Comparing cutaneous molecular improvement with different treatments in atopic dermatitis patients

Journal Pre-proof Comparing cutaneous molecular improvement with different treatments in atopic dermatitis patients Jacob W. Glickman, MA, Celina Dubin, BA, Joseph Han, BS, Dante Dahabreh, BA, Sandra Garcet, PhD, James G. Krueger, MD, PhD, Ana B. Pavel, PhD, Emma Guttman-Yassky, MD, PhD PII:

S0091-6749(20)30039-7

DOI:

https://doi.org/10.1016/j.jaci.2020.01.005

Reference:

YMAI 14369

To appear in:

Journal of Allergy and Clinical Immunology

Received Date: 5 December 2019 Revised Date:

9 January 2020

Accepted Date: 9 January 2020

Please cite this article as: Glickman JW, Dubin C, Han J, Dahabreh D, Garcet S, Krueger JG, Pavel AB, Guttman-Yassky E, Comparing cutaneous molecular improvement with different treatments in atopic dermatitis patients, Journal of Allergy and Clinical Immunology (2020), doi: https://doi.org/10.1016/ j.jaci.2020.01.005. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Published by Elsevier Inc. on behalf of the American Academy of Allergy, Asthma & Immunology.

Glickman 1

Title: Comparing cutaneous molecular improvement with different treatments in atopic

2

dermatitis patients

3 4

Authors:

5

Jacob W. Glickman MA1, Celina Dubin BA1, Joseph Han BS1, Dante Dahabreh BA1, Sandra

6

Garcet PhD2, James G. Krueger MD, PhD2, Ana B. Pavel PhD1,*, Emma Guttman-Yassky MD,

7

PhD1,*

8 9

1. Department of Dermatology and Laboratory of Inflammatory Skin Diseases, Icahn School of

10

Medicine at Mount Sinai, New York, NY

11

2. Laboratory for Investigative Dermatology, The Rockefeller University, New York, NY

12 13 14 15 16

*Equally contributing corresponding authors:

17

Ana B. Pavel, PhD; Emma Guttman-Yassky MD, PhD

18

Department of Dermatology and Laboratory of Inflammatory Skin Diseases

19

Icahn School of Medicine at Mount Sinai

20

5 E. 98th Street, New York, NY 10029

21

Emails: [email protected]; [email protected]

22

Telephone: 212-241-9728/3288; Fax: 212-876-8961

23

1

Glickman

2

24

Disclosures: EGY is an employee of Mount Sinai and has received research funds (grants paid

25

to the institution) from: Abbvie, Celgene, Eli Lilly, Janssen, Medimmune/Astra Zeneca,

26

Novartis, Pfizer, Regeneron, Vitae, Glenmark, Galderma, Asana, Innovaderm, Dermira, UCB.

27

EGY is also a consultant for Sanofi Aventis, Regeneron, Stiefel/GlaxoSmithKline, MedImmune,

28

Celgene, Anacor, AnaptysBio, Dermira, Galderma, Glenmark, Novartis, Pfizer, Vitae, Leo

29

Pharma, Abbvie, Eli Lilly, Kyowa, Mitsubishi Tanabe, Asana Biosciences, and Promius. JGK

30

has received research support (grants paid to his institution) and/or personal fees from Pfizer,

31

Amgen, Janssen, Lilly, Merck, Novartis, Kadmon, Dermira, Boehringer, Innovaderm, Kyowa,

32

BMS, Serono, BiogenIdec, Delenex, AbbVie, Sanofi, Baxter, Paraxel, Xenoport, and

33

Kineta. The rest of the authors declare that they have no conflicts of interest to disclose.

34 35 36

Word Count: 1114

37 38

Capsule summary: This is the first study that objectively compares the molecular cutaneous

39

improvement with different treatments in AD and may facilitate personalized medicine by

40

identifying the best treatments for individual phenotypes, based on immune/barrier profiles.

41 42

Keywords: Atopic dermatitis, precision medicine, ASN002, cyclosporine, crisaborole,

43

dupilumab, fezakinumab, ustekinumab, narrow-band ultraviolet type B, corticosteroids

Glickman 44

Abbreviations:

45

AD

atopic dermatitis

46

BID

bis in die (twice a day)

47

CE

cornified envelope

48

CsA

cyclosporine

49

EASI

Eczema Area and Severity Index

50

EDC

epidermal differential complex

51

JAK

Janus kinase

52

LS

lesional

53

MADAD

meta-analysis-derived atopic dermatitis

54

NB

narrow band

55

NL

nonlesional

56

OR

online repository

57

SYK

spleen tyrosine kinase

58

TH

T helper cell type

59

UVB

type B ultraviolet

3

Glickman 60 61

4

To the Editor: Atopic dermatitis/AD is a common inflammatory skin disease that affects up to 10% of

62

adults worldwide.1 AD is a heterogenous disease with primary TH2/TH22-skewing, and variable

63

TH1/TH17 contribution. Due to its multi-cytokine nature, many broad-acting and specific

64

systemic and topical treatments were developed targeting certain cytokine pathways, with

65

variable efficacy.1 Recently, broader agents (i.e cyclosporine/CsA, narrow-band UVB/NB-UVB,

66

Janus kinase/JAK inhibitors/ASN002, corticosteroids), and more specific antagonists targeting

67

TH2/IL-4R (dupilumab), TH22/IL-22 (fezakinumab), TH17/TH22/TH2 (crisaborole), and IL-12/IL-

68

23 (ustekinumab) have shown significant changes in lesional AD molecular profiles compared to

69

placebo.1-8 Nevertheless, a comprehensive, objective means to compare cutaneous effects of

70

various drugs and placebos is lacking. To better understand how different pathways contribute to

71

AD pathogenesis, it is important to objectively compare relative cutaneous changes in immune

72

and barrier mechanisms with different treatments.

73

We used a meta-analysis approach similar to a recent psoriasis study,9 to evaluate

74

molecular changes in skin with different AD therapeutics. We only included clinical trials in

75

which Affymetrix U133Plus 2.0 lesional and nonlesional skin gene-arrays were available from

76

AD patients treated with broad (CsA, NB-UVB, JAK/SYK inhibitor/ASN002, topical

77

corticosteroids) and specific (dupilumab, fezakinumab, ustekinumab, crisaborole) treatments

78

(Table E1). Seven publicly available microarray datasets (Six1-3, 5-7 in moderate-to-severe AD

79

and one4 in mild-to-moderate AD patients) were found in Gene Expression Omnibus/GEO

80

(GSE133385/GSE133477/GSE58558/GSE130588/GSE99802/GSE27887/GSE140684; Table

81

E1).1-7 The main outcome of interest was percent improvement by gene-signatures

82

(see Methods section in this article's Online Repository/OR at www.jacionline.org). The R limma

Glickman

5

83

package was used to model expression values by mixed-effect models from all mechanistic skin

84

studies derived from microarray trial data. Drug improvement percentages were then calculated

85

at available timepoints for each gene, and grouped by previously reported AD-related gene-

86

subsets.1 These signatures include upregulated/downregulated meta-analysis derived

87

AD/MADAD transcriptome1, 2, immune and barrier gene-subsets1, 2, negative regulators1, 2, and

88

TH1/TH2/TH17/IL-22/IL-34-specific genes. Due to the inherent normalization of percent drug

89

improvement, grouped AD-related gene signatures could be directly compared across treatments

90

by Wilcoxon test.

91

Overall, CsA showed the highest improvements in AD-related gene signatures (Figs 1,

92

2), significantly outperforming all other drugs/placebos in most gene-subsets (i.e up- and down-

93

regulated MADAD transcriptome, immune genes, epidermal differentiation and lipid genes,

94

negative regulators and TH1/TH17/TH22/IL-22-regulated genes; P<.01, Figs 1, 2; Tables E2,E3);

95

however, dupilumab shows the highest overall improvement in TH2-specific genes

96

(median=205%), and significantly outperforms other drug and placebo arms with available

97

evaluations at Wk16, across the MADAD, immune genes, negative regulators, and TH2 genes

98

(P<.05, Figs 1, 2; Table E2). Although dupilumab significantly outperforms placebo and

99

ustekinumab at Wk16 for TH1/TH17 genes (P<.05, Fig 2; Table E2), it shows higher median

100

improvement for TH17 (125%) compared to TH1 genes (74%). At Wk4, ASN002-40mg

101

significantly improved the upregulated MADAD profile and TH22/IL-22-specific genes more

102

than ASN002-20mg/ASN002-80mg/fezakinumab/placebo with median differences greater than

103

36% (P<.05, Fig 1; Table E2). While IL-22-targeting with fezakinumab showed modest

104

improvements on immune pathways, it showed comparable or higher improvement in MADAD

105

down-regulated and barrier genes to dupilumab, highlighting the role of IL-22 in the AD barrier

Glickman 106

dysregulation. As expected with an IL-22 inhibitor, fezakinumab has a median improvement of

107

84% on TH22/Il-22 genes, outperforming placebo by a median difference of 49% at Wk12

108

(P<.01); however, it shows no significant difference in improvement compared to NB-UVB at

109

Wk12 for TH22/IL-22 genes (P>.05, Fig 1; Table E2).

110

6

NB-UVB, a safe and efficacious albeit time-consuming treatment, shows significant

111

improvements on MADAD, negative regulators, TH1 and immune genes, with lesser efficacy on

112

barrier-associated markers; it outperforms fezakinumab with median differences greater than

113

20% across the upregulated MADAD transcriptome, negative regulators, TH1 and immune genes

114

(P<.01, Figs 1, 2; Table E2). Crisaborole improves AD-related signatures more than placebo

115

(P<.05), with greatest improvements in TH22/TH17-related genes, and shows significantly more

116

improvement than with systemic ASN002 across the upregulated MADAD transcriptome at Wk2

117

(P<.001, Fig 1; Table E2).

118

While ustekinumab does not have robust improvement for AD-related gene-subsets (Figs

119

1, 2), mild-to-moderate topical steroids (triamcinolone 0.025%) were used consistently twice

120

daily until Wk16 during the placebo-controlled portion, and published separately detailing

121

continuous improvement beyond Wk4 with topical corticosteroids.8 The topical steroid group

122

(“ustekinumab placebo”) significantly outperforms dupilumab’s placebo and ustekinumab at

123

Wk16 by a median difference of at least 13% across upregulated/downregulated MADAD

124

transcriptomes, immune genes, negative regulators and TH17 (P<.05, Figs 1, 2; Table E2). This

125

highlights the relevance of topical corticosteroid use when evaluating placebo and drug

126

responses and for trial design. It also serves as a reminder that caution is needed when comparing

127

drugs to “placebo” allowing topical medications. Finally, we compared improvements in IL-34, a

128

negative regulator of inflammation, that is normally down-regulated in AD, and a biomarker of

Glickman 129

response for different therapies.2, 4, 6, 8 Interleukin-34 is a cytokine that promotes the

130

differentiation and viability of monocytes and macrophages through the colony-stimulating

131

factor-1 receptor, and has been shown to play important role in autoimmune disorders,

132

inflammation and cancer. Interestingly, IL-34 discriminated molecular response between

133

efficacious drugs and placebo, using a 35% cut-off.

134

7

These results confirm CsA as the gold standard of cutaneous improvement in AD, but

135

also show that narrow-targeting with dupilumab achieves broad changes in skin, beyond those

136

expected with specific TH2 inhibition. These data also shed light onto the relative improvement

137

with different treatments in various immune and barrier mechanisms, highlighting unique

138

properties of targeting a particular axis. For example, targeting the TH22 axis with fezakinumab

139

showed major improvement on barrier dysfunction, but only modest improvement in

140

inflammatory signatures.

141

Our approach has the potential to facilitate future development of a personalized

142

medicine approach in AD, and identify the best treatments for individual phenotypes, based on

143

immune and barrier profile. For instance, a patient with TH2 driven disease may primarily benefit

144

from dupilumab, while for a patient with high TH2 and TH1 axes JAK targeting may be

145

preferable.

146

Our study has several limitations. It compares therapeutics with different modes of action

147

and their counterpart placebos. The actual “placebo” may vary across trials, as in the

148

ustekinumab study. Also, studies differ in their experimental design and patient numbers. For

149

instance, the crisaborole study incorporated an intrapatient design, the ASN002 study had 36

150

patients with 9 patients per arm, and the CsA and NB-UVB studies had no placebo. Further, the

151

timepoints for biopsy evaluations varied between studies, as well as the early and late biopsy

Glickman 152

timepoints, thus the data analysis was restricted to available timepoints. Almost all studies

153

included moderate-to-severe AD except the crisaborole study that included mild-to-moderate

154

patients. As a result, drug improvements may have been smaller for milder compared to more

155

severe patients. Finally, the therapeutics were administered disproportionally to white

156

populations making it more difficult to generalize the analysis. Overall, this approach may

157

determine which treatments best benefit various AD characteristics and can promote a precision

158

medicine strategy.

8

159

Jacob W. Glickman MA1

160

Celina Dubin BA1

161

Joseph Han BS1

162

Dante Dahabreh BA1

163

Sandra Garcet PhD2

164

James G. Krueger MD, PhD2

165

Ana B. Pavel PhD1

166

Emma Guttman-Yassky MD, PhD1

167 168

1. Department of Dermatology and Laboratory of Inflammatory Skin Diseases, Icahn School of

169

Medicine at Mount Sinai, New York, NY

170

2. Laboratory for Investigative Dermatology, The Rockefeller University, New York, NY

171 172

Glickman

9

173

Figure legends

174

Figure 1. Drug improvement percentages for AD-related gene signatures:

175

upregulated/downregulated MADAD transcriptomes, immune genes, EDC-CE and lipids,

176

TH2, and TH22/IL-22 specific genes at weeks 1, 2, 4, 12, and 16. Stars denote significant

177

differences in the improvement scores between drug and placebo groups from the same study by

178

Wilcoxon test. Bars superimposed on each point represent the median (scaled by study size) ±

179

IQR/2 and are colored by treatment. +P < .1, *P < .05, **P < .01, ***P<.001. EDC-CE,

180

epidermal differential complex—cornified envelope.

181 182

++

Subjects treated in the ustekinumab study were given a background of triamcinolone .025%

BID. The placebo group used more triamcinolone .025% than the drug group.5

183 184

Figure 2. Drug improvement percentages for AD-related gene signatures: TH1, TH17,

185

negative regulators, and IL-34 specific genes at weeks 1, 2, 4, 12, and 16. Stars denote

186

significant differences in the improvement scores between drug and placebo groups from the

187

same study by Wilcoxon test. Bars superimposed on each point represent the median (scaled by

188

study size) ± IQR/2 and are colored by treatment. +P < .1, *P < .05, **P < .01, ***P<.001.

189 190 191

++

Subjects treated in the ustekinumab study were given a background of triamcinolone .025%

BID. The placebo group used more triamcinolone .025% than the drug group.5

Glickman 10 192

References

193

1.

Guttman-Yassky E, Bissonnette R, Ungar B, Suarez-Farinas M, Ardeleanu M, Esaki H, et

194

al. Dupilumab progressively improves systemic and cutaneous abnormalities in patients

195

with atopic dermatitis. J Allergy Clin Immunol 2019; 143:155-72.

196

2.

Pavel AB, Song T, Kim HJ, Del Duca E, Krueger JG, Dubin C, et al. Oral Janus

197

kinase/SYK inhibition (ASN002) suppresses inflammation and improves epidermal

198

barrier markers in patients with atopic dermatitis. J Allergy Clin Immunol 2019.

199

3.

Khattri S, Shemer A, Rozenblit M, Dhingra N, Czarnowicki T, Finney R, et al.

200

Cyclosporine in patients with atopic dermatitis modulates activated inflammatory

201

pathways and reverses epidermal pathology. J Allergy Clin Immunol 2014; 133:1626-34.

202

4.

Bissonnette R, Pavel AB, Diaz A, Werth JL, Zang C, Vranic I, et al. Crisaborole and

203

atopic dermatitis skin biomarkers: An intrapatient randomized trial. Journal of Allergy

204

and Clinical Immunology 2019.

205

5.

Khattri S, Brunner PM, Garcet S, Finney R, Cohen SR, Oliva M, et al. Efficacy and

206

safety of ustekinumab treatment in adults with moderate-to-severe atopic dermatitis. Exp

207

Dermatol 2017; 26:28-35.

208

6.

Brunner PM, Pavel AB, Khattri S, Leonard A, Malik K, Rose S, et al. Baseline IL-22

209

expression in patients with atopic dermatitis stratifies tissue responses to fezakinumab. J

210

Allergy Clin Immunol 2019; 143:142-54.

211

7.

Tintle S, Shemer A, Suarez-Farinas M, Fujita H, Gilleaudeau P, Sullivan-Whalen M, et

212

al. Reversal of atopic dermatitis with narrow-band UVB phototherapy and biomarkers for

213

therapeutic response. J Allergy Clin Immunol 2011; 128:583-93 e1-4.

Glickman 11 214

8.

Brunner PM, Khattri S, Garcet S, Finney R, Oliva M, Dutt R, et al. A mild topical steroid

215

leads to progressive anti-inflammatory effects in the skin of patients with moderate-to-

216

severe atopic dermatitis. J Allergy Clin Immunol 2016; 138:169-78.

217

9.

Sawyer LM, Malottki K, Sabry-Grant C, Yasmeen N, Wright E, Sohrt A, et al. Assessing

218

the relative efficacy of interleukin-17 and interleukin-23 targeted treatments for

219

moderate-to-severe plaque psoriasis: A systematic review and network meta-analysis of

220

PASI response. PloS one 2019; 14.

221 222

MADAD Up

160

140

120

80

% Improvement

100 *** ***

60 40

*** ***

20

***

***

***

*** ***

0

***

***

100 80

40 20

***

***

1

2

0

Immune Genes

4 Weeks

12

*** ***

***

***

***

*** ***

40

16

*** ***

*** ***

20

40

1

2

4 Weeks

EDC-CE & Lipids

12

16

200

140

180 160

100

140

% Improvement

120

80 60

***

40

***

20

*********

***

0

***

***

**

20

***

**

***

120 100 80 60 +

40 20 0

***

1

2

TH2

4 Weeks

12

+

***

**

60

16

***

***

40

60

*** ***

*** ***

***

20

40

300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0 20 40 60

***

60

***

***

20

160

MADAD Down

140

120

1

2

TH22 & IL-22

4 Weeks

12

16

160 140 120

% Improvement

% Improvement

% Improvement

% Improvement

Figure 1

**

* +

80 *

60 40

***

*

20

** *

100

*** **

*

** **

**

0

**

***

**

20 1

2

4 Weeks

12

1

16

2

Treatment

4 Weeks

ASN002, 20mg

Dupilumab

ASN002, 40mg

Dupilumab, Placebo

ASN002, 80mg

Fezakinumab (anti-IL-22)

ASN002, Placebo

Fezakinumab (anti-IL-22), Placebo

Crisaborole

Narrow-band UVB (NB-UVB)

Crisaborole, Vehicle

Ustekinumab++

Cyclosporine

Triamcinolone 0.025% BID (Ustekinumab, Placebo)

12

16

Figure 2 TH1

TH17

180

160

160

140

140

120

% Improvement

% Improvement

120

100

100 80 60 20

*** ***

***

40

** ***

***

0

**

**

***

*

*** *

40

80

1

2

4 Weeks

12

16

Negative Regulators

*

* **

1

2

4 Weeks

12

16

1

2

4 Weeks

12

16

IL-34 160

120

140

100

120

% Improvement

140

% Improvement

*

20

60

100

80 60

***

40 0

** ***

***

*

0

40

20

***

60

20

20

160

***

80

***

* +

20

*

******* **

*** *

80 60 40 20 0

40

20

60 1

2

4 Weeks

12

40

16

Treatment ASN002, 20mg

Dupilumab

ASN002, 40mg

Dupilumab, Placebo

ASN002, 80mg

Fezakinumab (anti-IL-22)

ASN002, Placebo

Fezakinumab (anti-IL-22), Placebo

Crisaborole

Narrow-band UVB (NB-UVB)

Crisaborole, Vehicle

Ustekinumab++

Cyclosporine

Triamcinolone 0.025% BID (Ustekinumab, Placebo)