ARTHRITIS & RHEUMATISM
Vol. 43, No. 11, November 2000, pp 2391–2396
© 2000, American College of Rheumatology
SYSTEMIC ANTI–TUMOR NECROSIS FACTOR THERAPY
IN RHEUMATOID ARTHRITIS DOWN-REGULATES
SYNOVIAL TUMOR NECROSIS FACTOR SYNTHESIS
ANN-KRISTIN ULFGREN, ULF ANDERSSON, MARIANNE ENGSTRO
¨
M, LARS KLARESKOG,
RAVINDER N. MAINI, and PETER C. TAYLOR
Objective. To investigate the hypothesis that tu-
mor necrosis factor (TNF) blockade in rheumatoid
arthritis (RA) diminishes synovial synthesis of TNF,
interleukin-1 (IL-1), and IL-1.
Methods. Patients with active RA received a single
10 mg/kg infusion of infliximab. Multiple synovial bi-
opsy specimens were obtained from a knee the day
before infusion and 14 days later. A modified immuno-
histochemical method detecting cytokine-producing
rather than cytokine-binding cells was applied to deter-
mine synthesis of TNF, IL-1, and IL-1 in fixed,
cryopreserved sections. Computerized image analysis
using two different methodologies was performed by
independent observers blinded to the identity of samples.
Results. All 8 patients met the American College
of Rheumatology 20% improvement response criteria
(ACR 20) at 2 weeks, and half of these patients met the
ACR 50. With a few exceptions, there was concordance
between both image analysis methodologies regarding
the direction of change in immunopositive area fraction
for all cytokines analyzed. TNF synthesis was signifi-
cantly reduced after treatment (P 0.05 at the Karo-
linska Institute, Stockholm, Sweden; P 0.008 at the
Kennedy Institute, London, UK). Patients meeting the
ACR 50 were those with the highest baseline levels of
TNF synthesis. There was a significant correlation
between baseline levels of TNF expression and change
in TNF levels in response to therapy. Both IL-1 and
IL-1 synthesis were reduced in 3 patients; IL-1
synthesis alone was reduced in 2 patients and IL-1
synthesis alone was reduced in 2 patients. In 1 patient,
neither IL-1 nor IL-1 synthesis was reduced.
Conclusion. Analysis of synovial tissue by means
of immunomorphology and image analysis in a clinical
trial setting may allow the drawing of biologically mean-
ingful conclusions. Synovial TNF synthesis was re-
duced 2 weeks after infliximab treatment. Reductions in
IL-1 and IL-1 synthesis were demonstrated in a
subgroup of patients. High levels of synovial TNF
production prior to treatment may predict responsive-
ness to therapy.
Clinical trials of therapy for rheumatoid arthritis
(RA) based on tumor necrosis factor (TNF) neutral-
ization using TNF-specific antibodies or soluble recep-
tor fusion proteins demonstrate marked dose-dependent
improvement in both clinical and laboratory measures of
inflammation (1). Evidence to date indicates that there
are at least two major mechanisms of action. First, there
is diminished influx of inflammatory cells to the joints
(2,3). The reduced cell margination (4) and migration
are secondary to a down-regulation of both synovial
endothelial adhesiveness (2) and chemotactic gradient
(3). Second, there is deactivation of the proinflamma-
tory cytokine cascade (5), as demonstrated by reductions
in serum C-reactive protein and interleukin-6 (IL-6)
concentrations. By means of serum studies, however, it
has not been possible to test unequivocally the predic-
tion, based on ex vivo studies (6), that TNF blockade
will inhibit IL-1 and TNF production in vivo.
IL-1 is either undetectable or present in very low
concentrations in the peripheral blood of most patients.
The Kennedy Institute is supported by a core grant from the
Arthritis Research Campaign, UK, and holds a grant from Centocor
Inc. for the investigation of mechanisms of action of anti-TNF
therapy.
Ann-Kristin Ulfgren, PhD, Ulf Andersson, MD, Marianne
Engstro ¨m, BSc, Lars Klareskog, MD: Karolinska Hospital, Stockholm,
Sweden; Ravinder N. Maini, FRCP, Peter C. Taylor, MA, PhD, FRCP:
Kennedy Institute of Rheumatology, London, UK.
Address reprint requests to Ann-Kristin Ulfgren, PhD, Rheu-
matology Research Laboratory, CMM L8-004, Karolinska Hospital,
S-171 76 Stockholm, Sweden.
Submitted for publication December 31, 1999; accepted in
revised form June 27, 2000.
2391