Effect on Anti-TNF Agents on Gene Expression in Children with Juvenile Arthritis

[Guest guest]

Presentation Number:1713

Poster Board Number:327

Presentation Time:11/10/2007 8:00:00 AM

Title:Effect on Anti-TNF Agents on Gene Expression in Children with Juvenile

Arthritis

Category:12. Pediatric rheumatology pathogenesis and genetics

Author(s):Lakshmi Nandini Moorthy1, Gunter Schemmann2, K. Crow3,

Emmanuel Zachariah1, Margaret 3, Lehman3, Notterman1.

1 Wood Med Sch-Univ of Med and Dentistry of NJ, New Brunswick,

NJ; 2UMDNJ, New Brunswick, NJ; 3Hospital For Special Surgery, New York, NY

We hypothesize that gene expression patterns for childhood juvenile

idiopathic arthritis (JIA) and systemic lupus erythematosus (SLE) are

unique, will be altered by anti-cytokine/cytotoxic agents, and may predict

response to therapy.

Purpose: To examine the effect of etanercept in JIA and of

cyclophosphamide/rituximab combination (COMB) in SLE on PBMC gene expression

using microarray-based methods.

Methods: Paired blood samples were collected from 4 children with active JIA

[3 polyarticular (PoJIA), 1 systemic onset (SoJIA)] and one with active SLE

prior to and 90±30 days after initiation of therapy with etanercept for the

JIA patients and COMB therapy for the SLE patient. Paired blood samples from

2 age-matched controls (1 SoJIA, 1 post-streptococcal reactive arthritis)

with inactive disease on no therapy were also collected.

PBMCs were separated, total RNA isolated and microarray experiments were

conducted using Affymetrix chips (HGU133 plus 2). Pre and post treatment

data were compared. The t-test, fold change tests and change expression

analysis using Affymetrix MAS software were performed to compare the 2

groups.

Change analysis compares the pre and post treatment samples from the same

patient and is done on the level of probes (as opposed to probe-sets). In

addition, data from responders (2 of 3 PoJIA) and non-responders (1 PoJIA,

the SoJIA) and the SLE patient were compared.

Results: PoJIA, SoJIA, SLE and control subjects had distinct gene expression

signatures. Probes not scored as present in at least 3 of 4 JIA samples at

either time point were removed; 157 differentially expressed genes remained.

After etanercept, 80 genes were downregulated and 77 upregulated. Genes

involved in inflammatory pathways and known to be regulated by TNF (CEBPD,

SOCS3, and PBEF1) were decreased after treatment in 3 PoJIA patients,

consistent with functional inhibition of TNF in vivo.

UBE2V1, involved in the NFkB pathway; CD74, the MHC class II-associated

invariant chain; HSP90AB1, a regulator of PKR, a component of the type I

interferon pathway; and TLN1, a contributor to normal integrin function,

were increased after treatment.

Some genes over expressed in the SLE patient were also detected in the

non-responder with PoJIA (CALR, IFNG, STAT1) and the non-responder with

SoJIA (SLC16A3, MMP9, VSIG4, DEFA1 and 4, ARG1, CYP453, CEACAM8 and 6,

ANXA3, OLFM4). Many of these are components or targets of pro-inflammatory

cytokine pathways.

Conclusion: Our preliminary analysis suggests that gene expression is

modified by etanercept in JIA patients and separates responders from

non-responders. A lupus-like signature in JIA patients may predict lack of

response to anti-TNF therapy.

Disclosures: L.N. Moorthy, Arthritis Foundation Investigator Award

2007-2009; G. Schemmann, NIH support on colon cancer-related work

(unrelated), 2; UMDNJ, Princeton University in Mechanical Engineering Dept,

3; M.K. Crow, NIH, Alliance for Lupus Research, 2; Hospital For Special

Surgery, 3; E. Zachariah, None; M. , Co-investigator on about 3 NIH

grants, 2; Hospital For Special Surgery, 3; Dr. Moorthy's Pfizer clinical

scholars grant on QOL and SLE; Consultant on 2 NIH grants; Consultant Burke

Hospital, 5; Consultant on this project (no- fees), 9; T. Lehman, Hospital

For Special Surgery, 3; Genentech, Abbott, Wyeth, 8; D. Notterman, NIH (NCI)

support, 2.