Using transcriptomics and proteomics to gain insights in the development of psoriasiform skin lesions under anti-tumor necrosis factor (TNF) therapy, and predicting response to ustekinumab.

The investigators will prospectively include patients with Crohn's disease (CD) or ulcerative colitis (UC), who develop psoriasiform skin lesions (including psoriasiform eczema, psoriasis guttata, psoriasis inversa and pustulosis) under therapy with anti-TNF and refractory to at least 12 weeks of topical therapy. Since no scoring systems are available to describe (the severity of) anti-TNF induced skin lesions, the lesions will be followed up at different time points using 3 different methods. First, the Dermatology Life Quality Index (DLQI), which is a standardised questionnaire that is used in routine clinical practice to assess the impact of any type of skin lesion on the quality of life of the patient. Next a global visual analogue scale (VAS) to be filled out by the expert dermatologist based on the general appearance of the skin lesions, and finally a self-designed physician global assessment (PGA). Physician global assessment of paradoxical skin lesions will be based on the following:

• Number of body regions that are affected: 1 region (1 point), 2 or 3 regions (2 points) and more than 3 regions (3 points). The 8 regions to consider are: scalp, face, both axilla, groins, genital (including pubis and perineal), trunk (front and back + neck and collar), arms (including hands), legs (including feet)
• Location of skin lesions: face (1 point), palmoplantar (1 point), genital (1 point), flexural (1 point)
• Affected body surface area >10% (3 points)
• Presence of itching (1 point)
• Painful skin lesion (1 point)
• Is the lesion wet including erosions, pustules and maceration (2 points) or dry (0 points)?
• Is the lesion superinfected? Yes (1 point) or No (0 points)

The investigators plan to include 20 patients, namely the first 10 patients who will be switched to ustekinumab (UST, n=10, only available for patients with CD) and the first 10 patients who will be switched to anti-integrin therapy (VDZ, n=10). The latter will be used as a control group to evaluate whether starting ustekinumab or discontinuing anti-TNF leads to improvement of the lesions. Patients will be included through our outpatient clinic or infusion unit. The decision to start ustekinumab or vedolizumab will be based on routine clinical practice, and not be influenced by this study.

At the first evaluation (moment of treatment change), a blood draw (EDTA, serum, PAX gene) will be performed, the clinical CD or UC activity will be calculated by using patient reported outcome (PRO2) and Crohn's Disease Activity Index (CDAI) or Mayo score, and the patient will be asked to fill out a DLQI. Next, the patient will be send to one of two board certified dermatologists. The latter will document the skin lesions, take pictures of all lesions, state the clinical diagnosis and perform two punch biopsies of the affected skin ánd of unaffected skin at the contralateral sight. The dermatologist will also fill out the above mentioned VAS and PGA.

Patients switched to ustekinumab will receive standard induction with intravenous ustekinumab 6mg/kg, followed by subcutaneous ustekinumab 90mg every 8 weeks. Patients switched to vedolizumab will receive standard induction with intravenous vedolizumab 300mg at weeks 0, 2, 6, 10, 14 and every 8 weeks thereafter.

The patients will be re-evaluated 6-8 and 14-16 weeks after change in treatment. At week 6-8, patients will be clinically (including PRO2, CDAI/Mayo, and DLQI) and serologically (blood samples) re-evaluated by one of our three inflammatory bowel disease (IBD) specialists and all lesions will be photographed again. At week 14-16, re-evaluation by a gastroenterologist as well as dermatologist will take place. This will also include a new skin biopsy, but only at the initial site of the skin lesions (not the unaffected contralateral sight) (cfr. Study diagram). Dermatological response will be defined as a drop in VAS with at least 50%.

All blood samples will be analysed at the Translational Research in GastroIntestinal Disorders (TARGID) lab, Catholic University Leuven. One of the baseline skin biopsies will be send for histological examination by a certified pathologist (prof. dr. van den Oord) to obtain a correct histological diagnosis of the lesion. All the other biopsies will be stored in the TARGID-lab, Catholic University Leuven and will be used to perform transcriptomic and proteomic analyses (see further).

First, all patients included will be genetically characterize by using the Infinium Global Screening Array (GSA, Illumina). For the analyses the investigators will collaborate with the Department of Complex Genetics and Genomics Core Leuven. Next, a weighted polygenetic risk score will be calculated for all individuals. By performing a weighted risk score, the effect size and population frequency of the included genetic variants will be taken into account. The latter will be taken from already published genome-wide association studies. Using GSA will also enable the investigators to find new allele. Third, compare protein expression between the different patients groups will be compared by using serum tubes. Protein biomarkers in serum will be compared using the immunology panel (OLINK Proteomics). This panel is a high-throughput, multiplex immunoassay enabling analysis of 92 protein biomarkers across 96 samples simultaneously. The proteins will be analyzed following quality control and excluding markers with >75% missing data. To validate the found biomarkers the Mesoscale platform (Meso Scale Diagnostics) will be used. In addition, differences in gene expression in blood as well as skin biopsies will be compared. For the blood analyses PAX tubes will be used from the investigators' biobank. To extract RNA from PAX tubes the PAXgene Blood RNA Kit (Qiagen) will be used. For RNA extraction from skin biopsies AllPrep DNA/RNA Mini Kit (Qiagen) will be used. After extraction of RNA, next-generation single-end sequencing will be performed in collaboration with the Genomics core facility, Leuven. The high sequencing (4000NGS) platform from Ilumina will be used. Based on this proof of concept study a more limited panel of markers will be developed and evaluate in a confirmation cohort.

It is known that TNF, type I interferon (IFN) and interleukin-17 (IL-17) play an important role in the development of paradoxical psoriasiform skin lesions with the overproduction of IFN being the major driving force.(14, 33) However, the interplay between IL-17 and IFN is not clear. In addition, which immune cells are activated by IFN and through which cytokines is still unknown. To help solve these questions, skin biopsies will be used to further explore the interplay between IFN and other cytokines (IL-17, IL-22 and IL-23) and cells (innate lymphoid cells, ILC3) in patients treated with anti-TNF and how this interplay changes when treatment is changed to anti-IL12/23 therapy or vedolizumab.

Inclusion Criteria:

• Patients with CD or UC, who develop psoriasiform skin lesions (including psoriasiform eczema, psoriasis guttata, psoriasis inversa and pustulosis) under therapy with anti-TNF and refractory to at least 12 weeks of topical therapy.
• Written informed consent must be obtained and documented.

Exclusion Criteria:

• IBD patients not treated with anti-TNF therapy
• IBD patients with skin lesions under anti-TNF not refractory to topical therapy
• Patients who previously received anti-interleukin 12/23 (IL12/23) or anti-interleukin 23 therapy
• Patients with history of psoriasis prior to initiation of anti-TNF therapy

No individual participant data (IPD) sharing