitored according to local treatment protocols of the specific types of immunosuppressive drugs.

Exclusion criteria were age below 18 years, a history of organ transplantation, treatment with chemotherapy or insufficient documentation of either lymphocyte counts or medication. AD should have been the primary indication for the treatment with oral immunosuppressive drugs. Lymphopenia should have been measured during treatment with oral immunosuppressive drugs or within three months after cessation of medication. Lymphopenia that did not occur within this time frame was assumed not to be related to the oral immunosuppressive drug.

The following data were systematically collected from the medical records of the included patients: sex, age at initiation of systemic treatment, type and dose of oral immunosuppressive drugs used, start- and stop date of immunosuppressive drugs, date and value of all lymphocyte counts and total white blood cell counts at time of decreased lymphocyte counts . When available, lymphocyte immunophenotyping including CD4+ and CD8+ lymphocytes were analyzed. The lower limit of normal for CD4+ and CD8+ counts were identified as 560 and 216 cells per μl (C/ μl), respectively. The electronic patient files were screened for the documentation of severe (opportunistic) infections.

Statistical analysis Data were analyzed using SPSS 21.0 (Version 21.0.0.0, SPSS Inc., Chicago, IL, USA). Absolute numbers and percentages were presented. Median with interquartile range [IQR] was calculated in case of not normally distributed variables. LYMPHOPENIA AND ORAL IMMUNOSUPPRESSIVE DRUGS 61

Results Patient characteristics Patients were selected from our immunosuppressive drugs database, which contains 360 patients. In 84 patients no lymphocyte measurements were performed. Eleven AD patients (six male, 54.5%) met the inclusion criteria and were included for further analysis (table 1). Median age at start of the immunosuppressive therapy was 39.0 years [IQR 32.5 – 43.8]. Patients were treated with AZA, CsA, MTX, EC-MPS and oral tacrolimus, as monotherapy or in combination with prednisone, and prednisone monotherapy (figure 1a – 1k). The median amount of different types of immunosuppressive drugs used per patient was 3.0 [IQR 3.0 – 5.0]. Median nadir (lowest lymphocyte count) during treatment was 0.59 [IQR 0.46 – 0.68].

4 62 CHAPTER 4

A. Patient 1 B. Patient 2 AZA 2.5 AZA 2.5 EC-MPS

2.0 2.0 Prednisone

1.5 1.5

LLN 1.0 1.0 LLN

0.5 0.5 Lymphocyte count x10^9/L Lymphocyte count x10^9/L

0.0 0.0

Start Start 2 months 1 24 months 24 months 48 months 72 months 96 months

C. Patient 3 D. Patient 4 AZA AZA 2.5 2.5 CsA Myfortic Prednisone 2.0 2.0 Prednisone

1.5 1.5

1.0 LLN 1.0 LLN

0.5 0.5 Lymphocyte count x10^9/L Lymphocyte count x10^9/L

0.0 0.0

Start Start

6 months 12 months 12 months 24 months 36 months 48 months 60 months

E. Patient 5 F. Patient 6 AZA AZA 2.5 CsA 2.5 EC-MPS 2.0 2.0 Tacrolimus Prednisone 1.5 1.5

LLN 1.0 1.0 LLN

0.5 0.5 Lymphocyte count x10^9/L Lymphocyte count x10^9/L

0.0 0.0

ths ths art hs hs Start St

24 mon 48 mon 72 months 96 months 12 mont 24 mont 120 months LYMPHOPENIA AND ORAL IMMUNOSUPPRESSIVE DRUGS 63

G. Patient 7 H. Patient 8 Tacrolimus AZA 2.5 2.5 Prednisone CsA Prednisone 2.0 2.0

1.5 1.5

LLN LLN 1.0 1.0

0.5 0.5 Lymphocyte count x10^9/L Lymphocyte count x10^9/L

0.0 0.0

nths nths nhts Start Start

12 mo 24 mo 36 mo 12 months 24 months

I. Patient 9 J. Patient 10 AZA CsA 2.5 2.5 MTX EC-MPS Tacrolimus Prednisone 2.0 2.0 Prednisone 4 1.5 1.5

LLN 1.0 1.0 LLN

0.5 0.5 Lymphocyte count x10^9/L Lymphocyte count x10^9/L

0.0 0.0

art Start St

12 months 24 months 36 months 24 months 48 months 72 months 96 months

K. Patient 11 CsA 2.5 EC-MPS

2.0 Prednisone

1.5

1.0 LLN

0.5 Lymphocyte count x10^9/L

0.0 s s h h nths Start ont ont onths 3 m 6 m 9 m 12 mo

p Time of immunophenotyping

Figure 1 Lymphocyte counts and treatment characteristics; a) Patient 1 (female, 62 years old); b) Patient 2 (male, 48 years old); c) Patient 3 (female, age 25 ); d) Patient 4 (female, age 52); e) Patient 5 (male, age 55); f) Patient 6 (female, age 48); g) Patient 7 (male, age 49); h) Patient 8 (male, age 43); i) Patient 9 (female, age 55); j) Patient 10 (male, age 64); k) Patient 11 (male, age 43) AZA azathioprine, CsA cyclosporine A, EC-MPS enteric-coated mycophenolate sodium, MTX methotrexate, LLN Lower limit of normal (0.8x10^9/L) 64 CHAPTER 4

Table 1 Basic characteristics and lymphocyte measurements of included patients Total (n=11) Male n (%) 6 (54.5) Age at start therapy (years), median [IQR] 39.0 [32.5-43.8] Immunosuppressive drugs per patient, median [IQR] 3.0 [3.0-5.0] Nadir during treatment, median [IQR] 0.59 [0.46-0.68]

IQR Inter quartile range; Nadir lowest lymphocyte count

Analysis of lymphopenia Most of the patients had fluctuating lymphocyte counts during therapy. In one patient (figure 1k) lymphopenia was found before start of treatment with oral immunosuppressive drugs and persisted during the whole treatment period and after cessation. Fluctuations of lymphocyte counts during a period with continuous oral immunosuppressive treatment were carefully analyzed. No associations with dosage adjustment were found. In eight patients (patient 2, 3, 4, 5, 7, 8, 9 and 10) the oral immunosuppressive drugs were taken in combination with prednisone at time of lymphopenia. In patients 2, 4, 5 and 9 lymphocyte counts increased or normalized after cessation of prednisone, despite continuation of the combined oral immunosuppressive drug. Remarkably, almost all total white blood cell counts were within the normal range at the time a decreased lymphocyte count was measured (figure 2). No lymphopenia-related opportunistic infections or other serious infectious pathology were documented during the study period.

18 4.0 ULN

16 3.5 Patient 1 Patient 2 14 3.0 Patient 3 12 Patient 4 2.5 Patient 5 10 ULN 2.0 Patient 6 8 Patient 7 1.5 Patient 8 6 Patient 9 4 LLN 1.0 LLN Patient 10 2 Total lymphocyte count x10^9/L 0.5 Patient 11 Total white blood cell count x10^9/L 0 0.0

Figure 2 Total white blood cell counts and corresponding total lymphocyte counts LLN Lower limit of normal ULN Upper limit of normal LYMPHOPENIA AND ORAL IMMUNOSUPPRESSIVE DRUGS 65

CD4+ and CD8+ counts In seven of the eleven patients (patient 1, 5, 6, 7, 8, 9 and 10) CD8+ and CD4+ counts were measured as suggested by the consulted clinical immunologist (figure 3). These patients were treated with AZA (patient 1 and 6), AZA combined with prednisone (patient 5, 8 and 9), tacrolimus combined with prednisone (patient 7) and prednisone monotherapy (patient 10) at the moment of measurement. In patient 10 the immunophenotyping was performed twice. All CD4+ counts in these seven patients were decreased (figure 3). The median CD4+ count was 342 C/μl [IQR 126 – 470]. CD4+ counts were <200 C/μl in two cases, assuming an increased risk of infections. The number of CD8+ was decreased in six of the seven patients. The median CD8+ count was 98 C/μl [IQR 57 – 136].

CD4 CD8 Legend 600 Patient 1 4 ULN ULN Patient 5 1000 Patient 6

400 Patient 7 Patient 8 LLN Patient 9

500 LLN Patient 10 200 Absolute CD8 cells per mm3 Absolute CD4 cells per mm3

0 0

Figure 3 Immunophenotyping in 7 patients LLN Lower limit of normal ULN Upper limit of normal

Discussion In this retrospective study, we investigated the occurrence of lymphopenia in AD patients treated with oral immunosuppressive drugs. A persistent lymphopenia was seen in eleven patients. Patients were treated with different immunosuppressive drugs, but concomitant treatment with prednisone seems to be a risk factor for the development of lymphopenia. CD4+ counts were decreased in all patients in which immunophenotyping was performed. In two patients CD4+ counts reached critically low levels of <200 C/μl (one patient using AZA + prednisone and one patient using only prednisone). No lymphopenia-related opportunistic infections or other serious infectious pathology were documented during the study period.

Studies on the occurrence of lymphopenia in patients with AD are lacking and limited data is available on other chronic inflammatory skin diseases. Lehman et al analyzed 198 patients treated with immunosuppressive drugs for different dermatologic indications (AD not included). Total lymphocyte counts were not analyzed in this cohort, since this study was designed to define the incidence of pneumocystis pneumonia. However, CD4+ lymphocyte counts were tested in 12 patients, of whom 5 patients (41.7%) had low levels.19 Pneumocystis pneumonia developed in one patient in this study, but no CD4+ was measured in this patient. 66 CHAPTER 4

Most data on the risk of lymphopenia in patients treated with oral immunosuppressive drugs are derived from studies in non-dermatological diseases.5-9 Lymphopenia is common in patients with chronic inflammatory diseases treated with AZA, MTX, cyclophosphamide and prednisone and transplant recipients treated with several combinations of immunosuppressive agents, including AZA, MMF, CsA and prednisone.5,7,8 Combined immunosuppression appears to have a greater risk for the development of lymphopenia.5,7 Al Rifai et al found that in 66% of their IBD cohort treated with AZA , lymphopenia resolved spontaneously with no change of AZA dosing. In most of these cases, lymphopenia occurred within the first 3 months of treatment.7

In our study, most decreased lymphocyte counts were found during concomitant treatment with prednisone. This is in line with previous studies in patients receiving chronic immunosuppressive therapy for other conditions, that showed that the co-prescription of systemic corticosteroids can be an additional risk factor for the development of lymphopenia and opportunistic infections.5,7,20

Total white blood cell counts measured at time of a decreased lymphocyte count, were almost all within the normal range or even increased. This can be explained by the fact that eosinophil numbers, one of the other subsets of the total white blood cell count, are elevated in the peripheral blood in most AD patients.21 Additionally, prednisone can induce leukocytosis, predominantly due to the increase of neutrophils.22 Increased eosinophil and neutrophil counts may compensate a decreased lymphocyte count in making the total white blood cell count normal. As a result, a lymphopenia can be overlooked. It is therefore important to monitor white blood cell count differential instead of total white blood cell counts in AD patients treated with oral immunosuppressive drugs.

Immunosuppressive drugs can reduce all lymphocytes subsets, but the circulating CD4+ lymphocytes are generally the most affected.9 In seven of our patients immunophenotyping was performed, after a consultation with the clinical immunologist. CD4+ lymphocyte counts were decreased in all these patients. A low CD4+ count is associated with an increased risk of opportunistic infections especially when CD4+ cells reach a critically low level (<200 C/μl).23 In our study, two patients had CD4+ counts <200 C/μl; in these patients, no opportunistic infections or other serious infectious pathology were documented during the study period. However, current guidelines for the treatment of HIV infection recommend prophylaxis for certain opportunistic infections (such as Pneumocystis jirovecii pneumonia or toxoplasmosis) when CD4+ counts are <200 C/μl.24

Immunosuppressive treatment was continued despite a persistent lymphopenia in all included patients. This can be justified by the fact that an immunologist was consulted in most of the cases and patients were frequently controlled and monitored for signs and symptoms of viral or opportunistic infections. No serious infectious pathology was reported. In one of the included LYMPHOPENIA AND ORAL IMMUNOSUPPRESSIVE DRUGS 67

patients immunosuppressive therapy was even started while there was already an persistent lymphopenia after consulting of a clinical immunologist. The patient was screened for HIV infection, leukemia and lymphoma and the results were all negative.

This study has some limitations. In literature, there is discussion on the definition of lymphopenia, with lower limits of normal ranging from 0.8 up to 1.2 x10^9/L.9 In our study, we defined lymphopenia as an absolute lymphocyte count <0.8x10^9/L, according to the Common Terminology Criteria of Adverse Events.25 Total white blood cell count differential, which includes the total lymphocyte count, was not included in the standard monitoring guidelines for prednisone monotherapy and CsA treatment. In 84 patients out of the 360 patients in the screened database, no lymphocyte count was 4 measured. Therefore we expect an underestimation of patients with persistent lymphopenia in this study. It needs to be considered that many other factors can influence lymphocyte counts, such as comorbidities, physical or physiological stress, co-medication and age. This information was often lacking, due to the retrospective design. Besides, viral infections can induce lymphopenia and older age is associated with lower lymphocyte counts.26,27 Due to the tertiary character of our hospital, information about the incidence of viral infections (usually treated by the general practitioner or in secondary centres) was insufficiently documented. No severe infections were reported in the medical files in our hospital, but patients might have visited their general practitioner for mild infections in the meantime.

Conclusion and recommendation This study shows that persistent lymphopenia is occasionally seen in AD patients treated with oral immunosuppressive drugs. The co-prescription of prednisone seems to be a risk factor. We did not observe any major infectious pathology in the analyzed patients; therefore the clinical implication of a persistent lymphopenia still remains unclear.

We suggest to include lymphocyte counts in the standard follow-up monitoring for all AD patients treated with oral immunosuppressive drugs. Monitoring total white blood cell count is not sufficient, since lymphocyte counts can be decreased while total white blood cell count is normal. Patients should be monitored for symptoms of viral or opportunistic infections. In case of persistent lymphopenia, consulting a clinical immunologist should be considered to decide whether further analysis including immunophenotyping or antibiotic prophylaxis for opportunistic infections is required. Apart from that, the indication for immunosuppressive treatment needs to be reevaluated in these patients. 68 CHAPTER 4

References

1 DaVeiga SP. Epidemiology of atopic derma- 10 Duncan RA, von Reyn CF, Alliegro GM et al. titis: a review. Allergy Asthma Proc 2012; Idiopathic CD4+ T-lymphocytopenia--four 33: 227-34. patients with opportunistic infections and no evidence of HIV infection. N Engl J Med 2 Silverberg JI, Hanifin JM. Adult eczema 1993; 328: 393-8. prevalence and associations with asthma and other health and demographic factors: 11 Cunningham-Rundles C, Bodian C. a US population-based study. J Allergy Clin Common variable immunodeficiency: Immunol 2013; 132: 1132-8. clinical and immunological features of 248 patients. Clin Immunol 1999; 92: 34-48. 3 Bieber T. Atopic dermatitis. N Engl J Med 2008; 358: 1483-94. 12 van Oosten BW, Killestein J, Barkhof F et al. PML in a patient treated with dimethyl 4 Roekevisch E, Spuls PI, Kuester D et al. fumarate from a compounding pharmacy. Efficacy and safety of systemic treatments N Engl J Med 2013; 368: 1658-9. for moderate-to-severe atopic dermatitis: A systematic review. J Allergy Clin Immun 13 Ermis U, Weis J, Schulz JB. PML in a patient 2014; 133: 429-38. treated with fumaric acid. N Engl J Med 2013; 368: 1657-8. 5 Gluck T, Kiefmann B, Grohmann M et al. Immune status and risk for infection in 14 Sweetser MT, Dawson KT, Bozic C. patients receiving chronic immunosup- Manufacturer’s response to case reports of pressive therapy. J Rheumatol 2005; 32: PML. N Engl J Med 2013; 368: 1659-61. 1473-80. 15 Rosenkranz T, Novas M, Terborg C. PML 6 Vogelin M, Biedermann L, Frei P et al. in a patient with lymphocytopenia treated The Impact of Azathioprine-Associated with dimethyl fumarate. N Engl J Med 2015; Lymphopenia on the Onset of 372: 1476-8. Opportunistic Infections in Patients with Inflammatory Bowel Disease. PLoS One 16 Nieuwkamp DJ, Murk JL, van Oosten BW et 2016; 11: e0155218. al. PML in a patient without severe lymphocytopenia receiving dimethyl fumarate. N 7 Al Rifai A, Prasad N, Shuttleworth E et al. Engl J Med 2015; 372: 1474-6. Natural history of azathioprine-associated lymphopenia in inflammatory bowel 17 Longbrake EE, Naismith RT, Parks BJ et al. disease patients: a prospective observa- Dimethyl fumarate-associated lympho- tional study. Eur J Gastroenterol Hepatol penia: Risk factors and clinical significance. 2011; 23: 153-8. Mult Scler J Exp Transl Clin 2015; 1.

8 Hutchinson P, Chadban SJ, Atkins RC et al. 18 Hanifin JM, Rajka G. Diagnostic Features Laboratory assessment of immune function of Atopic-Dermatitis. Acta Derm-Venereol in renal transplant patients. Nephrol Dial 1980: 44-7. Transplant 2003; 18: 983-9. 19 Lehman JS, Kalaaji AN. Role of primary 9 Gergely P. Drug-induced lymphopenia: prophylaxis for pneumocystis pneumonia focus on CD4+ and CD8+ cells. Drug Saf in patients treated with systemic corti- 1999; 21: 91-100. costeroids or other immunosuppressive agents for immune-mediated dermatologic conditions. J Am Acad Dermatol 2010; 63: 815-23. LYMPHOPENIA AND ORAL IMMUNOSUPPRESSIVE DRUGS 69

20 Toruner M, Loftus EV, Jr., Harmsen WS et al. Risk factors for opportunistic infections in patients with inflammatory bowel disease. Gastroenterology 2008; 134: 929-36.

21 Simon D, Braathen LR, Simon HU. Eosinophils and atopic dermatitis. Allergy 2004; 59: 561-70.

22 Nakagawa M, Terashima T, D’Yachkova Y et al. Glucocorticoid-induced granulocytosis: contribution of marrow release and demar- gination of intravascular granulocytes. Circulation 1998; 98: 2307-13. 4

23 Okafor PN, Nunes DP, Farraye FA. Pneumocystis jiroveci pneumonia in inflammatory bowel disease: when should prophylaxis be considered? Inflamm Bowel Dis 2013; 19: 1764-71.

24 Kaplan JE, Benson C, Holmes KK et al. Guidelines for prevention and treatment of opportunistic infections in HIV-infected adults and adolescents: recommendations from CDC, the National Institutes of Health, and the HIV Medicine Association of the Infectious Diseases Society of America. MMWR Recomm Rep 2009; 58: 1-207; quiz CE1-4.

25 SERVICES USDOHAH. Common Terminology Criteria for Adverse Events (CTCAE) In, Vol. Version 4.0 2010.

26 Hussain T, Kulshreshtha KK, Yadav VS et al. CD4+, CD8+, CD3+ cell counts and CD4+/CD8+ ratio among patients with mycobacterial diseases (leprosy, tubercu- losis), HIV infections, and normal healthy adults: a comparative analysis of studies in different regions of India. J Immunoassay Immunochem 2015; 36: 420-43.

27 Giefing-Kroll C, Berger P, Lepperdinger G et al. How sex and age affect immune responses, susceptibility to infections, and response to vaccination. Aging Cell 2015; 14: 309-21.

Risk of non-melanoma skin cancer in patients with atopic dermatitis treated with oral immunosuppressive drugs

Floor M. Garritsen, Jorien van der Schaft, Juul M.P.A. van den Reek, Klaziena Politiek, Harmieke van Os-Medendorp, Marijke R. van Dijk, Dirk J. Hijnen, Marlies de Graaf, Carla A.F.M. Bruijnzeel-Koomen, Elke M.G.J. de Jong, Marie-Louise A. Schuttelaar, Marjolein S. de Bruin-Weller

Acta Derm Venereol 2017;97(6):724-730 72 CHAPTER 5

Abstract There is uncertainty about the risk of developing non-melanoma skin cancer (NMSC), including basal cell carcinoma and squamous cell carcinoma (SCC), in patients with atopic dermatitis (AD) treated with oral immunosuppressive drugs. A total of 557 patients with AD treated with these drugs in the University Medical Center Utrecht and Groningen, the Netherlands, were analyzed. NMSC after oral immunosuppressive treatment was reported in 18 patients (3.2%). The standardized incidence ratio for developing SCC was 13.1 [95% CI 6.5-19.7]. Patients developing NMSC were older at therapy start (p<0.001) and data lock (p<0.001) compared to patients without NMSC. No significant differences were found in sex, cumulative days of -oral immu nosuppressive drugs use and follow-up between these groups (p=0.42; p=0.88 and p=0.34, respectively). In interpreting these results it is important to include other factors, such as lack of association between treatment duration and tumor development and the long interval between treatment discontinuation and tumor development in some patients. NON-MELANOMA SKIN CANCER AND ORAL IMMUNOSUPPRESSIVE DRUGS 73

Introduction Atopic dermatitis (AD) is a chronic inflammatory skin disease with a prevalence of 1-3% in adults.1 Although AD can be controlled adequately with topical treatment and/or ultraviolet (UV) light therapy in the majority of patients, a subgroup of severe and difficult-to-treat patients remains. Furthermore, in some patients it is impossible to taper topical corticosteroid treatment to a safe maintenance scheme. Oral immunosuppressive drugs are indicated in all these patients.

Oral immunosuppressive drugs that are regularly used in the management of AD are cyclosporine A (CsA), azathioprine (AZA), methotrexate (MTX), mycophenolate mofetil (MMF), enteric-coated mycophenolate sodium (EC-MPS), (extended-release) tacrolimus and systemic glucocorticosteroids. Clinical efficacy and safety have been proven in clinical trials for most of 2-5these drugs. However, treatment duration in clinical trials is limited. Due to the chronic nature of AD, long- term treatment with oral immunosuppressive drugs is often necessary to maintain adequate 5 disease control. Recent drug survival studies demonstrate that oral immunosuppressive drugs are regularly used for many years in daily practice.6-8

An important barrier to long-term use of oral immunosuppressive drugs in AD patients is the possible increased risk of development of malignancies, especially non-melanoma skin cancer (NMSC), including basal cell carcinoma (BCC) and squamous cell carcinoma (SCC). Most data on the risk of developing malignancies in patients treated with oral immunosuppressive drugs are derived from transplant patients.9,10 Immunosuppressive agents may increase the risk of cancer development by causing DNA damage and diminishing DNA repair mechanisms. Tumour angiogenesis may be promoted and the susceptibility of viral infections may be increased. Finally, immune surveillance, which normally prevents the growth and development of malignancies, may be inhibited by immunosuppressive drugs.9,11,12 Recent studies also report an increased risk of NMSC and lymphoma in patients using AZA for autoimmune diseases, such as inflammatory bowel disease (IBD) and other non-rheumatic autoimmune diseases.13-15 To date, there has been a lack of data regarding the risk of NMSC in AD patients using oral immunosuppressive drugs.

The aim of this study was to estimate the incidence of NMSC in a large cohort of AD patients treated with oral immunosuppressive drugs in the Netherlands and to compare these findings with those for the Dutch general population. 74 CHAPTER 5

Materials & methods Design This study was exempted from review by our institutional review board. The data were collected from the departments of Dermatology of the University Medical Center Utrecht and the University Medical Center Groningen, the Netherlands, in the period from 1989 until 1 January 2014.

Study population All adult patients with AD receiving oral immunosuppressive drugs (CsA, AZA, MTX, MMF, EC-MPS and [extended release] tacrolimus) for more than 2 months were included. For all patients, the follow-up period ended 1 January 2014, independent of whether the treatment had already discontinued. AD should have been the primary indication for the treatment with oral immunosuppressive drugs. Patients aged less than 18 years at the start of the treatment were excluded. Medical records were screened for the following patient and treatment characteristics: sex, age at start of use of oral immunosuppressive drug treatment, age at data lock, duration of follow-up calculated from the first starting date of oral immunosuppressive treatment until data lock, type of medication and cumulative days of oral immunosuppressive drug use.

Outcome All patient files in the histopathology register (Pathologisch Anatomisch Landelijk Geautomatiseerd Archiefsysteem [PALGA]), a nation-wide database for pathology reports in the Netherlands with national coverage, were screened for NMSC until 16 May 2014.16 Tumors that developed within 6 months after the start of the treatment were considered as not related to drug treatment and were excluded from analysis.

In patients with a diagnosis of NMSC, the following additional information was collected: type of malignancy, age at time of NMSC diagnosis, cumulative days of oral immunosuppressive drug use until diagnosis, time between start of oral immunosuppressive drug use and diagnosis, time between stop of oral immunosuppressive drug use and diagnosis, history of UV light therapy and history of malignancies before treatment with oral immunosuppressive drugs.

> 10 years 16 > 5 years 59 4-5 years 29 3-4 years 39 2-3 years 62 1-2 years 129

Duration of treatment 0-1 years 223 0 50 100 150 200 250 Patients (n) Figure 1 Duration of treatment with oral immunosuppressive drugs (n=557) NON-MELANOMA SKIN CANCER AND ORAL IMMUNOSUPPRESSIVE DRUGS 75

Statistical analysis All statistical analyses were performed using SPSS statistics 21. Subgroup analyses for patients with and without NMSC were performed. The Mann-Whitney U test and the X2 test were used to calculate whether there was a statistically significant difference between the subgroups in terms of sex, age at data lock, age at start of treatment, the total duration of treatment and the duration of follow up. The incidence of NMSC (including both BCC and SCC) was compared between patients treated ≤2 years and >2 years and patients treated ≤5 years and >5 years. Separated analyzes for the incidence of only SCC were performed as well. Dependent on the number of patients treated with monotherapy of a specific oral immunosuppressive drug (without a history of other oral immunosuppressive drugs), subgroup analyses of the individual treatment groups were carried out.

The standardized incidence ratio (SIR) of SCC in our cohort was calculated by dividing the number of observed cases (number of newly diagnosed malignancies) by the number of expected cases in the general Dutch population in the same period, corrected for age.17 The 5 95% confidence interval (95% CI) of the SIR was calculated using the indirect method.18 This method was previously described by van den Reek et al.13 Due to the fact that BCCs are not systematically registered in the Netherlands, no SIR could be determined for BCCs.

Table 1 Patient characteristics All patients (n=557) Male, n (%) 299 (53.7) Age at data lock, median [IQR] 44.7 [33.4-55.2] Age at inclusion1 , median [IQR] 37.1 [25.5-48.7] Duration of follow-up in years2 , median [IQR] 6.0 [3.0-10.2] Total patients years of follow-up 3855.5 Treatment characteristics Cyclosporine A only, n (%) 281 (50.4) Azathioprine only, n (%) 13 (2.3) Enteric-coated mycophenolate sodium only, n (%) 8 (1.4) Methotrexate only, n (%) 15 (2.7) > one oral immunosuppressive drug, n (%) 240 (43.1) Two different drugs, n (%) 164 (29.4) Three different drugs, n (%) 55 (9.9) Four different drugs, n (%) 20 (3.6) Five different drugs, n (%) 1 (0.2)

IQR Interquartile range 1 Age at moment of starting treatment with oral immunosuppressive drugs 2 From starting oral immunosuppressive drugs until data lock 76 CHAPTER 5

Results Characteristics of the total group A total of 557 AD patients (299 male patients, 53.7%) with one or more treatment episodes with oral immunosuppressive drugs from 1 January 1989 until 1 January 2014 were included in this study (table 1). CsA was prescribed most frequently (770 episodes), followed by EC-MPS (157 episodes), AZA (139 episodes), MTX (69 episodes), MMF (15 episodes), tacrolimus (24 episodes) and extended release tacrolimus (13 episodes). There was a wide variation in treatment duration (figure 1).

Results from the histopathology database (PALGA) NMSC during or after oral immunosuppressive treatment was reported in 18 patients (3.2%) (figure 2). The individual results are shown in table 2. SCCs after oral immunosuppressive treatment were found in 10 patients (1.8%). Two of these patients had more than 1 SCC and 3 of these patients were already diagnosed with an SCC before the start of oral immunosuppressive treatment. One of the 10 patients also developed a BCC. BCCs after oral immunosuppressive treatment were found in 9 patients (1.6%). One of these patients also developed an SCC. One of these 9 patients developed 3 BCCs and two patients were already diagnosed with a BCC before the start of oral immunosuppressive treatment.

Patients treated with oral immunosuppressive drugs (n=557 patients) No NMSC (n=534 patients)

NMSC (n=23 patients) NMSC before start treatment, not after start treatment (n=5 patients) NMSC after start treatment (n=18 patients)*

SCC BCC (n=15 cases) (n=11 cases)

Figure 2 Flowchart BCC Basal cell carcinoma; NMSC Non-melanoma skin cancer; SCC Squamous cell carcinoma * Some patients had more than one malignancy Table 2 Characteristics of patients with non-melanoma skin cancer Patient Sex Tumor Treatment Age at Cumulative days of oral Time between start of oral Time between stop of oral History History of malignancy diagnosis immunosuppressive drug use immunosuppressive drug immunosuppressive drug use UV light (years) from start until diagnosis use and diagnosis (days) and diagnosis (days) therapy 1 F SCC CsA 51 554 3961 <34071 Unknown No 2 F SCC CsA 55 227 3278 3051 No No 3 F SCC CsA 56 489 808 319 Unknown No

4 M SCC AZA 40 185 185 Still on treatment at diagnosis No SCC, dysplastic papilloma NON-MELANOMA SKINCANCER AND ORAL IMMUNOSUPPRESSIVEDRUGS frenulum, non-Hodgkin lymphoma 5 F SCC (5x) CsA 68 (SCC 1) 344 (SCC 1) 3380 (SCC 1) Still on treatment at diagnosis Unknown No 69 (SCC 2 370 (SCC 2) 3644 (SCC 2) (SCC 1 and 2) and 3) 441 (SCC 3,4 and 5) 3715 (SCC 3) 62 (SCC 3) 70 (SCC 4 4038 (SCC 4) 385 (SCC 4) and 5) 4101 (SCC 5) 448 (SCC 5) 6 M SCC EC-MPS, CsA 68 2705 4855 1331 UVB SCC 7 M SCC CsA 48 2166 6286 4120 UVB and No PUVA 8 M SCC (2x) CsA 64 (SCC 1 632 (SCC 1 and 2) 2792 (SCC 1) 1267 (SCC 1) UVB No and 2) 2957 (SCC 2) 1432 (SCC 2) 9 F SCC MTX, CsA 59 385 406 Still on treatment at diagnosis UVB No 10 F SCC and MTX, CsA 70 (SCC and 249 (SCC) 328 (SCC) Still on treatment at diagnosis No SCC, breast cancer BCC BCC) 452 (BCC) 528 (BCC) (SCC and BCC) 11 M BCC CsA 44 210 1460 <1250 1 PUVA No 12 M BCC CsA 59 281 3094 2813 No No 13 F BCC CsA, EC-MPS 31 243 1574 1331 UVB No 14 F BCC CsA, MMF, EC-MPS, 65 437 611 Still on treatment at diagnosis UVB No advagraf (tacrolimus) 15 M BCC CsA 66 280 280 Still on treatment at diagnosis No BCC 16 F BCC CsA 41 259 350 61 UVB No 17 F BCC (3x) AZA, CsA, MTX 72 (BCC 1,2 462 (BCC 1) 462 (BCC 1) Still on treatment at diagnosis UVB No and 3) 560 (BCC 2 and 3) 560 (BCC 2 and 3) 18 M BCC CsA 74 218 1128 910 UVB and 8x BCC, SCC, adenocarci- PUVA noma esophagus AZA azathioprine; BCC basal cell carcinoma; CsA cyclosporine A;EC-MPS enteric-coated mycophenolate sodium; F female; M male; MMF mycophenolate mofetil; NMSC non-melanoma skin cancer; PUVA psoralen and ultraviolet A; SCC squamous cell carcinoma; UVA ultraviolet A; UVB ultraviolet B 77 1 Loss to follow-up 5 78 CHAPTER 5

Patients with a malignancy versus patients without a malignancy Patients who developed NMSC were significantly older compared to patients without a malignancy at start of therapy ( p<0.001) and at data lock ( p<0.001) (table 3a). Sex, cumulative days of oral immunosuppressive drugs use until data lock and duration of follow-up were not statistically significant different between the groups (p=0.42; p=0.88 and p=0.34 respectively). There was no significant difference in the incidence of NMSC between patients treated ≤ 2 years (n=352, incidence 3.4%) and patients treated >2 years (n=205, incidence 2.9%) (p=0.76) and no significant difference between patients treated ≤ 5 years (n=482, incidence 3.3%) and patients treated > 5 years (n=75, incidence 2.7%) (p=0.77).

Patients who developed an SCC were compared to patients who did not develop an SCC (table 3b). Patients developing an SCC were significantly older at data lock and start of treatment compared to patients who did not develop an SCC (p=0.001 and p=0.004 respectively). Sex, cumulative days of oral immunosuppressive drugs until data lock and duration of follow-up were not statistically significantly different between the groups (p=0.38; p=0.35 and p=0.09 respectively). There was no significant difference in the incidence of SCC between patients treated ≤ 2 years (n=352, incidence 2.0%) and patients treated > 2 years (n=205, incidence 1.5%) (p=0.65) and no significant difference between patients treated ≤ 5 years (n=482, incidence 1.7%) and patients treated > 5 years (n=75, incidence 2.7%) (p=0.54).

Due to the small number of events, multivariate analysis to evaluate the effect of multiple influences on the risk of development of NMSC could not be performed.

Table 3a Patient characteristics of total group (n=557) NMSC during or after No NMSC during or p-value treatment (n=18) after treatment (n=539) differences Male, n (%) 8 (44.4%) 291 (54.0%) 0.424 Age at data lock in years, median 61.8 [51.7-70.5] 44.1 [32.7-53.8] <0.001 [IQR] Age at start in years, median [IQR]1 54.7 [40.4-64.2] 36.4 [25.0-48.1] <0.001 Cumulative days of oral immunosup- 499.0 [255.0-1087.0] 505.0 [252.0-1123.0] 0.879 pressive drug use until last day of follow-up, median [IQR] Durations of follow-up, in years, 6.9 [3.9-12.5] 6.0 [2.9-10.1] 0.343 median [IQR]2

IQR Interquartile range; NMSC non-melanoma skin cancer 1 Age at moment of starting treatment with oral immunosuppressive drugs 2 From starting oral immunosuppressive drugs until data lock NON-MELANOMA SKIN CANCER AND ORAL IMMUNOSUPPRESSIVE DRUGS 79

Table 3b Patient characteristics of total group (n=557), specified for SCC SCC during or after No SCC during or after p-value treatment (n=10) treatment (n=547) differences Male, n (%) 4 (40.0) 295 (53.9) 0.381 Age at data lock in years, median 61.2 [52.2-70.5] 44.3 [33.0-54.7] 0.001 [IQR] Age at start in years, median [IQR]1 54.7 [40.5-58.3] 36.9 [25.1-48.3] 0.004 Cumulative days of oral immunosup- 531.5 [438.3-1663.5] 498.0 [252.0-1117.0] 0.351 pressive drug use until last day of follow-up, median [IQR] Durations of follow-up, in years, 11.7 [4.6-13.3] 6.0 [2.9-10.1] 0.087 median [IQR]2

IQR Interquartile range; SCC squamous cell carcinoma 1 Age at moment of starting treatment with oral immunosuppressive drugs 2 From starting oral immunosuppressive drugs until data lock

Subgroup analysis: CsA monotherapy 5 CsA monotherapy was used in 281 patients (table 4a). Patients with CsA monotherapy who developed NMSC were statistically significantly older at start of therapy (p=0.001) and at data lock (p<0.001) compared to patients with CsA monotherapy without malignancy. Sex, cumulative days of oral immunosuppressive drugs until data lock and duration of follow-up were not statistically significantly different between the groups (p=0.96; p=0.79 and p=0.10 respectively). Patients with CsA monotherapy who developed SCC were compared to patients who did not develop SCC (table 4b). Patients developing SCC were significantly older at data lock and start of treatment compared to patients who did not develop SCC (p=0.003 and p=0.02 respectively). Duration of follow-up was longer in patients who developed an SCC (p=0.01) versus patients who did not develop an SCC. Sex and the cumulative days of oral immunosuppressive drugs until data lock were not significantly different (p=0.31 and p=0.30) between patients with and without an SCC.

Subgroup analyses of the other treatment groups were not performed due to the small number of patients in these groups. 80 CHAPTER 5

Table 4a Patient characteristics of cyclosporine A monotherapy (n=281) CsA monotherapy CsA monotherapy p-value with NMSC during or without NMSC during or differences after treatment (n=11) after treatment (n=270) Male, n (%) 6 (54.5%) 145 (53.7 %) 0.956 Age at data lock in years, median 59.0 [52.3-69.5] 40.1 [28.8-51.0] <0.001 [IQR] Age at start in years, median [IQR]1 51.0 [40.2-59.7] 32.1 [23.6-45.1] 0.001 Cumulative days of oral immuno- 379.0 [227.0-632.0] 350.5 [210.8-772.3] 0.785 suppressive use until last day of follow-up, median [IQR] Durations of follow-up in years, 11.0 [4.0-12.5] 5.4 [2.5-9.7] 0.095 median [IQR]2

CsA cyclosporine A; IQR Interquartile range; NMSC non-melanoma skin cancer 1 Age at moment of starting treatment with oral immunosuppressive drugs 2 From starting oral immunosuppressive drugs until data lock

Table 4b Patient characteristics of cyclosporine A monotherapy (n=281), specified for SCC CsA monotherapy with CsA monotherapy p-value SCC during or after without SCC during or differences treatment (n=6) after treatment (n=275) Male, n (%) 2 (33.3) 149 (54.2) 0.311 Age at data lock in years, median 58.7 [52.2-68.8] 40.3 [29.1-51.8] 0.003 [IQR] Age at start in years, median [IQR]1 50.3 [38.3-57.5] 32.4 [23.7-45.3] 0.022 Cumulative days of oral immuno- 521.5 [341.0-1015.5] 345.0 [212.0-765.0] 0.304 suppressive drug use until last day of follow-up, median [IQR] Durations of follow-up, in years, 12.5 [9.3-14.0] 5.3 [2.6-9.7] 0.012 median [IQR]2

CsA cyclosporine A; IQR Interquartile range; SCC squamous cell carcinoma 1 Age at moment of starting treatment with oral immunosuppressive drugs 2 From starting oral immunosuppressive drugs until data lock NON-MELANOMA SKIN CANCER AND ORAL IMMUNOSUPPRESSIVE DRUGS 81

Comparison to the Dutch population The SIR for the risk of development of an SCC in this study population was 13.1 [95% CI 6.5-19.7]. One patient developed 5 SCCs; thereby increasing the SIR value. The calculated SIR for the development of an SCC, without this outlier, was 8.8 (95% CI 3.4-14.3). In addition, 3 of the 10 patients already had an SCC before the start of treatment and probably were more prone to develop another SCC. The calculated SIR for the development of an SCC, without these 3 patients, was 10.7 (95% 4.6-16.7). The SIR for the risk of development of an SCC in patients with CsA monotherapy was 25.3 [95% CI 10.3-40.2]. The calculated SIR for the development of SCC in patients with CsA monotherapy, without the aforementioned outlier who developed 5 SCCs, was 14.3 (95% CI 2.9-25.7).

Discussion This is the first study investigating the occurrence of NMSC in a large group of AD patients treated with oral immunosuppressive drugs. NMSC during or after oral immunosuppressive treatments were reported in 18 out of 557 patients (3.2%). The patients who developed NMSC 5 were significantly older than patients who did not develop these malignancies. Follow up did not differ significantly between these groups. However, it is noteworthy that in 4 out of the 18 patients the malignancy was detected under the age of 45 years, which is relatively young.

Literature concerning the risk of developing NMSC during or after oral immunosuppressive treatment in patients with AD are scarce. In a retrospective cohort study, Väkevä et al evaluated 272 patients with various skin diseases treated with CsA, with a median follow-up time of 10.9 years. No NMSC or lymphoma was found in the patients with AD.19 Berth-Jones et al evaluated the use of CsA in 100 patients with AD (mean follow up time 8 weeks). They reported one BCC, that developed in a nevus sebaceous; no SCC was reported.20 Furthermore, there are some case reports describing cases with (cutaneous) lymphoma in AD patients using oral immunosuppressive drugs.21-23

Most information on the development of NMSC after oral immunosuppressive treatment is derived from organ transplant studies. These patients have a markedly increased risk of NMSC.12,24 The cumulative incidence of malignancies is reported to increase in relation to the number of years since transplantation.25 A mean interval between transplantation and tumor diagnosis is reported in the literature: 8 years for patients who received transplants at approximately 40 years of age and about 3 years for those who receive transplants after the age of 60.24,26,27 These results are not entirely applicable to patients with AD, because organ transplant patients often use more than one oral immunosuppressive drug simultaneously and they more often have prolonged treatment, resulting in more long-term data. 82 CHAPTER 5

More recently, data relating oral immunosuppressive treatment to the risk of developing malignancies in other chronic inflammatory diseases have become available. Lymphomas are reported in patients with rheumatoid arthritis treated with methotrexate.28,29 In patients with IBD, various studies have shown that patients treated with thiopurines had an increased risk for the development of NMSC or lymphoproliferative disorders.30-32 In patients with psoriasis, different studies have shown an increased risk of NMSC.33-35 However, most of these patients were also treated with UV light for longer periods, which might have made a major contribution to the increased risk of SCC. Since AD and NMSC are both common diseases, it can be expected that AD patients will develop NMSC, irrespective of the immunosuppressive treatment. Confounding factors, such as other therapeutic interventions, life style factors and occupation (indoor or outdoor) are hard to eliminate and causal relationships are difficult to affirm. Two meta-analyses were found in literature. Deckert et al included 6 systematic reviews on the risk of cancer in patients with AD.36 They concluded that there are no data suggesting that AD itself is associated with an increased risk of NMSC. A more recent meta-analysis performed by Gandini et al included 18 studies (9 on NMSC).37 They concluded that AD patients may be at increased BCC risk, but methodological limitations prevented them from drawing a definitive conclusion.

In the present study, we compared our data concerning SCC with the general Dutch population with a correction for age and found an increased SIR for the development of an SCC. These findings were corrected for external, time dependent influences, by comparing our patients to a patient cohort of the same age in the same time period. The SIR for the risk of development of SCC in the patients included in this study was 13.1 [95% CI 6.5-19.7] (8.8 without outlier with 5 SCC’s). Earlier studies in organ transplantation (2561 patients) and autoimmune hepatitis (45 patients) reported an SIR of 65 [95% CI 53-79] and 28.5 [95% CI 9.9-43.1] respectively.38,39

The SIR of 13.1 suggests that patients with AD treated with oral immunosuppressive drugs are at risk of the development of an SCC. For interpretation of the results it is important to realize that the numbers of SCC’s were low. No significant association was found between the cumulative days of treatment and the risk of development of SCC (p=0.35). In a recent study investigating the incidence of SCC in 59 patients with auto-immune inflammatory rheumatic diseases treated with AZA, a higher cumulative dose and a treatment duration of at least 11 years were qualified as risk factors for the development of SCC.13 In our study only 16 patients had a treatment duration of > 10 years. None of these patients developed an SCC. In addition, no significant differences were seen in the SCC incidence between patients treated ≤ 5 years and patients treated > 5 years with oral immunosuppressive drugs. NON-MELANOMA SKIN CANCER AND ORAL IMMUNOSUPPRESSIVE DRUGS 83

For dermatologists prescribing oral immunosuppressive drugs in daily practice, it is important to know the risk of developing NMSC in individual drugs. A subgroup analysis in our study was only possible for CsA. The SIR for 281 patients treated with CsA monotherapy for development of an SCC was 25.3 [95% CI 10.3-40.2], suggesting an increased risk of developing an SCC during or after CsA treatment. However, one patient developed 5 SCCs; thereby exerting much influence on the SIR value. The calculated SIR for the development of SCC, without this outlier, was 14.3 (95% CI 2.9-25.7). Also in this group, lack of association with treatment duration and the sometimes long intervals between CsA discontinuation and the development of an SCC make the relationship doubtful in some patients.

For interpretation of the results it is important to realize that 5 out of 18 patients with NMSC during or after oral immunosuppressive treatment had a previous similar type of tumour before start of treatment. These patients are probably more prone to develop the tumour; it is not clear what the contribution of the immunosuppressive treatment was to the development of 5 new tumours. Robsahm et al showed that patients with a history of SCC were more at risk for developing another SCC (SIR of 9.88 in women and 10.1 in men).40 In our study, the calculated SIR for the development of an SCC, without these the three patients who already had an SCC in the past, was 10.7 (95% 4.6-16.7).

Limitations The median duration of follow-up in this study was 6.0 years [IQR 3.0-10.2], which is relatively short. However, the incidence of NMSC was comparable in the patients with follow-up ≤ 5 years (n=245) compared to the group with a follow-up > 5 years (n=312) (data not shown). Data concerning BCCs were collected with utmost care, but there will probably be an underestimation of the real incidence. This might be attributed to the fact that BCCs are regularly treated without histological confirmation. Data on a history of UV light therapy were not available for all evaluated patients, thus the influence of UV light therapy on the development of NMSC is unclear. However it is common in the Netherlands to prescribe UV light therapy only in short courses of up to 4 months. Psoralen combined with ultraviolet A light (PUVA), which is associated with NMSC, is rarely prescribed in AD patients in the Netherlands. In addition, data on skin type, phototype, naevi, hair and eye colour and history of sunburns were lacking. Data on tumour aggressivity were not available. Finally, the data of the general population that were used to calculate the SIR for SCC matched our cohort on age and calendar year, but not on sex. 84 CHAPTER 5

Conclusion NMSC during or after long-term treatment with oral immunosuppressive drugs was found in 18 out of 557 (3.2%) patients with AD with an SIR of 13.1 for SCC. For interpretation of the results it is important to include other factors: in this study we found a lack of association between treatment duration and the risk of developing a tumour, a history of a malignancy before treatment in 5 out of 18 patients and a long interval between treatment discontinuation and the development of the tumour in some patients. It is always important to balance benefit of treatment against the potential risks in each individual patient. Patients treated with oral immunosuppressive drugs should regularly visit the dermatologist for monitoring treatment effect and safety laboratory tests. Thorough inspection of the skin during each visit enables early detection and treatment of NMSC. As the occurrence of NMSC in our study was independent of treatment duration, skin inspection should start within the first year of treatment. NON-MELANOMA SKIN CANCER AND ORAL IMMUNOSUPPRESSIVE DRUGS 85

References

1 Ring J, Alomar A, Bieber T et al. Guidelines 9 Demir T, Ozel L, Gokce AM et al. for treatment of atopic eczema (atopic Cancer Screening of Renal Transplant dermatitis) part I. J Eur Acad Dermatol Patients Undergoing Long-Term Venereol 2012; 26: 1045-60. Immunosuppressive Therapy. Transplant Proc 2015; 47: 1413-7. 2 Roekevisch E, Spuls PI, Kuester D et al. Efficacy and safety of systemic treatments 10 Vajdic CM, van Leeuwen MT. Cancer inci- for moderate-to-severe atopic dermatitis: dence and risk factors after solid organ a systematic review. J Allergy Clin Immunol transplantation. Int J Cancer 2009; 125: 2014; 133: 429-38. 1747-54.

3 Haeck IM, Knol MJ, Ten Berge O et al. 11 Kasiske BL, Snyder JJ, Gilbertson DT et al. Enteric-coated mycophenolate sodium Cancer after kidney transplantation in the versus cyclosporin A as long-term treat- United States. Am J Transplant 2004; 4: ment in adult patients with severe atopic 905-13. dermatitis: a randomized controlled trial. J Am Acad Dermatol 2011; 64: 1074-84. 12 Geissler EK. Post-transplantation malignancies: here today, gone tomorrow? Nat Rev 5 4 Schram ME, Roekevisch E, Leeflang MM Clin Oncol 2015; 12: 705-17. et al. A randomized trial of methotrexate versus azathioprine for severe atopic 13 van den Reek JM, van Lumig PP, Janssen eczema. J Allergy Clin Immunol 2011; 128: M et al. Increased incidence of squamous 353-9. cell carcinoma of the skin after long-term treatment with azathioprine in patients 5 Schmitt J, Schakel K, Folster-Holst R et al. with auto-immune inflammatory rheumatic Prednisolone vs. ciclosporin for severe diseases. J Eur Acad Dermatol Venereol adult eczema. An investigator-initiated 2014; 28: 27-33. double-blind placebo-controlled multicentre trial. Br J Dermatol 2010; 162: 661-8. 14 Khan N, Abbas AM, Lichtenstein GR et al. Risk of lymphoma in patients with 6 van der Schaft J, Politiek K, van den Reek ulcerative colitis treated with thiopurines: JM et al. Drug survival for azathioprine and a nationwide retrospective cohort study. enteric-coated mycophenolate sodium Gastroenterology 2013; 145: 1007-15 e3. in a long-term daily practice cohort of adult patients with atopic dermatitis. Br J 15 Setshedi M, Epstein D, Winter TA et al. Use Dermatol 2016; 175: 199-202. of thiopurines in the treatment of inflammatory bowel disease is associated with 7 Politiek K, van der Schaft J, Coenraads PJ et an increased risk of non-melanoma skin al. Drug survival for methotrexate in a daily cancer in an at-risk population: a cohort practice cohort of adult patients with severe study. J Gastroenterol Hepatol 2012; 27: atopic dermatitis. Br J Dermatol 2016; 174: 385-9. 201-3. 16 Public pathology database. Available at 8 van der Schaft J, Politiek K, van den Reek http://www.palga.nl/en/ Last consulted on JM et al. Drug survival for ciclosporin A 19 August 2016. in a long-term daily practice cohort of adult patients with atopic dermatitis. Br J 17 Comprehensive Cancer Centre the Dermatol 2015; 172: 1621-7. Netherlands. Available at http://www. cijfersoverkanker.nl Last consulted on 22 July 2016. 86 CHAPTER 5

18 Boyle P, Parkin DM. Statistical methods for 28 Mariette X, Cazals-Hatem D, Warszawki J registries. Available at http://www.iarc.fr/ et al. Lymphomas in rheumatoid arthritis en/publications/pdfs-online/epi/sp95/ patients treated with methotrexate: a 3-year sp95-chap11.pdf. Last consulted on 29 prospective study in France. Blood 2002; April 2016. 99: 3909-15.

19 Vakeva L, Reitamo S, Pukkala E et al. Long- 29 Buchbinder R, Barber M, Heuzenroeder term follow-up of cancer risk in patients L et al. Incidence of melanoma and other treated with short-term cyclosporine. Acta malignancies among rheumatoid arthritis Derm Venereol 2008; 88: 117-20. patients treated with methotrexate. Arthritis Rheum 2008; 59: 794-9. 20 Berth-Jones J, Graham-Brown RA, Marks R et al. Long-term efficacy and safety of cyclo- 30 Peyrin-Biroulet L, Khosrotehrani K, Carrat sporin in severe adult atopic dermatitis. Br J F et al. Increased risk for nonmelanoma Dermatol 1997; 136: 76-81. skin cancers in patients who receive thiopurines for inflammatory bowel disease. 21 Sinha A, Velangi S, Natarajan S. Gastroenterology 2011; 141: 1621-28 e1-5. Non-Hodgkin’s lymphoma following treatment of atopic eczema with cyclosporin A. 31 Beaugerie L, Brousse N, Bouvier AM et al. Acta Derm Venereol 2004; 84: 327-8. Lymphoproliferative disorders in patients receiving thiopurines for inflammatory 22 Nakamura S, Takeda K, Hashimoto Y et al. bowel disease: a prospective observational Primary cutaneous CD30+ lymphoprolifera- cohort study. Lancet 2009; 374: 1617-25. tive disorder in an atopic dermatitis patient on cyclosporine therapy. Indian J Dermatol 32 Kandiel A, Fraser AG, Korelitz BI et al. Venereol Leprol 2011; 77: 253. Increased risk of lymphoma among inflammatory bowel disease patients treated with 23 Mougel F, Dalle S, Balme B et al. Aggressive azathioprine and 6-mercaptopurine. Gut CD30 large cell lymphoma after cyclospo- 2005; 54: 1121-5. rine given for putative atopic dermatitis. Dermatology 2006; 213: 239-41. 33 Paul CF, Ho VC, McGeown C et al. Risk of malignancies in psoriasis patients treated 24 Euvrard S, Kanitakis J, Claudy A. Skin with cyclosporine: a 5 y cohort study. J cancers after organ transplantation. N Engl Invest Dermatol 2003; 120: 211-6. J Med 2003; 348: 1681-91. 34 Pouplard C, Brenaut E, Horreau C et al. Risk 25 Wisgerhof HC, van der Geest LG, de of cancer in psoriasis: a systematic review Fijter JW et al. Incidence of cancer in and meta-analysis of epidemiological kidney-transplant recipients: a long-term studies. J Eur Acad Dermatol Venereol cohort study in a single center. Cancer 2013; 27 Suppl 3: 36-46. Epidemiol 2011; 35: 105-11. 35 Chiesa Fuxench ZC, Shin DB, Ogdie Beatty 26 Euvrard S, Kanitakis J, Pouteil-Noble C et A et al. The Risk of Cancer in Patients With al. Comparative epidemiologic study of Psoriasis: A Population-Based Cohort Study premalignant and malignant epithelial cuta- in the Health Improvement Network. JAMA neous lesions developing after kidney and Dermatol 2016; 152: 282-90. heart transplantation. J Am Acad Dermatol 1995; 33: 222-9. 36 Deckert S, Kopkow C, Schmitt J. Nonallergic comorbidities of atopic eczema: an over- 27 Webb MC, Compton F, Andrews PA et al. view of systematic reviews. Allergy 2014; Skin tumours posttransplantation: a retro- 69: 37-45. spective analysis of 28 years’ experience at a single centre. Transplant Proc 1997; 29: 828-30. NON-MELANOMA SKIN CANCER AND ORAL IMMUNOSUPPRESSIVE DRUGS 87

37 Gandini S, Stanganelli I, Palli D et al. Atopic dermatitis, naevi count and skin cancer risk: A meta-analysis. J Dermatol Sci 2016; 84: 137-43.

38 Jensen P, Hansen S, Moller B et al. Skin cancer in kidney and heart transplant recipients and different long-term immunosuppressive therapy regimens. J Am Acad Dermatol 1999; 40: 177-86.

39 Leung J, Dowling L, Obadan I et al. Risk of non-melanoma skin cancer in autoimmune hepatitis. Dig Dis Sci 2010; 55: 3218-23.

40 Robsahm TE, Karagas MR, Rees JR et al. New malignancies after squamous cell carcinoma and melanomas: a population-based study from Norway. BMC Cancer 2014; 14: 210. 5

Is there an increased risk of cervical neoplasia in atopic dermatitis patients treated with oral immunosuppressive drugs?

Floor M. Garritsen, René H.M. Verheijen, Kees G. Gerestein, Arjan D. van Zuilen, Jart A.F. Oosterhaven, Marijke R. van Dijk, Carla A.F.M. Bruijnzeel-Koomen, Marie-Louise A. Schuttelaar, Marjolein S. de Bruin-Weller

Accepted for publication in J Eur Acad Dermatol Venereol 2017 90 CHAPTER 6

Abstract Background: Oral immunosuppressive drugs are frequently prescribed in young women with atopic dermatitis (AD). Immunocompromised patients may have a higher risk of developing high risk HPV infections, CIN and cervical carcinoma. Most literature on patients using oral immunosuppressive drugs is available in organ transplant patients. Literature on the risk of developing cervical carcinoma in AD patients treated with oral immunosuppressive drugs is lacking. At this moment there is no clear guideline/consensus on this topic, but in daily practice, questions arise concerning whether this risk is increased and whether more intensive screening in women using immunosuppressive drugs should take place. Objective: To investigate the occurrence of cervical carcinoma in women with AD treated with oral immunosuppressive drugs. Methods: In this retrospective cohort study in two university medical centres in the Netherlands, all female adult AD patients receiving oral immunosuppressive drugs (cyclosporine A, azathioprine, methotrexate, mycophenolate mofetil, enteric-coated mycophenolic acid and extended release tacrolimus) for more than 2 months between 1989 and January 1st 2014 were included. Patient files in the national histopathology register were screened for PAP3a, CIN I, CIN II, CIN III and cervical carcinoma. Results: A total of 257 female AD patients with one or more treatment episodes from 1989 until January 1st 2014 were identified and included in this study. In 189 patients (73.5%) results of cervical examination were reported in the national histopathology database. Median total duration of treatment in these 189 women was 407.0 days (IQR 243.0-940.0). No cervical carcinoma during or following immunosuppressive therapy was found in our patient group. Conclusions: No intensified screening programme for cervical neoplasia seems necessary for women with AD using oral immunosuppressive drugs. CERVICAL NEOPLASIA AND ORAL IMMUNOSUPPRESSIVE DRUGS 91

Background Oral immunosuppressive drugs including cyclosporine A (CsA), azathioprine (AZA), methotrexate (MTX), mycophenolic acid (MPA) and mycophenolate mofetil (MMF) are increasingly prescribed to patients with severe and difficult to treat atopic dermatitis (AD). It is known from transplantation medicine that immunosuppression is associated with an increased risk of infections, but also with malignancies like non-melanoma skin cancer, lymphoma and cervical carcinoma.1,2

Cervical carcinoma is a common cause of death from cancer in women. Cervical intra-epithelial neoplasia (CIN) is a precursor of cervical carcinoma. CIN is divided into CIN I, CIN II and CIN III based on the presence of mitotic activity and nuclear atypia within respectively the basal third, two thirds and whole thickness of the epithelium. After years, CIN can progress into cervical carcinoma.3 There are national screening programmes in various countries to detect the asymp- tomatic, preinvasive cervical lesions.

The most important risk factor for developing cervical carcinoma is a persistent infection with oncogene types of the Human Papillomavirus (HPV). HPV type 16 and 18 are the most common HPV types in cervical cancer. Immunocompromised women are thought to have a higher risk of 6 developing high risk HPV infections, CIN and cervical carcinoma.4

Most literature on patients using oral immunosuppressive drugs is available in organ transplant patients. Some of these studies conclude that it is not clear whether the risk of high risk HPV infections and CIN is increased compared to healthy women5,6, while other studies found increased risks of CIN in renal transplant patients7 and an increased risk of infection related tumors.8

Oral immunosuppressive drugs are frequently prescribed in young women with AD, also in women who are not yet included in the national screening programme for cervical neoplasia. In daily practice, questions arise concerning whether the risk of cervical carcinoma is increased and whether more intensive screening should take place. Therefore the aim of this study was to investigate the occurrence of cervical carcinoma and CIN in women with AD treated with oral immunosuppressive drugs. 92 CHAPTER 6

Methods Design and study population Data were collected from the dermatology departments of the University Medical Center Utrecht and Groningen, the Netherlands, from 1989 until January 1st 2014. All female adult AD patients receiving oral immunosuppressive drugs (CsA, AZA, MTX, MMF, EC-MPS and [extended release] tacrolimus) for more than 2 months were included. Medical records were screened for type of medication, age and duration of follow-up from the first starting date of oral immunosuppressive treatment.

Cervical screening in the Netherlands Until January 1st 2017, all women aged 30-60 years were called to participate in a screening programme every five years. Pap smear test results are interpreted according to the CISOE-A classification, which can be translated into the Bethesda 2001 or Papanicolaou classification.9

Outcome All patient files in the histopathology register (Pathologisch Anatomisch Landelijk Geautomatiseerd Archiefsysteem [PALGA]), a database for pathology reports in the Netherlands with national coverage, were screened for PAP3a or higher, CIN I, CIN II, CIN III and cervical carcinoma until May 16th 2014. Carcinomas that developed within six months after the start of the treatment were considered not being related to drug treatment and were excluded from analysis. In patients with CIN, additional information on treatment duration and medical history was collected.

Comparison with the general population The total number of cervical carcinomas was compared to the number of expected cases in the general Dutch population in the same period, corrected for age, by using data of the National Dutch Cancer Registry.

Results Characteristics of the study population A total of 257 female using immunosuppressive drugs from 1989 until January 1st 2014 were included (table 1). There was a wide variation in treatment duration (figure 1). CERVICAL NEOPLASIA AND ORAL IMMUNOSUPPRESSIVE DRUGS 93

Table 1 Patient characteristics All patients (n=257) Age at data lock (median [IQR]) 44.3 [32.3-55.7] Age at inclusion* (median [IQR]) 37.9 [24.5-48.7] Duration of follow-up in years** (median [IQR]) 5.6 [2.7-9.2] Total patients years of follow-up 1636.6 Treatment characteristics CsA only (%) 133 (51.8) AZA only (%) 6 (2.3) EC-MPS only (%) 5 (1.9) MTX only (%) 9 (3.5) > one oral immunosuppressive drug (%) 104 (40.5)

AZA azathioprine; CsA cyclosporine A; EC-MPS enteric-coated mycophenolate sodium; IQR Interquartile range; MTX methotrexate

* Age at moment of starting treatment with oral immunosuppressive drugs ** From starting of oral immunosuppressive drugs until data lock

> 10 years 7 6 > 5 years 16 4-5 years 10 3-4 years 19 2-3 years 24 1-2 years 59

Duration of treatment 0-1 years 122 0 50 100 150 Patients (n)

Figure 1 Duration of treatment

PALGA results Fifty-one out of 257 patients (19.8%) were < 30 years old at the time of the PALGA pathology database search; it can be expected that they did not yet receive a call for the national screening programme. However, in six of them, results of cervical examination were reported, indicating that these patients underwent opportunistic cervical screening. Indications were not recorded and hence unknown. In 183 out of the 206 patients aged >30 years old (88.8%), results of cervical examination were reported in the PALGA database. Together with the aforementioned 6 women aged <30 years old, a total of 189 women (73.5%) was screened (figure 2). Median total duration of treatment in these 189 women was 407.0 days (IQR 243.0-940.0). In 5 women, one or more abnormal PAP result was found during or after immunosuppressive treatment. In one of these patients the follow-up Pap smears were normal (Pap 1). In one patient a CIN II was found and in three other patients a CIN III was detected (table 2). No cervical carcinoma was found. 94 CHAPTER 6

Table 2 Characteristics of patients with CIN II and III after treatment Patient Diagnosis Treatment Age at Cumulative Time Time History diagnosis days of oral between between of malig- (years) immunosup- start of oral stop of oral nancy pressive drug immunosup- immunosup- use until pressive drug pressive drug diagnosis use and diag- use and diagnosis (years) nosis (years) 1 CIN III CsA, 50 1125 12.1 5.5 Breast EC-MPS carcinoma 2 CIN II CsA, MTX 42 381 13.1 Treatment No continued after diagnosis 3 CIN III CsA 31 356 4.4 3.4 No 4 CIN III CsA 30 579 11.7 0.2 No

CIN cervical intraepithelial neoplasia; CsA cyclosporine A; EC-MPS enteric-coated mycophenolate sodium; MTX methotrexate

Comparison to the general population The number of expected cervical carcinomas in the general Dutch population in the same time period corrected for age was 0.09 cases. The occurrence of cervical carcinoma in our study population is not increased, compared to the general population. Since cervical premalignan- cies are not systematically registered in the Netherlands, no comparison could be made for Pap3a or higher and CIN.

Patients treated with oral immunosuppressive drugs (n=257 patients) No cervical examinations in pathology database (n=68 patients) Cervical examinations in pathology database (n=189 patients)

PAP 3a PAP 3b (n=4 patients) (n=1 patient)

Following pap Following cervical Following cervical Following cervical smears: Pap 1 biopsy CIN II biopsy CIN III biopsy CIN III (n=1 patient) (n=1 patient) (n=2 patient) (n=1 patient)

Following pap Following pap Following pap No more Pap smears: Pap 2 smears: Pap 2 smears: Pap 1 smears performed (n=1 patient) (n= 1 patient) (n= 1 patient) (n=1 patient)

Figure 2 Further evaluation of patients with a PAP 3a or PAP 3b CERVICAL NEOPLASIA AND ORAL IMMUNOSUPPRESSIVE DRUGS 95

Discussion Literature on the risk of cervical carcinoma in patients with dermatological diseases treated with systemic immunosuppressive drugs is scarce. In 2013, Clowry et al undertook an audit of 104 female patients on systemic immunosuppressive agents for a chronic inflammatory dermatologic condition.10 They reported an incidence of women referred for further treatment, indicating a history of dysplasia, of 12.5% (corresponding with the national prevalence of cervical intraepithelial neoplasia).

Most previously published information on changed HPV status, the risk of CIN and cervical carcinoma during or after oral immunosuppressive treatment is reported in organ transplant recipients. Literature on this topic is contradicting. Some studies report an increased incidence of cervical cancer.11,12 Other studies showed no increased HPV infections and no increased cancer risk.5,6,13,14 Several publications and guidelines have suggested intensifying the cervical cancer screening in organ transplant patients. Some other guidelines, including the Dutch guideline for CIN, state that no extra screening is needed for organ transplant women taking oral immunosuppressive drugs.17

Data in transplant patients could not simply be applied to AD patients, because transplant 6 patients often use more than one drug at the same time and the treatment is lifelong. In the current study, we found CIN II and CIN III in 2% of the screened women. Based on literature, this is in line with the general population in the Netherlands: in 2010, 5000 women (approximately 1% of the total amount of screened women) had been treated for a CIN II or CIN III.15,16

Primary prevention of HPV-16 and HPV-18 induced CIN-lesions and cervical carcinoma is possible by vaccination with a prophylactic HPV-vaccine. Since September 1st 2009, this HPV vaccination is part of the Dutch National vaccination programme and offered to all twelve year old girls. Unfortunately, the percentage of girls participating in this vaccination programme in the Netherlands is low (on January 1st 2017, 53.4% of girls born in 2002 had taken part in the vaccination program) so women are still advised to take part in the screening programme.

Based on our results and results of Clowry et al10, no intensified screening programme for women with AD treated with oral immunosuppressive drugs, aged >30 years, seems to be indicated. These women can participate in the general screening programme. 96 CHAPTER 6

However, both our studies evaluated a small patient population with a limited follow up period and therefore an increased risk of cervical carcinoma in these women still cannot be ruled out completely, especially not in the case of long term treatment. Therefore, for women aged <30 years at the start of the oral immunosuppressive treatment who receive long-term treatment with oral immunosuppressive drugs early screening is a matter of shared decision making. In these patients one or two extra cervical screening moments with a maximal time interval of five years can be considered, though this is for now not based on evidence in AD patients. In these cases, it should be taken into account that cytological screening in these younger patients results in the detection a higher amount of insignificant, low risk abnormalities.18 Besides, when using the HPV-screening, it should be kept in mind that the prevalence of HPV in particular this young age group is very high (approximately 20%), potentially leading to many unnecessary interventions. Finally, the HPV screening has a lower specificity in this age-group.

Conclusion and recommendations for daily practice In this study evaluating cervical (pre)malignancies in AD patients treated with oral immunosuppressive drugs, no cervical carcinoma was reported in a group of 257 women, of who 189 patients (median total duration of treatment 407.0 days) underwent cervical screening. The incidences of clinically relevant CIN II and CIN III lesions seem to be not different from literature data of the general Dutch population. Therefore no intensified screenings programme seems necessary for women with AD using oral immunosuppressive drugs. CERVICAL NEOPLASIA AND ORAL IMMUNOSUPPRESSIVE DRUGS 97

References

1 Vajdic CM, van Leeuwen MT. Cancer inci- 11 Vajdic CM, McDonald SP, McCredie MR dence and risk factors after solid organ et al. Cancer incidence before and after transplantation. Int J Cancer 2009; 125: kidney transplantation. JAMA 2006; 296: 1747-54. 2823-31.

2 Hernandez Mdel P, Martin P, Simkins 12 Porreco R, Penn I, Droegemueller W et al. J. Infectious Complications After Liver Gynecologic malignancies in immuno- Transplantation. Gastroenterol Hepatol (N Y) suppressed organ homograft recipients. 2015; 11: 741-53. Obstet Gynecol 1975; 45: 359-64.

3 Snijders PJ, Steenbergen RD, Heideman DA 13 Mazanowska N, Pietrzak B, Kaminski P et et al. HPV-mediated cervical carcinogen- al. Prevalence of cervical high-risk human esis: concepts and clinical implications. J papillomavirus infections in kidney graft Pathol 2006; 208: 152-64. recipients. Ann Transplant 2013; 18: 656-60. 4 Wieland U, Kreuter A, Pfister H. Human papillomavirus and immunosuppression. 14 Madeleine MM, Finch JL, Lynch CF et al. Curr Probl Dermatol 2014; 45: 154-65. HPV-related cancers after solid organ transplantation in the United States. Am J 5 Morrison EA, Dole P, Sun XW et al. Low Transplant 2013; 13: 3202-9. prevalence of human papillomavirus infection of the cervix in renal transplant 15 van der Veen N, Carpay MEM, van Delden recipients. Nephrol Dial Transplant 1996; JA, Brouwer E, Grievink L, Hoebee 6 11: 1603-6. B, et al. Uitvoeringstoets wijziging bevolkingsonderzoek baarmoederhal- 6 Pietrzak B, Mazanowska N, Ekiel AM et al. skanker 2013. Available at http://www. Prevalence of high-risk human papilloma- rivm.nl/Documenten_en_ publicaties/ virus cervical infection in female kidney Wetenschappelijk/Rapporten/2013/ graft recipients: an observational study. oktober/Uitvoeringstoets_wijziging_ Virol J 2012; 9: 117. bevolkingsonderzoek_baarmoederhal- skanker_2013. 7 Hinten F, Meeuwis KA, van Rossum MM et al. HPV-related (pre)malignancies of the 16 van Ballegooijen M, Weerding WJ. De female anogenital tract in renal transplant praktijk van het bevolkingsondezoek naar recipients. Crit Rev Oncol Hematol 2012; baarmoederhals kanker in Nederland in 84: 161-80. 2001. Rotterdam : Erasmus Universiteit Rotterdam, Instituut Maatschappelijke 8 Grulich AE, van Leeuwen MT, Falster MO et Gezondheidszorg, 2003. al. Incidence of cancers in people with HIV/ AIDS compared with immunosuppressed 17 Dutch National Guideline CIN, AIS en VAIN transplant recipients: a meta-analysis. 2015. Available at http://www.oncoline.nl/ Lancet 2007; 370: 59-67. cin-ais-en-vain. Last accessed 5 July 2017.

9 Bulk S, Van Kemenade FJ, Rozendaal L et al. 18 Daily LR, Erickson BK, Smith HJ et al. High The Dutch CISOE-A framework for cytology Rates of Cervical Intraepithelial Neoplasia 2 reporting increases efficacy of screening and Cervical Intraepithelial Neoplasia 3 in upon standardisation since 1996. J Clin High-Risk Young Women With Low-Grade Pathol 2004; 57: 388-93. Cervical Cytology Screening. Obstetrics and Gynecology 2014; 123: 91s-s. 10 Clowry J, Callanan I, Eustace K et al. Cervical screening uptake in a dermatology cohort on immunosuppressive agents. Br J Dermatol 2015; 172: 1682-3.

Pregnancy and fetal outcomes after paternal exposure to azathioprine, methotrexate or mycophenolic acid: a critically appraised topic

Floor M. Garritsen, Marcel P.H. van den Broek, Arjan D. van Zuilen, Herma H. Fidder, Marjolein S. de Bruin-Weller, Phyllis I. Spuls

Br J Dermatol 2017;176(4):866-877 100 CHAPTER 7

Clinical scenario A 35-year old male patient with atopic dermatitis (AD) is being treated with mycophenolate mofetil (MMF). He and his female partner wish to have children. They wonder whether the use of MMF, or any of the other alternative oral immunosuppressive drugs to treat AD, will be a risk for the fetus and whether this medication should be discontinued. He searched on the internet and found that the use of MMF was contraindicated.

Clinical question What are the risks of adverse pregnancy outcomes (preterm birth, stillbirth, spontaneous abortion) and fetal outcomes (low birth weight and congenital malformations) after paternal exposure to azathioprine (AZA), methotrexate (MTX) or mycophenolic acid (MPA)?

Background Oral immunosuppressive drugs including cyclosporine A (CsA), AZA, MTX, MPA and MMF are increasingly being prescribed to patients with chronic skin diseases. These drugs are also commonly prescribed in young patients who wish to conceive. Information concerning pregnancy outcomes after paternal drug use is predominantly available from transplant recipients who commonly use these drugs. However, the adverse pregnancy and fetal outcomes of the individual agents often remain unclear due to the fact that treatment in these transplant patients often consists of a combination of these oral immunosuppressive drugs.

Pregnancy outcomes depend on both maternal and paternal factors. Paternal drug use may affect ability to conceive, pregnancy and fetal outcomes.

Based on human physiology, drug distribution to semen is expected to have two pathways: 1) Its presence in seminal fluid in the prostate or seminal vesicles (major contribution) and 2) its presence in the sperm in the testes or epididymis (minor contribution due to the blood/testis barrier). Because seminal fluid and sperm are mixed together at the moment of ejaculation, the duration of exposure of the spermatocytes to drugs and metabolites from seminal fluid to spermatocytes is shorter than exposure during production and maturation in the testes and epididymis.

The mechanisms behind a paternal drug effect on pregnancy outcomes may include: (I) genetic or chromosomal damage of the spermatocytes; (II) drugs or metabolites might influence sperm maturation; (III) drugs or metabolites in seminal fluid might have a direct effect on the uterus; (IV) a systemic effect of the drug or its metabolites on the female through absorption of the seminal fluid by the vaginal mucosa.1,2 RISKS OF PATERNAL EXPOSURE TO ORAL IMMUNOSUPPRESSIVE DRUGS 101

What is the aim of this critically appraised topic? The opinions regarding the use of CsA in men who wish to have children are fairly straightforward (i.e. CsA can be used), but there are different conceptions about the use of the other oral immunosuppressive drugs.3,4 In the summary of product characteristics (SmPc), for men who wish to conceive, it is recommended that men using AZA, MTX, MMF and MPA take adequate contraceptive measures (for MTX, MMF and MPA even for three months after treatment).5 However, the existing reviews and guidelines are contradictory. Some guidelines mention the possibility of cryopreservation, while others do not even mention the risks of taking these drugs in men who wish to conceive.5-17 The British Association of Dermatologists’ guideline for the use of AZA concludes that several studies indicate that male patients receiving AZA father healthy children and that AZA at standard doses does not appear to affect male fertility.7 A recent guideline for AD from the American Academy of Dermatology8, guidelines for treatment of AD from the European Academy of Dermatology and Venereology9, the Asia-Pacific consensus guideline10 and a recent Japanese guideline on AD do not give information about men using oral immunosuppressive drugs who wish to conceive.11 The psoriasis guideline of the National Clinical Guideline Centre (UK) advises offering CsA to men who are considering conception and in whom systemic treatment cannot be avoided.12 Finally, the U.S. Food and Drug Administration evaluates drugs but does not provide an evidence table with recommendations for the paternal use of the investigated drugs in this CAT.15,16

Therefore, there remains concern about prescribing these drugs to young men in daily practice. 7 Because confusion still exists, the purpose of this CAT was to evaluate the literature dealing with the effect of paternal use of azathioprine, methotrexate, mycophenolic acid or mycophenolate mofetil on pregnancy and fetal outcomes. 102 CHAPTER 7

Methods Search and selection A systematic search was conducted in PubMed, Embase, GREAT and the Cochrane Central Register of Controlled Trials (CENTRAL) on the 19th of October 2015. No language restrictions were used. An additional hand search for references and recent reviews was performed to find missed relevant publications (figure 1). Titles and abstracts were screened for relevance regarding our clinical question. Studies on the effects of paternal use of AZA, MTX, MPA and MMF on pregnancy and fetal outcomes were included. Studies in which immunosuppressants were given for non-dermatological conditions were included as well.

Studies that were not available as full text, reviews, conference abstracts and animal studies were excluded. Studies evaluating only the quality and quantity of the sperm and studies evaluating the ability to conceive without looking at the pregnancy or fetal outcomes were also excluded.

Search 19-10-2015 (‘male’ OR ‘man’ OR ‘men’ OR ‘paternal’) AND (‘pregnancy’ OR ‘conception’) AND (‘azathioprine’ OR ‘Imuran’ OR ‘AZA’ OR ‘MTX’ OR ‘methotrexate’ OR ‘metotrexate’ OR mycophenolate’ OR ‘mycophenol’ OR ‘mycophenolic’ OR ‘mycophenolaat’ OR ‘cellcept’ OR ‘celcept’ OR ‘myfortic’)

Outcomes The outcomes of this study were adverse pregnancy outcomes (preterm birth, stillbirth, spontaneous abortion) and fetal outcomes, such as low birth weight and congenital malformations.

Critical appraisal The articles were appraised with respect to validity, according to predetermined criteria (table 1). Particular attention was given to a complete view of factors that may affect the pregnancy outcome, such as comorbidities and concomitant treatments in both mothers and fathers. A comparison of the results of the study patients to a control population existing of men not using oral immunosuppressive drugs, fathering children, was noted if present. The results of all articles were further evaluated.

Results Selection of articles and critical appraisal Titles and abstracts of 1741 articles were screened (figure 1). After screening the full text of 82 studies, 20 were selected for critical appraisal (table 1). Six new relevant articles were found by searching the references of related articles and citations of the 20 included studies. RISKS OF PATERNAL EXPOSURE TO ORAL IMMUNOSUPPRESSIVE DRUGS 103

Search on October 19th 2015

Pubmed Embase Cochrane CENTRAL GREAT Database 505 1023 149 64

Title and abstract screening Other treatment 812 1741 Review 381 Not in human beings 40 Not in pregnancy 205 Not in men 221

Full text screening 82 No full text available 15 Review 15 Conference abstract 13 Not in men 8 Not in human beings 3 Other outcome 1 Other treatments 3 20 Letter to editor 1 Double publication 3

Screening reviews and references (6 extra) 26 7

AZA 9 MTX 10 MPA/MMF 2 Various 5

Figure 1 Flowchart 104 Table 1 Critical appraisal Year Study Design Randomisation Dose and Evaluation of Standardi- Selective Paternal and Medical Maternal Maternal Interventions Complete Comparisson and observer durationb mono-therapyc sation of reportinge maternal historyg age and medication during data with healthy blindinga therapyd comorbiditiesf smokingh usei pregnancyj outcomek populationl CHAPTER 7 1968 Frank18 POS - + + - +/------+ - 1974 Lingardh19 POS - + - - +/- +/- - - - - + - 1977 Kroner20 CR - + - + +/------+ - 1980 Blatt21 RQ - - - - + ------1980 Matthews22 CR - +/- - + +/------+ - 1983 Perry23 CR - + + + +/------+ - 1994 Wagoner24 RQ - - - +/- + +/------2000 Rajapakse25 RQ - + + - + +/- + +/- + - - + 2001 Ben Neriah26 CR - + + + +/- - - +/- - + + - 2003 Oefferlbauer- CR - + + + +/- +/- + +/- + - + - Ernst27 2003 Francella28 RMCR - + + - + - - + - - + + 2004 NØrgård1 RMCR - +/- + +/- + + - + + - +/- + 2006 Griggs29 CR - + + + +/------+ - 2006 Østensen30 Review - - - +/- +/------2006 Hasani32 * POS - - - +/- +/------+ 2007 Østensen31 RQ - - + +/- + +/------2009 Paschou33 RMCR - +/- - + +/------+ - 2009 Lamboglia34 CR - + - + +/- +/- - +/- - - + - 2010 Teruel35 RQ - + + - + - + + - - - + 2011 Beghin36 POS - + + - + - - +/- - + + - 2011 Xu37 RMCR - + ------+ - 2012 Viktil38 RMCR - - + +/- + - - +/- - - + - 2012 Hoeltzenbein39 POS - +/- + - + +/- + + - + - + 2013 Jones41 RQ - - - +/- +/------+ + 2014 Weber- POS - + + - + +/- + + + + + + Schoendorfer40 2015 Wallenius42 RMCR - +/- + - - - - +/- - - - + CR case report; POS prospective observational study; RMCR retrospective medical chart review; RQ retrospective questionnaires * The original article of Hasani et al was not found but the data are cited in full in the review by Armentiet al.43 RISKS OF PATERNAL EXPOSURE TO ORAL IMMUNOSUPPRESSIVE DRUGS 105

Table 1 Critical appraisal a - non randomized and no observer blinding b + known dosage and duration of treatment; +/- known dosage or known duration of treatment; - unknown dosage and duration of treatment c + evaluation of oral immunosuppressive monotherapy; - no evaluation of monotherapy or unknown d + standardized therapy for all patients; +/- unknown; - no standardized therapy e + all prespecified outcomes have been addressed and reported; +/- no prespecified outcomes; - differences between predefined methods and final outcomes f + mention of comorbidities of both fathers and mothers; +/- mention of comorbidities of fathers or mothers; - not mentioned g + mention of the existence of familiar diseases, previous miscarriages or previous children with birth defects; - not mentioned h + mention of maternal smoking behavior and maternal age; +/- mention of maternal smoking behavior or maternal age; - not mentioned i + mention of maternal medication use; - not mentioned j + mention of the use of (medical) interventions during pregnancy (like invasive prenatal diagnostics); - not mentioned k + loss to follow up until delivery < 10%; +/- loss to follow up until delivery 10-20%; - loss to follow up until delivery >20% or not mentioned l + comparison with healthy population; - no comparison with healthy population 7 106 CHAPTER 7

Overall study quality Twenty-six articles were included for data extraction: 9 studies on AZA, 10 on MTX, 2 on MPA/ MMF and 5 on different oral immunosuppressive drugs.1,18-42 These included six retrospective questionnaire-based studies, six prospective observational studies, six retrospective medical chart reviews, one review with original data and seven case reports. None of the studies were randomized or conducted blinded. In 14 studies, dose and duration of treatment were described. In the other studies, this information was (partly) missing. A (sub) analysis of treatment with AZA, MTX, MPA or MMF would be valuable, but was not provided in 11 studies. An effort was made to collect information on comorbidities of fathers and mothers, age and (co)medication of both parents, previous pregnancies and known familiar genetic disorders. However, this information was rarely mentioned. Nine of the 26 studies had a control group.1,25,28,32,35,39-42 An overall evaluation of the critical appraisal shows that the studies of Blatt21, Ostensen30 and Xu37 were rated as the least reliable. They did not compare their results to the general population and did not evaluate monotherapy. Furthermore, information on comorbidities, family history, maternal smoking behavior and maternal drug use was lacking in these studies, making the results less reliable. The study of Weber-Schoendorfer40 had the greatest validity. They evaluated monotherapy and provided complete information on dose an durations of treatment. Information on medical history and maternal age and smoking behavior was available. They compared their results to the general population, making it easier to interpret.

Summary of study results The studies utilized a variety of different designs and methods, making it impossible to carry out a meta-analysis. Most studies lacked information on dosage, duration of therapy, comorbidities and the use of concomitant medication in both fathers and mothers and there was heterogeneity in outcome measurement across studies. Based on the results of the included studies, the following summary regarding pregnancy and fetal outcomes can be made (table 2 and 3).

Azathioprine en 6-mercaptopurine AZA is a purine analogue that, after oral intake, is immediately converted in the body into the active metabolite 6-mercaptopurine (6-MP).39 Both drugs were evaluated in several studies (tables 2 and 3) and there is no data suggesting that the effect of AZA and 6-MP varies. Fourteen studies were identified, incorporating 642 pregnancies. Adverse pregnancy outcomes included preterm birth (n=16) and spontaneous abortions (n=40). Adverse fetal outcomes included low birth weight (n=6) and congenital malformations (n=33). The best performed study was the study of Hoeltzenbein et al. They concluded that the number of serious malformations was not increased in the treatment group compared to the control group (3.0% versus 2.2% respectively).39 Other studies are of lower quality (table 1-3). RISKS OF PATERNAL EXPOSURE TO ORAL IMMUNOSUPPRESSIVE DRUGS 107

Methotrexate MTX is a folic acid antagonist. It inhibits reduction of dihydrofolic acid to tetrahydrofolic acid, a key step in the synthesis of nucleic acids, which are necessary in the synthesis of DNA, RNA and proteins.5,36 Thirteen studies were identified, incorporating 279 pregnancies. Adverse pregnancy outcomes included preterm birth (n=10) and spontaneous abortions (n=19). Adverse fetal outcomes included congenital malformations (n=13). The best performed study was the study of Weber-Schoendorfer et al.40 They described 113 pregnancies. No increased risk was seen in the MTX group for major birth defects (OR 1.02, 95% CI 0.05-7.0) or spontaneous abortion (hazard ratio 1.19 95% CI 0.65-2.17) compared to pregnancies not exposed to MTX. Although there were more preterm infants in the control group, (15.1% vs 9.2%), there was no significant difference in the gestational age at delivery (HR 0.94, 95% CI 0.74, 1.19), and the weight of the children was comparable. The other studies were of lower quality (table 1-3).

Mycophenolic acid and mycophenolate mofetil MMF is a prodrug, which, after resorption, is rapidly converted by the body into the active MPA. Enteric-coated mycophenolate sodium (EC-MPS) is a mycophenolate product that has been on the market since 2004.5,41 Four studies were identified, incorporating 392 pregnancies. Adverse pregnancy outcomes included preterm birth (n=21) and spontaneous abortions (n=30). Adverse fetal outcomes 7 included low birth weight (n=8) and congenital malformations (n=16). The best performed study is the study of Jones et al, which is of moderate quality.41 The authors conclude that their results did not differ from the results in the healthy population. The results of the other three (poorer quality) studies are shown in table 1-3.

Comparison to general population Nine out of 26 studies compared their results with a healthy control population.1,25,28,32,35,39-42 Seven studies concluded that the risk of adverse pregnancy and/or fetal outcomes did not differ from the risk in the healthy population. One study found an adjusted risk of congenital abnormalities in children fathered by men using AZA or 6-MP of 1.8 (95% CI 0.7-5.0).1 Another study concluded that the risk of complications was significantly higher in the group using 6-MP at least 3 months prior to conception compared to (1) men who used 6 -MP until a maximum of 3 months prior to conception (p=0.013, OR 16.0; 95% CI 1.6-161) or (2) men who had never used 6-MP (p<0.002, OR 19.6; 95% CI 3.1-122).25 108 Table 2 Results Study Disease (amount of Pregnancies Medication during Pregnancy Fetal outcomes (n) Comparison to controle group patients) (n) conception (n) outcomes (n) CHAPTER 7 Frank18 Psoriasis 2 MTX Not mentioned Healthy babies No control group Lingardh19 Renal transplantation 5 AZA+pred Legal abortion (1) Healthy babies (4) No control group

Kroner20 Acute lymphoblastic 2 6-MP+ MTX Spontaneous abortion Healthy baby (1) No control group leukemia (1), premature (1) Blatt21 Acute lymphocytic 4 MTX and 6-MP Elective abortion (1) Pilonidal dimpling without spina bifida No control group leukemia (1) Matthews22 Acute leukemia 1 MTX Full-term Healthy baby No control group Perry23 Reiter syndrome 1 MTX Full-term Healthy baby No control group Wagoner24 Heart (35) and 42 CsA+pred+AZA (25) Preterm (3) Cleft palate and lip (1), died from inter- No control group heart-lungtransplan- CsA+pred (6) ruption of umbilical cord circulation (1), tation (1) CsA+AZA (4) The exact treatment cardiomyopathy (1), pyloric stenosis AZA+pred (2) of the fathers in these (1), low birth weight (1), increased hair children was not growth (2), psoriasis (1) reported. The exact treatment of the fathers in these children was not reported. Rajapakse25 IBD 13 pregnan- 6-MP Spontaneous abortion Missing thumb (1), multiple anomalies Control group: n= 127 cies fathered (2) (1) Group 1B: IBD men who used 6-MP until a by men maximum of 3 months prior to conception using 6-MP Group 2: IBD men who never used 6-MP until at least 3 months The risk of complications was significantly higher in prior to group 1A, as compared to group 1B (p=0.013, OR conception 16.0; 95% CI 1.6-161) and group 2 (p<0.002, OR (group 1A) 19.6; 95% CI 3.1-122). Ben Neriah26 Crohn’s disease 3 AZA Spontaneous abortion WAGR syndrome (1) No control group (1) Oefferlbauer- Crohn’s disease 1 6-MP Full-term Healthy baby No control group Ernst27 Francella28 IBD 37 6-MP Spontaneous abortion Low birth weight (2), neoplasia (1), Control group: n =73 male patients who fathered (6), preterm (3) meningomyelocele and induced pregnancies prior to treatment with 6-MP abortion (1) No significant difference in spontaneous abortions, major congenital abnormalities, neoplasm or increased numbers of infections among male or female patients taking 6-MP, compared to healthy controls (RR= 0.85 [0.47-1.55], p=0.59). Separate tests of successful pregnancy outcomes of the male patients compared to the control group did not differ significantly (p=1.00) NØrgård1 Organ transplant 54 AZA (52), 6-MP (2) Not mentioned Polysyndactylia (1), oesophageal atresia Control group: n =57195 healthy fathers from the (19), Crohn’s (1), hydronephrosis and megaloureter Central Person Registry Denmark disease/ulcerative (1), VSD (1) colitis (17), skin Adjusted risk of congenital abnormalities in children disease (8), rheu- fathered by men using AZA or 6-MP of 1.8 (95% CI matic diseases 0.7-5.0). (7), lymphoma (2), chronic hepatitis (1) Griggs29 Crohn’s disease 1 MTX Not mentioned Healthy baby No control group Østensen30 Rheumatic diseases 69 MMF Not mentioned Outcome unknown (24). No control group Data from Roche Six children with congenital abnor- Pharma malities: hand- and feet abnormalities, RISKS OFPATERNAL EXPOSURE TO ORAL IMMUNOSUPPRESSIVEDRUGS bladder anomalies and chromosomal changes Østensen31 Rheumatic diseases 11 MTX Induced abortion (1) Atrophy of the hand and a small fistula No control group beneath the ear (1), anomalies of the toes (1) Hasani32,43 * Renal transplantation 142 CsA+MMF+pred Spontaneous abor- Polydactyly (1 in CsA+MMF+pred), Control group: US general population, number of (43), CsA+MMF (2), tions (11) undescended testis (1 in patients unclear. Tacr+MMF+pred tacr+MMF+pred), tongue tied (3 in (23), Tacr+MMF The exact treatment CsA+AZA and/or pred) The birth defects in pregnancy were not higher than (1), MMF+pred (4), of the fathers in these in the general population. MMF+sirolimus (1) children was not CsA, AZA and/or reported. pred (47) Tacrolimus, AZA and/or pred (11) Paschou33 Ankylosing 2 MTX Not mentioned Healthy babies No control group spondylitis Lamboglia34 Crohn’s disease 1 MTX Not mentioned Healthy baby No control group Teruel35 IBD 46 AZA (37) or 6-MP (9) Spontaneous abor- Low birth weight (3), persistent ductus Control group: n= 84 IBD fathers who had never tions (5), premature (2) arteriosus (1) been exposed to thiopurines or had stopped them > 3 months before conception

No significant differences between treatment and control group regarding unsuccessful pregnancies (spontaneous abortion, ectopic pregnancies, anem- bryonic pregnancies and fetal death) (10.9% vs. 13.1% control group; OR 0.79, CI 0.22-2.85), premature birth (4.3% vs. 2.4%, OR 1.3, CI 0.22-7.61), low birth weight (6.5% vs. 6%; OR 1.06, CI 0.25-4.54) or congenital malformations (2.2% vs. 2.4%; OR 0.82, CI 0.08-9) 109 7 110 Table 2 Results Study Disease (amount of Pregnancies Medication during Pregnancy Fetal outcomes (n) Comparison to controle group patients) (n) conception (n) outcomes (n) CHAPTER 7 Beghin36 Rheumatoid arthritis 42 MTX Spontaneous abortion No congenital malformations No control group (10), Psoriasis (9), (3), induced abortion Ankylosing spondy- due to personal litis (7), leukemia (6), reasons [not due to Crohn’s disease (2), malformations] (3), adrenal tumor (2), preterm (1) multiple sclerosis (2), lymphomatoid papulosis (2), Still’s disease (1), Sarcoidosis (1) Xu37 Renal transplantation 212 CsA+AZA+pred Spontaneous abortion Soft toenails (1) No control group (148), (5), voluntarily induced CsA+MMF+pred abortion (16) (28), tacr+MMF+pred (14), tacr+MMF (2), AZA+pred (1) Viktil38 Rheumatic disease 174 AZA (124), MTX (50) Not mentioned Most results given for the entire group No control group including other anti-rheumatic drugs. Only on AZA/MTX applied outcome showed 2 orofacial malformations in MTX. Hoeltzenbein39 Crohn’s disease (49), 115 AZA (108) or 6-MP Spontaneous abortion In AZA: Small muscular VSD + Control group: n = 340 pregnant women not ulcerative colitis (30), (7) (2 6MP, 7 AZA), prema- horsehoe kidney + hemangioma (1), exposed to known teratogens with partners without renal transplantation ture (7 AZA/6-MP). VSD + ASD II + pulmonary stenosis reported immunosuppressive or otherwise risky (5), myasthenia (1), motor retardation + dyspraxia + treatment. gravis (5), other auto persistent foramen ovale (1), umbilical immune disease(26) hernia (3), small hemangioma (2), The number of serious malformations was not persistent foramen ovale (1), mild hip increased in the treatment group compared to the dysplasia (1), xanthoma upper lip (1) control group (3.0% versus 2.2%). Weber- Rheumatoid arthritis 113 MTX Initiated abortion (11), trisomy 16 (1), cardiomyopathy + lung Control group: n = 412 pregnancies not exposed Schoendorfer40 (57), psoriasis (33), spontaneous abortion hypoplasia + intrauterine growth retar- to MTX ankylosing spon- (15), preterm (8) dation (1), persistent foramen ovale + dylitis (7), Crohn’s slight pyelectasia + intrauterine growth No increased risk was seen in the MTX group for disease(6), other (10) retardation (1), hypospadias glandis (1), major birth defects (OR 1.02, 95% CI 0.05-7.0) or microcephaly (1), ureteral dilatation + spontaneous abortion (hazard ratio 1.19 95% CI dermal sinus (1) 0.65-2.17). No significant difference in the gestational age at delivery (HR 0.94, 95% CI 0.74, 1.19). Weight of the children was comparable. Jones41 Organ 205 MPA or MMF Spontaneous abortion Low birth weight (8), clubfoot (1), US general population, number of patients unclear transplantation (14), preterm (21) tongue-tied (2), Prader-Willi syndrome Data in this study are not different from the results in (1), pyloric stenosis (1), congenital healthy population. diaphragmatic hernia (incompatible with life) (1), VSD (1) Wallenius42 Inflammatory joint 49 MTX Not mentioned VSD (1), pes equinovarus (1) Control group: n =230 fathers who never received disease DMARDs between the time of diagnosis and conception. Paternal exposure to DMARDs was not associated with an increase in adverse pregnancy outcomes or congenital malformations. RISKS OFPATERNAL EXPOSURE TO ORAL IMMUNOSUPPRESSIVEDRUGS

6-MP 6 mercaptopurine; ALL acute lymphocytic leukemia; ASD atrial septal defect; AZA azathioprine; CI confidence interval;CsA cyclosporine A; DMARDs disease modifying antirheumatic drugs; HR hazard ratio; IBD inflammatory bowel disease;MMF mycophenolate mofetil; MPA mycophenolic acid; MTX methotrexate; OR Odds ratio; pred prednisone; RR relative risk; tacr tacrolimus; VSD ventricular septal defect; WAGR Wilms tumor, aniridia, genital abnormalities and retardation

* The original article of Hasaniet al was not found but the data are cited in full in the review by Armenti et al.43 111 7 112 CHAPTER 7

Table 3 Summary of outcomes Drug Studies Pregnancies Preterm birth Spontaneous Low birth Congenital abortion weight malformations AZA and 6-MP 14 642 16 40 6 33 MTX 13 279 10 19 0 13 MPA and MMF 4 392 21 30 8 16

6-MP 6 mercaptopurine; AZA azathioprine; MMF mycophenolate mofetil; MPA mycophenolic acid; MTX methotrexate

Discussion Oral immunosuppressive drugs are increasingly used in dermatology for chronic skin diseases, including in patients who may wish to conceive. It can be a challenge for the physician to treat the disease with medications that are proven to be successful and healthy for conception. Therefore, dermatologists need to be aware of the possible risks of the use of oral immunosuppressive drugs during pregnancy and conception.

The review of human pregnancy drug exposure data and assessment of pregnancy and fetal outcomes requires consideration of human embryology, teratology, pharmacology, obstetrics and epidemiology.44,45 Whether a drug causes abnormal development or not depends not only on the physical and chemical nature of the drug, but also on the dose, duration, frequency, route of exposure and gestational timing involved.44 Besides, when evaluating whether a particular drug exposure increases the risk of abnormal pregnancy and fetal results, we should be aware of the background prevalence in a similar patient population. A low birth weight for instance, which is mentioned often in the included studies in this CAT, is considered to occur in 1 of 12 live births in the general population.46 A ventricular septal defect, that was reported 5 times in this CAT, was found to range between 8.41-79.18 per 10,000 live births in the general population.44,47 The FDA provided a guideline for the evaluation of risks of drug exposure, in which they discuss the above mentioned epidemiological problems in detail and provide a list of background incidences of different adverse pregnancy and fetal outcomes.44

For female patients, the risks of the use of oral immunosuppressive drugs during conception have been studied and described in a variety of patient populations.43 Reported adverse pregnancy outcomes are spontaneous abortion, prematurity and low birth weight. Reported adverse fetal outcomes are adrenal insufficiency due to maternal corticosteroid use, infections and tera- togenicity leading to congenital abnormalities.24 A recent published review of Murase et al, on the effects of different dermatological medications on pregnancy, concludes that the maternal use of MTX is an absolute contra-indication, the dose of AZA should be halved (if the mother’s leukocyte count was <1 standard deviation below the mean) and that MMF should be avoided during pregnancy.48 RISKS OF PATERNAL EXPOSURE TO ORAL IMMUNOSUPPRESSIVE DRUGS 113

The relationship of the paternal use of oral immunosuppressive drugs and congenital abnormalities is less clear than the effects in maternal use. Ethical rules discourage randomized studies. Therefore, clinical decisions should be made based on observational studies. However, the studies that we found and systematically appraised were predominantly small and of poor quality. Additionally, in some studies, fathers used different concomitant oral immunosuppressive drugs, making it impossible to elucidate the relationship between the individual drugs and pregnancy or fetal outcomes.

Immunosuppressive drugs might interfere with spermatogenesis, as mentioned previously. From a pharmacokinetics perspective, the presence of a drug in the prostate and seminal fluid will be negligible after five times the (terminal) plasma half-life. In contrast, intracellular half-lives of immunosuppressive drugs (e.g. in spermatocytes in the testis) are often found much longer than in plasma and depend on the cell cycle. This results in a prolonged exposure to the drug and possibly increased incidence and / or severity of detrimental effects.

Most studies suggest an interval of three to four months, when considering discontinuation of oral immunosuppressive drugs before conception, because spermatogenesis takes 70-90 days. When a therapy-free interval of three months before conception is adhered to, the risk of complications might be reduced.25 Controversially, the data of NØrgård et al show a possible long term teratogenic effect of AZA.1 Congenital abnormalities were seen until 38 months after the last prescription. It remains unclear how many capsules were prescribed and when patients exactly discontinued treatment. 7

Limitations of the studies Because studies did not provide information on fetal abnormalities in spontaneous abortions, the real rate of incidence of congenital anomalies remains unknown. Furthermore, in most studies, paternity was assumed but not proven.24 Some studies did not mention the fathers’ age at conception, or the fathers’ and mother’s comorbidities and relevant medical and familial history. Furthermore, most studies did not mention paternal use of other medication at time of the conception. These factors might contribute to the incidence of congenital abnormalities and should be evaluated, even though the maternal contribution is probably higher.49 Besides, most studies did not compare their results to a healthy control population. In these studies, it remained unclear whether the risk of adverse pregnancy or fetal outcomes is increased. Studies that did use a comparison group to find out whether the risk of adverse pregnancy and fetal outcomes is increased, used different methods to compare their data. One should be aware of the fact that a comparison to the general population does not take into account the influence of relevant maternal attributes such as disease state or age.44 As well, it is necessary to know that it is easier to detect an increased risk for an abnormal outcome that occurs at a relatively high background than at a relatively low background rate. 114 CHAPTER 7

Finally, it should be taken into consideration that MTX is dosed once a week. Therefore, the time between administering medications and conception may not be the same in all patients. This might influence the pregnancy or fetal outcomes, but the relatively long intracellular half-life of MTX makes this less likely.

Limitations of this critically appraised topic Our search was conducted carefully. Nevertheless, some studies were missed by our systematic search strategy. Of 15 articles, that we wanted to check for inclusion based on title and sometimes abstract, the full text was not available. We expect not to have missed crucial articles as these references were not cited in previously published studies and overviews. In the additional manual search, six further articles were identified. Publication bias may have occurred as well: cases without adverse pregnancy or fetal outcomes will probably not always be published in medical literature, resulting in underreporting of successful cases. Some studies used questionnaires to collect their information and possible recall bias should be considered. It is possible that oral immunosuppressive drugs may influence the quality and quantity of the sperm and testicular function , but no conclusions can be drawn on this topic, because studies evaluating the ability to conceive without looking at the pregnancy or fetal outcomes were excluded in this CAT. While preparing this manuscript, the European Medicine Agency (EMA) gave a warning for MMF for its use in pregnancy. They state that ‘if a pregnant woman is exposed to mycophenolate, either by taking it herself or through unprotected sex with a man taking the medicine, it is likely to harm the developing baby’ and ‘sexually active (including vasectomised) men taking mycophenolate are recommended to use condoms for sex during treatment and for 90 days thereafter’.50 This update of the product information followed a regular re-assessment of the benefit-risk balance of the medicine, carried out by EMA’s Pharmacovigilance Risk Assessment Committee.50 In the four studies in this CAT concerning MMF and EC-MPS (392 pregnancies), the conclusions were such limited that, purely based on this literature, we could not confirm and/or support these strong recommendations of the EMA.

In conclusion In this CAT, most of the included studies in AZA, MTX and MPA that compared their outcomes to a healthy population conclude that there was no increase in adverse pregnancy or fetal outcomes. Nevertheless, preterm birth, spontaneous abortion, low birth weight and congenital malformations have been described with a varying incidence. Therefore, a possible risk of adverse pregnancy of fetal outcomes when using these drugs cannot be excluded. Based on the studies that compared their outcomes to a control population, AZA/6-MP especially, should be used with caution. RISKS OF PATERNAL EXPOSURE TO ORAL IMMUNOSUPPRESSIVE DRUGS 115

In the clinical situation, we should emphasize to physicians that they ask their male patients starting AZA, MTX, MPA or MMF if they wish to conceive. In this clinical scenario in particular, we could discuss with the patient the results of this CAT, highlight the discrepancies between studies in reviews and guidelines and discuss the risk/benefit of oral immunosuppressive treatment. If possible, preference should be given to the use of CsA.

In future, a daily practice registration of pregnancy and fetal outcomes of men conceiving while taking oral immunosuppressive drugs could be of great value for guiding further recommendations.

Acknowledgements We thank prof. dr. M.A. de Rie, dr. M.A. Middelkamp-Hup and prof. dr. R. Hoekzema (Academic Medical Center Amsterdam, the Netherlands) and prof. C.A.F.M. Bruijnzeel-Koomen (University Medical Center Utrecht, the Netherlands) for their help with the discussion section.

7 116 CHAPTER 7

References

1 Norgard B, Pedersen L, Jacobsen J et al. 11 Katayama I, Kohno Y, Akiyama K et al. The risk of congenital abnormalities in Japanese Guideline for Atopic Dermatitis children fathered by men treated with 2014. Allergol Int 2014; 63: 377-98. azathioprine or mercaptopurine before conception. Aliment Pharmacol Ther 2004; 12 National Clinical Guideline Centre. 19: 679-85. Psoriasis: Assessment and Management of Psoriasis. London: Royal College of 2 Joffe JM. Influence of drug exposure of the Physicians (UK); 2012. father on perinatal outcome. Clin Perinatol 1979; 6: 21-36. 13 Grunewald S, Jank A. New systemic agents in dermatology with respect to fertility, 3 Millsop JW, Heller MM, Eliason MJ et al. pregnancy, and lactation. J Dtsch Dermatol Dermatological medication effects on male Ges 2015; 13: 277-89; quiz 90. fertility. Dermatol Ther 2013; 26: 337-46. 14 The Netherlands Pharmacovigilance Centre 4 Bouloux PM, Wass JA, Parslow JM et al. Lareb. Available at: http://www.lareb.nl (last Effect of cyclosporin A in male autoimmune consulted 3 September 2015) infertility. Fertil Steril 1986; 46: 81-5. 15 Personal correspondence with 5 Summary of product characteristics. College of Medicine Evaluation the Available at http://www.cbg-med.nl (last Netherlands (College ter Beoordeling van consulted 6 June 2016). Geneesmiddelen, CBG) on September 1st 2015 6 Farmacotherapeutisch kompas. Available at: http://www.farmacotherapeutisc- 16 U.S. Food and Drug Administration. hkompas.nl (last consulted 31 March 2015). Available at: http://www.fda.gov. (last consulted 11 september 2015). 7 Meggitt SJ, Anstey AV, Mohd Mustapa MF et al. British Association of Dermatologists’ 17 Personal correspondence with the Dutch guidelines for the safe and effective Association of Gynecology and Obstetrics prescribing of azathioprine 2011. Br J on September 15th 2015 Dermatol 2011; 165: 711-34. 18 Frank L, Lichtman H, Biro L et al. 8 Sidbury R, Davis DM, Cohen DE et al. Experiences with methotrexate in psoriasis. Guidelines of care for the management of Dermatologica 1968; 137: 87-96. atopic dermatitis: section 3. Management and treatment with phototherapy and 19 Lingardh G, Andersson L, Osterman B. systemic agents. J Am Acad Dermatol Fertility in men after renal transplantation. 2014; 71: 327-49. Acta Chir Scand 1974; 140: 494-7.

9 Ring J, Alomar A, Bieber T et al. Guidelines 20 Kroner TH, Tachumi A. Conception of for treatment of atopic eczema (atopic normal child during chemotherapy of acute dermatitis) Part II. J Eur Acad Dermatol lymphoblastic leukaemia in the father. Br Venereol 2012; 26: 1176-93. Med J 1977; 1: 1322-3.

10 Rubel D, Thirumoorthy T, Soebaryo RW et 21 Blatt J, Mulvihill JJ, Ziegler JL et al. al. Consensus guidelines for the manage- Pregnancy outcome following cancer ment of atopic dermatitis: an Asia-Pacific chemotherapy. Am J Med 1980; 69: perspective. J Dermatol 2013; 40: 160-71. 828-32. RISKS OF PATERNAL EXPOSURE TO ORAL IMMUNOSUPPRESSIVE DRUGS 117

22 Matthews JH, Wood JK. Male fertility during 32 Hasani A SN, Coscia LA, et al. Incidence of chemotherapy for acute leukemia. N Engl J structural malformations in pregnancies Med 1980; 303: 1235. fathered by male kidney transplant recipients. Am J Transplant 2006;7(suppl 2):456. 23 Perry WH. Methotrexate and teratogenesis. Arch Dermatol 1983; 119: 874-5. 33 Paschou S, Voulgari PV, Vrabie IG et al. Fertility and reproduction in male patients 24 Wagoner LE, Taylor DO, Price GD, Sr. et al. with ankylosing spondylitis treated with Paternity by cardiac transplant recipients. infliximab. J Rheumatol 2009; 36: 351-4. Transplantation 1994; 57: 1337-40. 34 Lamboglia F, D’Inca R, Oliva L et al. Patient 25 Rajapakse RO, Korelitz BI, Zlatanic J et al. with severe Crohn’s disease became a Outcome of pregnancies when fathers are father while on methotrexate and infliximab treated with 6-mercaptopurine for inflam- therapy. Inflamm Bowel Dis 2009; 15: matory bowel disease. Am J Gastroenterol 648-9. 2000; 95: 684-8. 35 Teruel C, Lopez-San Roman A, Bermejo F 26 Ben-Neriah Z, Ackerman Z. WAGR et al. Outcomes of pregnancies fathered syndrome in a baby--the result of 6-MP by inflammatory bowel disease patients treatment in a father affected by Crohn’s exposed to thiopurines. Am J Gastroenterol disease? Am J Gastroenterol 2001; 96: 251. 2010; 105: 2003-8.

27 Oefferlbauer-Ernst A, Reinisch W, Miehsler 36 Beghin D, Cournot MP, Vauzelle C et al. W et al. Healthy offspring in parents both Paternal exposure to methotrexate and receiving thiopurines. Gastroenterology pregnancy outcomes. J Rheumatol 2011; 2004; 126: 628. 38: 628-32. 7 28 Francella A, Dyan A, Bodian C et al. The 37 Xu LG, Yang YR, Wang HW et al. safety of 6-mercaptopurine for child- Characteristics of male fertility after renal bearing patients with inflammatory bowel transplantation. Andrologia 2011; 43: disease: a retrospective cohort study. 203-7. Gastroenterology 2003; 124: 9-17. 38 Viktil KK, Engeland A, Furu K. Outcomes 29 Griggs LR, Schwartz DA. Successful after anti-rheumatic drug use before and paternity of a healthy child while taking during pregnancy: a cohort study among methotrexate for Crohn’s disease. Am J 150,000 pregnant women and expectant Gastroenterol 2006; 101: 2893-4. fathers. Scand J Rheumatol 2012; 41: 196-201. 30 Ostensen M, Khamashta M, Lockshin M et al. Anti-inflammatory and immunosuppres- 39 Hoeltzenbein M, Weber-Schoendorfer C, sive drugs and reproduction. Arthritis Res Borisch C et al. Pregnancy outcome after Ther 2006; 8: 209. paternal exposure to azathioprine/6-mercaptopurine. Reprod Toxicol 2012; 34: 31 Ostensen M, von Esebeck M, Villiger PM. 364-9. Therapy with immunosuppressive drugs and biological agents and use of contra- 40 Weber-Schoendorfer C, Hoeltzenbein ception in patients with rheumatic disease. M, Wacker E et al. No evidence for an J Rheumatol 2007; 34: 1266-9. increased risk of adverse pregnancy outcome after paternal low-dose methotrexate: an observational cohort study. Rheumatology (Oxford) 2014; 53: 757-63. 118 CHAPTER 7

41 Jones A, Clary MJ, McDermott E et al. 49 Jenkins KJ, Correa A, Feinstein JA et al. Outcomes of pregnancies fathered by Noninherited risk factors and congenital solid-organ transplant recipients exposed cardiovascular defects: current knowl- to mycophenolic acid products. Prog edge: a scientific statement from the Transplant 2013; 23: 153-7. American Heart Association Council on Cardiovascular Disease in the Young: 42 Wallenius M, Lie E, Daltveit AK et al. No endorsed by the American Academy excess risks in offspring with paternal of Pediatrics. Circulation 2007; 115: preconception exposure to disease-mod- 2995-3014. ifying antirheumatic drugs. Arthritis Rheumatol 2015; 67: 296-301. 50 European Medicine Agency (EMA). Available at: http://www.ema.europa.eu/ 43 Armenti VT, Constantinescu S, Moritz MJ ema/index.jsp?curl=pages/news_and_ et al. Pregnancy after transplantation. events/news/2015/10/news_detail_002418. Transplant Rev (Orlando) 2008; 22: 223-40. jsp&mid=WC0b01ac058004d5c1 (last consulted 5 June 2016). 44 U.S. Department of Health and Human Services Food and Drug Administration (FDA), Center for Drug Evaluation and Research (CDER), Center for Biologics Evaluation and Research (CBER). Reviewer guidance: Evaluating the risk of drug exposure in human pregnancies. April 2005.

45 Trasler JM, Doerksen T. Teratogen update: paternal exposures-reproductive risks. Teratology 1999; 60: 161-72.

46 March of Dimes Birth Defects Foundation, 2001, Fact Sheets: Birth defects, miscarriages, and stillbirth; leading categories of birth defects; and perinatal profiles: Statistics for monitoring State Maternal and Infant Health. http://www.modimes.org.

47 National Center on Birth D, Developmental Disabilities CDC. Birth defects surveillance data from selected states, 1995-1999. Teratology 2002; 66 Suppl 1: S129-211.

48 Murase JE, Heller MM, Butler DC. Safety of dermatologic medications in pregnancy and lactation: Part I. Pregnancy. J Am Acad Dermatol 2014; 70: 401 e1-14; quiz 15. RISKS OF PATERNAL EXPOSURE TO ORAL IMMUNOSUPPRESSIVE DRUGS 119

7 Part III Optimizing the therapeutic potential of thiopurines

Thiopurine metabolite levels in patients with atopic dermatitis and/or chronic hand/foot eczema treated with azathioprine

Floor M. Garritsen, Jorien van der Schaft, Carla A.F.M. Bruijnzeel-Koomen, Ron H.N. van Schaik, Marlies de Graaf, Marcel P.H. van den Broek, Marjolein S. de Bruin-Weller

Accepted for publication in J Dermatolog Treat 2017 124 CHAPTER 8

Abstract Background: Azathioprine is frequently used in severe eczema. It is converted in the liver into active metabolites, including 6-thioguanine nucleotide (6-TGN) and methylated 6-methylmercaptopurine (6-MMP). In the past, the therapeutic potential of azathioprine may have not been fully utilized. Recent investigations on inflammatory bowel disease have led to a better understanding of azathioprine metabolism and optimizing treatment. Objective: To investigate whether measuring thiopurine metabolites in circulation can improve the effectiveness and safety of azathioprine treatment in patients with atopic dermatitis and/or chronic hand/foot eczema. Methods: Azathioprine metabolite levels were measured in eczema patients during maintenance treatment (Part I) and dose escalation (Part II). Clinical effectiveness, hepatotoxicity and bone marrow suppression were analyzed and TPMT genotype was assessed. Results: A wide variation in metabolite levels in all dose groups was observed. In part I (32 patients), there were no significant differences in 6-TGN levels between clinical responders and non-responders (p=0.806). No hepatoxicity or myelotoxicity was observed. In Part II, all 6-TGN and 6-MMP levels increased during dose escalation. Hypermethylation was observed in 2/8 patients. Conclusion: For individual eczema patients treated with azathioprine, routinely measuring 6-TGN and 6-MMP can be helpful in optimizing azathioprine dose, improving clinical effectiveness and preventing side effects. MEASURING THIOPURINE METABOLITE LEVELS 125

Introduction Atopic dermatitis (AD) and chronic hand/foot eczema are common chronic inflammatory skin diseases. Although the majority of patients can be treated adequately with topical corticosteroids and/or phototherapy, oral immunosuppressive drugs are indicated in more severe cases. Azathioprine (AZA) is an important off-label drug in the management of severe and difficult to treat AD. Randomized clinical trials report a decrease in clinical scores, varying from 26 to 37% decrease in SASSAD (Six Area, Six Sign Atopic Dermatitis) score and 39% decrease in SCORAD (Scoring Atopic Dermatitis) score after 12 weeks. The percentage of patients reporting adverse events in these short-term (12 weeks) randomized clinical trials ranged between 5.6% and 22.9%.1 Abnormalities in cell count, such as lymphocytopenia, were most frequently reported. In 9% of the patients these adverse events led to withdrawal and in 9% of patients dose adjustments were made. However, in daily practice, results are less favorable. In an AD population treated with AZA in daily practice, Thomsen et al found more than 30% treatment failure after one year of AZA treatment: 9/60 patients (15%) had to discontinue due to a lack of clinical benefit and 12/60 patients (20%) due to side effects.2

In a retrospective daily practice study in two university hospitals in the Netherlands, also 30% (14/46) of the adult AD patients treated with AZA discontinued treatment due to side effects and/ or ineffectiveness.3 In a recent drug survival study of 94 AD patients treated with AZA in daily practice, 57% discontinued treatment due to ineffectiveness (19%), side effects (36%) or both (2%).4 Little data are available on the efficacy of AZA in chronic hand/foot eczema. Agarwal et al performed a randomized controlled trial in which the use of AZA (n=45) was compared to topical clobetasol proprionate 0.05% (n=46) in patients with chronic hand eczema. After 24 weeks of treatment, the improvement of itch and the improvement of the HECSI (hand eczema scoring

5 index) was significantly higher in the AZA group (p=0.003 and p=0.001, respectively). 8

In the past, the therapeutic potential of AZA may have not been fully utilized.6 The AZA metabolism is very complex due to involvement of various enzymes; it is studied thoroughly in patients with inflammatory bowel disease (IBD). AZA is a thiopurine pro-drug and has no immunosuppressive activity itself. In the liver AZA is converted into 6-mercaptopurine (6-MP). Subsequently, 6-MP undergoes metabolic transformations, via a complex enzymatic pathway, resulting in a variety of pharmacologically active metabolites (figure 1).7 The most important metabolites are 6-thioguanine nucleotide (6-TGN) and methylated 6-methylmercaptopurine (6-MMP). The immunosuppressive effect of AZA is caused by the 6-TGN metabolites. Intracellular 6-TGN is incorporated into DNA instead of guanine nucleotides, which, after a strand breakage, triggers cell-cycle arrest and apoptosis. This results in the inhibition of nucleotide and protein synthesis, and ultimately in the inhibition of lymphocyte proliferation.8-10 However, the main mechanism of immunomodulation is by inducing T-cell apoptosis by modulation cell (Rac1) signaling.11 The therapeutic onset after thiopurine initiation is delayed as it takes 12 to 17 weeks for TGNs to be incorporated into DNA.12,13 This explains the delayed clinical efficacy in AD, which is usually reached after 3-4 months. 126 CHAPTER 8

Thiouric acid

HHGGPPRRTT Mercaptopurine Azathioprine Mercaptopurine 6-TGN nucleotide

TPMT TPMT

Methylmercaptopurine 6-MMP

Figure 1 Azathioprine metabolism (simplified) 6-MMP methylated 6-methylmercaptopurine; 6-TGN 6-thioguanine nucleotide; HGPRT hypoxanthine-guanine phosphoribosyl transferase; TPMT thiopurinemethyltransferase; XO xanthine oxidase

Recent investigations, especially in IBD, have led to a better understanding of the AZA metabolism and optimization of treatment.6,14,15 Therapeutic drug monitoring can reveal non-compliance or under-dosing and can be used as a practical tool to optimize AZA therapy.6 Data on the AZA metabolism and therapeutic drug monitoring of AZA metabolites in patients with chronic inflammatory skin diseases are scarce. Recently, Reynolds et al suggested, based on their unpublished experience in individual patients with AD, that measuring 6-TGN levels particularly in patients with heterozygote thiopurine S-methyltransferase (TPMT) activity may be useful.16 The aim of the present study is to investigate whether measurement of thiopurine metabolites is helpful in improving effectiveness and safety of AZA treatment in patients with AD and/or chronic hand/foot eczema.

Methods Study population This study was exempted from review by our institutional review board. In this prospective mono-center study, AZA metabolite levels were measured in adult patients with AD and/or hand/foot eczema during maintenance treatment (≥ 100mg/day, for more than three months) (part I). The criteria of Hanifin and Rajka and of Williams et al were used for the diagnosis of AD. 17,18 The following patient characteristics were collected: age, sex, type of eczema (AD or hand/foot eczema), AZA dose and the concomitant use of oral corticosteroids. Additionally, AZA metabolites were measured structurally in another small group of patients from start of AZA treatment and during dose escalation (part II). MEASURING THIOPURINE METABOLITE LEVELS 127

Measurements of azathioprine metabolites 6-TGN and 6-MMP levels in red blood cells (RBC) were measured in erythrocytes, as a surro- gate marker of the levels in leukocytes, using the Dervieux-method.19 The majority of previously published studies used the Lennard method (comparison of the values shown in table 1). Metabolites were analyzed by the Clinical Pharmaceutical and Toxicological laboratory of the department of Pharmacy, University Medical Center Utrecht, the Netherlands.

Table 1 Measurement of azathioprine metabolites: Dervieux versus Lennard method Analysis method Lennard Dervieux Unit Therapeutic 6-TGN 230-450 600-1200 pmol/8*108 RBC 6-MMP / 6-TGN Ratio 5-25 2-10 Toxic 6-TGN >450 >1200 pmol/8*108 RBC 6-MMP >5700 >5700 pmol/8*108 RBC

Values used by the Dutch Society of Clinical Pharmarcy, based on studies using Lennard method in inflammatory bowel disease patients and the conversion factor 2.6 for deriving the Dervieux values 20

Outcomes Part I 6-TGN and 6-MMP steady-state levels were measured and ratios between these metabolites were calculated in all patients. The correlation between the AZA dose and the metabolite levels was investigated. 6-TGN and 6-MMP levels and 6-MMP/6-TGN ratios between responders and non-responders were compared. Responsiveness was based on the Investigator Global Assessment (IGA,

21 6-point scale). A responder was defined as IGA 0-2. Non-responders were defined as an IGA 8 3-5 or concomitant treatment with oral corticosteroids. The optimal 6-TGN level to achieve disease control was defined as the median 6-TGN level in the responder group. TPMT genotype was determined. Safety results, including hepatotoxicity and bone marrow suppression, were analyzed. Clinically significant hepatotoxicity was defined as an alanine aminotransferase (ALT) or aspartate aminotransferase (AST) 2 times the upper limit of normal (ALT: >45 U/L in men and >35 U/L in women; AST: >35 U/L in men, >30 U/L in women).22 Bone marrow suppression was defined as white blood cell count <4x109/L and/or thrombocytopenia (platelet count <150x109/L). 128 CHAPTER 8

Part II AZA metabolites were measured in a small group of patients from the start of treatment and during dose escalation. TPMT genotype was determined using TaqMan analysis. Patients started with AZA 50mg/day, followed by metabolite measurement after 2 weeks (T1). When hepatotoxicity or bone marrow suppression was not observed, AZA dose was increased to 100mg/ day, followed by metabolite measurement 2 weeks later (T2). Finally, the dose was increased to 150mg/day, again followed by metabolite measurement 2 weeks later (T3). The following patient characteristics were collected in these patients: age, sex, type of eczema (AD or hand/foot eczema), 6-TGN and 6-MMP levels at T1, T2 and T3. Because the therapeutic effect after AZA initiation is delayed up to four months, clinical efficacy was not evaluated at this early time point.

Statistical analysis The data of the total group were evaluated. Subgroup analyses for responders and non-responders were made. All statistical analyses were performed using SPSS statistics 21. Frequencies, percentages and medians with inter quartile ranges (IQR) were calculated. The Mann-Whitney U test and the Chi Square test were used to calculate whether there were statistically significant differences between responders and non-responders, regarding sex, age, dose, 6-TGN levels and 6-MMP levels. The ANOVA test, corrected for multiple testing by Bonferroni, was used to calculate whether there were statistically significant differences in 6-TGN and 6-MMP levels between the different AZA doses.

Results Part I Patient characteristics Thirty-five patients were treated with AZA. Three patients (8.6%) had 6-TGN levels <50 pmol/8x108 RBC and 6-MMP levels <250 pmol/8x108 RBC, suggesting non-compliance to AZA. These patients were excluded from further analysis: 32 patients remained for further analysis (table 2). The median age at metabolite measurement was 48.0 years [IQR 43.0-56.7]. Twenty-three patients (71.9%) had the diagnosis AD. The other nine patients had isolated hand- and/or foot-eczema (eight dyshidrotic eczema, one hyperkeratotic eczema). Twenty-four patients (75.0%) used monotherapy AZA (dose range 100-200mg/day). Eight patients (25.0%) used concomitant oral corticosteroids (dose range 5-30 mg/day) MEASURING THIOPURINE METABOLITE LEVELS 129

Table 2 Patient characteristics (Part I) All patients (n=32) Male, n (%) 15 (46.9) Age at AZA metabolite measurement (years), median [IQR] 48.0 [43.0-56.7] Patients with AD, n (%) 23 (71.9) Patients with isolated hand/foot eczema, n (%) 9 (28.1) AZA dose in mg/day, median [IQR] 150.0 [100.0-187.5] Concomitant use of oral corticosteroids, n (%) 8 (25.0)

AD atopic dermatitis, AZA azathioprine, IQR inter quartile range

Metabolite levels (figure 2) AZA dose varied from 100 to 200 mg/day with a median dose of 150 mg/day [IQR 100.0-187.5]. Twenty-eight patients had an extensive metabolizer TPMT genotype, three patients had an intermediate metabolizer TPMT genotype (*1/*3A). In one patient the TPMT genotype determination failed. There was a wide variation in metabolite levels in all dose groups. 6-TGN levels varied between 42 to 696 pmol/8x108 RBC in patients using 100 mg/day, between 99 to 656 pmol/8x108 RBC in patients using 150 mg/day and between 142 to 861 pmol/8x108 RBC in patients using 200mg/ day. 6-MMP levels varied between 205 to 3411 pmol/8x108 RBC in patients using 100 mg/day, between 275 to 26027 pmol/8x108 RBC in patients using 150 mg/day and between 126 to 4173 pmol/8x108 RBC in patients using 200 mg/day. In six (18.8%) patients (all using 150 mg/day) the 6-MMP level was >5700 pmol/8x108 RBC (toxic level). The steady state 6-MMP/6-TGN ratio varied from 0.9 to 81.2 in patients using 100 mg/day, from 1.2 to 262.9 in patients using 150 mg/day and from 0.5 to 18.9 in patients using 200 mg/day. The metabolite levels of the three TPMT*1/*3A patients are highlighted in figure 2 and 3. 8

6-MMP levels were statistically significantly different among patients with various AZA doses (p=0.0431), but not after correction for body weight (data not shown). 6-TGN levels were significantly higher in the three patients with an intermediate metabolizer TPMT genotype (p=0.01) compared to patients with an extensive metabolizer TPMT genotype, but no 6-TGN levels above the upper limit of 1200 were observed. 130 CHAPTER 8

1000 27000 26000 800 25000 10000 600

400 5000 6-TGN (pmol/8×10^8 RBC)

200 6-MMP (pmol/8×10^8 RBC)

0 0 100 150 200 100 150 200mg Azathioprine dose (mg/day) Azathioprine dose (mg/day)

Figure 2 Azathioprine dose (100, 150 and 200 mg/day) and steady-state metabolite (6-TGN and 6-MMP) levels (Part I) (median with IQR) ◊ = Intermediate metabolizer TPMT genotype 6-MMP methylated 6-methylmercaptopurine, 6-TGN 6-thioguanine nucleotide, IQR inter quartile range, RBC red blood cell 6-MMP levels were statistically significantly different among patients with various AZA dose (p=0.0431); 6-TGN levels were not significantly different among patients with different AZA dose (p=0.0682).

Responders versus non-responders In total, 16 patients were characterized as responder and 16 patients as non-responder (table 3, figure 3). Responders were significantly older (median 50.9 years versus 44.3 years in the non-responder group) at the time of the metabolite measurement (p=0.029). There were no significant differences in AZA dose (mg/day) (p=0.628), 6-TGN levels (p=0.806), 6-MMP levels (p=0.763), or 6-MMP/6-TGN ratios (p=0.940) between responders and non-responders. The median 6-TGN level in the responder group was 197.0 pmol/8x108 RBC.

Table 3 Responders versus non-responders (Part I) Responders (n=16) Non-responders (n=16) p-value Male, n (%) 7 (43.8) 8 (50) 0.723 Age at AZA metabolite measurement 50.9 [46.7-61.3] 44.3 [26.0-52.1] 0.029 (years), median [IQR] AZA dose (mg/day), median [IQR] 150.0 [100.0-187.5] 150.0 [112.5-187.5] 0.628 Steady-state 6-TGN level (pmol/8x108 197.0 [110.3-619.5] 219.0 [167.3-281.3] 0.806 RBC), median [IQR] Number of patients with 6-TGN levels 0 (0.0) 0 (0.0) - >1200, n (%) Steady-state 6-MMP level (pmol/8x108 2324.0 [829.8-3228.5] 2156.5 [668.3-6982.5] 0.763 RBC), median [IQR] Number of patients with 6-MMP levels 2 (12.5) 4 (25.0) 0.365 >5700, n (%) Steady-state 6-MMP/6-TGN ratio, 5.6 [3.2-27.7] 7.0 [3.2-25.5] 0.940 median [IQR]

AZA azathioprine, 6-MMP methylated 6-methylmercaptopurine, 6-TGN 6-thioguanine nucleotide, IQR inter quartile range MEASURING THIOPURINE METABOLITE LEVELS 131

1000 27000 800 25000 10000 600 400 5000 200

6-TGN (pmol/8×10^8 RBC) 0 6-MMP (pmol/8×10^8 RBC) 0 Responders Non-responders Responders Non-responders

Figure 3 Responders versus non-responders and steady-state metabolite levels (Part I) (median with IQR) ◊ = Intermediate metabolizer TPMT genotype 6-MMP methylated 6-methylmercaptopurine, 6-TGN 6-thioguanine nucleotide, IQR inter quartile range, RBC red blood cell There was no significant difference in 6-TGN levels (p=0.806) and 6-MMP levels (p=0.763) between responders and non-responders.

Safety No patients met the criteria of hepatotoxicity. Six patients had 6-MMP levels >5700 pmol/8×108 RBC (toxic level) with a median duration of treatment until metabolite measurement of 0.82 years [IQR 0.27-2.16]. Two of these patients (al extensive metabolizer TPMT genotype) had slightly increased ALT levels at time of metabolite measurement (83 U/L in one male patient and 42 U/L in one female patient). None of the 32 patients met the criteria of bone marrow depression. None of the patients had 6-TGN levels >1200 pmol/8×108 RBC (toxic level).

8 132 CHAPTER 8

Part II AZA metabolites were measured consecutively in eight patients (three men, five women) from start of AZA treatment and during dose escalation (table 4, figure 4). Median age at start of AZA treatment was 46.2 years [39.9-50.9]. Five patients were diagnosed with AD, three patients with isolated hand/foot eczema. Seven patients had an extensive metabolizer TPMT genotype. The TPMT genotype was missing in one patient. 6-TGN and 6-MMP values at the different measurements varied widely. All levels of 6-TGN and 6-MMP increased during dose escalation up to T3 (150 mg/day). In one patient (patient 5) AZA dose was not increased to 150 mg/day due to subjective side effects. Thiopurine hypermethylation (low 6-TGN with high 6-MMP) was observed in patients 2 and 3.

Table 4 6-TGN and 6-MMP levels in eight patients during dose escalation (Part II) All patients (n=8) Male, n (%) 3 (37.5) Patients with AD, n (%) 5 (62.5 Patients with isolated hand/foot eczema, n (%) 3 (37.5) 6-TGN level (pmol/8x108 RBC), median [IQR] T1 (2 weeks, 50mg/day) 70.0 [50.3-91.8] T2 (4 weeks, 100mg/day) 170.0 [130.5-193.3] T3* (6 weeks, 150mg/day) 262.5 [164.5-359.3] 6-MMP level (pmol/8x108 RBC), median [IQR] T1 (2 weeks, 50mg/day) 284.0 [250.0-301.3] T2 (4 weeks, 100mg/day) 788.5 [393.5-1154.5] T3* (6 weeks, 150mg/day) 1384.0 [932.5-3253.3]

AZA azathioprine, 6-MMP methylated 6-methylmercaptopurine, 6-TGN 6-thioguanine nucleotide, IQR inter quartile range * In one patient, dose was 125mg/day due to subjective side effects MEASURING THIOPURINE METABOLITE LEVELS 133

Patient 1 Patient 2 -M po/×08RBC) (pmol/8×10^8 6-MMP 500 2,000 RBC) (pmol/8×10^8 6-MMP 500 4000

400 400 3500 1,500 2,000 300 300 1,500 1,000 200 200 1,000 500 100 100 500 6-TGN (pmol/8×10^8 RBC) 0 0 6-TGN (pmol/8×10^8 RBC) 0 0

(T2) (T3) g (T1) mg (T1) 0mg 0mg 50 50m 10 15 100mg (T2) 150mg (T3) Dose escalation Dose escalation

Patient 3 Patient 4 -M po/×08RBC) (pmol/8×10^8 6-MMP 500 11000 500 2,000 RBC) (pmol/8×10^8 6-MMP

400 10000 400 2,000 1,500 300 1,500 300 1,000 200 1,000 200 500 100 500 100 6-TGN (pmol/8×10^8 RBC) 6-TGN (pmol/8×10^8 RBC) 0 0 0 0 ) 2) 3)

50mg (T1) 50mg (T1 100mg (T2) 150mg (T3) 100mg (T 150mg (T Dose escalation Dose escalation

Patient 5 Patient 6 -M po/×08RBC) (pmol/8×10^8 6-MMP RBC) (pmol/8×10^8 6-MMP 500 2,000 500 2,000

400 400 1,500 1,500 300 300 1,000 1,000 200 200 500 500 100 100

6-TGN (pmol/8×10^8 RBC) 0 0 6-TGN (pmol/8×10^8 RBC) 0 0 8 (T2) (T3) mg (T1) 0mg 0mg 50mg (T1) 50 100mg (T2) 125mg (T3) 10 15 Dose escalation Dose escalation

Patient 7 Patient 8 -M po/×08RBC) (pmol/8×10^8 6-MMP 500 2,000 RBC) (pmol/8×10^8 6-MMP 500 2,000

400 400 1,500 1,500 300 300 1,000 1,000 200 200 500 500 100 100 6-TGN (pmol/8×10^8 RBC) 6-TGN (pmol/8×10^8 RBC) 0 0 0 0

1) 3

50mg (T1) 50mg (T 100mg (T2) 150mg (T3) 100mg (T2) 150mg (T Dose escalation Dose escalation

Figure 4 6-TGN and 6-MMP levels in patients treated with AZA treatment during dose escalation 134 CHAPTER 8

Discussion This study aimed to investigate whether measurement of thiopurine metabolites is helpful in improving effectiveness and safety of AZA treatment in patients with AD and/or chronic hand/ foot eczema. A wide inter- and intra-individual variation in both 6-TGN and 6-MMP levels in patients with AD or chronic hand/foot eczema during AZA maintenance treatment was observed. Metabolite levels between individuals were not significantly related to the AZA dose, but within individuals higher AZA dose led to higher metabolite levels. Metabolite levels were not significantly different between responders and non-responders.

Several past studies investigated thiopurine metabolites and their relation to clinical efficacy and side effects in IBD. Although it is hypothesized that 6-TGN is responsible for the immunosuppressive effect of AZA, several studies in IBD patients failed to find a statistically significant correlation between metabolite levels, and clinical response.6,23-26 A prospective study of steroid-dependent patients with IBD showed that dose optimization of AZA to achieve 6-TGN levels ≥250 pmol/8×108 RBC was significantly associated with a higher rate of disease remission (Lennard method).27 Other studies in IBD suggest a therapeutic range in 6-TGN concentration between 235 and 490 pmol/8×108 RBC (Lennard method) as significantly more exacerbations were found in patients below the threshold of 235 pmol/8×108.6,28-30 The risk for leucocytopenia increased with 6-TGN levels above 490 pmol/8×108 RBC.6,30,31

6-TGN and 6-MMP levels are not routinely measured during AZA treatment in dermatological diseases. However, attempts have been made in a small number of studies to optimize clinical efficacy of AZA by measuring metabolite levels. Recently, Reynolds et al suggested, based on their unpublished experience in individual cases, that measuring 6-TGN levels particularly in patients with heterozygote TPMT activity may be useful.16 In 2009, el Azhary et al investigated the optimal levels of 6-TGN for disease remission in patients with immunobullous disease treated with AZA (Dervieux method). They found a mean optimal level of 6-TGN of 190.7 pmol/8x108 RBCs.32 Limited disease (involving only the oral or genital mucosa, the face, or limited areas on the chest) was found to require less 6-TGN to achieve clinical improvement in their study, with a mean of 145.3 pmol/8x108 RBCs. In 2013, Caufield et al reported a case series of 12 children with recalcitrant AD treated with AZA.33 They found in 11 responders, 6-TGN levels ranging from 45 to 358 pmol/8x108 RBC during stable disease (not clear whether the Dervieux or Lennard method was used). The above mentioned studies on AZA treatment in dermatological disease and our results (showing a median 6-TGN level in responders 197.0), suggest that 6-TGN levels needed to induce clinical efficacy might be lower compared to IBD. Though, differences in measurement methods (Dervieux versus Lennard) may have contributed to these differences as well, because some studies using Dervieux methods in IBD patients also reported lower 6-TGN levels for therapeutic response.20 MEASURING THIOPURINE METABOLITE LEVELS 135

The majority of the non-responders in our study had low 6-TGN levels compared to responders (figure 2), but this difference was not significantly different. This may be a consequence of the small sample size and the large variation in 6-TGN levels in the responders. Dose escalation was not performed in the non-responder group for a number of reasons, such as subjective side effects and the potential risk of hepatotoxicity. Literature on the use of AZA in IBD patients reports an increased risk for hepatotoxicity in case of 6-MMP levels higher than 5700 pmol/8x108 RBC. In the present study, none of the patients met the criteria of hepatotoxicity, though high 6-MMP levels (>5700 pmol/8x108 RBC) were found in 6/32 patients (two clinical responders and four non-responders). These patients may be at risk of developing hepatotoxicity when AZA is continued for longer periods.

The metabolism of AZA is partly influenced by genetic polymorphisms in thiopurine S-methyltransferase (TPMT). This results in inter-individual variation in the metabolism of thiopurines; both with respect to side effects and efficacy.13 High TPMT activity is associated with lower 6-TGN levels and high 6-MMP levels, which may result in decreased immunosuppressive activity and an increased risk of hepatotoxicity. Patients with high TPMT activity may be under- dosed with standard AZA dose, which could explain, at least partially, non-responsiveness to AZA. Increasing daily AZA dose in these patients will not lead to higher effectiveness because of a preferential 6-MMP production. TPMT activity is not the only predictor for the risk of hepatotoxicity. Other metabolites, other enzymatic conversions within the metabolic pathway or interference medication (like aminosalicylates, non-steroidal anti-inflammatory drugs, anticoagulant, diuretics and infliximab) may also influence the production of 6-MMP metabolites.34 Furthermore, low TPMT activity is associated with high 6-TGN levels, resulting in an increased risk of myelotoxicity. However TPMT 8 activity does not predict myelotoxicity in all patients. Therefore, monitoring haematological parameters remains crucial.35 Reports indicate that between 50% and 75% of thiopurine-related leucopenia occurs in patients with normal TPMT activity.13,36,37 Finally, although TPMT activity is generally measured only once before starting therapy, some studies have shown that TPMT levels are not static and induction of activity can occur during treatment.32,33 Therefore measurement of 6-TGN and 6-MMP instead of TPMT activity, seems to be more relevant to predict/ monitor the risk of hepato- and myelotoxicity. In this study, information on TPMT genotype was missing in two patients (one in part I and one in part II). Since these patients displayed normal complete blood counts and AZA metabolite levels, the TPMT genotype in these patients was assumed to be normal. 136 CHAPTER 8

Recent studies on AZA in IBD and AD have reported high percentages of treatment failure due to sides effects and/or inefficacy.2-4,38 We suggest that this may, at least partly, be related to a skewed thiopurine metabolism profile. The so-called ‘ultramethylators’ (hypermethylators) preferentially produce excessive amounts of 6-MMP, whereas 6-TGN levels remain very low (6-MMP/6-TGN range above 20).38-40 We observed this phenomenon in both parts of this study. In part I of this study, 6/32 patients (18.8%), all using AZA 150 mg/day, had high 6-MMP levels (>5700 pmol/8x108 RBC). Three of them had 6-TGN levels <235 pmol/8x108 RBC, suggesting thiopurine hypermethylation or preferential 6-MMP production. In part II of this study, thiopurine hypermethylation was observed in patients 2 and 3 (25%). These hypermethylators are at risk of side effects and non-responsiveness, resulting in AZA discontinuation after several months of treatment. TPMT activity does not predict the risk of hypermethylation. Thus, metabolite levels should be routinely monitored from start of treatment and during dose escalation in order to detect hypermethylators early in AZA treatment. Studies in IBD patients showed that concomitant use of allopurinol can result in a shift of the AZA metabolism towards the production of 6-TGN, reducing the relative formation of 6-MMP.41-44

Three patients were excluded because they had 6-TGN levels <50 pmol/8x108 RBC and 6-MMP levels <250 pmol/8x108 RBC, indicating that they had not been taking their medication. The (suspected) noncompliance rate was 8.5%. One patient confirmed noncompliance; in the other two patients compliance remained unclear. One of these two patients still had an active eczema; the other was in clinical remission. It should be taken into consideration that low 6-TGN and 6-MMP levels may be a result of noncompliance; however, a patient-dependent variation in the AZA metabolism (e.g. malabsorption, yet unknown enzyme defects or extremely high enzyme activity in the thiopurine pathway) cannot be excluded.6

Limitations Concomitant medication (including aminosalicylates, non-steroidal anti-inflammatory drugs, anticoagulants, diuretics and infliximab) may interfere with the thiopurine metabolism, but in this study no information on concomitant medication (except for oral corticosteroids) was collected.34 In order to use a clinical parameter that was useful in both AD and chronic hand/foot eczema, the IGA was selected to define response to treatment. This clinical score is less detailed than for instance the EASI (Eczema Area and Severity Index) score that is commonly used as a score for AD severity.

Conclusion For individual AD patients treated with AZA, routinely measuring 6-TGN and 6-MMP levels can be helpful in optimizing AZA dose, improving clinical effectiveness, and preventing side effects. MEASURING THIOPURINE METABOLITE LEVELS 137

References

1 Roekevisch E, Spuls PI, Kuester D et al. 9 Lennard L. TPMT in the treatment of Crohn’s Efficacy and safety of systemic treatments disease with azathioprine. Gut 2002; 51: for moderate-to-severe atopic dermatitis: 143-6. a systematic review. J Allergy Clin Immunol 2014; 133: 429-38. 10 Fairchild CR, Maybaum J, Kennedy KA. Concurrent unilateral chromatid damage 2 Thomsen SF, Karlsmark T, Clemmensen KK and DNA strand breakage in response et al. Outcome of treatment with azathio- to 6-thioguanine treatment. Biochem prine in severe atopic dermatitis: a 5-year Pharmacol 1986; 35: 3533-41. retrospective study of adult outpatients. Br J Dermatol 2015; 172: 1122-4. 11 Tiede I, Fritz G, Strand S et al. CD28- dependent Rac1 activation is the molecular 3 Garritsen FM, Roekevisch E, van der Schaft target of azathioprine in primary human J et al. Ten years experience with oral CD4+ T lymphocytes. J Clin Invest 2003; immunosuppressive treatment in adult 111: 1133-45. patients with atopic dermatitis in two academic centres. J Eur Acad Dermatol 12 Prefontaine E, Macdonald JK, Sutherland Venereol 2015; 29: 1905-12. LR. Azathioprine or 6-mercaptopurine for induction of remission in Crohn’s 4 van der Schaft J, Politiek K, van den Reek disease. Cochrane Database Syst Rev 2010: JM et al. Drug survival for azathioprine and CD000545. enteric-coated mycophenolate sodium in a long-term daily practice cohort of 13 Goel RM, Blaker P, Mentzer A et al. adult patients with atopic dermatitis. Br J Optimizing the use of thiopurines in inflam- Dermatol 2016; 175: 199-202. matory bowel disease. Ther Adv Chronic Dis 2015; 6: 138-46. 5 Agarwal US, Besarwal RK. Topical clobetasol propionate 0.05% cream alone 14 Lee MN, Kang B, Choi SY et al. Relationship and in combination with azathioprine in between azathioprine dosage, 6-thiogua- patients with chronic hand eczema: an nine nucleotide levels, and therapeutic observer blinded randomized comparative response in pediatric patients with IBD trial. Indian J Dermatol Venereol Leprol treated with azathioprine. Inflamm Bowel 8 2013; 79: 101-3. Dis 2015; 21: 1054-62.

6 Gilissen LP, Wong DR, Engels LG et al. 15 Wong DR, Coenen MJ, Vermeulen SH et al. Therapeutic drug monitoring of thiopurine Early Assessment of Thiopurine Metabolites metabolites in adult thiopurine tolerant IBD Identifies Patients at Risk of Thiopurine- patients on maintenance therapy. J Crohns induced Leukopenia in Inflammatory Bowel Colitis 2012; 6: 698-707. Disease. J Crohns Colitis 2017; 11: 175-84.

7 Lennard L. The clinical pharmacology of 16 Reynolds NJ, Sinha A, Elias MS et al. 6-mercaptopurine. Eur J Clin Pharmacol Translating translation into patient benefit 1992; 43: 329-39. for atopic eczema. Br J Dermatol 2016; 175 Suppl 2: 8-12. 8 Chouchana L, Narjoz C, Beaune P et al. Review article: the benefits of pharmaco- 17 Hanifin J, Rajka G. Diagnostic features of genetics for improving thiopurine therapy atopic dermatitis. Acta Derm Venereol in inflammatory bowel disease. Aliment 1980; 92: 44-7. Pharmacol Ther 2012; 35: 15-36. 138 CHAPTER 8

18 Williams HC, Burney PG, Pembroke 27 Roblin X, Serre-Debeauvais F, Phelip JM et AC et al. The U.K. Working Party’s al. 6-tioguanine monitoring in steroid-de- Diagnostic Criteria for Atopic Dermatitis. pendent patients with inflammatory bowel III. Independent hospital validation. Br J diseases receiving azathioprine. Aliment Dermatol 1994; 131: 406-16. Pharmacol Ther 2005; 21: 829-39.

19 Dervieux T, Boulieu R. Simultaneous deter- 28 Cuffari C, Seidman EG, Latour S et al. mination of 6-thioguanine and methyl Quantitation of 6-thioguanine in peripheral 6-mercaptopurine nucleotides of azathi- blood leukocyte DNA in Crohn’s disease oprine in red blood cells by HPLC. Clin patients on maintenance 6-mercaptopurine Chem 1998; 44: 551-5. therapy. Can J Physiol Pharmacol 1996; 74: 580-5. 20 Shipkova M, Armstrong VW, Wieland E et al. Differences in nucleotide hydrolysis 29 Osterman MT, Kundu R, Lichtenstein GR et contribute to the differences between al. Association of 6-thioguanine nucleotide erythrocyte 6-thioguanine nucleotide levels and inflammatory bowel disease concentrations determined by two widely activity: a meta-analysis. Gastroenterology used methods. Clin Chem 2003; 49: 260-8. 2006; 130: 1047-53.

21 Futamura M, Leshem YA, Thomas KS et al. 30 Dubinsky MC, Lamothe S, Yang HY et A systematic review of Investigator Global al. Pharmacogenomics and metabolite Assessment (IGA) in atopic dermatitis (AD) measurement for 6-mercaptopurine trials: Many options, no standards. J Am therapy in inflammatory bowel disease. Acad Dermatol 2016; 74: 288-94. Gastroenterology 2000; 118: 705-13.

22 Dubinsky MC, Yang H, Hassard PV et 31 Derijks LJ, Gilissen LP, Engels LG et al. al. 6-MP metabolite profiles provide a Pharmacokinetics of 6-mercaptopurine in biochemical explanation for 6-MP resis- patients with inflammatory bowel disease: tance in patients with inflammatory bowel implications for therapy. Ther Drug Monit disease. Gastroenterology 2002; 122: 2004; 26: 311-8. 904-15. 32 el-Azhary RA, Farmer SA, Drage LA et al. 23 Goldenberg BA, Rawsthorne P, Bernstein Thioguanine nucleotides and thiopurine CN. The utility of 6-thioguanine metabolite methyltransferase in immunobullous levels in managing patients with inflam- diseases: optimal levels as adjunctive tools matory bowel disease. Am J Gastroenterol for azathioprine monitoring. Arch Dermatol 2004; 99: 1744-8. 2009; 145: 644-52.

24 Belaiche J, Desager JP, Horsmans Y et al. 33 Caufield M, Tom WL. Oral azathioprine for Therapeutic drug monitoring of azathio- recalcitrant pediatric atopic dermatitis: clin- prine and 6-mercaptopurine metabolites in ical response and thiopurine monitoring. J Crohn disease. Scand J Gastroenterol 2001; Am Acad Dermatol 2013; 68: 29-35. 36: 71-6. 34 Al Hadithy AF, de Boer NK, Derijks LJ et al. 25 Lowry PW, Franklin CL, Weaver AL et al. Thiopurines in inflammatory bowel disease: Measurement of thiopurine methyltrans- pharmacogenetics, therapeutic drug moni- ferase activity and azathioprine metabo- toring and clinical recommendations. Dig lites in patients with inflammatory bowel Liver Dis 2005; 37: 282-97. disease. Gut 2001; 49: 665-70.

26 Gupta P, Gokhale R, Kirschner BS. 6-mercaptopurine metabolite levels in children with inflammatory bowel disease. J Pediatr Gastroenterol Nutr 2001; 33: 450-4. MEASURING THIOPURINE METABOLITE LEVELS 139

35 Gearry RB, Barclay ML, Burt MJ et al. 43 Ansari A, Patel N, Sanderson J et al. Thiopurine S-methyltransferase (TPMT) Low-dose azathioprine or mercaptopurine genotype does not predict adverse drug in combination with allopurinol can bypass reactions to thiopurine drugs in patients many adverse drug reactions in patients with inflammatory bowel disease. Aliment with inflammatory bowel disease. Aliment Pharmacol Ther 2003; 18: 395-400. Pharmacol Ther 2010; 31: 640-7.

36 Colombel JF, Ferrari N, Debuysere H 44 Sparrow MP. Use of allopurinol to optimize et al. Genotypic analysis of thiopurine thiopurine immunomodulator efficacy in S-methyltransferase in patients with inflammatory bowel disease. Gastroenterol Crohn’s disease and severe myelosup- Hepatol (N Y) 2008; 4: 505-11. pression during azathioprine therapy. Gastroenterology 2000; 118: 1025-30.

37 Ansari A, Hassan C, Duley J et al. Thiopurine methyltransferase activity and the use of azathioprine in inflammatory bowel disease. Aliment Pharmacol Ther 2002; 16: 1743-50.

38 Kreijne JE, Seinen ML, Wilhelm AJ et al. Routinely Established Skewed Thiopurine Metabolism Leads to a Strikingly High Rate of Early Therapeutic Failure in Patients With Inflammatory Bowel Disease. Ther Drug Monit 2015; 37: 797-804.

39 Hoentjen F, Seinen ML, Hanauer SB et al. Safety and effectiveness of long-term allopurinol-thiopurine maintenance treatment in inflammatory bowel disease. Inflamm Bowel Dis 2013; 19: 363-9.

40 van Egmond R, Chin P, Zhang M et al. High 8 TPMT enzyme activity does not explain drug resistance due to preferential 6-methylmercaptopurine production in patients on thiopurine treatment. Aliment Pharmacol Ther 2012; 35: 1181-9.

41 Smith MA, Blaker P, Marinaki AM et al. Optimising outcome on thiopurines in inflammatory bowel disease by co-prescription of allopurinol. J Crohns Colitis 2012; 6: 905-12.

42 Pavlidis P, Stamoulos P, Abdulrehman A et al. Long-term Safety and Efficacy of Low-dose Azathioprine and Allopurinol Cotherapy in Inflammatory Bowel Disease: A Large Observational Study. Inflamm Bowel Dis 2016; 22: 1639-46.

Improving outcome of azathioprine treatment in chronic eczema by allopurinol co-prescription

Floor M. Garritsen, Jorien van der Schaft, Marlies de Graaf, Dirk J. Hijnen, Carla A.F.M. Bruijnzeel-Koomen, Marcel P.H. van den Broek, Marjolein S. de Bruin-Weller

Accepted for publication in Acta Derm Venereol 2017 142 CHAPTER 9

Abstract Background: Azathioprine (AZA) is frequently used in atopic dermatitis (AD). However, high discontinuation rates due to side effects and/or inefficacy are reported. Studies in inflammatory bowel disease led to strategies to reduce toxicity risk and optimize efficacy by adding allopurinol. This co-prescription shifts the AZA metabolism towards 6-thioguanine nucleotide (6-TGN) production, resulting in reduction of hepatotoxicity and increase of clinical effect. Objective: To investigate the effect of allopurinol co-prescription in patients with AD and/ or chronic hand/foot eczema treated with AZA on metabolite levels (6-TGN and methylated 6-methylmercaptopurine [6-MMP]), side effects and clinical efficacy. Methods: AZA metabolite levels were measured in adult patients with AD and/or chronic hand/ foot eczema, during AZA monotherapy and after co-prescription of allopurinol. Clinical effectiveness (Investigator Global Assessment) and side effects were analyzed. Results: Fifteen patients were enrolled. Reasons for allopurinol co-prescription were inefficacy during AZA monotherapy, side effects or skewed metabolism. After allopurinol addition, 6-MMP levels decreased and 6-TGN levels increased in all patients. Prior to allopurinol addition, four patients (26.7%) were classified as responder, compared to seven patients (56.7%) after allopurinol co-prescription (p=0.013). Conclusion: Co-prescription of allopurinol may optimize AZA treatment outcome in chronic eczema by increasing 6-TGN and decreasing 6-MMP levels. AZATHIOPRINE AND ALLOPURINOL CO-PRESCRIPTION 143

Introduction Azathioprine (AZA) is frequently used in the treatment of atopic dermatitis (AD). Efficacy of AZA treatment in AD has been demonstrated in a number of randomized controlled trials (RCTs) and open label studies1, but recent daily practice studies show less favourable results.2,3 AZA is a thiopurine prodrug and has no immunosuppressive activity itself. It is converted in the liver and undergoes metabolic transformations resulting in a variety of pharmacologically active metabolites. The most important metabolites are 6-thioguanine nucleotide (6-TGN) and methylated 6-methylmercaptopurine (6-MMP). The immunosuppressive effect of AZA is caused by the 6-TGN metabolites. Highly elevated 6-TGN concentrations are associated with development of myelotoxicity; highly elevated 6-MMP concentrations are associated with development of hepatotoxicity. New insights obtained by assessment of 6-TGN and 6-MMP metabolites in inflammatory bowel disease (IBD) patients on AZA treatment led to strategies to reduce the risk of toxicity and optimize effectiveness and safety. Raised 6-MMP/6-TGN ratios (especially with 6-MMP levels > 5700 pmol/8x108 RBC) indicate that the patient preferentially metabolizes thiopurine to 6-MMP at the expense of the therapeutically active 6-TGN. This phenomenon of skewed drug metabolism is known as thiopurine hypermethylation, which appears to occur in up to 20% of the population.4 In IBD patients, the combination of AZA and allopurinol has gained popularity for bypassing thiopurine-related side effects.5,6 Allopurinol can shift the AZA metabolism towards 6-TGN production, resulting in a decrease of the incidence of hepatotoxicity and improvement of remission rates (figure 1). Although AZA is a frequently used drug in chronic inflammatory skin diseases, data on metabolite levels and attempts to influence these levels are scarce. The aim of this study is to investigate the effect of allopurinol co-prescription in patients with AD and chronic hand/foot eczema treated with AZA on metabolite levels (6-TGN, 6-MMP), side effects and clinical effectiveness.

Thiouric acid 9 XO

HHGGPPRRTT Mercaptopurine Azathioprine Mercaptopurine 6-TGN nucleotide

TPMT TPMT

Methylmercaptopurine 6-MMP

Figure 1 Azathioprine metabolism (summarized) and the shift towards 6-TGN production after the addition of allopurinol (red crosses and green arrows). 6-MMP methylated 6-methylmercaptopurine; 6-TGN 6-thioguanine nucleotide; HGPRT hypoxanthine-guanine phosphoribosyl transferase; TPMT thiopurinemethyltransferase; XO xanthine oxidase 144 CHAPTER 9

Methods Study population In this prospective observational study between January 1st 2015 and October 1st 2016, adult patients with AD and/or chronic hand/foot eczema, who failed AZA monotherapy, starting combination therapy of AZA and allopurinol were included. The criteria of Hanifin and Rajka and of Williams et al were used for the diagnosis of AD. Reasons for AZA failure were lack of clinical effectiveness on therapeutic AZA doses, side-effects and/or a skewed metabolism. A fixed dose of allopurinol 100 mg/day was used for all patients. In order to prevent myelotoxic levels of 6-TGN after allopurinol co-prescription, AZA dose was decreased by at least 50% in all patients.

Measurements of azathioprine metabolites 6-TGN and 6-MMP metabolites were analyzed in whole blood by the Clinical Pharmaceutical and Toxicological laboratory of the department of Pharmacy, University Medical Center Utrecht, the Netherlands, using liquid chromatography-tandem mass spectrometry (LC-MS/MS). The Dervieux-method was used (table 1).7 Values were expressed in picomole/8x108 red blood cells (RBCs).

Table 1 Measurement of azathioprine metabolites: Dervieux versus Lennard method Analysis method Lennard Dervieux Unit Therapeutic 6-TGN 230-450 600-1200 pmol/8*108 RBC 6-MMP/6-TGN ratio 5-25 2-10 Toxic 6-TGN >450 >1200 pmol/8*108 RBC 6-MMP >5700 >5700 pmol/8*108 RBC

Based on studies in inflammatory bowel disease patients. Conversion factor 2.6.

Outcomes 6-TGN and 6-MMP levels were measured before and at least three weeks after allopurinol addition, assuming steady-state levels. 6-MMP/6-TGN ratios were calculated. Clinical responsiveness before and after allopurinol co-prescription was determined, using the Investigator Global Assessment (IGA, 6-point scale).8 A responder was defined as an IGA score of 0-2 or a decrease of IGA with 2 points. Non-responders were patients with IGA score of 3-5 (except those patients with an IGA decrease of 2 points, those were responders) or with concomitant oral corticosteroids treatment. Clinical responsiveness and therapeutic drug monitoring were assessed at the same visit. AZATHIOPRINE AND ALLOPURINOL CO-PRESCRIPTION 145

Hepatotoxicity was defined as an alanine aminotransferase (ALT) or aspartate aminotransferase (AST) two times the upper limit of normal (ALT: >45 U/L in men and >35 U/L in women; AST: >35 U/L in men, >30 U/L in women). Bone marrow suppression was defined as white blood cell count <4x109/L and/or thrombocytopenia (platelet count <150x109/L). Subjective side effects before and after allopurinol co-prescription were evaluated.

Statistical analysis All statistical analyses were performed using SPSS statistics 21. Frequencies, percentages and medians with inter quartile ranges (IQR) were calculated. The Mann-Whitney U test was used to test whether differences in 6-TGN level, 6-MMP level or 6-MMP/6-TGN ratio before and after the addition of allopurinol were significantly different.

Results Patient characteristics Fifteen patients were enrolled, including nine AD patients (60.0%) and six patients (40%) with isolated hand- and/or foot-eczema (table 2 and 3). Five patients (33.3%) were male. Median age at the start of the allopurinol was 45.7 years (IQR 38.3-53.0). Patients used combination treatment with allopurinol because of ineffectiveness of AZA monotherapy (n=10), side effects (n=4) or a skewed metabolism (n=12). In 10 patients, there was a combination of reasons. AZA dose varied between 100-200 mg/day before and between 25-100 mg/day after the addition of allopurinol. Median time between start of AZA and the introduction of allopurinol was 0.5 years (IQR 0.3-1.7).

Table 2 Patient characteristics Total (n=15) Male, n (%) 5 (33.3) Type of eczema AD, n (%) 9 (60.0) 9 Chronic hand/foot eczema, n (%) 6 (40.0) Age at start allopurinol (years), median [IQR]) 45.7 [38.3-53.0] Dose of allopurinol (mg/day) 100 in all patients Time between start of AZA and start of allopurinol (years), median [IQR] 0.5 [0.3-1.7] Reasons for starting allopurinol Ineffectiveness of AZA treatment (%) 10 (66.7) Side effects (objective and/or subjective) (%) 4 (26.7) Skewed metabolism 12 (80.0) Combination of reasons 10 (66.7)

AD atopic dermatitis, AZA azathioprine; IQR inter quartile range 146 CHAPTER 9

Metabolite levels The median time between the addition of allopurinol and the measurement of the metabolite levels was 36.0 days [IQR 27.0-56.0]. There was a large variation in AZA metabolite levels before and after starting allopurinol (figure 2). 6-TGN levels increased after the addition of allopurinol in all patients. 6-MMP levels and 6-MMP/6-TGN ratios decreased in all patients after the addition of allopurinol. The median 6-MMP level was 3916.0 [IQR 1735.0-10028.0]) before and 234.0 [IQR 148.0-676.0]) after allopurinol addition (p<0.001). The median 6-TGN level was 207.0 [IQR 127.0-285.0]) before and 429.0 [IQR 190.0-543.0]) after allopurinol addition (p=0.009). Median 6-MMP/6-TGN ratio was 247.0 [IQR 77.0-341.0]) before and 10.0 [IQR 4.0-18.0]) after allopurinol addition (p<0.001).

30000 300

1000

25000 250 15000 100

750

10000 500 50 6-MMP/6-TGN ratio 6-TGN (pmol/8×10^8 RBC) 6-MMP (pmol/8×10^8 RBC) 5000 250

0 0 0 ol rin opurinol opurinol opurinol opu ll l llopurinol l llopurinol l al a n a n a n on al tio ion al io diti r addit r addition r ad e e e fore addi ft fore additio ft fore addit ft e A e A e A B B B

Figure 2 6-TGN levels, 6-MMP levels and 6-MMP/6-TGN ratio before and after the addition of allopurinol (Same colors indicating the same patient) AZATHIOPRINE AND ALLOPURINOL CO-PRESCRIPTION 147

Clinical responsiveness Before the addition of allopurinol, four patients (patient 1-4; 26.7%) were classified as responder, which increased to seven patients (patient 1-7;46.7%) after allopurinol addition (p=0.013). Prednisone was tapered successfully in three patients. Noteworthy, in patient 12, a high dose of prednisone was added after the addition of allopurinol by the patient’s general practitioner, because of ineffectiveness, without our knowledge.

Safety One patient (patient 3) met the criteria of hepatotoxicity before the addition of allopurinol. Liver values nearly normalized after AZA dose reduction and the addition of allopurinol. Patient 6 and 11 met the criteria of myelotoxicity after the addition of allopurinol due to extremely high 6-TGN levels. These abnormalities normalized after temporary discontinuation of AZA and allopurinol treatment and remained within safety limits after a resumed treatment with lower AZA dose in combination with allopurinol.

Side effects No side effects of allopurinol were seen. Four patients started co-prescription of allopurinol because of gastro-intestinal side effects on AZA monotherapy. In two patients, these side effects resolved after addition of allopurinol and dose reduction of AZA.

9 148 Table 3 Metabolite levels before and after the addition of allopurinol to treat ment with azathioprine Before the addition of allopurinol Reason for addition of After the addition of allopurinol allopurinol

Patient Sex AD or AZA dose 6-MMP 6-TGN 6-MMP/ Clinical AZA dose 6-MMP 6-TGN 6-MMP/ Clinical CHAPTER 9 HE (mg/day) pmol/8x108 pmol/8x108 6-TGN ratio effect (IGA) (mg/day) pmol/8x108 pmol/8x108 6-TGN ratio effect (IGA) RBC RBC RBC RBC 1 F HE 100 2242 74 30.3 2 (R) Skewed metabolism 50 335 123 2.7 1 (R) 2 F HE 150 26027 99 262.9 2 (R) Skewed metabolism 50 316 173 1.8 2 (R) 3 F HE 150 10526 141 74.7 1 (R) Subjective side effects, 50 865 211 4.1 1 (R) skewed metabolism 4 F AD 100 6233 207 30.1 2 (R) Skewed metabolism 25 195 261 0.7 2 (R) 5 ¢ M AD 100 439 76 5.8 3 (N) Ineffectiveness 50 146 139 1.1 2 (R) 6 F AD 150 7919 232 34.1 3 (N) Ineffectiveness, skewed 50 721 478 1.5 1 (R) metabolism 7 F AD 150 3712 245 15.2 3 (N) Ineffectiveness, skewed 50 178 543 0.3 2 (R) metabolism, subjective side effects 8 M HE 100**** 547 227 2.4 4 (N) Ineffectiveness, subjective side 50* 95 616 0.2 3 (N) effects 9 M HE 150**** 1735 127 13.7 3 (N) Ineffectiveness, skewed 50 198 190 1.0 3 (N) metabolism 10 M AD 200** 2762 357 7.7 3 (N) Ineffectiveness, skewed 75** 127 525 0.2 3 (N) metabolism 11 F AD 200*** 13758 340 40.5 2 (N) Skewed metabolism 100** 2192 1083 2.0 2 (N) 12 F HE 150 8231 285 28.9 5 (N) Ineffectiveness, skewed 50***** 676 466 1.5 2 (N) metabolism 13 p M AD 150 10028 406 24.7 3 (N) Ineffectiveness, skewed 50 234 593 0.4 3 (N) metabolism 14 ¿ F AD 150 3916 172 22.8 3 (N) Ineffectiveness, skewed 50 148 418 0.4 3 (N) metabolism 15 p F AD 150 1240 163 7.6 3 (N) Ineffectiveness, subjective side 50 396 429 0.9 3 (N) effects AD atopic dermatitis; AZA azathioprine, HE chronic hand/foot eczema; IGA investigator global assessment; (N) non-responder; (R) responder * Concomitant use of prednisone 5mg/day, ** Concomitant use of prednisone 10mg/day, *** Concomitant use of prednisone 12.5mg/day, **** Concomitant use of prednisone 20mg/day, ***** Concomitant use of prednisone 30mg/day AZATHIOPRINE AND ALLOPURINOL CO-PRESCRIPTION 149

Discussion In this study in 15 chronic eczema patients, the addition of allopurinol resulted in an improvement of hepatotoxicity and subjective sided effects. The amount of responders increased and prednisone dose was tapered successfully in some patients. This is all in line with earlier findings in studies in IBD patients. 4-6,9-11 In patients with inadequate clinical efficacy during AZA monotherapy, simply increasing AZA dose will not always result in increased 6-TGN levels and better clinical outcome. Hypermethylation may lead to an increase of 6-MMP levels and risk of adverse reactions. The addition of allopurinol enables a better balance between 6-TGN and 6-MMP, in order to improve clinical effectiveness and reduce side effects during azathioprine treatment. Although the exact mechanism of action of allopurinol co-prescription remains unclear, there are a few proposed mechanisms (figure1).4,12,13 Competitive inhibition of xanthine oxidase with allopurinol results in an increase of the bioavailability of the active metabolites.12 It has also been suggested that allopurinol may inhibit TPMT by the production of metabolite 6-thioxanthine (6-TX).12,14,15 The last mechanism involves increased activity of HGPRT towards the active pathway.12,14,16 The patients included in our study represent a relative negative selection of patients: patients who were successfully treated with AZA monotherapy were not included. This explains why clinical efficacy in our group before starting allopurinol is very low; only 4/15 patients were classified as responders. In these responders, allopurinol co-prescription was started because of skewed metabolism.

Allopurinol co-prescription effectuates a shift of the AZA metabolism towards the production of 6-TGN. High TGN-levels are associated with an increased myelotoxicity risk.4 Therefore AZA dose reduction of at least 50% is mandatory. However, in this study we reported two patients with decreased lymphocyte counts after allopurinol addition, despite an adequate AZA dose reduction. Close monitoring of blood counts remains essential during this combination therapy. Limitations of this study include the small size and the use of the IGA score instead of other clinical (such as EASI) and patient oriented scores. Finally, because dose reduction and the addition of allopurinol have been done simultaneously, it is not clear which of those improved 9 the liver function test.

Conclusion This study shows encouraging results of allopurinol co-prescription in a small, selected group of difficult to treat patients with AD and/or hand/foot eczema, who failed on AZA monotherapy. Allopurinol co-prescription resulted in an improvement of treatment outcome and an increase of the 6-TGN levels and a decrease of the 6-MMP levels. It should be considered in patients with a skewed metabolism or in patients with subjective side effects. AZA dose should be reduced by at least 50% before start of allopurinol (100 mg/day for adults) and regular laboratory monitoring for myelotoxicity and hepatotoxicity is needed. 150 CHAPTER 9

References

1 Roekevisch E, Spuls PI, Kuester D et al. 9 Ansari A, Elliott T, Baburajan B et al. Efficacy and safety of systemic treatments Long-term outcome of using allopurinol for moderate-to-severe atopic dermatitis: co-therapy as a strategy for overcoming a systematic review. J Allergy Clin Immunol thiopurine hepatotoxicity in treating inflam- 2014; 133: 429-38. matory bowel disease. Aliment Pharmacol Ther 2008; 28: 734-41. 2 Thomsen SF, Karlsmark T, Clemmensen KK et al. Outcome of treatment with azathio- 10 Hoentjen F, Seinen ML, Hanauer SB et al. prine in severe atopic dermatitis: a 5-year Safety and effectiveness of long-term allo- retrospective study of adult outpatients. Br purinol-thiopurine maintenance treatment J Dermatol 2015; 172: 1122-4. in inflammatory bowel disease. Inflamm Bowel Dis 2013; 19: 363-9. 3 van der Schaft J, Politiek K, van den Reek JM et al. Drug survival for azathioprine and 11 Leung Y, Sparrow MP, Schwartz M et al. enteric-coated mycophenolate sodium Long term efficacy and safety of allopurinol in a long-term daily practice cohort of and azathioprine or 6-mercaptopurine in adult patients with atopic dermatitis. Br J patients with inflammatory bowel disease. J Dermatol 2016; 175: 199-202. Crohns Colitis 2009; 3: 162-7.

4 Goel RM, Blaker P, Mentzer A et al. 12 Amin J, Huang B, Yoon J et al. Update Optimizing the use of thiopurines in inflam- 2014: advances to optimize 6-mercaptopu- matory bowel disease. Ther Adv Chronic rine and azathioprine to reduce toxicity and Dis 2015; 6: 138-46. improve efficacy in the management of IBD. Inflamm Bowel Dis 2015; 21: 445-52. 5 Smith MA, Blaker P, Marinaki AM et al. Optimising outcome on thiopurines in 13 Blaker PA, Arenas-Hernandez M, Smith MA inflammatory bowel disease by co-prescrip- et al. Mechanism of allopurinol induced tion of allopurinol. J Crohns Colitis 2012; 6: TPMT inhibition. Biochem Pharmacol 2013; 905-12. 86: 539-47.

6 Ansari A, Patel N, Sanderson J et al. 14 Min MX, Weinberg DI, McCabe RP. Low-dose azathioprine or mercaptopurine Allopurinol enhanced thiopurine treatment in combination with allopurinol can bypass for inflammatory bowel disease: safety many adverse drug reactions in patients considerations and guidelines for use. J with inflammatory bowel disease. Aliment Clin Pharm Ther 2014; 39: 107-11. Pharmacol Ther 2010; 31: 640-7. 15 Duley JA, Chocair PR, Florin TH. 7 Dervieux T, Boulieu R. Simultaneous deter- Observations on the use of allopurinol in mination of 6-thioguanine and methyl combination with azathioprine or mercap- 6-mercaptopurine nucleotides of azathi- topurine. Aliment Pharmacol Ther 2005; 22: oprine in red blood cells by HPLC. Clin 1161-2. Chem 1998; 44: 551-5. 16 Seinen ML, de Boer NK, Smid K et al. 8 Futamura M, Leshem YA, Thomas KS et al. Allopurinol enhances the activity of hypox- A systematic review of Investigator Global anthine-guanine phosphoribosyltransferase Assessment (IGA) in atopic dermatitis (AD) in inflammatory bowel disease patients trials: Many options, no standards. J Am during low-dose thiopurine therapy: Acad Dermatol 2016; 74: 288-94. preliminary data of an ongoing series. Nucleosides Nucleotides Nucleic Acids 2011; 30: 1085-90. AZATHIOPRINE AND ALLOPURINOL CO-PRESCRIPTION 151

First experience with thioguanine in the treatment of atopic dermatitis

Floor M. Garritsen, Marcel P.H. van den Broek, Carla A.F.M. Bruijnzeel-Koomen, Marjolein S. de Bruin-Weller

Manuscript in preparation 154 CHAPTER 10

Abstract Background: Azathioprine (AZA) is commonly used in the treatment of atopic dermatitis (AD), but high discontinuation rates due to side effects and ineffectiveness are reported. It has been hypothesized that thioguanine therapy instead of AZA reduces generation of potentially toxic metabolites, because it is primarily converted into the therapeutically aimed metabolite 6-TGN by bypassing several rate-limiting metabolic steps. Aim: To investigate whether treatment with thioguanine is effective and safe in patients with very difficult to treat AD. Method: Patients with severe and difficult to treat AD who failed on AZA treatment were treated with thioguanine. The start dose was often 10mg/day and doses were escalated within the first month. Clinical efficacy, side effects and laboratory safety values were regularly measured during the first three months of treatment. Results: Five adult patients with AD were enrolled. After three months, disease activity decreased in most patients and one patient could stop oral steroids. No myelotoxicity or other clinically relevant changes in laboratory results were seen. Some mild subjective side effects were reported, but only one patient discontinued treatment due to headache. Conclusion: Thioguanine may be an alternative treatment option in patients with difficult to treat, severe AD who failed on AZA treatment. TREATMENT WITH THIOGUANINE 155

Introduction Atopic dermatitis (AD) is one of the most common chronic inflammatory skin diseases. Although the majority of patients can be treated adequately with topical corticosteroids and/or phototherapy, oral immunosuppressive drugs are indicated in the more severe cases. In most countries, only cyclosporine A (CsA) is registered for the treatment of severe AD patients, not responding to topical treatment and/or UV light therapy. Although CsA has proven to be very effective in the treatment of AD, side effects such as hypertension, kidney function disturbances and other objective side effects such as gingival hyperplasia and hypertrichosis may limit its use. Other treatments that showed efficacy in prospective (randomized) trials and open label studies are azathioprine (AZA), methotrexate (MTX), mycophenolate mofetil (MMF) and enteric-coated mycophenolate sodium (EC-MPS).1 Unfortunately, daily practice data on the use of these oral immunosuppressive drugs show less favorable results with high discontinuation rates due to ineffectiveness and/or side effects. Treatment failure was seen in 44.7% for CsA, 40% for MTX , 57% for AZA and 56% for enteric-coated mycophenolate sodium (EC-MPS).2-6

AZA is a frequently used off-label drug in the management of AD. However, its therapeutic potential may have not been fully utilized. The AZA metabolism is very complex due to involvement of various enzymes. AZA is a thiopurine pro-drug and has no immunosuppressive activity itself. In the liver AZA is converted into mercaptopurine (6-MP). Subsequently, 6-MP undergoes metabolic transformations, via a complex enzymatic pathway, resulting in a variety of pharmacological active metabolites (figure 1). The most important metabolites are 6-thioguanine nucleotide (6-TGN) and methylated 6-methylmercaptopurine (6-MMP). The immunosuppressive effect of AZA is caused by the 6-TGN metabolites. Increased values of 6-MMP may cause (subjective) side effects and hepatotoxicity, frequently being the reason for discontinuation of treatment.

Thiouric acid

HHGGPPRRTT HHGGPPRRTT Mercaptopurine Azathioprine Mercaptopurine 6-TGN Thioguanine nucleotide 10 TPMT TPMT

Methylmercaptopurine 6-MMP

Figure 1 Azathioprine and Thioguanine metabolism (simplified) HGPRT Hypoxanthine-guanine phosphoribosyl transferase; TPMT Thiopurinemethyltransferase; XO Xanthine oxidase; 6-TGN 6-thioguanine nucleotide; 6-MMP methylated-6-methylmercaptopurine 156 CHAPTER 10

In the past, especially in inflammatory bowel disease (IBD), attempts have been made to influence the metabolism of AZA. It is known that the co-prescription of allopurinol can shift the production towards 6-TGN and decrease the production of 6-MMP. In several studies, this resulted in a clinical improvement and a decrease of (subjective) side effects and hepatotoxicity. We used this treatment approach in a group of 15 patients with AD and chronic hand/foot eczema (manuscript accepted for publication).

Using allopurinol influences the balance between 6-TGN and 6-MMP in favor of 6-TGN ina rather indirect way. The metabolism of thioguanine leads more directly towards the intended pharmacologically active products: 6-thioguanine is converted directly into 6-TGN using hypoxanthine-guanine phosphoribosyl transferase (HGPRT) without producing the potentially toxic metabolite 6-MMP (figure 1).

Therefore, thioguanine may be an interesting new drug in patients with AD. The aim of the present study is to investigate whether treatment with thioguanine is effective and safe in patients with difficult to treat AD who failed on AZA treatment.

Methods Patient selection This study was exempted from review by our institutional review board. This open study was performed between September 1st 2016 and September 1st 2017 at the department of Dermatology in the University Medical Center Utrecht, the Netherlands. Adult patients with severe AD, uncontrolled by topical corticosteroid treatment and failure on AZA treatment were candidates for treatment with thioguanine. Exclusion criteria were: patients < 18 years, concomitant use of other oral immunosuppressive drugs and/or UV light therapy, preexisting abnormalities in liver function, kidney function or hematological abnormalities, history of malignancy (except basal cell carcinoma) within the last 5 years, uncontrolled hypertension, pregnancy or lactation, wish for pregnancy (both men and women) during the treatment period, infections requiring continued therapy, known positivity for HIV, evidence of drug and/or alcohol abuse, recent vaccinations with living virus in the past 12 weeks and the concomitant use of olsalazine, mesalazine or sulfasalazine.

Treatment protocol Before starting with thioguanine, preference was given to a wash-out period of 2 weeks for previously used CsA and oral corticosteroids. Patients using AZA prior to the use of thioguanine were able to switch directly, without a wash-out period. Patients were allowed to use topical corticosteroids and oral antihistamines. Before start, the thiopurinemethyltransferase (TPMT) genotype was determined. Patients with a poor metabolizer TPMT genotype were not indicated to start with thioguanine. TREATMENT WITH THIOGUANINE 157

A standardized protocol for dosing and laboratory controls was provided, but it was possible for the physician to deviate from this protocol, based on clinical response, side effects or laboratory findings (figure 2).

T=0 T= 2 weeks T= 4 weeks T= 6 weeks T= 8 weeks T= 12 weeks

Before start Start of treatment Consider Consider - Check for contraindications - Start 10mg/day dose increase dose increase - TPMT genotype - In patients with a body to 15mg/day to 20mg/day determination weight <60kg, consider to - Complete blood count, halve the tablet and to start leukocyte differentials, with 5mg/day - Complete blood count, leukocyte differentials, creatinine, GGT, ALT, AST, creatinine, GGT, ALT, AST, bilirubin, 6-TGN levels bilirubin - Evaluation of side effects - EASI - EASI

Figure 2 Study protocol ALT alanine aminotransferase; AST aspartate aminotransferase; EASI eczema area and severity index; GGT gamma-glutamyl transferase; TPMT Thiopurinemethyltransferase; 6-TGN 6-thioguanine nucleotide

According to the protocol, the start dose thioguanine was 10mg/day. This dose was based on experience with thioguanine in IBD patients, in which the registered dose is 0.3mg/kg. In the case of AD, a start dose of 10mg/day was chosen, because we know from clinical experience with azathioprine, that AD patients tend to have more toxicity compared to IBD patients using the same dose. In patients with a body weight <60kg, the start dose was 5mg/day. After two weeks, the dose was increased to 15mg/day. After four weeks, the dose could be further increased to 20mg/day, dependent on clinical response, side effects and laboratory findings. In patients who recently used AZA without side effects or laboratory deviations, the thioguanine start dose could be 15mg/day.

Clinical efficacy and safety Clinical efficacy was preferably evaluated every two weeks during the first two months and after three months by using the Eczema Area and Severity Index (EASI). Side effects and changes in co-medication were evaluated every visit. Safety controls include a complete blood count, leukocyte differentials, serum creatinine, gamma-glutamyl transferase (GGT), alanine aminotransferase (ALT), aspartate aminotransferase (AST) and bilirubin. Levels of 6-TGN were determined in red blood cells (RBC). 10 Myelotoxicity was defined as white blood cell count <4x109/L and/or thrombocytopenia (platelet count <150x109/L). Additionally, hemoglobin level (Hb) and lymphocyte counts were analyzed.

Statistical analysis All statistical analyses were performed using SPSS statistics 21. Frequencies, percentages and medians with inter quartile ranges (IQR) were calculated. 158 CHAPTER 10

Results Patient characteristics Five patients were enrolled (table 1). All patients had an extensive metabolizer TPMT genotype. Median age at the start of the thioguanine treatment was 54.0 years [IQR 49.8-62.2]. Most patients used more than one type of oral immunosuppressive drugs before. All patients used azathioprine before, but had to discontinue due to side effects, skewed metabolism or ineffectiveness (table 1).

Clinical responsiveness EASI scores before start ranged from 2.7 to 27.5 (median 11.7 [IQR 6.1-23.6]). EASI scores in patient 3 and 4 were remarkably low at the start of thioguanine, because these patients had been treated with azathioprine shortly before the start of thioguanine. Their treatment was changed to thioguanine, because of subjective side effects or a skewed metabolism on AZA treatment. After three months, the EASI scores ranged from 4.6 to 14.1 (median 10.2 [IQR 5.1-14.1]). Patient 2 had low EASI values after treatment, but it should be noted that this patient still used prednisone, which may influence the clinical outcome. In patient 3 the prednisone dose was successfully discontinued in the meantime.

Dosing Thioguanine doses were increased to 20mg/day in most patients. Patient 1 used 22.5mg/day, because at the time she received treatment, tablets of 10mg were not yet available. Back then, 7.5mg thioguanine capsules were compounded by our pharmacist and the patients received 3 capsules daily. In patient 5 the thioguanine dose was not further increased than 15mg/day, because of slightly increased liver enzymes and serum creatinine.

Laboratory examinations Levels of 6-TGN varied widely among the patients, but all patients showed an increase of the 6-TGN level after escalation of the thioguanine dose. After three months, 6-TGN levels ranged between 289 and 1443 (median 585.0 [IQR 310.0-1258.5]). No 6-TGN levels higher than 2600 pmol/8x108 RBC (toxic level) were found. None of the patients met the criteria of myelotoxicity. In patient 5, after three months the dose of thioguanine had to be decreased from 20mg/day to 15mg/day because of an increased serum creatinine and a slightly decreased hemoglobin value. After this dose decrease, both values improved to (nearly) normal. In patient 6, out of precaution, the thioguanine dose was not further increased than 15mg/day, because of slightly increased liver enzymes and serum creatinine. No other clinically relevant changes were found in the laboratory results. Table 1 Patient characteristics

Patient Sex History or oral immunosuppres- Reason for discontin- Thioguanine Thioguanine dose Clinical score before Clinical score Concomitant use of oral prednisone sive drugs uation of previously dose at start after 3 months start thioguanine after 3 months used azathioprine Before start After 3 months thioguanine

1 F Azathioprine, mycophenolate Not effective 7.5mg/day 22.5mg/day EASI 27.5 EASI 14.1 Yes, 20mg/day No mofetil, methotrexate, cyclosporine A

2 M Azathioprine, methotrexate, Not effective 10mg/day 20mg/day EASI 19.7 EASI 6.4 Yes, 10mg/day Yes, 10mg/day prednisone, tacrolimus, cyclosporine A

3 M Azathioprine with allopurinol, Subjective side effects 10mg/day Thioguanine EASI 2.7 Thioguanine Yes, 7.5mg/day NA cyclosporine A, enteric-coated discontinued after 2 discontinued mycophenolate sodium, months because of after 2 months prednisone, methotrexate side effects because of side effects

4 F Azathioprine, cyclosporine A Skewed metabolism 10mg/day 20mg/day EASI 9.4 EASI 4.6 No No

5 M Azathioprine with allopurinol, Skewed metabolism 10mg/day 15mg/day EASI 11.7 EASI 13.9 No No cyclosporine A, enteric-coated mycophenolate sodium, tacrolimus

EASI Eczema area and severity score;NA not applicable TREATMENTWITH THIOGUANINE 159 10 160 CHAPTER 10

Subjective side effects Some mild side effects were reported (table 2) that were acceptable for most patients and did not require a dose reduction. Only patient 3 had to discontinue thioguanine treatment after two months because of severe headache. It is not known if this headache was related to the thioguanine, because this patient was familiar with headache symptoms and had to discontinue other immunosuppressive treatments (including AZA) in the past because of the same reason. Besides, an uncontrolled hypertension, that was possibly present before initiation of thioguanine therapy, was detected in this patient.

Table 2 Number of patients with side effects during thioguanine treatment and dose reduction due to side effects Side effect Number of patients (%) Number of patients with dose with side effects reduction due to side effects Headache 2 (33.3) 1 Abdominal pain 1 (16.7) 0 Facial hair growth 1 (16.7) 0 Oral candida 1 (16.7) 0

Discussion This is the first study using thioguanine in five patients with severe, difficult to treat AD who failed on AZA treatment. In difficult to treat AD patients thioguanine can be an alternative treatment option. Disease activity decreased in most patients and one patient could even stop oral steroids. Side effects during thioguanine treatment were limited: only one patient had to discontinue treatment because of headache symptoms.

Thioguanine has recently been registered for patients with IBD in the Netherlands. In 2014, there were approximately 1500 thioguanine users in the Netherlands, with no serious toxicity reported.7 Meijer et al performed a systematic review in 2016 on the use of thioguanine in inflammatory bowel disease.8 In this review, 228 out of 353 IBD-patients (65%) had clinical improvement on thioguanine therapy. Short-term results were based on 268 treatment years (median follow-up 9 months, range 3-22 months) with a median daily dose of 20 mg (range 10-80 mg). Discontinuation, mostly due to adverse events, was reported in 72 patients (20%), compromising mainly gastrointestinal complaints, hypersensitivity reactions and elevated liver enzymes. TREATMENT WITH THIOGUANINE 161

No studies on AD using thioguanine have been published before. Five publications on the use of thioguanine in psoriasis patients were found. A review of these studies was published in 2002 by Sherer et al.9-14 The included studies showed an improvement of the psoriasis, but bone marrow suppression was the most reported side effect. These side effects are probably related to the dose of thioguanine, that was higher (up to 160mg/day) than the dose that we used in our study population.

Thioguanine is thought to have a more favorable side effect profile compared to AZA. In this study, only a few mild subjective side effects have been reported, but this study is too small to compare the incidence of side effects of AZA to thioguanine. One out of the five patients treated with thioguanine in this study needed to discontinue treatment due to headache, but the question remains whether this was related to the treatment with thioguanine, because this patient was familiar with headaches, even on previously AZA treatment. In one other patient, the thioguanine dose was reduced because of increasing serum creatinine values. No other clinically relevant side effects or laboratory results were reported.

Limitations of the present study are the open observational design that resulted in an inter- observer variability in EASI scores, the small sample size of only five patients and the limited follow up period. Besides it should be highlighted that these five patients were categorized as difficult to treat AD patients: all of them had failed on AZA treatment and most of them had also failed on several other oral immunosuppressive drugs. Randomized controlled studies on thioguanine with longer follow up periods and larger numbers of patients are needed to further investigate the efficacy and safety of thioguanine in the treatment of patients with of moderate to severe AD who failed on AZA.

Conclusion Treatment with thioguanine may be an alternative treatment option in patients with difficult to treat, severe AD, who failed on treatment with AZA.

10 162 CHAPTER 10

References

1. Roekevisch E, Spuls PI, Kuester D, Limpens 9. Sherer DW, Lebwohl MG. 6-thioguanine J, Schmitt J. Efficacy and safety of systemic in the treatment of psoriasis: a case report treatments for moderate-to-severe atopic and literature review. J Cutan Med Surg dermatitis: a systematic review. J Allergy 2002;6:546-50. Clin Immunol 2014;133:429-38. 10. Molin L, Thomsen K. Thioguanine treat- 2. Van der Schaft J, Politiek K, van den Reek ment in psoriasis. Acta Derm Venereol JM, et al. Drug survival for azathioprine and 1987;67:85-8. enteric-coated mycophenolate sodium in a long-term daily practice cohort of 11. Zackheim HS, Maibach HI. Treatment of adult patients with atopic dermatitis. Br J psoriasis with 6-thioguanine. Australas J Dermatol 2016;175:199-202. Dermatol 1988;29:163-7.

3. Politiek K, van der Schaft J, Christoffers WA, 12. Zackheim HS, Glogau RG, Fisher DA, et al. Drug survival of methotrexate treat- Maibach HI. 6-Thioguanine treatment of ment in hand eczema patients: results from psoriasis: experience in 81 patients. J Am a retrospective daily practice study. J Eur Acad Dermatol 1994;30:452-8. Acad Dermatol Venereaol 2016;30:1405-7. 13. Silvis NG, Levine N. Pulse dosing of 4. Politiek K, van der Schaft J, Coenraads PJ, thioguanine in recalcitrant psoriasis. Arch et al. Drug survival for methotrexate in a Dermatol 1999;135:433-7. daily practice cohort of adult patients with 14. Mason C, Krueger GG. Thioguanine for severe atopic dermatitis. Br J Dermatol refractory psoriasis: a 4-year experience. J 2016;174:201-3. Am Acad Dermatol 2001;44:67-72. 5. Christoffers WA, Politiek K, Coenraads PJ, et al. Drug survival of cyclosporine in the treatment of hand eczema: a multicentre, daily use study. J Eur Acad Dermatol Venereol 2016;30:63-6.

6. Van der Schaft J, Politiek K, van den Reek JM, et al. Drug survival for ciclosporine A in a long term daily practice cohort of adult patients with atopic dermatitis. Br J Dermaol 2015;172:1621-7.

7. Mulder CJ, van Asseldonk DP, de Boer NK. Drug rediscovery to prevent off-label prescription reduces health care costs: the case of thioguanine in the Netherlands. J Gastrointestin Liver Dis 2014;23:123-125.

8. Meijer B, Mulder CJ, Peters GJ, et al. Efficacy of thioguanine treatment in inflammatory bowel disease: A systematic review. World J Gastroenterol 2016;22:9012-9021. TREATMENT WITH THIOGUANINE 163

General discussion

166 CHAPTER 11

General discussion In this thesis we evaluated the experience with and safety aspects of maintenance treatment of oral immunosuppressive drugs in atopic dermatitis (AD) in a daily practice setting. Besides, the possibilities of the use of thiopurines were further explored, in the search for alternative treatment options for moderate to severe AD. This general discussion will consist of three parts, followed by concluding remarks and a look at future aspects.

1. Qualitative and quantitative exploration of oral immunosuppressive drug use for AD in daily practice in the Netherlands. How often are oral immunosuppressive drugs prescribed for AD and which compounds are used? This evaluation gives an impression of the percentage of difficult to treat AD patients in the Netherlands. 2. Open questions concerning safety of long-term oral immunosuppressive drugs in AD; gaps in the field of safety are studied with an attempt to answer open questions in the guidelines. 3. Improving performance of thiopurines in AD in order to explore new treatment options for difficult to treat AD patients.

1. Qualitative and quantitative exploration of oral immunosuppressive drug use for atopic dermatitis in daily practice in the Netherlands.

1.1 How often are oral immunosuppressive drugs prescribed in atopic dermatitis patients and which drugs are regularly used in daily practice?

Use of oral immunosuppressive drugs in the treatment of atopic dermatitis: a qualitative exploration In the studies described in chapter 2 and 3 of this thesis, the treatment of AD with oral immunosuppressive drugs in the Netherlands was evaluated. In chapter 2, the prescription behavior in 334 patients in the University Medical Center Utrecht and Amsterdam, two centres with expertise in the treatment of severe AD, in the period from 2001-2011 was analyzed. Cyclosporine A was given in 80% of the patients, mycophenolate mofetil or enteric-coated mycophenolate sodium in 31%, azathioprine in 14%, methotrexate in 11%, systemic glucocorticosteroids in 7% and systemic tacrolimus in 5%. Cyclosporine A was the first choice treatment in 252 patients. In chapter 3, the prescription behavior throughout the Netherlands was evaluated, based on a large database containing information of prescriptions from 556 public pharmacies. The database also contains a small number (<10) of outpatient pharmacies located in hospitals. Surprisingly, this study showed that between 2012 and 2017, mostly methotrexate and azathioprine were prescribed to patients with AD. GENERAL DISCUSSION 167

Although both studies in chapter 2 and 3 analyzed the prescription behavior of oral immunosuppressive drugs in the Netherlands, a striking difference in the preference was seen: in chapter 2 cyclosporine A was the most commonly prescribed drug, while chapter 3 shows that azathioprine and methotrexate were more popular. The preference for azathioprine and methotrexate in the study in chapter 3 is striking, because cyclosporine A is still the only registered oral immunosuppressive drug for the treatment of AD in the Netherlands. Several factors may have contributed to the differences between the studies in chapter 2 and 3. An important difference between the studies is the time frame in which the analyses were performed: the study in chapter 2 was performed between 2001-2011, while the study in chapter 3 evaluated prescription behavior between 2012-2016. An explanation for the fact that many azathioprine and methotrexate prescriptions were found in chapter 3 may be that these treatments were second line options and that patients already used cyclosporine A before 2012. Since the data in the used NControl database were not complete for a longer time period (than 2012-2017), this assumption could not be checked. Another explanation for the increased prescription frequency of azathioprine and methotrexate in chapter 3 may be that after 2011 several publications including a randomized controlled trial on azathioprine and methotrexate use in AD patients, were published. This may have led to the increased popularity of these drugs. Both studies were performed in a different setting: the study in chapter 2 included two university hospitals, while chapter 3 focused on the prescriptions in general hospitals throughout the country. Dermatologists working in general hospitals probably have more experience with methotrexate and azathioprine than cyclosporine A, as these drugs are more commonly prescribed for other inflammatory skin disorders like cutaneous lupus and bullous diseases. Concerns about the (long-term) safety of cyclosporine A and questions according monitoring of side effects may be present in these general hospitals and may be an explanation for the fact that methotrexate and azathioprine were more commonly prescribed in the study in chapter 3. In the study in chapter 3, remarkably, 31% of the patients discontinued treatment within the first months after start: the time between the first and last prescription in these patients was less than 30 days. Sometimes even no repeat prescription was disposed. The data described in chapter 3 include mostly non-specialized centres. Because experience with (long-term) use of oral immunosuppressive drugs is often lacking in these centres, treatments are often discontinued early. This may be due to the fact that unexperienced dermatologists are not aware of the fact that it takes at least three months for azathioprine and methotrexate to achieve clinical improvement. Therefore treatments may be discontinued early, in the absence of clinical effectiveness in the initial phase. On the other hand, treatments may be discontinued early because of the fear of 11 the development of side effects after long-term treatment. 168 CHAPTER 11

Finally, different methods were used in the studies: chapter 2 describes a retrospective medical chart review. This method is very accurate, but the population is small and biased as only patients in two university hospitals with a focus on AD were included. Data are probably not applicable to all Dutch university hospitals, since the expertise of treating severe AD is generally clustered in a few hospitals. In chapter 3, data were derived from pharmacy prescriptions based on an algorithm, in the absence of documented diagnoses in the used database. The algorithm was based on the use of different drugs that were supposed to be used by AD patients and was made with most care. Prescriptions that were supposed to be made for rheumatic diseases or organ transplant patients, were excluded from the analysis as much as possible. Nevertheless, since diagnoses could not have been checked, it is still possible that other (skin) diseases than AD were included in this study. A nationwide database including disease diagnosis would have been preferable, but was regrettably not available. However, despite all differences between chapter 2 and 3, the study in chapter 3 represents the common practice concerning immunosuppressive treatment for AD in the Netherlands probably better.

In the past years, a few other studies on the prescribing behavior of oral immunosuppressive drugs in the treatment of AD have been published.1-4 Most studies were monocentric and only a small number of patients was included. The time frame in which those studies were performed differed and different methods were used, varying from an online survey among dermatologists, a retrospective medical chart review and a prospective registry of AD treatment. Therefore interpretation of these data should be made with care. Overall, cyclosporine A was the first choice treatment in these studies. This is in line with the fact that in most European countries, cyclosporine A is the only licensed oral immunosuppressive drug for the treatment of AD. An exception is the study of Taylor et al. In this study azathioprine was the most common drug reported being used as first line in the United Kingdom. Methotrexate was the most commonly used second-line agent and mycophenolate mofetil the most commonly used third-line drug.

Conclusion 1: In the University Medical Center Utrecht and Amsterdam, two centres with experience in the treatment of eczema, cyclosporine A is most frequently prescribed as first line oral immunosuppressive treatment in atopic dermatitis. In general hospitals, azathioprine and methotrexate were more frequently prescribed as first or second line treatment.

Use of oral immunosuppressive drugs in the treatment of atopic dermatitis: a quantitative exploration Since oral immunosuppressive drugs are only used in more severe and difficult to treat AD patients, evaluation of prescription of these drugs may give an impression of the number of patients with difficult to treat AD. GENERAL DISCUSSION 169

In chapter 3, we performed a study in which we investigated the information from 556 pharmacies throughout the Netherlands, by using an algorithm in the NControl database. Based on the algorithm, 65 943 AD patients were identified, of which 943 patients (1.4%) were treated with oral immunosuppressive drugs in a five year period. If we correct this number for the atopic factor and for the total number of pharmacies in the Netherlands, a total of 5070 patients using oral immunosuppressive drugs would be present during this period. On December 31st 2016, 411 patients (0.6%) were defined as ‘active’ patients, using oral immunosuppressive drugs. When we correct this number for the atopic factor and for the total number of pharmacies in the Netherlands, a total of 2210 patients using oral immunosuppressive drugs on December 31st 2016 was found. In this study, (short) courses of prednisone were not included, in order to give a correct estimate of the number of patients with difficult to treat AD.

In literature, most publications refer to the study of Schmitt et al when trying to answer the question how many AD patients have a difficult to treat disease. Nevertheless, questions arise whether the percentage given in this study is a correct reproduction of the actual number. Schmitt et al evaluated the medical treatment of 11,555 AD patients between 2003 and 2004 and concluded that 10.2% of the adults were treated with systemic corticosteroids and <0.1% with cyclosporine A.5 These patients were supposed to have moderate to severe eczema, or at least eczema that could not be controlled with topical corticosteroids. This publication is often cited and many publications state that up to 10% of the adult AD patients actually use, or are candidates for oral immunosuppressive drugs. Until now, this percentage has not been confirmed in other studies. Schmitt et al concluded that, although their cohort was a representative sample of patients with AD, the percentage of 10% is rather high for general hospitals without a focus on AD treatment. No explanation was given for this high percentage in this publication, but the mean number of systemic corticosteroids prescribed in the study period of 2 years was 58.5 g (range 1-1,770 g). Although the duration of treatment was not mentioned in this publication, the range of treatment dosing indicates that short courses of prednisone were also included in this analysis. Short courses of prednisone are often given to patients with AD to treat an acute exacerbation of symptoms, also in Germany. This is often not restricted to patients with moderate to severe, difficult to treat AD: patients with mild to moderate AD with occasional flares may be treated with prednisone courses as well. The percentage of 10% is most likely an overestimation of the real number of AD patients in need for oral immunosuppressive drugs, and may not be a realistic reflection of the group of patients with really difficult to treat AD. This population commonly needs maintenance treatment (> 3 months) with oral immunosuppres- 11 sive drugs. 170 CHAPTER 11

In addition, it should be noted that the method that Schmitt et al used differed from our studies in chapter 2 and 3 and that it is difficult to directly compare the found percentages of patients using oral immunosuppressive drugs.

Conclusion 2: Oral immunosuppressive drugs are prescribed in the treatment of moderate to severe and difficult to treat AD. The cumulative percentage of patients using these drugs is supposed to be 1.4% over a five year period (2012-2017). Cross-sectional, 0.6% of the patients used these drugs on December 31st 2016.

2. Open questions concerning safety of long-term oral immunosuppressive drug use in atopic dermatitis

2.1 Current guidelines and their gaps in the field of safety Before starting oral immunosuppressive drugs, dermatologists have to discuss the advantages (potential efficacy) and disadvantages (potential risks or adverse events) of this therapy with their patients. The risk-benefit ratio may change over time when treatment is continued for several years. In daily practice, questions arise on whether a prolonged treatment with oral immunosuppressive drugs is safe, for instance concerning the risk of infections, the risk of malignancies and pregnancy outcomes. Information on these aspects of immunosuppressive drugs is often incomplete in guidelines.

Risk of lymphopenia and serious infections General (inter)national guidelines on oral immunosuppressive drugs recommend regular monitoring of organ related toxicity using clinical laboratory tests. These tests vary for the different drugs, but often include liver- and renal function and a complete blood count. Monitoring an elevated risk of infection is more difficult and in the absence of a standardized questionnaire, guidelines often advise to monitor the white blood cell count (WBC) (normal range 4.0-10.0x109/L). However, in our study in chapter 4 we showed that measurement of the WBC as an indicator for an increased risk of infection is insufficient. We evaluated a large group of AD patients treated with oral immunosuppressive drugs and found eleven patients with lymphopenia. It is notable that in most of these 11 patients a prolonged lymphopenia occurred during maintenance treatment with prednisolone. Surprisingly, we noticed that in some patients there was a lymphopenia, while the WBC was within the normal range. These patients would have been missed when only the WBC was measured. Because prolonged lymphopenia is associated with an increased risk of opportunistic infections, including the John Cunningham (JC) virus that potentially might lead to progressive multifocal leukoencephalopathy (PML), and pneumocystis jiroveci pneumonia (PJP), it is important to be aware of this risk and to regularly control the lymphocyte count. GENERAL DISCUSSION 171

Our study described in chapter 4 has some limitations that need to be highlighted. The reported number of eleven patients with a persistent lymphopenia will probably be an underestimation of the real number of patients with lymphopenia, because lymphocyte counts were not measured regularly in all patients. The exact number of patients with a lymphopenia therefore remains unclear. Based on this study, it also remains unclear what the consequence of a detected lymphopenia in daily practice should be. No serious or prolonged infections were reported in this group, but the number of eleven patients with persistent lymphopenia is small. It is remarkable that these patients did not develop infections, since some of their laboratory measurements (including T-cell subsets) were quite alarming. This finding makes it questionable whether long-time antibiotic prophylaxis in all these patients is mandatory. Larger, prospective observational studies are needed to find out what the exact incidence of lymphopenia is in AD patients treated with oral immunosuppressive drugs and what the risk for developing infections is in these patients.

Nevertheless, since persistent lymphopenia was occasionally seen in our population, routinely measurement of WBC subsets, including the lymphocyte count, should be considered in AD patients using (long-term) oral immunosuppressive drugs. In patients with a prolonged lymphopenia during treatment (defined as five or more lymphocyte counts below9 0.8x10 /L), there should be attention for the possible risk of infectious diseases. Therefore we suggest to consult a clinical immunologist to discuss patients using oral immunosuppressive drugs in which a prolonged lymphopenia is observed. In collaboration with a clinical immunologist the risk of serious infections can be evaluated and the need for antibiotic prophylaxis or discontinuation of oral immunosuppressive treatment discussed.

Conclusion 3: Lymphopenia is occasionally seen in AD patients treated with oral immunosuppressive drugs. Larger, prospective observational studies are needed to find out what the exact incidence of lymphopenia is in these patients and what the clinical consequences are. Measuring lymphocyte counts in these patients can be considered, and in case of a prolonged lymphopenia, it is suggested to consult a clinical immunologist to discuss the risk of infections and the need for antibiotic prophylaxis.

Risk of HPV related tumors Studies in patients with inflammatory bowel disease, rheumatic diseases and organ transplant patients under immunosuppressive treatment suggest an increased risk of human papilloma 11 virus (HPV) associated tumors. The actual risk of HPV associated tumors in AD patients is unknown. General guidelines on oral immunosuppressive drugs in AD do not give practical recommendations how to monitor on HPV infection-related tumors. 172 CHAPTER 11

Non melanoma skin cancer In the study described in chapter 5, an increased risk of squamous cell carcinoma was found in 557 AD patients treated with oral immunosuppressive drugs (standardized incidence ratio 13.1 for the total group and 25.3 for cyclosporine A monotherapy. Comparison made to the general Dutch population). This increase is also seen in other diseases treated with oral immunosuppressive drugs, although the standardized incidence ratios in organ transplant patients and rheumatic diseases are even higher. There are some reasons to question if the increased incidence of a squamous cell carcinoma in chapter 5 was causally related to the intake of the oral immunosuppressive drugs. First, we found no association between the duration of treatment and the risk of a squamous cell carcinoma: some patients already developed the tumor within the first year. Besides, some patients developed a squamous cell carcinoma more than ten years after the discontinuation of their oral immunosuppressive treatment. Finally, some of the patients who developed a squamous cell carcinoma during or after the use of oral immunosuppressive drugs, already had another squamous cell carcinoma before they started their treatment. It is possible that these patients are just more prone to develop skin tumors, independently of their treatment.

One of the reasons for the increased prevalence of squamous cell carcinoma can be that in AD patients using oral immunosuppressive drugs skin inspection is more often performed, compared to patients without a skin disease. There are also studies showing that the risk of non-melanoma skin cancer (including basal cell carcinoma and squamous cell carcinoma) in AD patients, independently of their drug use, is increased.6-8 Possible reasons for altered risks in AD patients include exposure to ultraviolet irradiations and AD-related pathomechanistic factors, including impaired cell-mediated immunity.8 However, greater cutaneous surveillance associated with regular care for AD may also be accountable.6

Based on our findings, guidelines should be supplemented with the recommendation to start skin inspection directly after the start of the oral immunosuppressive treatment. In addition, since some tumors developed even years after discontinuation of treatment, patients should be instructed to contact their general practitioner or dermatologist in case of new skin abnormalities. Finally, the use of sunscreens should be highly recommended to all patients using oral immunosuppressive drugs.

Conclusion 4: An increased risk of squamous cell carcinoma is seen in AD patients treated with oral immunosuppressive drugs, although it is not clear whether there is a causal relation with the intake of the drugs. Nevertheless, patient education, sun protection and regular skin inspections by a dermatologist at every follow up visit are advised in these patients, directly from the start of the treatment. GENERAL DISCUSSION 173

HPV related cervical carcinoma Data on the risk of cervical carcinoma in patients receiving oral immunosuppressive drugs are conflicting: some studies report an increased incidence of cervical cancer9,10 while other studies showed no increased HPV infections and increased cancer risk.11-14 According to the Dutch CIN guideline, no intensified screening is suggested in women treated with oral immunosuppressive drugs. Based on our results from chapter 6 and based on the study of Clowry et al, we also conclude that no intensified screening program for women with AD treated with oral immunosuppressive drugs, aged >30 years, is indicated.15 However, both studies evaluated a small patient population with a limited follow up period and therefore an increased risk of cervical carcinoma in these women still cannot be ruled out completely, especially not in the case of long term treatment. Therefore, the question arises if women aged <30 years who are not part of a national screening programme for cervical cancer should be screened during oral immunosuppressive treatment. Since there are no data to answer this question, the answer will be given by shared decision making. In these patients one or two extra cervical screening moments with a maximal time interval of five years may be considered.

It is notable that since September 1st 2009, prophylactic HPV vaccination is part of the Dutch national vaccination program and offered to all twelve year old girls, in order to decrease the risk of developing cervical carcinoma. Nevertheless, since the percentage of girls participating in this vaccination program in the Netherlands is low (on January 1st 2017, 53.4% of girls born in 2002 had taken part in the vaccination program), all women are still advised to take part in the national screening programme.

Based on the findings of chapter 6, future guidelines on AD should be supplemented with some information on the risk of cervical cancer in women taking oral immunosuppressive drugs, in order to increase the awareness regarding this topic. The conclusions from literature should be discussed with the patient and risks and benefits of starting treatment with oral immunosuppressive drugs should be weighted.

Conclusion 5: There is no increased risk of cervical carcinoma in AD patients treated with oral immunosuppressive drugs. No intensification of the national cervical screening program is necessary. In women aged <30 years one or two extra cervical screening moments within a maximal time interval of five years may be considered. 11 174 CHAPTER 11

Oral immunosuppressive drug use at the moment of conception Drug use in men may influence the quantity and quality of the sperm and may influence the change for a successful pregnancy and its outcome. In most dermatological guidelines on the treatment of skin diseases with oral immunosuppressive drugs, no attention is paid to this aspect. In chapter 7 of this thesis, we performed a systematic search and evaluated studies on this topic. Most of the studies on azathioprine, methotrexate and mycophenolic acid that compared their outcomes to a healthy population conclude that there is no increase in adverse pregnancy or fetal outcomes. The exceptions are the thiopurines. One study found an adjusted risk of congenital abnormalities in children fathered by men using azathioprine or 6-mercaptopurin of 1.8 (95% CI 0.7-5.0).16 Another study concluded that the risk of complications was significantly higher in the group using 6-MP at least 3 months prior to conception compared to (1) men who used 6 -MP until a maximum of 3 months prior to conception (p=0.013, OR 16.0; 95% CI 1.6-161) or (2) men who had never used 6-MP (p<0.002, OR 19.6; 95% CI 3.1-122).17 These immunosuppressive drugs should be used with caution and risks of continuing these drugs during conception should be discussed with the patient.

In summary, according to the presented data in chapter 7, an increased risk of negative pregnancy outcomes seems to be rare. Nevertheless, forming clear recommendations is tough. The studies utilized a variety of different designs and methods and most studies lacked information on dosage, duration of therapy, comorbidities and the use of concomitant medication in both fathers and mothers. In addition, there was heterogeneity in outcome measurement across studies.

Future guidelines should be supplemented with the information that all young men should be informed that we know little about the risk that azathioprine, methotrexate and mycophenolic acid may pose during conception. Preterm birth and congenital malformations have been described in the evaluated studies and an increased risk due to paternal drug use cannot be ruled out completely. Since cyclosporine A can be used safely during conception, this drug should get the preference in all young AD patients with an indication for oral immunosuppressive drugs.

The publication of the data in chapter 7 resulted in more awareness for this subject, and recommendations have already been included in national and international guidelines. Although the suggestions in the guidelines are still summarily, it is the first step towards an increased awareness regarding this topic.

Conclusion 6: Adverse pregnancy and fetal outcomes in pregnancies fathered by men taking azathioprine, methotrexate and mycophenolic acid are rare. Since the risk cannot be ruled out completely, the awareness regarding these risk among dermatologists should be increased and patients should be informed. GENERAL DISCUSSION 175

3. Improving performance of thiopurines in atopic dermatitis in order to explore new treatment options for difficult to treat atopic dermatitis patients

3.1 Measurement of thiopurine metabolite levels in atopic dermatitis patients treated with azathioprine Most guidelines suggest to determine the thiopurine S-methyltransferate (TPMT) activity in each AD patient before initial treatment with azathioprine, in order to guarantee a safe dose escalation. The individual enzyme activity varies depending on the genetic polymorphisms of TPMT gene located at chromosome 6.18 Based on the TPMT genotype, three phenotypical groups can be distinguished: poor metabolizers (severly decreased enzyme activity), intermediate metabolizers (decreased enzyme activity) and extensive metabolizers (normal or wild-type enzyme activity).

The azathioprine dose is often suggested to be matched with the TPMT enzyme activity; patients with lower enzyme capacities require lower doses. However, in chapter 8 of this thesis, we showed that, although most patients had the same TPMT genotype (corresponding to the extensive metabolizer phenotype), a wide range was observed in the azathioprine metabolite levels. For instance, levels of 6-TGN ranged from 99 to 656 pmol/8x108 RBC in all patients treated with 150mg/day azathioprine. The TPMT enzyme activity was not a predictor for the serum level of the metabolites. This is in line with the view that several other enzymes contribute to the azathioprine metabolism and the formation of 6-TGN and 6-MMP. It is suggested that TPMT status only explains 30-60% of all cases with thiopurine intolerance.19,20 Recently several publications appeared on other contributing enzymes, such as hypoxanthine guanine phosphoribosyl transferase (HGPRT), glutathiones-transferases (GSTs), xanthine oxidase (XO), inosine triphosphate pyrophosphatase (ITPase) and nucleoside diphosphate-linked moiety X motif 15 (NUDT15).21,22 It is now postulated that the variability in the azathioprine drug response is, at least partly, determined by genetic polymorphisms of at least one and probably more of these enzymes.19

In both part I and II of the study described in chapter 8, a few patients being extensive metabolizers with a skewed metabolism were detected (raised 6-MMP/6-TGN ratios with very high 6-MMP levels). It is thought that this occurs in up to 20% of the population.23,24 These patients are at risk for developing hepatotoxicity after prolonged treatment and measuring metabolite levels may identify these patients in an early stage, enabling pre-emptive dose adjustments. As 11 confirmed in the study in chapter 8, it is known that the risk of developing a skewed metabolism cannot be predicted by only measuring the TPMT enzyme activity. 176 CHAPTER 11

Conclusion 7: Routinely monitoring of 6-TGN and 6-MMP metabolite levels during azathioprine treatment is more effective in displaying interpatient variability than TPMT genotyping before treatment, probably because of the involvement of the multiple enzymes in the metabolic pathways of thiopurines.

No clear cut therapeutic range of 6-TGN in steady-state could be defined based on the results of the study in chapter 8. Levels of 6-TGN did not differ significantly between clinical responders and non-responders. However, the results from chapter 8 (showing a median 6-TGN level in responders of 197.0), suggest that 6-TGN levels needed to induce clinical efficacy might be lower compared to inflammatory bowel disease where the therapeutically range is 600-1200 mpol/8x108 RBC. This has also been described in studies on azathioprine treatment in other dermatological disease. El Azhary et al determined optimal levels of 6-TGN for disease remission in patients with immunobullous disease treated with AZA. They found a mean optimal level of 6-TGN of 190.7 pmol/8x108 RBCs for all patients. Limited disease requires less 6-TGN to achieve clinical improvement in this study, with a mean of 145.3 pmol/8x108 RBCs.25 Caufieldet al treated twelve children with recalcitrant AD with azathioprine. For the 11 responders, 6-TGN levels ranged from 45 to 358 pmol/8x108 RBC at the time their improvement stabilized. There was no difference in clinical response between patients with 6-TGN levels ranging 45 to 54 and patients with levels ranging from 207 to 358 pmol/8x108 RBC.26 Future research in a larger AD population may complement our findings and might lead to the definitions of cut-off values for the therapeutically range in patients with AD treated with azathioprine.

Conclusion 8: A large spread is seen in 6-TGN levels in AD patient among clinical responders. No clear correlation between 6-TGN levels and clinical responsiveness was seen.

Conclusion 9: Based on a comparison to literature, levels of 6-TGN that are needed to induce clinical efficacy seem to be lower in atopic dermatitis patients compared to patients with inflammatory bowel disease.

Measuring the metabolite levels is a more direct way to monitor the azathioprine metabolism in the patients and seems to be more safe than only investigating the TPMT enzyme activity, especially in the 20% of patients that develops a skewed metabolism. Skipping the TPMT genotyping may be considered, when regular laboratory controls (including the measurement of 6-TGN and 6-MMP soon after the start of the treatment) are guaranteed. However, since the measurement of the azathioprine metabolites is more expensive than TPMT analysis (€ 60,- per metabolite measurement and €120,- for a one-off TPMT determination) and only possible in a limited number of (mostly university) clinical laboratories, preference is still given to TPMT measurement for patients treated in daily practice. Nevertheless, in patients who do not respond to azathioprine treatment after treatment with adequate dosing for period of GENERAL DISCUSSION 177

at least three months, measuring the azathioprine metabolite levels should be considered, as it can give an insight in the drug metabolism and may explain non-responsiveness.

In conclusion, measuring 6-TGN and 6-MMP may be useful in the individual patient. It can give the clinician an insight in the inter-individual variation of the azathioprine metabolism and can give an answer to the question whether dose escalation is possible and safe. Besides, it can indicate patients who are not taking their medication and patients at risk of developing myelotoxicity or hepatotoxicity. Therefore it may give the clinician the opportunity to adequately monitor these patients or to adjust the azathioprine dose, in order to improve the performance of azathioprine.

Conclusion 10: Measuring thiopurine metabolite levels in atopic dermatitis patients can improve the performance of azathioprine.

3.2 Influencing the azathioprine metabolism by co-prescription of allopurinol Co-prescription of other drugs may positively influence the azathioprine metabolism and can be an interesting way to improve the performance of azathioprine. This is especially interesting in patients with a skewed metabolism.

The co-prescription of allopurinol has gained popularity for bypassing thiopurine-related (subjective) side effects (mostly gastro-intestinal) and hepatotoxicity in patients with inflammatory bowel disease.24,27-30 Most of these studies report that allopurinol can cause a shift of the azathioprine metabolism towards the 6-TGN production, resulting in a decrease of thiopurine hepatotoxicity. Additionally, the remission rate is reported to improve in inflammatory bowel disease patients.

In chapter 9, we reported our study in which we treated 15 patients with both azathioprine and allopurinol. Data of this study were promising. 6-MMP levels and the 6-MMP/6-TGN ratio decreased in all patients after the addition of allopurinol; 6-TGN levels increased in all patients. The number of responders increased from four to seven patients after co-prescription of allopurinol. In two of the four patients who started co-prescription of allopurinol because of gastro- intestinal side effects on azathioprine monotherapy, these side effects resolved after the addition of allopurinol and dose reduction of azathioprine. We suggested that the co-prescription of allopurinol should be considered in patients with a skewed metabolism or in patients with subjective side effects in which the azathioprine dose 11 cannot be increased. It is important to reduce the azathioprine dose by at least 50% before the start of allopurinol, in order to prevent myelotoxicity due to (too) high levels of 6-TGN. 178 CHAPTER 11

Conclusion 11: Azathioprine with allopurinol co-prescription is an effective treatment option for patients with AD who have a skewed metabolism or experience (subjective) side effects to azathioprine monotherapy.

Although the concomitant use of allopurinol seems to be successful, some comments should be posted. In some patients in our study we found extremely high 6-TGN levels after the addition of allopurinol, although the azathioprine dose in these patients was carefully tapered with at least 50%. There is some evidence that the addition of allopurinol 50mg/day may be sufficient to optimize azathioprine therapy.31,32 In our daily practice, we have limited experience with this reduced allopurinol dose, but this can be considered and might be successful in patients who develop side effects to allopurinol or who develop largely increased 6-TGN levels although the azathioprine dose was adequately reduced.

Treatment with allopurinol is not without risks. In our study, no side effects of allopurinol have been observed. However, allopurinol is known for a variety of side effects including especially allergic reactions with skin rash (1-10%). It is always important to discuss possible side effects with your patients before starting allopurinol. Finally, the prescription of two different drugs requires some commitment of the patient. Patients need to understand the function of allopurinol and need to be aware of the fact that the dose of azathioprine (that most of them will be using for quite some time), needs to be reduced. When the azathioprine dose is not reduced with at least 50%, patients are at risk for developing increased levels of 6-TGN, probably resulting in myelotoxicity. Therefore regular contact with the patient and laboratory monitoring remains mandatory.

3.3 Bypassing the azathioprine metabolism by using thioguanine Thioguanine (a thiopurine drug just as azathioprine and mercaptopurine) is a new player in the management of inflammatory bowel disease and skin diseases. Less enzymes are involved in the metabolism of thioguanine, leading more directly towards the pharmacologically active products.33 In the past, physicians were reluctant to introduce thioguanine as a standard treatment of inflammatory bowel disease, because one study reported hepatotoxicity (nodular regenerative hyperplasia and sinusoidal obstruction syndrome), especially in high doses (median 40mg/ day).33,34 Meijer et al performed a systematic review on the experiences with thioguanine and concluded that these findings were not confirmed in subsequent studies in which lower dosages of thioguanine (20mg/day) were prescribed.33 In their review, they conclude that the efficacy of thioguanine in inflammatory bowel disease patients being intolerant to azathioprine therapy is 65%, with short term adverse events in 20% of the patients. GENERAL DISCUSSION 179

Experience with thioguanine for skin diseases is scarce. Studies from the eighties and nine- ties showed that thioguanine can be a treatment option for psoriasis, but that myelotoxicity is the most common side effect.35-38 Remarkably, dosages in these studies were quite high (up to 160mg/day), which may be an explanation for the presence of myelotoxicity. Zakheim et al treated 81 patients with psoriasis with thioguanine. Thirty-seven patients (48.8%) were effectively maintained (free from treatment failure) for 3 to 145 months (median 33 months).35 Molin et al reported treatment of 9 psoriasis patients with thioguanine.36 Most patients demonstrated some degree of clearing after 3-4 weeks. Mason et al retrospectively evaluated 18 psoriasis patients treated with thioguanine.37 Fourteen patients (78%) had dramatic improvement (>90%), three patients had less improvement (50-90%) and 1 patient had less than 50% improvement.

In 2002, an American guideline on the treatment of psoriasis suggested in the absence of FDA approval to use thioguanine in psoriasis. The starting dose they suggested was 80mg two times/week, increased by 20mg every 2-4 weeks with a maximum dose of 160mg three times per week.

In the study described in chapter 10 of this thesis, we investigated the clinical effect and safety of low doses thioguanine in five patients with AD. Although the follow up period is short (around 12 weeks) and the study size is small, these first results are interesting in this selected group of difficult to treat AD patients, who previously failed on oral immunosuppressive drugs (including azathioprine treatment). The AD scores improved in most patients. One patient had to discontinue treatment due to subjective side effects. No serious side effects were seen, especially no myelotoxicity. Therefore thioguanine may be an alternative option in AD patients who failed on azathioprine monotherapy. But, it should be kept in mind that thioguanine is more expensive than azathioprine (thioguanine 20mg/day €33,93 for one week, azathioprine 150mg/day €2,60 per week). Whether thioguanine will also be effective and safe for long-term treatment in patients with AD who failed on AZA has to be further investigated in a randomized trial with a larger group of AD patients with longer follow-up periods. Whether a higher dose of thioguanine (higher than 20mg/day) will be safe and more effective in the treatment of AD, is yet unknown.

Conclusion 12: Low doses of thioguanine (up to 20mg/day) may be an effective and safe treatment in AD patients who failed on azathioprine. 11 180 CHAPTER 11

4. Concluding Remarks In this thesis, long term safety of the use oral immunosuppressive drugs was investigated, as well as attempts to optimize the treatment with thiopurines. The studies in this thesis focused on relevant questions that arose in daily practice and the conclusions of this thesis will be very valuable for updating and completing existing national and international guidelines. The following conclusions can be drawn from this thesis:

Prescription behavior of oral immunosuppressive drugs in atopic dermatitis patients in the Netherlands - Cyclosporine A was most often used as first choice treatment in two Dutch University hospitals, but in the past years azathioprine and methotrexate had gained popularity. - Based on analysis of a large pharmacy database, it is estimated that 2210 patients (0.6% of the AD patients) was treated with oral immunosuppressive drugs on December 31st 2016.

Long term safety of oral immunosuppressive drugs - Lymphopenia is occasionally seen in AD patients treated with oral immunosuppressive drugs. Larger, prospective observational studies are needed to find out what the exact incidence of lymphopenia is in these patients and what the clinical consequences are. Measuring lymphocyte counts in these patients can be considered; only measuring leukocyte count is insufficient. - An increased risk for the development of a cutaneous squamous cell carcinoma is observed in AD patients treated with oral immunosuppressive drugs. Whether this increased risk has a causal relationship with the drug remains unclear, but an extensive skin inspection and sun protection are advised, directly from the start of the treatment. - There is no increased risk for the development of cervical neoplasia in female AD patients treated with oral immunosuppressive drugs. No intensification of the national cervical screening programme is needed. In women younger than 30 years who start long-term treatment with oral immunosuppressive drugs, extra cervical screening may be considered. - Adverse pregnancy and fetal outcomes in men using oral immunosuppressive drugs during conception are rare, but this risk cannot be excluded completely, based on the current available literature. Risks and benefits should always be discussed with the patient and preference is given to cyclosporine A.

Optimizing the use of thiopurines - For individual AD patients treated with azathioprine, especially patients who do not respond to adequate azathioprine dose, measuring 6-TGN and 6-MMP levels may be helpful in optimizing azathioprine dose, improving clinical effectiveness and preventing side effects. - The combination of azathioprine and allopurinol should be considered in patients with a skewed metabolism or in patients with subjective side effects. Then azathioprine dose should be reduced by at least 50% before the start of allopurinol (100mg/day). - Thioguanine may be a safe and effective treatment alternative in difficult to treat AD patients, who failed on azathioprine monotherapy. GENERAL DISCUSSION 181

5. Future aspects

5.1 Dupilumab entering the market In 2017 several new drugs are under investigation for AD, including biologics and small molecules. Dupilumab is the first biologic that is approved for the treatment of moderate to severe AD, both in the United States and Europe. Dupilumab is a fully human monoclonal antibody, that binds specifically to the shared α chain subunit of the interleukin-4 and interleukin-13 receptors (interleukin 4R-α), inhibiting the signaling of interleukin 4 and interleukin 13, type 2/Th2 inflammatory cytokines implicated in numerous allergic diseases ranging from asthma to atopic dermatitis.39 In two early-phase, randomized placebo-controlled trials (16 weeks) and one larger long-term trial (52 weeks), treatment with dupilumab resulted in significant improvements in clinical signs and symptoms of AD, with an acceptable safety profile.39-41

With dupilumab entering the European market in 2018, guidelines and recommendations on the treatment of moderate to severe AD need to be updated. The position of dupilumab compared to the classic oral immunosuppressive drugs need to be carefully considered.

The following suggestion was made for the position of dupilumab (figure 1).

In the case of moderate to severe AD that is not sufficiently controlled by adequate topical treatment (if possible supplemented with structured education or during an inpatient clinical setting) and/or UV light treatment, oral immunosuppressive drugs need to be considered. The choice for cyclosporine A, methotrexate, azathioprine or mycophenolic acid will vary among different European countries and is dependent on several factors, including doctor’s experience, patient dependent factors, like age and comorbidities and license for AD. In case of treatment failure, for instance because of side effects or non-responsiveness, despite the use of an adequate dose during a period of at least 16 weeks, a second oral immunosuppressive drug can be considered or dupilumab can be started.

11 182 CHAPTER 11

Cyclosporine A 2-3mg/kg/day (preferable start dose 5 mg/kg/day during 3-6 weeks)

Consider second oral immunosuppressive Methotrexate 15-22.5mg/wk drug Treatment failure Indications to start despite adequate systemic maintenance dose for treatment at least 16 weeks Azathioprine 2-3mg/kg/day Start Dupilumab

- Mycophenolate mofetil 200mg/day - Enteric coated mycophenolate sodium 1440mg/day

Figure 1 Suggestion for the position of dupilumab

5.2 New developments in the treatment of atopic dermatitis Several other new drugs are currently investigated for the treatment of AD. Some of these drugs have extracellular targets while others target intracellular pathways. One of these drugs is nemolizumab, which is a humanized antibody against interleukin (IL)-31 receptor A. A recent phase 2, randomized, double-blind, placebo-controlled, multicenter, multidose study showed significant improvements compared to placebo, especially for pruritus.42 Tralokinumab targets IL-13. The press release of a phase 2b, randomized, double-blind, placebo controlled study showed significant reductions from baseline in EASI score and a greater number of patients achieving IGA (0 or 1) with 150mg and 300mg tralokinumab versus placebo.43 Lebrikizumab also targets IL-13. The data of this phase 2 randomized controlled trial showed a significant increase in the proportion of patients achieving EASI-50 for lebrikizumab compared to placebo.44

Recently, small molecules are investigated for the treatment of AD. Interesting data have been released on the Janus Kinase-Signal Transducer and Activator of Transcription (JAK-STAT) inhibition. Many inflammatory dermatoses are driven by soluble inflammatory mediations that rely on this intracellular signaling JAK-STAT pathway. Inhibition of this pathway using JAK inhibitors might be a useful therapeutic strategy for these diseases. A press release on upadacitinib showed that 69% of the patients treated with 30mg reached EASI-75 and 50% reached EASI-90. A 69% pruritus change in this dose group was seen on the pruritus numeric ratings scale.45 GENERAL DISCUSSION 183

5.3 The march of biologics: indicating the end of the oral immunosuppressive era? It is interesting to speculate whether dupilumab and other new agents will replace the use of oral immunosuppressive drugs. Pros and cons of the different oral immunosuppressive drugs and dupilumab should be carefully weighted to answer this question. Most important, Dupilumab is registered for the treatment of AD while most oral immunosuppressive drugs (except cyclosporine A) are not licensed for this indication. In addition, the study programs of dupilumab were all well designed, including a prospective analysis of long-term efficacy and safety, while studies on oral immunosuppressive drugs are less well designed, only evaluate small patient groups and lack long-term follow up data.

Nevertheless, there are several reasons why oral immunosuppressive treatment will be used in the next years for the treatment of moderate to severe AD. Not all patients will respond adequately to dupilumab and efficacy and safety data from daily practice are lacking at this moment. From earlier daily practice studies with oral immunosuppressive drugs we know that clinical results are less positive, compared to trial data, because patients in daily practice often have more comorbidities, are older and are sometimes less compliant. This may result in a higher number of non-responders and patients developing side effects when the effect of dupilumab is evaluated in a daily practice setting. In the absence of other treatment options, these patients will be eligible again for oral immunosuppressive drugs. Daily practice data in M. Crohn show that 27% of the patients discontinue treatment with biologics and are again treated with classical drugs.

In addition, in psoriasis, inflammatory bowel disease and rheumatic diseases it is sometimes seen that biologics are combined with methotrexate or azathioprine, in order to improve the performance. Whether this will also be necessary in future with the treatment of biologics in AD treatment, is yet unknown, but that possibility should be considered.

Last but not least: biologics will be far more expensive compared to oral immunosuppressive drugs (table 1).

Table 1 Costs of oral immunosuppressive drugs, based on www.medicijnkosten.nl Drug Dose Costs per 30 days Cyclosporine A Twice daily 100mg € 169,64 Azathioprine Twice daily 75 mg € 17,84 Methotrexate Once weekly 15mg € 4,28 11 Mycophenolic acid Twice daily 1000 mg € 30,86 Dupilumab Once every 2 weeks 300mg € 1250,10*

* Data from Sanofi 184 CHAPTER 11

5.4 Towards personalized medicine in the treatment of atopic dermatitis AD is a very heterogeneous disease, but treatment protocols are still universal (‘one size fits all’ approach). Preferable, treatment options should be adapted to the individual patient, to prevent discontinuation due to side effects or ineffectiveness. In order to improve the drug performance, the choice for a specific drug should be based on demographic factors such as age, co-morbidities and the metabolism and biological availability of the drug in each individual patient. One of the newer insights to individualize AD treatment, is looking at the biological diversity of patients. A better understanding of the different genetic and immunologic mechanisms underlying the wide spectrum of disease, will able the development of a roadmap for individualized treatments. The measurement of biomarker levels in AD patients may able the dermatologist to provide a more tailored management. Recently, Thijs et al found four potential clusters of AD patients, based on clinical AD severity and different biomarkers.46 They concluded that AD is a heterogeneous disease, both clinically and biologically. For future research, it is interesting to investigate whether clinical responders and non- responders to oral immunosuppressive drugs and biologics have different immunologic endotypes and whether these different endotypes can be targets for new treatments.

These findings will be especially interesting for the above mentioned new treatment agents: the biologic and small molecules. These new agents will allow a more personalized treatment of AD, based on their different modes of actions. It is supposed that the ‘one size fits all’ approach will become less popular.

In the next years, it will be an interesting challenge to find the optimal indication and positioning for all these new drugs, in order to improve the performance and safety of the treatment of patients with AD. GENERAL DISCUSSION 185

References

1 Taylor K, Swan DJ, Affleck A et al. Treatment 10 Porreco R, Penn I, Droegemueller W et al. of moderate-to-severe atopic eczema in Gynecologic malignancies in immuno- adults within the U.K.: results of a national suppressed organ homograft recipients. survey of dermatologists. Br J Dermatol Obstet Gynecol 1975; 45: 359-64. 2017; 176: 1617-23. 11 Mazanowska N, Pietrzak B, Kaminski P et 2 Hegazy S, Tauber M, Bulai-Livideanu C et al. al. Prevalence of cervical high-risk human Systemic treatment of severe adult Atopic papillomavirus infections in kidney graft dermatitis in clinical practice: analysis of recipients. Ann Transplant 2013; 18: prescribing pattern in a cohort of 241 656-60. patients. J Eur Acad Dermatol Venereol 2017; 31: e423-e4. 12 Pietrzak B, Mazanowska N, Ekiel AM et al. Prevalence of high-risk human papilloma- 3 Vedie AL, Ezzedine K, Amazan E et al. virus cervical infection in female kidney Long-term Use of Systemic Treatments for graft recipients: an observational study. Moderate-to-Severe Atopic Dermatitis in Virol J 2012; 9: 117. Adults: A Monocentric Retrospective Study. Acta Derm Venereol 2016; 96: 802-6. 13 Morrison EA, Dole P, Sun XW et al. Low prevalence of human papillomavirus 4 Schmitt J, Abraham S, Trautmann F et al. infection of the cervix in renal transplant Usage and effectiveness of systemic treat- recipients. Nephrol Dial Transplant 1996; ments in adults with severe atopic eczema: 11: 1603-6. First results of the German Atopic Eczema Registry TREATgermany. J Dtsch Dermatol 14 Madeleine MM, Finch JL, Lynch CF et al. Ges 2017; 15: 49-59. HPV-related cancers after solid organ transplantation in the United States. Am J 5 Schmitt J, Schmitt NM, Kirch W et al. Transplant 2013; 13: 3202-9. Outpatient care and medical treatment of children and adults with atopic eczema. J 15 Clowry J, Callanan I, Eustace K et al. Dtsch Dermatol Ges 2009; 7: 345-51. Cervical screening uptake in a dermatology cohort on immunosuppressive agents. Br J 6 Tennis P, Gelfand JM, Rothman KJ. Dermatol 2015; 172: 1682-3. Evaluation of cancer risk related to atopic dermatitis and use of topical calcineurin 16 Norgard B, Pedersen L, Jacobsen J et al. inhibitors. Br J Dermatol 2011; 165: 465-73. The risk of congenital abnormalities in children fathered by men treated with 7 Jensen AO, Svaerke C, Kormendine Farkas azathioprine or mercaptopurine before D et al. Atopic dermatitis and risk of skin conception. Aliment Pharmacol Ther 2004; cancer: a Danish nationwide cohort study 19: 679-85. (1977-2006). Am J Clin Dermatol 2012; 13: 29-36. 17 Rajapakse RO, Korelitz BI, Zlatanic J et al. Outcome of pregnancies when fathers are 8 Ruff S, Egeberg A, Andersen YMF et al. treated with 6-mercaptopurine for inflam- Prevalence of Cancer in Adult Patients with matory bowel disease. Am J Gastroenterol Atopic Dermatitis: A Nationwide Study. 2000; 95: 684-8. 11 Acta Derm Venereol 2017. 18 Asadov C, Aliyeva G, Mustafayeva K. 9 Vajdic CM, McDonald SP, McCredie MR Thiopurine S-Methyltransferase as a et al. Cancer incidence before and after Pharmacogenetic Biomarker: Significance kidney transplantation. JAMA 2006; 296: of Testing and Review of Major Methods. 2823-31. Cardiovasc Hematol Agents Med Chem 2017. 186 CHAPTER 11

19 Citterio-Quentin A, Moulsma M, Gustin 28 Ansari A, Elliott T, Baburajan B et al. MP et al. ITPA Activity in Adults and Long-term outcome of using allopurinol Children Treated With or Without co-therapy as a strategy for overcoming Azathioprine: Relationship Between TPMT thiopurine hepatotoxicity in treating inflam- Activity, Thiopurine Metabolites, and matory bowel disease. Aliment Pharmacol Co-medications. Ther Drug Monit 2017; 39: Ther 2008; 28: 734-41. 483-91. 29 Ansari A, Patel N, Sanderson J et al. 20 Atanasova S, Shipkova M, Svinarov D et Low-dose azathioprine or mercaptopurine al. Analysis of ITPA phenotype-genotype in combination with allopurinol can bypass correlation in the Bulgarian population many adverse drug reactions in patients revealed a novel gene variant in exon 6. with inflammatory bowel disease. Aliment Ther Drug Monit 2007; 29: 6-10. Pharmacol Ther 2010; 31: 640-7.

21 Meijer B, Mulder CJ, de Boer NK. NUDT15: 30 Pavlidis P, Stamoulos P, Abdulrehman A a novel player in thiopurine metabolism. J et al. Long-term Safety and Efficacy of Gastrointestin Liver Dis 2016; 25: 261-2. Low-dose Azathioprine and Allopurinol Cotherapy in Inflammatory Bowel Disease: 22 Al-Judaibi B, Schwarz UI, Huda N et al. A Large Observational Study. Inflamm Genetic Predictors of Azathioprine Toxicity Bowel Dis 2016; 22: 1639-46. and Clinical Response in Patients with Inflammatory Bowel Disease. J Popul Ther 31 Amin J, Huang B, Yoon J et al. Update Clin Pharmacol 2016; 23: e26-36. 2014: advances to optimize 6-mercaptopurine and azathioprine to reduce toxicity and 23 Goel RM, Blaker P, Mentzer A et al. improve efficacy in the management of IBD. Optimizing the use of thiopurines in inflam- Inflamm Bowel Dis 2015; 21: 445-52. matory bowel disease. Ther Adv Chronic Dis 2015; 6: 138-46. 32 Curkovic I, Rentsch KM, Frei P et al. Low allopurinol doses are sufficient to optimize 24 Smith MA, Blaker P, Marinaki AM et al. azathioprine therapy in inflammatory bowel Optimising outcome on thiopurines in disease patients with inadequate thiopu- inflammatory bowel disease by co-prescrip- rine metabolite concentrations. Eur J Clin tion of allopurinol. J Crohns Colitis 2012; 6: Pharmacol 2013; 69: 1521-31. 905-12. 33 Meijer B, Mulder CJ, Peters GJ et al. Efficacy 25 el-Azhary RA, Farmer SA, Drage LA et al. of thioguanine treatment in inflammatory Thioguanine nucleotides and thiopurine bowel disease: A systematic review. World J methyltransferase in immunobullous Gastroenterol 2016; 22: 9012-21. diseases: optimal levels as adjunctive tools for azathioprine monitoring. Arch Dermatol 34 Dubinsky MC, Vasiliauskas EA, Singh H et al. 2009; 145: 644-52. 6-thioguanine can cause serious liver injury in inflammatory bowel disease patients. 26 Caufield M, Tom WL. Oral azathioprine for Gastroenterology 2003; 125: 298-303. recalcitrant pediatric atopic dermatitis: clinical response and thiopurine monitoring. J 35 Zackheim HS, Glogau RG, Fisher DA et al. Am Acad Dermatol 2013; 68: 29-35. 6-Thioguanine treatment of psoriasis: experience in 81 patients. J Am Acad Dermatol 27 Sparrow MP, Hande SA, Friedman S et al. 1994; 30: 452-8. Effect of allopurinol on clinical outcomes in inflammatory bowel disease nonre- 36 Molin L, Thomsen K. Thioguanine treatment sponders to azathioprine or 6-mercapto- in psoriasis. Acta Derm Venereol 1987; 67: purine. Clin Gastroenterol Hepatol 2007; 5: 85-8. 209-14. GENERAL DISCUSSION 187

37 Mason C, Krueger GG. Thioguanine for 42 Ruzicka T, Hanifin JM, Furue M et al. Anti- refractory psoriasis: a 4-year experience. J Interleukin-31 Receptor A Antibody for Am Acad Dermatol 2001; 44: 67-72. Atopic Dermatitis. N Engl J Med 2017; 376: 826-35. 38 Sherer DW, Lebwohl MG. 6-thioguanine in the treatment of psoriasis: a case report 43 Wollenberg A et al. Presented at: American and literature review. J Cutan Med Surg Academy of Dermatology Meeting; March 2002; 6: 546-50. 3–7, 2017; Orlando, FL, USA. Poster 4496.

39 Blauvelt A, de Bruin-Weller M, Gooderham 44 Simpson EL. Presented at: European M et al. Long-term management of Association of Dermatology and moderate-to-severe atopic dermatitis Venereology Congress; September with dupilumab and concomitant topical 28–October 2, 2016; Vienna, Austria. Late- corticosteroids (LIBERTY AD CHRONOS): a breaking oral presentation D3T01.1F. 1-year, randomised, double-blinded, placebo-controlled, phase 3 trial. Lancet 2017; 45 AbbVie. 2017. [Press Release]. Available at: 389: 2287-303. https://news.abbvie.com/news/abbvies- upadacitinib-abt-494-meets-primary-end- 40 Simpson EL, Bieber T, Guttman-Yassky E et point-in-phase-2b-study-in-atopic-derma- al. Two Phase 3 Trials of Dupilumab versus titis.htm. Accessed September 2017. Placebo in Atopic Dermatitis. N Engl J Med 2016; 375: 2335-48. 46 Thijs JL, Strickland I, Bruijnzeel-Koomen C et al. Moving toward endotypes in atopic 41 Thaci D, Simpson EL, Beck LA et al. Efficacy dermatitis: Identification of patient clusters and safety of dupilumab in adults with based on serum biomarker analysis. J moderate-to-severe atopic dermatitis inad- Allergy Clin Immunol 2017; 140: 730-7. equately controlled by topical treatments: a randomised, placebo-controlled, dose- ranging phase 2b trial. Lancet 2016; 387: 40-52.

Summary Samenvatting 190 CHAPTER 12

Summary In this thesis the treatment of atopic dermatitis (AD) by the use of oral immunosuppressive drugs is discussed. In the first part we evaluated the qualitative and quantitative prescribing of oral immunosuppressive drugs in the Netherlands. In part two, open questions concerning safety of long-term treatment with oral immunosuppressive drugs in AD were answered. In part three, we described how treatment of AD patients with thiopurines could be improved, in order to explore new treatment options for difficult to treat AD patients.

1. Qualitative and quantitative exploration of oral immunosuppressive drug use for atopic dermatitis in daily practice in the Netherlands Since oral immunosuppressive drugs are only used in more severe and difficult to treat AD patients, evaluation of the prescribing of these drugs may give an impression of the number of patients with difficult to treat AD. Studies on this topic are scarce. Therefore, in the studies described in chapter 2 and chapter 3 of this thesis, we evaluated the treatment of AD with oral immunosuppressive drugs in the Netherlands.

In chapter 2 we described our ten years experience with oral immunosuppressive drugs in the treatment of AD in the University Medical Center Utrecht and the Academic Medical Center Amsterdam. We evaluated 334 patients. Cyclosporine A was given in 80% of the patients, mycophenolate mofetil or enteric-coated mycophenolate sodium in 31%, azathioprine in 14%, methotrexate in 11%, systemic glucocorticosteroids in 7% and systemic tacrolimus in 5%. In most cases, cyclosporine A was the first choice treatment in these two academic centers. Unfortunately, in contrast with previously published prospective controlled trials, we found that many patients had to discontinue treatment due to ineffectiveness and adverse events.

The data from chapter 2 were only derived in two academic hospitals with a focus on AD, and therefore may not be applicable for the general Dutch population. Therefore, in chapter 3, we performed a study in which we investigated the prescription of oral immunosuppressive drugs in the Netherlands by using information from a pharmaceutical database (NControl) containing data of 557 million prescriptions and 7.2 million patients. An algorithm, based on the WHO Anatomical Therapeutic Chemical (ATC) codes, was used to identify patients with AD. Based on the algorithm, 65 943 patients with AD were selected. 943 AD patients (1.4%) used cyclosporine A, methotrexate, azathioprine or mycophenolic acid between January 1st 2012 and January 1st 2017. In the used algorithm, only atopic patients were included. Besides, the NControl database did not include all Dutch pharmacies. We corrected for this and found a total of 5070 patients using oral immunosuppressive drugs during this period. On December 31st 2016, 411 patients (0.6%) were defined as ‘active’ patients, using oral immunosuppressive drugs. When we correct this number for the atopic factor and for the total number of pharmacies, a total of 2210 patients using oral immunosuppressive drugs on December 31st 2016 was found. SUMMARY 191

In this study, methotrexate was most commonly used. A switch in medication was rarely seen. In the evaluation period a decrease in the prescription of cyclosporine A was seen, together with an increase of the prescription of methotrexate. In 31% of the patients who stopped treatment, the discontinuation took place within the first months of treatment.

2. Open questions concerning safety of long-term oral immunosuppressive drug use in atopic dermatitis Long-term treatment with oral immunosuppressive drugs is often necessary to control moderate to severe AD. However, in daily practice, dermatologists fear long-term treatment because of the possible risks of (irreversible) side effects and the development of malignancies. Existing guidelines do not always answer all questions concerning safety of long-term treatment. Therefore, in chapter 4, 5, 6 and 7, we tried to find answers to guideline gaps in the field of safety when treating AD patients with oral immunosuppressive drugs.

Risk of lymphopenia and serious infections Oral immunosuppressive drugs have been associated with lymphopenia and the presence of lymphopenia is related to an increased risk of opportunistic infections. The incidence of lymphopenia in patients with AD treated with oral immunosuppressive drugs is yet unknown. In the study in chapter 4, we evaluated the occurrence of persistent lymphopenia in patients with AD treated with oral immunosuppressive drugs. In this study we systematically screened medical records of all adult AD patients treated with oral immunosuppressive drugs and found 11 patients with persistent lymphopenia (i.e. > 5 times lymphocyte count below 0.8x109/L). Co-prescription of prednisone seemed to be a risk factor. No serious infections were observed. Interestingly, total white blood cell counts were normal in all patients at the time of decreased lymphocyte counts. These patients would have been missed when only a white blood cell count was measured. Based on the results of this study, we suggest to monitor a leukocyte differenti- ation including the lymphocyte count, instead of only a white blood cell count, in AD patients using oral immunosuppressive drugs. In patients with a prolonged lymphopenia a referral to a clinical immunologist can be considered, to discuss what the risk of serious infections is and to discuss whether starting antibiotic prophylaxes is needed.

12 192 CHAPTER 12

Risk of HPV related tumors In chapter 5, we investigated the risk of developing NMSC (including basal cell carcinoma and squamous cell carcinoma [SCC]) in 557 AD patients using oral immunosuppressive drugs in the Netherlands from January 1st 1989 until January 1st 2014. NMSC during or after oral immunosuppressive treatments was reported in 18 patients (3.2%). Patients who developed NMSC were significantly older compared to patients without a malignancy at start of therapy (p<0.001) and at data lock ( p<0.001). We compared the incidence of SCC in this study to the general Dutch population. The standardized incidence ratio (SIR) for the risk of development of an SCC in this study population was 13.1 [95% CI 6.5-19.7]. The SIR for the risk of development of an SCC in patients with cyclosporine A monotherapy was 25.3 [95% CI 10.3-40.2]. We concluded that an increased risk of SCC was found in this study, but that it is not clear whether this increased risk can be completely explained by the use of oral immunosuppressive drugs, because we found a lack of association between treatment duration and the risk of developing a tumor and a long interval between treatment discontinuation and the development of the tumor in some patients. Nevertheless, we suggest that dermatologists should always balance the benefit of treatment against the potential risks in each individual patient. A thorough inspection of the skin and the use of sunscreens is recommended, directly from the start of the treatment.

In chapter 6 we investigated the occurrence of cervical carcinoma in women with AD treated with oral immunosuppressive drugs in the University Medical Center Utrecht and Groningen between January 1st 1989 and January 1st 2014. A total of 257 female AD patients was identified. In 189 patients (73.5%) results of cervical examination were reported in the national histopathology database. No cervical carcinoma during or following immunosuppressive therapy was found in our patient group. This number is in line with the general Dutch population in the same time period. CIN II was found in one patient, CIN III was found in three patients. We conclude that, as confirmed by the Dutch Guideline for cervix carcinoma, no intensified screenings program seems necessary for women with AD using oral immunosuppressive drugs.

Oral immunosuppressive drugs at the moment of conception In chapter 7, we performed a review of the existing literature in order to answer the question whether it is safe or not for men who wish to conceive, to continue their immunosuppressive treatment. After a systematic search, we selected 26 studies for data analysis. Most of the included studies on azathioprine, methotrexate and mycophenolic acid that compared their outcomes to a healthy population (9 out of 26) conclude that there was no increase in adverse pregnancy or fetal outcomes. Nevertheless, preterm birth, spontaneous abortion, low birth weight and congenital malformations have been described with a varying incidence. Therefore, a possible risk of adverse pregnancy of fetal outcomes when using these drugs cannot be excluded. Therefore we emphasize physicians to ask their male patients starting azathioprine, SUMMARY 193

methotrexate or mycophenolic acid if they wish to conceive. Dermatologists should discuss with the patient the results of this CAT, highlight the discrepancies between studies in reviews and guidelines and discuss the risk/benefit of oral immunosuppressive treatment. If possible, preference should be given to the use of cyclosporine A.

3. Improving performance of thiopurines in atopic dermatitis in order to explore new treatment options for difficult to treat atopic dermatitis patients Azathioprine is frequently used off-label in the management of AD but its therapeutic potential may have not been fully utilized. The azathioprine metabolism is very complex due to involvement of various enzymes. Azathioprine is a thiopurine pro-drug and has no immunosuppressive activity itself. In the liver, it is converted into 6-mercaptopurin (6-MP). Subsequently, 6-MP undergoes metabolic transformations, via a complex enzymatic pathway, resulting in a variety of pharmacological active metabolites. The most important metabolites are 6-thioguanine nucleotide (6-TGN) and methylated 6-methylmercaptopurine (6-MMP). The immunosuppressive effect of azathioprine is caused by the 6-TGN metabolites, while high levels of 6-MMP may cause hepatotoxicity.

Azathioprine metabolite levels In an approach to individualize treatment, the azathioprine dose can be based on the levels of 6-TGN and 6-MMP (measured in red blood cells). This method, that is called therapeutic drug monitoring, is already commonly used in patients with inflammatory bowel disease and organ transplant patients, but not in AD. In chapter 8 we investigated whether measuring thiopurine metabolites in red blood cells can improve the effectiveness and safety of azathioprine treatment in patients with AD and/or chronic hand/foot eczema. Azathioprine metabolite levels were measured in eczema patients during maintenance treatment (Part I) and dose escalation (Part II). Clinical effectiveness, hepatotoxicity and bone marrow suppression were analyzed. A wide variation in metabolite levels in all dose groups was observed. In part I (32 patients), there were no significant differences in 6-TGN levels between clinical responders and non-responders (p=0.806). No hepatoxicity or myelotoxicity was observed. In Part II, all 6-TGN and 6-MMP levels increased during dose escalation. Hypermethylation was observed in 2/8 patients.

We concluded that for individual patients, routinely measuring 6-TGN and 6-MMP levels can be helpful in optimizing azathioprine dose, improving clinical effectiveness and preventing side effects.

12 194 CHAPTER 12

Allopurinol co-prescription In studies in patients with inflammatory bowel disease, attempts have been made to influence the metabolism of azathioprine by the co-prescription of allopurinol. This co-prescription shifts the azathioprine metabolism towards 6-TGN production. In several studies, this resulted in a clinical improvement and a decrease of (subjective) side effects and hepatotoxicity, but experience with allopurinol co-prescription in eczema patients was lacking.

In chapter 9, de effect of allopurinol co-prescription in eczema patients on metabolite levels, side effects and clinical efficacy was studied. Fifteen patients were enrolled. Reasons for allopurinol co-prescription were inefficacy during azathioprine monotherapy, side effects or skewed metabolism. After allopurinol addition, 6-MMP levels decreased and 6-TGN levels increased in all patients. Prior to allopurinol addition, four patients (26.7%) were classified as responder, compared to seven patients (56.7%) after allopurinol co-prescription (p=0.013). No side effects of allopurinol were seen. We concluded that co-prescription of allopurinol may optimize azathioprine treatment outcome in chronic eczema by increasing 6-TGN and decreasing 6-MMP levels.

Thioguanine It has been hypothesized that thioguanine therapy instead of azathioprine reduces generation of potentially toxic metabolites, because it is primarily converted into the therapeutically aimed metabolite 6-TGN by bypassing several rate-limiting metabolic steps. In chapter 10, we investigated whether treatment with thioguanine is effective and safe in patients with difficult to treat AD. Patients with severe AD who failed on azathioprine treatment were treated with thioguanine. The start dose was often 10mg/day and doses were escalated within the first month. Clinical efficacy, side effects and laboratory safety values were regularly measured during the first three months of treatment. Five adult patients were enrolled. After three months, disease activity decreased in most patients and one patient could stop oral steroids. No myelotoxicity or other clinically relevant changes in laboratory results were seen. Some mild subjective side effects were reported and one patient discontinued treatment due to headache.

We concluded that thioguanine may be an alternative treatment option in patients with difficult to treat, severe AD who failed on azathioprine treatment. SUMMARY 195

Samenvatting In dit proefschrift wordt de behandeling van patiënten met constitutioneel eczeem (CE) door middel van orale immunosuppressiva besproken. In het eerste deel evalueren we het kwali- tatieve en kwantitatieve voorschrijfgedrag van orale immunosuppressiva bij CE in Nederland. In deel twee worden openstaande vragen ten aanzien van veiligheid van langdurige behandeling met orale immunosuppressiva bij CE patiënten beantwoord. In deel drie beschrijven we hoe de behandeling van CE patiënten met thiopurines geoptimaliseerd kan worden, in een zoektocht naar nieuwe therapeutische opties voor moeilijk te behandelen CE.

1. Kwalitatief en kwantitatief onderzoek naar het gebruik van orale immunosuppressiva voor de behandeling van constitutioneel eczeem in de dagelijkse praktijk in Nederland

Orale immunosuppressiva zijn gereserveerd voor de behandeling van ernstig en moeilijk behandelbaar CE. Het bestuderen van het voorschrijfgedrag van deze medicijnen kan daarom een indruk geven van de omvang van de totale populatie met moeilijk te behandelen CE. Er zijn echter weinig studies gepubliceerd met betrekking tot dit onderwerp. In hoofdstuk 2 en hoofdstuk 3 van dit proefschrift is daarom het voorschrijfgedrag van orale immunosuppressiva in Nederland geëvalueerd.

In hoofdstuk 2 beschrijven we onze tien jaar ervaring met het behandelen van CE patiënten met orale immunosuppressiva in het Universitair Medisch Centrum Utrecht en het Academisch Medisch Centrum Amsterdam. We evalueerden 334 patiënten. Ciclosporine A werd gebruikt door 80% van de patiënten, mycofenolaat mofetil of mycofenolzuur door 31%, azathioprine door 14%, methotrexaat door 11%, systemische glucocorticosteroiden door 7% en orale tacrolimus door 5%. Bij de meeste patiënten was ciclosporine A de behandeling van eerste keuze in deze twee academische ziekenhuizen. Helaas werd gezien dat, in tegenstelling tot eerder gepubliceerde prospectieve gecontroleerde trials, veel patiënten hun behandeling moesten staken door bijwerkingen of door ineffectiviteit.

De informatie uit hoofdstuk 2 werd verzameld in twee academische ziekenhuizen met veel ervaring in de behandeling van eczeem. Deze data zijn daarom waarschijnlijk niet repre- sentatief voor de algehele Nederlandse populatie. Hoofdstuk 3 beschrijft een studie waarin we het voorschrijfgedrag van orale immunosuppressiva in Nederland hebben bestudeerd. We maakten hiervoor gebruik van een farmaceutische database (NControl) met daarin 557 miljoen recepten en 7.2 miljoen patiënten. Een algoritme, gebaseerd op de WHO Anatomical Therapeutic Chemical (ATC) codes, werd gebruikt om patiënten met CE te identificeren binnen deze database. Met dit algoritme werden 65.943 CE patiënten geselecteerd. 943 van hen (1.4%) gebruikten ciclosporine A, methotrexaat, azathioprine of mycofenolzuur tussen 1 januari 12 196 CHAPTER 12

2012 en 1 januari 2017. Omdat dit algoritme alleen atopische patiënten heeft meegenomen en omdat de NControl database slechts een deel van alle apotheken in Nederland bevat, werd hiervoor gecorrigeerd. Na correctie werd een totaal aantal van 5070 CE patiënten gevonden die in deze vijf jaar periode tussen 2012 en 2017 orale immunosuppressiva gebruikten. Op 31 december 2016 werden 411 patiënten (0.6%) als ‘actieve’ gebruikers van orale immunosuppressiva gedefinieerd. Wanneer dit aantal patiënten gecorrigeerd wordt voor de atopische factor en voor het aantal apotheken in Nederland, wordt een totaal van 2210 actieve patiënten gevonden op 31 december 2016.

De studie in hoofdstuk 3 laat zien dat methotrexaat het meest wordt gebruik. Een wisseling van medicatiesoort werd nauwelijks gezien. In de vijfjarige evaluatieperiode werd een afname van het aantal voorschrijvingen van ciclosporine A beschreven, vergezeld door een toename van het aantal methotrexaat voorschrijvingen. Bij 31% van alle patiënten die hun behandeling binnen de evaluatie periode stopten, vond deze beëindiging plaats binnen de eerste maanden.

2. Open vragen ten aanzien van veiligheid van langdurige behandeling van patiënten met constitutioneel eczeem met orale immunosuppressiva

Langdurige behandeling met orale immunosuppressiva is vaak nodig om matig tot ernstig CE goed onder controle te kunnen houden. In de dagelijkse praktijk wordt echter regelmatig angst bemerkt voor langdurige behandeling vanwege het risico op (irreversibele) bijwerkingen en de ontwikkeling van maligniteiten. Richtlijnen over het gebruik van orale immunosuppressiva bij CE geven nog niet op alle vragen omtrent het langdurig gebruik van orale immunosuppressiva antwoord. In hoofdstuk 4, 5, 6 en 7 hebben we getracht antwoorden te vinden op deze openstaande vragen met betrekking tot veiligheid van langdurige behandeling van CE patiënten met orale immunosuppressiva.

Risico op lymfopenie en ernstige infecties Orale immunosuppressiva worden geassocieerd met lymfopenie. De aanwezigheid van lymfopenie is gerelateerd aan een verhoogd risico op opportunistische infecties. De incidentie van lymfopenie in CE patiënten die worden behandeld met orale immunosuppressiva is echter vooralsnog onbekend.

In de studie in hoofdstuk 4 evalueerden we het vóórkomen van persisterende lymfopenie in patiënten met CE die behandeld worden met orale immunosuppressiva in het Universitair Medisch Centrum Utrecht. Alle CE patiënten die behandeld werden met orale immunosuppressiva werden systematisch bestudeerd. In totaal werden 11 patiënten met een persisterende lymfopenie (gedefinieerd als > 5 maal een lymfocyten aantal onder 0.8x109/L) gevonden. Het SUMMARY 197

gebruik van prednison leek een risicofactor te zijn voor het ontwikkelen van een lymfopenie. Er werden bij de patiënten met persisterende lymfopenie geen ernstige infecties gezien. Opvallend was dat het totaal aantal leukocyten in deze patiënten vaak normaal was op het moment dat het lymfocyten aantal verlaagd was. De lymfopenie in deze patiënten zou gemist zijn wanneer alleen een leukocyten aantal gemeten was en geen differentiatie. Gebaseerd op de resultaten van deze studie, suggereren we daarom om een leukocyten differentiatie (inclusief het lymfocyten aantal) te monitoren, in plaats van alleen het leukocyten aantal, in CE patiënten die behandeld worden met orale immunosuppressiva. In het geval van een aanhou- dende lymfopenie kan een verwijzing naar een klinisch immunoloog of internist overwogen worden, om de risico’s voor een ernstige infectie en het nut van profylactische antibiotische therapie te bespreken.

Risico op HPV gerelateerde tumoren In hoofdstuk 5 onderzochten we het risico op het ontwikkelen van non-melanoma skin cancer (NMSC, gedefinieerd als het basaalcelcarcinoom en plaveiselcelcarcinoom [PCC]) in 557 CE patiënten die tussen 1 januari 1989 en 1 januari 2014 werden behandeld met orale immunosuppressiva in het Universitair Medisch Centrum Utrecht en Groningen. NMSC tijdens of na behandeling met orale immunosuppressiva werd gerapporteerd bij 18 patiënten (3.2%). Patiënten met NMSC waren significant ouder vergeleken met patiënten zonder NMSC bij aanvang van de therapie en datalock. We vergeleken de incidentie van het PCC in deze studie met de algehele Nederlandse populatie en vonden een standardized incidence ratio (SIR) voor het krijgen van een PCC in deze studie populatie van 13.1 [95% CI 6.5-19.7]). De SIR voor het krijgen van een PCC bij patiënten die alleen behandeld waren met cyclosporine A was 25.3 [95% CI 10.3-40.2]. We concludeerden in deze studie dat er een verhoogd risico werd gevonden voor het krijgen van een PCC, maar dat het niet duidelijk is of dit verhoogde risico in zijn totaliteit kan worden verklaard door het gebruik van orale immunosuppressiva. Zo vonden we geen associatie tussen behandelduur en het risico op NMSC en vonden we bij sommige patiënten een erg lang interval tussen het staken van de therapie en het ontwikkelen van de maligniteit. Desalniettemin adviseren we dat dermatologen altijd de voordelen en risico’s van langdurige behandeling met immunosuppressiva afwegen en met hun patiënten bespreken. Tevens wordt regelmatige inspectie van de huid en het gebruik van antizonnebrandmiddelen geadviseerd, direct vanaf de start van de immunosuppressieve behandeling.

In hoofdstuk 6 onderzochten we het vóórkomen van een cervix carcinoom bij vrouwen met CE die behandeld werden met orale immunosuppressiva in het Universitair Medisch Centrum Utrecht en Groningen tussen 1 januari 1989 en 1 januari 2014. Een totaal van 257 vrouwen werd geëvalueerd. Bij 189 patiënten (73.5%) waren resultaten van cervix onderzoek gerapporteerd in het nationale histopathologie register. In deze patiëntengroep werd geen cervixcarcinoom tijdens of na behandeling met orale immunouppressiva gevonden. Dit komt overeen met 12 198 CHAPTER 12

de cijfers in de algehele Nederlandse populatie in dezelfde tijdsperiode. CIN II werd bij één patiënt gezien; CIN III werd gevonden bij drie patiënten. We concluderen op basis van deze gegevens dat er geen geïntensiveerde screening voor cervixcarcinoom noodzakelijk is voor CE patiënten die behandeld worden met orale immunosuppressiva. Dit sluit aan bij de huidige aanbevelingen in de Nederlandse richtlijn voor cervixcarcinoom.

Gebruik van orale immunosuppressiva ten tijden van conceptie In hoofdstuk 7 beschrijven we een review van de bestaande literatuur over de veiligheid van het gebruik van orale immunosuppressiva door mannen ten tijden van conceptie. Na een systematische zoekstrategie werden 26 artikelen geïncludeerd voor data analyse. Het meren- deel van de studies over azathioprine, methotrexaat en mycofenolzuur die hun data vergeleken met een gezonde controle populatie (dat gebeurde bij 9 van de 26 studies) concludeerde dat er geen verhoogd risico werd gezien op negatieve zwangerschapsuitkomsten of afwijkingen bij het kind. Desalniettemin werden vroeggeboorte, spontane abortus, laag geboortegewicht en congenitale malformaties gerapporteerd met wisselende incidenties. Een mogelijk risico op negatieve zwangerschapsuitkomsten of afwijkingen bij het kind bij paternaal gebruik van deze medicatie ten tijde van de conceptie kan daarom niet geheel worden uitgesloten. We adviseren artsen daarom om voorafgaand aan de behandeling met hun mannelijke patiënten te bespreken of er een actuele kinderwens is. De resultaten van deze review, de discrepanties in studies in de literatuur en de voor- en nadelen van immunosuppressieve therapie moeten besproken worden met de patiënt. Indien mogelijk, dan gaat de voorkeur uit naar behandeling met ciclosporine A, dat veilig gebruikt kan worden.

3. Het verbeteren van de effectiviteit van thiopurines; de zoektocht naar nieuwe behandelopties voor moeilijk te behandelen constitutioneel eczeem

Azathioprine wordt regelmatig ‘off-label’ gebruikt voor de behandeling van CE, maar de therapeutische mogelijkheden van azathioprine worden waarschijnlijk nog niet optimaal benut. Het azathioprine metabolisme is complex door de betrokkenheid van verschillende enzymen. Azathioprine is een thiopurine ‘pro-drug’ en heeft zelf geen immunosuppressieve activiteit. In de lever wordt azathioprine omgezet in 6-mercaptopurine (6-MP). Vervolgens ondergaat 6-MP verschillende metabole transformaties, via een complex enzymatische cascade, hetgeen resul- teert in verschillende farmacologisch actieve metabolieten. De meest belangrijke metabolieten zijn 6-thioguanine nucleotide (6-TGN) en methylated 6-methylmercaptopurine (6-MMP). Het immunosuppressieve effect van azathioprine wordt bewerkstelligd door de 6-TGN metabolieten, terwijl hoge spiegels van 6-MMP hepatotoxiciteit kunnen veroorzaken. SUMMARY 199

Azathioprine metaboliet spiegels In een poging om behandeling van patiënten te individualiseren, kan de azathioprine dosering worden afgestemd op de spiegels van 6-TGN en 6-MMP (gemeten in rode bloedcellen). Deze methode, die ‘therapeutische drug monitoring’ wordt genoemd, wordt al regelmatig gebruikt bij patiënten met inflammatoire darmziekten en bij patiënten na orgaantransplantatie. Bij CE is hiermee echter nauwelijks ervaring.

In hoofdstuk 8 onderzochten we of het meten van thiopurine metaboliet spiegels in de circu- latie de effectiviteit en veiligheid van het gebruik van azathioprine behandeling bij patiënten met CE en/of chronisch hand/voet eczeem kan verbeteren. Azathioprine metaboliet spiegels werden gemeten bij eczeem patiënten gedurende onderhoudsbehandeling (deel I van de studie) en gedurende ophoging van de dosis (deel II). Klinische effectiviteit, hepatotoxiciteit en beenmerg suppressie werden geanalyseerd. Een brede spreiding in metaboliet spiegels werd gezien in alle doseringsgroepen. In deel I van de studie (32 patiënten) werd geen significant verschil gezien in 6-TGN spiegels tussen klinische responders en non-responders. Er werd geen hepatotoxiciteit of myelotoxiciteit geobserveerd. In deel II werd aangetoond dat bij alle patiënten de spiegels van 6-TGN en 6-MMP toenamen wanneer de dosering azathioprine werd verhoogd.

We concluderen dat voor individuele patiënten routinematige metingen van 6-TGN en 6-MMP spiegels behulpzaam kunnen zijn bij het optimaliseren van de azathioprine dosis en bij het verbeteren van het klinisch effect en het verminderen van bijwerkingen.

De toevoeging van allopurinol In studies in patiënten met inflammatoire darmziekten zijn al eerder pogingen gedaan om het metabolisme van azathioprine te beïnvloeden door het toevoegen van allopurinol. Het idee achter deze redenering is dat de toevoeging van allopurinol zorgt voor een verschuiving van het metabolisme in de richting van de productie van 6-TGN, ten koste van 6-MMP. Verschillende studies over dit onderwerp toonden een verbetering van het klinisch effect en een afname van (subjectieve) bijwerkingen en hepatotoxiciteit. Ervaring met de toevoeging van allopurinol aan de behandeling van CE patiënten met azathioprine ontbreekt vooralsnog. In hoofdstuk 9 onderzochten we het effect van het toevoegen van allopurinol aan de behandeling met azathioprine op de metaboliet spiegels, bijwerkingen en het klinisch effect. Vijftien patiënten werden in deze studie geëvalueerd. De reden voor het starten van allopurinol in deze patiënten was: ineffectiviteit van azathioprine monotherapie behandeling, bijwerkingen of een ‘skewed’ metabolisme (waarbij er vooral 6-MMP wordt gevormd en minder 6-TGN). Na de toevoeging van allopurinol stegen de 6-TGN spiegels en daalden de 6-MMP spiegels in alle patiënten. Voorafgaand aan de toevoeging van allopurinol waren vier patiënten (26.7%) gedefinieerd als een klinische responder; na de toevoeging van allopurinol steeg dit aantal 12 naar zeven patiënten (56.7%, p=0.013). Er werden geen bijwerkingen gezien. 200 CHAPTER 12

We concluderen dat de toevoeging van allopurinol de behandeling met azathioprine kan optimaliseren in patiënten met chronisch eczeem door het verhogen van de 6-TGN spiegels en het verlagen van de 6-MMP spiegels.

Thioguanine De hypothese bestaat dat behandeling met thioguanine in plaats van azathioprine de vorming van potentieel toxische metabolieten reduceert, doordat thioguanine primair wordt omgezet in het therapeutisch actieve metaboliet 6-TGN. In hoofdstuk 10 onderzochten we of behandeling met thioguanine in patiënten met moeilijk behandelbaar CE effectief en veilig is. Vijf patiënten met ernstig CE, die eerder niet succesvol werden behandeld met azathioprine, werden behandeld met thioguanine. Het klinisch effect, bijwerkingen en laboratorium bloedcontroles werden regelmatig gecontroleerd gedurende de eerste drie maanden. Na drie maanden was bij de meeste patiënten een klinische verbetering te zien. Een van de patiënten kon stoppen met gelijk- tijdige behandeling met prednison. Er werd geen myelotoxiciteit en geen andere relevante laboratorium afwijkingen gezien. Enkele milde subjectieve bijwerkingen werden gerapporteerd. Eén patiënt stopte voortijdig met de behandeling door hoofdpijnklachten.

We concluderen dat thioguanine een alternatieve behandeloptie kan zijn voor patiënten met moeilijk te behandelen CE, die eerder niet succesvol werden behandeld met azathioprine. SUMMARY 201

Appendices

List of co-authors List of publications Acknowledgement Curriculum vitae 204 APPENDICES

List of co-authors

Daphne S. Bakker Department of Dermatology and Allergology, University Medical Center Utrecht, the Netherlands

Marcel P.H. van den Broek Department of Clinical Pharmacy, University Medical Center Utrecht, the Netherlands Department of Clinical Pharmacy, St Antonius Hospital Nieuwegein, the Netherlands

Carla A.F.M. Bruijnzeel-Koomen Department of Dermatology and Allergology, University Medical Center Utrecht, the Netherlands

Marjolein S. de Bruin-Weller Department of Dermatology and Allergology, University Medical Center Utrecht, the Netherlands

Joukje Deinum Department of Dermatology and Allergology, University Medical Center Utrecht, the Netherlands

Marijke R. van Dijk Department of Pathology, University Medical Center Utrecht, the Netherlands

Herma H. Fidder Department of Gastroenterology, University Medical Center Utrecht, the Netherlands

Kees G. Gerestein Department of Gynaecological Oncology, University Medical Center Utrecht, the Netherlands

Marlies de Graaf Department of Dermatology and Allergology, University Medical Center Utrecht, the Netherlands

J. Maurik van den Heuvel Department of respiratory disease, Academic Medical Center Amsterdam, the Netherlands

Dirk J. Hijnen Department of Dermatology and Allergology, University Medical Center Utrecht, the Netherlands

Elke M.G.J. de Jong Department of Dermatology, Radboud University Nijmegen Medical Centre, the Netherlands

Helen L. Leavis Department of Reumatology, University Medical Center Utrecht, the Netherlands

Anke H. Maitland-van der Zee Department of respiratory disease, Academic Medical Center Amsterdam, the Netherlands

Jart A.F. Oosterhaven Department of Dermatology, University Medical Center Groningen, the Netherlands

Harmieke van Os-Medendorp Department of Dermatology and Allergology, University Medical Center Utrecht, the Netherlands

Klaziena Politiek Department of Dermatology, University Medical Center Groningen, the Netherlands APPENDICES 205

Juul M.P.A. van den Reek Department of Dermatology, Radboud University Nijmegen Medical Centre, the Netherlands

Evelien Roekevisch Department of Dermatology, Academic Medical Center Amsterdam, the Netherlands

Jorien van der Schaft Department of Dermatology and Allergology, University Medical Center Utrecht, the Netherlands

Ron H.N. van Schaik Department of Clinical Chemistry, Erasmus University Medical Center Rotterdam, the Netherlands

Marie-Louise A. Schuttelaar Department of Dermatology, University Medical Center Groningen, the Netherlands

Phyllis I. Spuls Department of Dermatology, Academic Medical Center Amsterdam, the Netherlands

René H.M. Verheijen Department of Gynaecological Oncology, University Medical Center Utrecht, the Netherlands

Arjan D. van Zuilen Department of Nephrology and Hypertension, University Medical Center Utrecht, the Netherlands 206 APPENDICES

List of publications (Thesis)

Garritsen FM, Roekevisch E, van der Schaft J, Deinum J, Spuls PI, de Bruin-Weller MS. Ten years experience with oral immunosuppressive treatment in adult patients with atopic dermatitis in two academic centres. J Eur Acad Dermatol Venereol 2015;29:1905-12.

Garritsen FM, van den Heuvel JM, Bruijnzeel-Koomen CAFM, Maitland-van der Zee AH, van den Broek MPH, de Bruin-Weller MS. Use of oral immunosuppressive drugs in the treatment of atopic dermatitis in the Netherlands. Manuscript submitted.

Bakker DS, Garritsen FM, Leavis HL, van der Schaft J, Bruijnzeel-Koomen CAFM, van den Broek MPH, de Bruin-Weller MS. Lymphopenia in atopic dermatitis patients treated with oral immunosuppressive drugs. Manuscript in preparation.

Garritsen FM, van der Schaft J, van den Reek JM, Politiek K, van Os-Medendorp H, van Dijk MR, Hijnen DJ, de Graaf M, Bruijnzeel-Koomen CAFM, de Jong EM, Schuttelaar MLA, de Bruin- Weller MS. Risk of non-melanoma skin cancer in patients with atopic dermatitis treated with oral immunosuppressive drugs. Acta Derm Venereol 2017;97:724-730.

Garritsen FM, Verheijen RHM, Gerestein CG, van Zuilen AD, Oosterhaven JAF, van Dijk MR, Bruijnzeel-Koomen CAFM, Schuttelaar MLA, de Bruin-Weller MS. Is there an increased risk of cervical neoplasia in atopic dermatitis patients treated with oral immunosuppressive drugs? Accepted for publication in J Eur Acad Dermatol Venereol 2017.

Garritsen FM, van de Broek MPH, van Zuilen AD, Fidder HH, de Bruin-Weller MS, Spuls PI. Pregnancy and fetal outcomes after paternal exposure to azathioprine, methotrexate or mycophenolic acid: a critically appraised topic. Br J Dermatol 2017;176:866-877.

Garritsen FM, van der Schaft J, Bruijnzeel-Koomen CAFM, van Schaik RH, de Graaf M, van den Broek MPH, de Bruin-Weller MS. Thiopurine metabolite levels in patients with atopic dermatitis and/or chronic hand/foot eczema treated with azathioprine. Accepted for publication in J Dermatolog Treat 2017.

Garritsen FM, van der Schaft J, de Graaf M, Hijnen DJ, Bruijnzeel-Koomen CAFM, van den Broek MPH, de Bruin-Weller MS. Improving outcome of azathioprine treatment in chronic eczema by allopurinol co-prescription. Accepted for publication in Acta Derm Venereol 2017.

Garritsen FM, van den Broek MPH, Bruijnzeel-Koomen CAFM, de Bruin-Weller MS. First experience with thioguanine in the treatment of atopic dermatitis. Manuscript in preparation. APPENDICES 207

List of publications (other)

Garritsen FM, ter Haar NM, Spuls PI. House dust mite reduction in the management of atopic dermatitis. A critically appraised topic. Br J Dermatol 2013; 168:688-91.

Garritsen FM, Brouwer MW, Limpens J, Spuls PI. Photo(chemo)therapy in the management of atopic dermatitis: an updated systematic review with implication for practice and research. Br J Dermatol 2014;170:501-13.

Garritsen FM, Spuls P.I De evidence voor fototherapie bij constitutioneel eczeem: de GRADE methodiek. NTDV, 24(3), 144-147.

Garritsen FM, Toonstra J. Halo Scalp Ring bij twee neonaten. NTDV, 24(5), 298-301.

Garritsen FM, van Dijk M, Bruinzeel-Koomen CAF. Subcutane gekapselde vetnecrose. NTDV, 24(5), 301-303.

Garritsen FM, Sanders CJG, Sillevis Smitt JH, Toonstra J. De klassieke efflorescentieleer: nog steeds geen eenduidige zaak. NTDV, 24(6), 390-395.

Garritsen FM, Thijs J, Hijnen DJ, de Bruin-Weller M, Bruijnzeel-Koomen CAFM. Constitutioneel eczeem bij kinderen. Pathofysiologie, diagnostiek en behandeling. Praktische Pediatrie 2014, 129-135. (Een tweede druk in 2017 verschenen in Dé verpleegkundig specialist)

Garritsen FM, Kraag DE, de Graaf M. Guttate psoriasis triggered by perianal streptococcal infection. Clin Exp Dermatol 2017;42:536-538.

Garritsen FM, van den Broek MPH, van Zuilen AD, Fidder HH, de Bruin-Weller MS, Spuls PI. Het effect van paternaal gebruik van azathioprine, methotrexaat of mycofenolzuur op het verloop van de zwangerschap en de gezondheid van het kind: een critically appraised topic (CAT). NTDV, 27(7), 358-359.

Hoep YE, Heinze ZCM, Garritsen FM, Sigurdsson V. De behandeling van scabiës: Ivermectine oraal of permetrine lokaal? Een critically appraised topic (CAT). Accepted for publication in NTDV 2017. 208 APPENDICES

Acknowledgement (dankwoord) Tijdens mijn opleiding tot dermatoloog behield ik de interesse voor het wetenschappelijk onderzoek. Vier dagen per week opleiding en één dag onderzoek: ‘We zien wel hoever je komt’ zei men. Dat er nu, na vier jaar, dit proefschrift ligt, is iets wat ik niet had durven dromen. Alleen was mij dit nooit gelukt. Ik wil dan ook graag iedereen bedanken die heeft bijgedragen aan dit proefschrift, op welke manier dan ook. Een aantal personen wil ik in het bijzonder bedanken.

Mijn promotor prof. dr. C.A.F.M. Bruijnzeel-Koomen. Carla, ik wil je bedanken voor je vertrouwen, je steun, je begeleiding en je altijd kritische blik. Stiekem had je altijd gelijk: artikelen werden beter door jouw opmerkingen. Het is fijn en bewonderenswaardig dat je altijd betrokken bent gebleven, ook op de momenten dat je even niet in de buurt was. Carla, ik vind het een eer om de indrukkende lijst van jouw promovendi te mogen afsluiten.

Mijn copromotor dr. M.S. de Bruin-Weller. Marjolein, jij bent de drijvende kracht achter dit proefschrift geweest. Het optimistische plan om parttime te gaan promoveren werd door jou al snel als een grote uitdaging gezien. Jij zag de voordelen en zocht altijd naar mogelijkheden. Je was op de hoogte van de laatste ontwikkelingen op het gebied van constitutioneel eczeem en had altijd een duidelijk doel voor ogen. Je wist welke vraagstukken er leefden en aan welke antwoorden en onderzoeken er behoefte was. Bijna maandelijks kwam je weer met nieuwe ideeën. Marjolein, ontzettend bedankt voor alle hulp, voor de inspiratie, voor je enthousiasme en je steun. Promoveren samen met jou was iets wat ik niet had willen missen!

Mijn copromotor dr. M.P.H. van den Broek. Marcel, een apotheker als copromotor, wat leuk en bijzonder! Naarmate de samenwerking tussen jou en de afdeling dermatologie intensiever werd, ontstond het idee voor jouw copromotor schap. Jouw expertise, jouw enthousiasme en je contacten met andere vakgebieden hebben de basis gevormd voor deel drie van dit proefschrift. Maar ook bij de andere studies was je vaak nauw betrokken. Geen vraag was jou te gek: binnen de kortste keren had jij steeds de nieuwste publicaties gevonden of dingen nog even nagevraagd bij je collega’s of andere specialismen. Ik heb ontzettend veel van je geleerd en wil je heel erg bedanken voor de leuke samenwerking de afgelopen jaren.

Leden van de promotie beoordelingscommissie: prof. dr. A.C.G. Egberts, prof. dr. A.H. Maitland-van der Zee, prof dr. J.M. van Laar, dr. M.L.A. Schuttelaar en prof. dr. P.I. Spuls. Ik dank u voor het kritisch beoordelen van het manuscript. APPENDICES 209

Mijn opleider, dr. V. Sigurdsson. Vigfús, toen ik in 2014 aangaf dat ik met het wetenschappelijk onderzoek door wilde gaan, bedacht jij direct een constructie waarin dat mogelijk was. Je dacht met me mee over een opleidingsschema met een goede balans tussen kliniek en wetenschap. Daar wil ik je graag voor bedanken. André, Anna-Marijke, Chris, Deepak, DirkJan, Feiko, Heike, Johan, Koos, Marlies, Sagar en Thuy- My: stafleden en oud-stafleden dermatologie in het UMC Utrecht. Bedankt voor jullie interesse in mijn onderzoek de afgelopen jaren. Ik ga met heel veel plezier mijn opleiding afmaken.

Prof. dr. P.I. Spuls, Phyllis. Ik denk dat ik wel kan zeggen dat jij aan de basis van mijn dermatolo- gische carrière hebt gestaan. Ik werd met open armen ontvangen toen ik in 2010 op zoek was naar een wetenschappelijke stage van zes weken. Jij zag geen enkel probleem: ‘Het is kort, maar we verzinnen wel wat!’. Je vertrouwde in mijn kunnen en zorgde voor goede begeleiding, niet alleen tijdens mijn stage maar ook in de jaren die volgden. Ik denk nog vaak aan jouw opmerking ‘Als je er niet aan begint, dan kom je er nooit’. Je had gelijk, bedankt!

Lieve mede onderzoeks-collega’s, Anna, Ans, Barbara, Daphne, Florine, Jorien, Judith, Harmieke, Henrike, Ischa, Jos, Lieneke, Marieke, Mignon, Mark, Sarah en Stans. Ik kijk met veel plezier terug op alle gezellige lunches, koffies, speerpunten en hulp bij SPSS/Graphpad/Visio! Fijn dat ik altijd welkom was bij jullie op de kamer als ik weer eens een computer nodig had. Jullie interessante onderzoeken zullen de komende jaren ook resulteren in prachtige proefschriften, daar twijfel ik niet over! Jorien, de ‘future perspectives’ van jouw proefschrift vormden de basis en inspiratie voor mijn boekje en jouw kritische blik heeft menig artikel verbeterd. Ik hoop nog heel lang met je te kunnen samenwerken en congressen te bezoeken! Daphne en Lieneke, als studenten hebben jullie me erg geholpen de afgelopen jaren tijdens jullie wetenschappelijke stage. Jullie enthousiasme is niet onopgemerkt gebleven en ik vind het ontzettend leuk dat ook jullie nu deel uitmaken van het onderzoeksteam en bezig zijn met een eigen promotie traject!

Lieve AIOS en oud AIOS uit het UMC Utrecht, wat fijn dat jullie al jaren mijn collega’s zijn. Dank voor de flexibiliteit als er weer eens wat geschoven moest worden met mijn ‘onderzoeksdag’, dank voor alle gezelligheid, borrels en de fantastische weekendjes weg inclusief de hilarische verkleedpartijtjes. 210 APPENDICES

Stafsecretaresses Jantine en Miranda. Jullie zorgden altijd voor gezelligheid en toonden regelmatig interesse in de voortgang van mijn onderzoek. Bedankt voor de talrijke keren dat ik de vergaderzaal wilde reserveren of met meerdere personen een afspraak probeerde in te plannen. Jullie wilden me altijd graag helpen! Collega’s van het secretariaat en de verpleegkundigen en dokters-assistenten van de polikliniek en verpleegafdeling dermatologie en allergologie: dank voor de prettige samenwerking de afgelopen jaren!

Collega AIOS van de trial kamer uit het AMC: Anne, Dominique, Evelien, Marijke en Stef, dank voor jullie gezelligheid en hulp tijdens de wetenschappelijke stages in het AMC. Mede door jullie ben ik enthousiast geworden voor het onderzoek! Evelien. In 2010 was jij de directe begeleider van mijn wetenschappelijke stage bij de dermatologie in het AMC en heb je mij direct geïnspireerd: je was een enthousiaste begeleider met interesse voor het gebruik van immunosuppressiva bij constitutioneel eczeem en met passie voor onderzoek. Ik wil je ontzettend bedanken voor de leuke tijd tijdens mijn onderzoeksstages in het AMC en voor de fijne samenwerking daarna! Marijke, bedankt dat je me cappuccino hebt leren drinken, ik kan nu niet meer zonder!

Collega’s uit Nijmegen en Groningen: prof. dr. Elke de Jong, dr. Juul van den Reek, dr. Marie- Louise Schuttelaar, drs. Klaziena Politiek en drs. Jart Oosterhaven. Heel erg bedankt voor de prettige samenwerking. Juul, wat leuk dat je als oude ‘brugplusser’ betrokken was bij een van mijn eerste publicaties!

Drs. Joukje Deinum, prof. dr. Marijke van Dijk, dr. Herma Fidder, dr. Kees Gerestein, dr. Helen Leavis, dr. Harmieke van Os-Medendorp, prof. dr. Ron van Schaik, prof. dr. René Verheijen, dr. Arjan van Zuilen: bedankt voor jullie hulp bij het schrijven van de verschillende manuscripten. Jullie ervaringen met het gebruik van orale immunosuppressiva bij andere ziekten dan eczeem zijn zeer waardevol geweest.

Fenno en Joep, Die Jongens. Ontzettend bedankt voor de lay-out van dit proefschrift. Wat is het mooi geworden, wat een eer dat jullie me hierbij wilden helpen.

Agonisten: Arja, Francine, Joppe, Nienke en Ruben. Bij elkaar gezet om in 2007 het bestuur van MSFU “Sams” te vormen en nu vrienden voor het leven. Na jullie proefschriften kon ik als benjamin van de groep natuurlijk niet achter blijven. Met jullie is alles zilver, voorzien van een gouden randje. APPENDICES 211

Mixed grill, vriendinnetjes van het eerste uur: Anna, Célinde, Floor en Inge. Ieder een eigen weg gegaan en allemaal een andere specialisatie op het oog. Maar wat is het leuk om jullie nog zo vaak te zien. Ik kijk uit naar heel veel NS-wandelingen en dinertjes die nog gaan volgen.

Jaarclub Yacht: Anneke, Charlotte, Dieke, Eveline, Janna, Marieke, Mariëlle, Marije, Rosa, Sanne en Zjoske. Het is heerlijk om bij jullie niet over geneeskunde te praten maar over andere belangrijke zaken van het leven. Ik kijk met veel plezier terug op onze etentjes, borrels en weekendjes weg (inclusief de huisfeestjes!) en hoop jullie nog heel veel te zien!

Lieve vrienden uit Helmond (Anne, Ingeborg), Utrecht (Marcelle, Willemijn en de trimhockeyers van Kampong), Bunnik (Maarten en Maaike) en Groningen (Hanneke, Mirjam): bedankt voor de altijd welkome en gezellige avondjes!

Gerben, Mariëlle, Casper, Martijn en Anne Sophie. Maar natuurlijk ook Barbara en Joost. Wat heb ik het getroffen met jullie als schoonfamilie. Ontzettend bedankt voor jullie interesse de afgelopen jaren en voor de ontspannen weekendjes in Friesland. Wie van jullie gaat er vanavond met de hapjes rond?

Lieve papa en mama, Fons en Henriëtte. Ondanks het feit dat jullie geen medische achtergrond hebben, zijn jullie altijd geïnteresseerd geweest in mijn onderzoek. Ik vind het bijzonder om nu mijn eindresultaat aan jullie te kunnen laten zien. Bedankt dat jullie me altijd gestimuleerd hebben om de wereld te verkennen.

Lieve zusjes, Linde, Imke en Heike. Ieder een andere richting op gegaan, maar altijd geïnter- esseerd in elkaars bezigheden. Linde, jouw eeuwige enthousiasme en doorzettingsvermogen werken aanstekelijk. Imke, dat jij je drukke baan met een opleiding combineert is echt bewonderenswaardig. Heike, wat ontzettend leuk dat ook jij interesse hebt voor het onderzoek en nu met een promotie traject gaat starten! Lieve Els, Floris, Jelle en Max, wat fijn dat jullie bij onze familie zijn.

Lieve Thomas, de laatste woorden van dit dankwoord zijn voor jou. Wat is het leven leuk met jou! We zijn beiden de afgelopen jaren druk geweest met onze opleiding en onderzoek, maar hebben gelukkig voldoende tijd gevonden om regelmatig te reizen en samen de mooiste plekken van de wereld te zien. Ik wil je bedanken voor je liefde, je grappen, je interesse, je ideeën, je relativeringsvermogen en je steun. Lieve Tho, bedankt dat je er voor me bent. 212 APPENDICES

Curriculum vitae Floor Garritsen werd geboren op 22 oktober 1986 te Beek en Donk als oudste van vier meiden. Na in 2005 cum laude haar gymnasium diploma behaald te hebben aan het Carolus Borromeus College te Helmond, begon zij in datzelfde jaar met de opleiding Geneeskunde aan de Universiteit Utrecht. In 2007-2008 onderbrak zij haar studie tijdelijk, om als commis- saris Externe Betrekkingen & PR deel uit te maken van het bestuur van de Medische Studenten Faculteitsvereniging Utrecht “Sams”. In het vijfde jaar van haar studie deed zij onderzoek naar het gebruik van orale immunosuppressiva bij de behandeling van constitutioneel eczeem bij de afdeling dermatologie in het Academisch Medisch Centrum Amsterdam, onder begeleiding van drs. Evelien Roekevisch en prof. dr. Phyllis Spuls. In het zesde jaar zette zij dit onderzoek aldaar voort en volgde zij tevens een bijzondere semi-arts stage bij de afdeling dermatologie/ allergologie in het UMC Utrecht, onder begeleiding van drs. Koos Sanders en dr. Marjolein de Bruin-Weller. Na het behalen van haar artsexamen in december 2012, is zij aansluitend begonnen als arts klinisch onderwijs (AKO) bij de afdeling dermatologie/allergologie in het UMC Utrecht. In deze functie was zij gedurende één jaar verantwoordelijk voor het dermatologie onderwijs in de opleiding geneeskunde en voor de directe begeleiding van de coassistenten dermatologie. In dit jaar behaalde zij haar Basis Kwalificatie Onderwijs (BKO). Aansluitend is zij op 1 januari 2014 gestart met de opleiding tot dermatoloog in het UMC Utrecht met dr. Vigfús Sigurdsson als opleider. Tijdens haar opleiding behield zij haar interesse voor het wetenschappelijk onderzoek, hetgeen heeft geleid tot dit proefschrift. Floor woont samen met Thomas Klein Nulent in Utrecht.