AZD2315 in Patients w/ Active RA: Results of 3-Month Safety & Efficacy Study

[Guest guest]

Specific Blockade of Shared Epitope Antigen Presentation by the HLA-DRB1

Antagonist, AZD2315, in Patients with Active Rheumatoid Arthritis (RA):

Results of a 3-Month Double-blind, Placebo-controlled, Safety and

Efficacy Study.

Background: HLA-DRB1 alleles with QKRAA, QRRAA and RRRAA (shared

epitope, SE) at codons 70–71 are associated with increased prevalence,

persistence, severity and adverse outcome in RA.

A leading hypothesis involves presentation of an arthritogenic

peptide(s). We evaluated the safety and efficacy of AZD2315, a high

affinity peptide based ligand which blocks the RA-associated HLA-DRB1

molecules.

Methods: 88 patients (aged 23-81yrs) with active RA (>12 tender & >10

swollen joints with either: ESR >28 mm/hr, CRP >1.5 mg/dl or >45 min.

morning stiffness), who had failed 1 to 4 DMARDs entered this double

blind, placebo-controlled, parallel group study.

In Cohort I (safety), patients received 3-months treatment with AZD2315,

0.75 mg/day [n=12] or placebo [n=3] and in Cohort II 4.2 mg/day [n=37]

or placebo [n=36]. Study drugs were administered by continuous

subcutaneous infusion via a mini-pump (MedipadTM). The proportion of

patients achieving ACR20 responses in Cohort II was the primary outcome

variable.

Results: In Cohort II, mean AZD2315 serum levels were consistent with

those shown to block HLA-DRB1*0401-restricted T cell responses in

pre-clinical studies, and 70% had at least one SE allele. The primary

efficacy analysis was performed in 55 patients (31, AZD2315 vs 24,

placebo) which included completing patients (23 vs 20), treatment

failures (4 vs 6) and patients discontinuing due to AEs (4 vs 0) but

excluded patients discontinuing for other reasons (6 vs 10) and 2

placebo patients not meeting eligibility criteria.

Safety discontinuations with AZD2315 related to application site

reaction (n=2), nausea and vomiting (n=1) and rash (n=1). ACR 20, 50 and

70 responses were observed with AZD2315 in 2, 3 and 1 patients (ACR >20

= 6/31, 19%) compared to 4, 1 and 0 with placebo (ACR >20 = 5/24, 21%).

Detailed analysis of response according to SE zygosity, binding of

AZD2315 to each subjects’ specific HLA-DRB1 molecule, and other baseline

clinical and laboratory variables did not reveal significant responses

in particular subsets.

Conclusion: Over a 3-month period, AZD2315 was well tolerated but

response rates were not different from placebo. AZD2315 serum levels

obtained were consistent with those required to block

HLA-DRB1-restricted T cell responses ex vivo. However, the levels and

pharmacodynamic actions of AZD2315 within the synovial compartment were

not determined and may have been inadequate to fully block pathogenic

antigen presentation. These findings mitigate against, but do not

invalidate, a role for HLA-DRB1 SE mediated interactions in established RA.

E. Furst1, Roy Fleischmann2, Arthur Kavanaugh3, Ruderman4,

E. Beatty5, Tom Uryniak5, Gibson6, Mike Heath6, Lachy

McLean6, J. Culbert6. 1UCLA Medical Center, Los Angeles, CA; 2UT

Southwestern, Dallas, TX; 3UCSD, La Jolla, CA; 4Northwestern University,

Chicago, IL; 5AstraZeneca, Wilmington, DE; 6AstraZeneca, Alderley Park,

United Kingdom