Articles

Effect of interleukin-6 receptor inhibition with tocilizumab

in patients with rheumatoid arthritis (OPTION study):
a double-blind, placebo-controlled, randomised trial
Josef S Smolen, Andre Beaulieu, Andrea Rubbert-Roth, Cesar Ramos-Remus, Josef Rovensky, Emma Alecock, Thasia Woodworth, Rieke Alten, for
the OPTION Investigators*

Summary
Background Interleukin 6 is involved in the pathogenesis of rheumatoid arthritis via its broad effects on immune and Lancet 2008; 371: 987–97
inflammatory responses. Our aim was to assess the therapeutic effects of blocking interleukin 6 by inhibition of the See Comment page 961
interleukin-6 receptor with tocilizumab in patients with rheumatoid arthritis. See Articles page 998
*Listed at end of paper
Methods In this double-blind, randomised, placebo-controlled, parallel group phase III study, 623 patients with Division of Rheumatology,
moderate to severe active rheumatoid arthritis were randomly assigned with an interactive voice response system, Department of Internal
stratified by site with a randomisation list provided by the study sponsor, to receive tocilizumab 8 mg/kg (n=205), Medicine III, Medical University
of Vienna, Vienna, Austria
tocilizumab 4 mg/kg (214), or placebo (204) intravenously every 4 weeks, with methotrexate at stable pre-study doses (Prof J S Smolen MD); 2nd
(10–25 mg/week). Rescue therapy with tocilizumab 8 mg/kg was offered at week 16 to patients with less than 20% Department of Medicine,
improvement in both swollen and tender joint counts. The primary endpoint was the proportion of patients with 20% Hietzing Hospital, Vienna,
improvement in signs and symptoms of rheumatoid arthritis according to American College of Rheumatology criteria Austria (J S Smolen); Faculty of
Medicine, Laval University,
(ACR20 response) at week 24. Analyses were by intention to treat. This trial is registered with ClinicalTrials.gov, Laval, Quebec, Canada
number NCT00106548. (A Beaulieu MD); Medical
Clinic I, University of Cologne,
Findings The intention-to-treat analysis population consisted of 622 patients: one patient in the 4 mg/kg group did not Cologne, Germany
(A Rubbert-Roth MD); Research
receive study treatment and was thus excluded. At 24 weeks, ACR20 responses were seen in more patients receiving Unit for Chronic Diseases,
tocilizumab than in those receiving placebo (120 [59%] patients in the 8 mg/kg group, 102 [48%] in the 4 mg/kg Guadalajara, Mexico
group, 54 [26%] in the placebo group; odds ratio 4·0 [95% CI 2·6–6·1], p<0·0001 for 8 mg/kg vs placebo; (Prof C Ramos-Remus MD);
and 2·6 [1·7–3·9], p<0·0001 for 4 mg/kg vs placebo). More people receiving tocilizumab than those receiving placebo National Institute of
Rheumatic Diseases, Piestany,
had at least one adverse event (143 [69%] in the 8 mg/kg group; 151 [71%] in the 4 mg/kg group; 129 [63%] in the Slovakia (Prof J Rovensky MD);
placebo group). The most common serious adverse events were serious infections or infestations, reported by six Roche Products Ltd, Welwyn,
patients in the 8 mg/kg group, three in the 4 mg/kg group, and two in the placebo group. UK (E Alecock MSc,
T Woodworth MD); and
Department of Internal
Interpretation Tocilizumab could be an effective therapeutic approach in patients with moderate to severe active Medicine II, Rheumatology,
rheumatoid arthritis. Schlosspark Clinic, University
Medicine Berlin, Berlin,
Germany (R Alten MD)
Funding F Hoffmann-La Roche, Chugai Pharmaceutical.
Correspondence to:
Prof Josef S Smolen, Division of
Introduction rheumatoid arthritis, with raised concentrations in Rheumatology, Department of
Rheumatoid arthritis is a systemic autoimmune inflam- serum and synovial fluid.9,10 Interleukin 6 affects the Internal Medicine III, Medical
matory disease associated with progressive joint damage, function of neutrophils, T cells, B cells, monocytes, and University of Vienna, Vienna,
pain, fatigue, and disability. Although the exact cause of osteoclasts—cells that are highly activated in rheumatoid Austria
josef.smolen@meduniwien.
rheumatoid arthritis is still unknown, insights into its arthritis—and is the major inducer of the hepatic acute ac.at
pathogenesis have confirmed the role of proinflammatory phase response,11 which is also a key feature of rheumatoid
cytokines—eg, tumour necrosis factor α (TNFα), arthritis that is correlated with disease activity and joint
interleukin 1, and interleukin 6—in disease pathways.1,2 destruction.12,13 The effects of interleukin 6 are mediated
Current treatments target the inflammatory system with by binding to the interleukin-6 receptor (CD126, IL6Rα
disease modifying antirheumatic drugs (DMARDs; eg, chain), which is expressed on cell surfaces and as a
methotrexate) or biological agents. The available biological circulating soluble form.
agents inhibit the action of cytokines (eg, TNFα or Thus, targeting interleukin 6 is an attractive therapeutic
interleukin 1) or limit B-cell function or T-cell costimulation, option in rheumatoid arthritis. Tocilizumab, a humanised
especially in combination with methotrexate.1,3–6 Despite monoclonal antibody that binds to both forms of the
their efficacy, none of these agents leads to response in all interleukin-6 receptor, showed clinical efficacy in a
patients, and even among responders the improvement is Japanese phase II study in rheumatoid arthritis.14 The
often limited.1,7,8 results of the 16-week, phase II, double-blind, multicentre,
Interleukin 6 is a pleiotropic cytokine that is European dose-ranging study CHARISMA15 confirmed
over-expressed in synovial tissue in patients with the Japanese monotherapy data. In view of these

www.thelancet.com Vol 371 March 22, 2008 987

 Articles

prednisone or equivalent) and non-steroidal

812 patients screened anti-inflammatory drugs (NSAIDs) were permitted if
doses were stable for 6 weeks or more before inclusion.
189 patients excluded Main exclusion criteria were other autoimmune
diseases or significant systemic involvement secondary
to rheumatoid arthritis (eg, vasculitis, pulmonary fibrosis,
623 patients enrolled or Felty’s syndrome), functional class IV rheumatoid
arthritis, previous or current inflammatory joint disease
other than rheumatoid arthritis, currently active or
previous recurrent bacterial, viral, fungal, or other
204 patients randomly 214 patients randomly 205 patients randomly infections including, but not limited to, tuberculosis and
assigned to receive assigned to receive assigned to receive atypical mycobacterial disease, clinically significant
placebo tocilizumab 4 mg/kg* tocilizumab 8 mg/kg
abnormalities on chest radiograph, hepatitis B and C,
and recurrent herpes zoster. Investigators were
12 withdrew 25 withdrew† 13 withdrew encouraged to exclude potentially eligible patients if, in
6 safety 14 safety 12 safety
6 non-safety 11 non-safety 1 non-safety
their judgment, they had a history of unacceptably
frequent recurrent infections.
Patients were also excluded if they had active liver
68 on rescue 31 on rescue 19 on rescue
disease, indicated by screening and baseline
therapy‡ therapy‡ therapy‡
concentrations of alanine or aspartate aminotransferase
of 1·5 times the upper limit of normal or more, or
3 withdrew 3 withdrew 1 withdrew previous unsuccessful treatment with an anti-TNF agent
2 safety 2 safety 1 non-safety
1 non-safety 1 non-safety (ie, lack of efficacy or significant safety issues;
terminations due to cost or injection discomfort were not
excluded).
189 completed the study 186 completed the study 191 completed the study
124 on placebo 158 on tocilizumab 4 mg/kg 173 on tocilizumab 8 mg/kg
All participants gave written informed consent before
65 on rescue therapy 28 on rescue therapy 18 on rescue therapy screening. The study was approved by the ethics
committees for each site, and complied with the
Figure 1: Trial profile principles of Good Clinical Practice and the Declaration
*Includes one patient who was randomised to tocilizumab 4 mg/kg but received tocilizumab 8 mg/kg throughout of Helsinki.
the study. †Includes one patient who did not receive study treatment and was subsequently withdrawn. ‡Patients
who did not achieve >20% improvement from baseline in both tender and swollen joint count at week 16 were
offered rescue therapy.
Procedures
This phase III, three arm, randomised, double-blind,
promising early phase data, we undertook the tOcilizumab placebo-controlled, parallel group, international study
Pivotal Trial in methotrexate Inadequate respONders was done in 73 centres in 17 countries; patients were
(OPTION) to assess the efficacy of tocilizumab in patients enrolled and treated between Feb 16, 2005, and
with active rheumatoid arthritis who were receiving Nov, 13, 2006. On the basis of the CHARISMA study,15
background methotrexate therapy. patients were randomly assigned to receive placebo,
tocilizumab 4 mg/kg, or tocilizumab 8 mg/kg
Methods intravenously at baseline and thereafter every 4 weeks for
Patients 24 weeks in combination with weekly administration of
Adult patients with moderate to severe active rheumatoid their stable dose of methotrexate (10–25 mg).
arthritis (diagnosed according to American College of Randomisation was done centrally with an interactive
Rheumatology [ACR] criteria16) of more than 6 months’ voice response system, stratified by site with a
duration who had an inadequate response to methotrexate randomisation list provided by Roche. To minimise
were recruited. Active disease was defined by a swollen methotrexate-related toxicity, all patients received a stable
joint count of 6 or more plus a tender joint count of 8 or dose of folic acid (≥5 mg/week).
more and C-reactive protein (CRP) over 10 mg/L or ESR Patients were scheduled for routine clinic visits at
of 28 mm/h or more. To be eligible, patients had to have weeks 2, 4, 8, 12, 16, 20, and 24 for efficacy-related
received methotrexate for 12 weeks or longer before the assessments; safety parameters and change in
start of the study (stable dose of 10–25 mg/week for medications were assessed at weeks 2, 4, 6, 8, 12, 14, 16,
8 weeks or longer). All other DMARDs were discontinued 20, 24, 28, and 32. To maintain the double-blind status, a
before the start of the study: leflunomide for 12 weeks or dual assessor approach for efficacy and safety
more (or ≥4 weeks after 11 days of standard colestyramine assessments was used. Swollen and tender joint counts
washout), anakinra for 1 week or more, etanercept for based on 66/68 joints were done by trained assessors
2 weeks or longer, and infliximab or adalimumab for who had no access to patient data. A physician blinded
8 weeks or longer. Oral glucocorticoids (≤10 mg/day to patient’s treatment made all treatment decisions on

988 www.thelancet.com Vol 371 March 22, 2008

 Articles

the basis of the patient’s clinical response and safety

Tocilizumab 4 mg/kg Tocilizumab 8 mg/kg Placebo
data. Patients who had not achieved at least 20% (n=213) (n=205) (n=204)
improvement in both swollen joint count and tender
Demographic data
joint count by week 16 were eligible for rescue therapy
Age (years) 51·4 (12·8) 50·8 (11·8) 50·6 (12·1)
with tocilizumab 8 mg/kg and, if necessary,
Sex (male) 38 (18%) 30 (15%) 45 (22%)
intra-articular steroids or an increase in oral
Weight (kg) 69·9 (17·5) 68·0 (15·6) 71·6 (17·0)
corticosteroid dose (maximum 10 mg/day).
Disease duration (years) 7·4 (7·4) 7·5 (7·3) 7·8 (7·2)
Complete blood counts and hepatic function tests were
Methotrexate dose (mg/week) 14·7 (4·3) 14·5 (4·4) 14·8 (4·2)
monitored at each visit and study treatment was stopped
Number of DMARDs used before study* 1·5 (1·4) 1·5 (1·4) 1·7 (1·5)
temporarily in patients with increases in alanine or
Concomitant steroids 117 (55%) 112 (55%) 111 (54%)
aspartate aminotransferase concentrations of three times
the upper limit of normal or more. Patients with such Previous anti-TNF treatment 21 (10%) 11 (5%) 19 (9%)

increases on two consecutive visits, an increase in alanine Concomitant NSAIDs 145 (68%) 135 (66%) 139 (68%)

or aspartate aminotransferase concentrations of more than RF positive (RF≥15 U/mL) 167 (78%) 171 (83%) 144 (71%)
five times the upper limit of normal, or a neutrophil count Laboratory tests
below 0·5×10⁹ cells per L were discontinued from study RF concentration (U/mL) 227 (370) 245 (938) 210 (328)
treatment permanently. Patients with infections could CRP (mg/L) 28 (34) 26 (26) 24 (28)
temporarily interrupt treatment for up to two infusions, if ESR (mm/h) 49·2 (26·8) 51·2 (26·6) 49·7 (26·3)
deemed necessary by the investigator. Lipid concentrations Haemoglobin (g/L) 136 (17) 136 (16) 136 (16)
were assessed in a fasted state. Transient changes in Serum amyloid A (μg/mL) 69·2 (90·7) 70·4 (89·1) 64·4 (85·2)
laboratory abnormalities were those that self-corrected Total cholesterol (mmol/L) 5·08 (0·97) 5·11 (1·19) 4·98 (0·95)
between two dose administrations without treatment. To HDL cholesterol (mmol/L) 1·49 (0·42) 1·47 (0·41) 1·44 (0·40)
maintain the double-blind status, acute phase reactant data LDL cholesterol (mmol/L) 3·02 (0·81) 2·93 (0·99) 2·86 (0·78)
from the central laboratory were blinded to site and Disease activity and quality of life
sponsor personnel after the start of the study. DAS28 score 6·8 (0·9) 6·8 (0·9) 6·8 (0·9)
The primary efficacy endpoint was the proportion of Swollen joint count 20·0 (10·9) 19·5 (11·3) 20·7 (11·7)
patients with a 20% improvement in rheumatoid arthritis Tender joint count 33·2 (15·6) 31·9 (15·5) 32·8 (16·1)
signs and symptoms according to ACR criteria (ACR20 Patient pain VAS† (mm) 60·7 (21·0) 59·9 (22·4) 57·3 (22·2)
response)17 at 24 weeks. Secondary efficacy endpoints Patient global VAS (mm) 65·6 (20·8) 64·8 (22·1) 63·6 ( 21·8)
included the proportion of patients with an ACR50 and Physician global VAS (mm) 63·6 (15·8) 64·0 (15·3) 63·7 (14·8)
ACR70 response at 24 weeks, change from baseline in HAQ-DI score 1·6 (0·6) 1·6 (0·6) 1·5 (0·6)
disease activity score using 28 joint counts (DAS28)18,19 at FACIT-Fatigue score 27·0 (11·5) 27·7 (10·6) 26·7 (11·1)
24 weeks, the proportion of patients in DAS28 remission SF36 score (physical) 31·5 (7·5) 32·1 (7·0) 32·3 (7·0)
(DAS28 <2·6) at 24 weeks, and categorical DAS28 SF36 score (mental) 40·1 (11·8) 40·9 (10·6) 39·1 (11·0)
(European League Against Rheumatoid Arthritis
[EULAR]) response by 24 weeks.20 Haemoglobin Data are mean (SD) or n (%). RF=rheumatoid factor. *Not counting previous use of methotrexate. †VAS scales were 0 to
100 mm.
concentrations were also assessed. Patient’s pain and
patient’s or physician’s global assessment of disease Table 1: Characteristics at baseline
activity was assessed on a 0 to 100 mm horizontal visual
analogue scale (VAS), with 0 mm corresponding to no (higher scores indicating better health). The
pain or no disease activity and 100 mm corresponding to FACIT-Fatigue scale ranges from 0 to 52 and has been
unbearable pain or maximum disease activity. The Health validated in the general population, patients with cancer,23
Assessment Questionnaire-Disability Index (HAQ-DI) and patients with rheumatoid arthritis;24 higher scores
was used for assessment of physical function.21 HAQ-DI represent less fatigue.
assesses the ability to do daily activities by use of
20 questions in eight domains; the final HAQ score is Statistical analysis
the mean of the highest scores across the eight domains A sample size of 210 patients per arm (630 patients) was
and ranges from 0 to 3, with higher levels reflecting calculated to provide 90% power to detect a difference
greater disability. Medical Outcomes Study 36-Item between tocilizumab and placebo (two-sided test,
Short-Form General Health Survey (SF36),22 completed corrected for multiple comparisons), assuming ACR20
at baseline and weeks 8, 16, and 24, and Functional responses of 60% with the study drug versus 40% with
Assesment of Chronic Illness Therapy (FACIT)-Fatigue placebo, allowing for a 15% dropout rate.
assessment,23,24 done at baseline and every 4 weeks, were To control for false positive conclusions for the primary
used to assess health-related quality-of-life. SF36 consists endpoint, the tocilizumab 8 mg/kg arm was first
of 36 items and eight scales, from which two summary compared with the placebo arm, and a p value was
scores can be extracted: the physical component and the derived. Only if this comparison gave a p value of 0·05 or
mental component scores. The scores range from 0 to 100 less was a comparison of the tocilizumab 4 mg/kg with

www.thelancet.com Vol 371 March 22, 2008 989

 Articles

forward (LOCF) was used for tender and swollen joint

Tocilizumab 4 mg/kg Tocilizumab 8 mg/kg Placebo (n=204)
(n=213) (n=205) counts; no imputation was used for missing HAQ, CRP,
ESR, and global VAS data. Patients completing the
ACR20
6-month trial were allowed to transfer to an open-label
Number of patients 102 (48%) 120 (59%) 54 (26%)
extension trial for assessment of long-term safety and
p value vs placebo* p<0·0001 p<0·0001 NA
efficacy durability.
Odds ratio (vs placebo)† 2·6 (1·7–3·9) 4·0 (2·6–6·1) NA
Changes in FACIT-Fatigue, SF36 questionnaire
ACR50
scores, and haemoglobin concentrations were
Number of patients 67 (31%) 90 (44%) 22 (11%)
summarised descriptively and compared between the
p value vs placebo* p<0·0001 p<0·0001 NA
treatment groups with an ANOVA model with site
Odds ratio (vs placebo)† 3·8 (2·3–6·5) 6·6 (3·9–11·2) NA included in the model. HAQ-DI scores were compared
ACR70 with respect to changes from baseline and proportions
Number of patients 26 (12%) 45 (22%) 4 (2%) of patients achieving clinically relevant benefit
p value vs placebo* p<0·0001 p<0·0001 NA exceeding the minimal clinically important difference,25
Odds ratio (vs placebo)† 7·0 (2·4–20·4) 14·2 (5·0–40·4) NA as reflected by improvement from baseline of 0·3 or
DAS28 <2·6 more.21 Although partly incorrect translations in some
Number of patients 21/156 (13%) 47/171 (27%) 1/121 (0·8%) patient questionnaires led to exclusion of HAQ-DI data
p value vs placebo* p=0·0002 p<0·0001 NA for 105 patients (35 patients in each group), and
Odds ratio (vs placebo)† 18·8 (2·5–142·2) 45·0 (6·1–331·6) NA SF36 data for 32 patients (12, 11, and nine patients in
EULAR the placebo, 4 mg/kg, and 8 mg/kg groups, respectively)
Good response 45 (21%) 78 (38%) 6 (3%) the overall result was not affected significantly (data not
Moderate response 87 (41%) 85 (41%) 65 (32%) shown).
No response 81 (38%) 42 (20%) 133 (65%) The safety analysis was unadjusted for multiplicity.
p value vs placebo* p<0·0001 p<0·0001 NA χ² or Fisher’s exact tests were used where appropriate;
Odds ratio (vs placebo)†‡ 3·1 (2·1–4·7) 7·6 (4·9–12·0) NA reported 95% CI are asymptotic 95% CI. Continuous
endpoints are presented with the adjusted means, unless
*Calculated with Cochran-Mantel-Haenszel χ2 test with adjustment for site. †Calculated with logistic regression,
including region in the model (as algorithm fails to converge using site). ‡Good and moderate EULAR response
otherwise stated.
compared with no response. SAS version 8.2 was used for all analyses. This trial is
registered with ClinicalTrials.gov, number NCT00106548,
Table 2: Clinical response to treatment
and with EudraCT, number 2004/00374/40.

the placebo arm made. For the secondary endpoints, a
prespecified fixed sequence approach was applied,
allowing us to test each of the null hypotheses at the
same significance level of α without any adjustment.
ACR20 response at 24 weeks was compared between
treatment groups with a Cochran-Mantel-Haenszel χ² test
with adjustment for site. The same method was used for
ACR50, ACR70, and EULAR response analyses. An
analysis of variance (ANOVA) model was used to compare
mean change from baseline in each of the individual
variables of the ACR core set and in DAS28. A secondary
efficacy analysis to show the effect size for ACR response,
DAS remission, and EULAR response was done with
logistic regression, including site if the model converged,
or region if not.
The primary efficacy analysis was done on the
intention-to-treat population (ITT)—ie, all patients
randomised who received at least one infusion of study
drug. The safety population included all randomised
patients who received at least one infusion of study
medication and who had at least one assessment of safety
after randomisation. Patients who withdrew before
week 24, patients who received rescue therapy, and
patients whose week 24 categorical endpoints could not
be determined due to insufficient data were deemed to be
non-responders in the analysis. Last observation carried
Role of the funding source
F Hoffmann-La Roche funded the study and was
responsible for developing the protocol in collaboration
with the corresponding author, for all aspects of study
conduct, including distribution of study medication, data
processing and management, statistical analysis, and
reporting of results. As a co-development partner, Chugai
Pharmaceutical was involved in aspects of protocol
development and supplied the study medication. JSS,
EA, and TW had full access to all the data. The
corresponding author had final responsibility for the
decision to submit for publication.
Results
The trial profile is shown in figure 1. One patient in the
tocilizumab 4 mg/kg group did not receive study
treatment and was withdrawn; therefore, 622 patients
were included in the ITT population. Baseline patient
characteristics were much the same in all three groups
(table 1). There were no important differences between
the groups with respect to comorbidities (data not
shown). Most patients had begun to take folic acid before
the start of the study; only nine (4%) patients in the
4 mg/kg group, ten (5%) in the 8 mg/kg group, and eight
(4%) in the placebo group began folic acid supplemen-
tation when they entered the study.

990 www.thelancet.com Vol 371 March 22, 2008

 Articles

More patients in the 4 mg/kg group than in the
8 mg/kg group or the placebo group withdrew from the
study prematurely (figure 1). Major reasons for
withdrawal were adverse events (14 patients in the
4 mg/kg group, 12 in the 8 mg/kg group, and six in the
placebo group), insufficient response (two patients in
the 4 mg/kg group, none in the 8 mg/kg group, and
three in the placebo group), and refusal of treatment
(six patients in the 4 mg/kg group, one in the 8 mg/kg
group, and two in the placebo group). More patients in
the placebo group than in the 4 mg/kg group switched
to rescue therapy; likewise, more patients in the 4 mg/kg
group switched to rescue therapy than did those in the
8 mg/kg group (figure 1). The proportion of patients
who completed the study, including those who switched
to rescue therapy, was much the same in all three groups
(figure 1).
By week 24, significantly more patients receiving
tocilizumab had an ACR20 response than did those
receiving placebo (table 2; p<0·0001 for both individual
tocilizumab groups vs placebo). Significantly greater
numbers of patients receiving tocilizumab showed
ACR50 and ACR70 responses at week 24 than did those
receiving placebo (table 2; p<0·0001 for both individual
tocilizumab groups vs placebo for both ACR50 and
ACR70). Numerical differences between the placebo
and tocilizumab 8 mg/kg group were observed at week 2
for ACR20, at week 4 for ACR50, and at week 8 for
ACR70 (figure 2).
By week 24, significantly better responses in all core
set variables, whether physician, patient, or laboratory
derived, were seen with both doses of tocilizumab than
with placebo (table 3). DAS28 decreased rapidly with
tocilizumab therapy, with the mean improvement
corresponding to at least moderate EULAR response at
the first assessed time point of 2 weeks (figure 2);
DAS28 continued to improve over time. DAS28
remission at 24 weeks was seen in significantly more
patients in the tocilizumab groups than in the placebo
group (table 2).
Mean CRP concentrations normalised by week 2 of
treatment with tocilizumab 8 mg/kg (figure 2) and
remained below the upper limit of normal until the end of
the study (p<0·0001 vs baseline). By contrast, the

ACR50 tocilizumab 8 mg/kg

Tocilizumab 4 mg/kg ACR70 tocilizumab 8 mg/kg
Tocilizumab 8 mg/kg ACR50 placebo
A B 80
80 Placebo ACR70 placebo

60 60
Patients (%)

Patients (%)

40 40

20 20

0 0

C Tocilizumab 4 mg/kg D Tocilizumab 4 mg/kg

8 Tocilizumab 8 mg/kg 40 Tocilizumab 8 mg/kg
Placebo Placebo
7
Mean CRP levels (mg/L)
Mean DAS28 scores

30
6

5 20
4
10
3

2 0

E Tocilizumab 4 mg/kg F Tocilizumab 4 mg/kg

1·8 Tocilizumab 8 mg/kg 16 Tocilizumab 8 mg/kg
FACIT-Fatigue score by visit

Placebo 14 Placebo
HAQ-DI score by visit

1·6
12
1·4 10
8
1·2 6
4
1
2
0·8 0
0 4 8 12 16 20 24 0 4 8 12 16 20 24
Time (weeks) Time (weeks)

Figure 2: Change from baseline in disease signs and symptoms over time
(A) ACR20 response; (B) ACR50 and ACR70 responses; (C) DAS28 scores; (D) CRP concentration (upper limit of normal shown by broken line); (E) HAQ-DI scores;
(F) FACIT-Fatigue score. Data are mean (SE).

www.thelancet.com Vol 371 March 22, 2008 991

 Articles

improvement with tocilizumab 4 mg/kg was less striking differences observed from week 2 onward with both doses
and CRP concentrations fluctuated over the dosing interval (data not shown). Mean levels of complement proteins
(table 3 and figure 2). ESR normalised with tocilizumab C3 and C4 decreased from baseline in the tocilizumab
8 mg/kg but not with the 4 mg/kg dose, in much the same groups, but remained within the normal range in all
way as CRP (data not shown). By week 24, serum amyloid groups (data not shown).
A concentrations had fallen more from baseline levels with Mean haemoglobin concentrations increased from
both doses of tocilizumab than with placebo (table 4), with baseline by 6–7 g/L by 4 weeks in both tocilizumab
groups, but not in the placebo group, and continued to
Tocilizumab 4 mg/kg Tocilizumab 8 mg/kg Placebo
increase until week 24 (table 4). By contrast, mean
haemoglobin concentrations in the placebo group did
Swollen joint count –8·5 –10·5 –4·3
not change during the study (data not shown).
Difference from placebo group –4·2 (–6·1 to –2·3) –6·2 (–8·1 to –4·2) NA
p<0·0001 p<0·0001
Rheumatoid factor (RF) concentration had decreased
from baseline more in both tocilizumab groups than in
Tender joint count –14·5 –17·1 –7·4
the placebo group by week 24 (table 4). IgG and IgM
Difference from placebo group –7·0 (–10·0 to –4·1) –9·6 (–12·6 to –6·7) NA
p<0·0001 p<0·0001 serum levels were stable in all three groups at all time
Patient pain VAS (mm) –25·0 –29·8 –14·0 points (data not shown).
Difference from placebo group –11·0 (–17·0 to –5·0) –15·8 (–21·7 to –9·9) NA There were greater improvements in physical function,
p=0·0004 p<0·0001 as judged by mean difference in HAQ-DI score from
Patient global VAS (mm) –28·8 –32·7 –17·8 baseline, with both doses of tocilizumab than with
Difference from placebo group –10·9 (–17·1 to –4·8) –14·9 (–20·9 to –8·9) NA placebo (table 3). A clear difference between the
p=0·0005 p<0·0001 two tocilizumab groups from placebo was noted by
Physician global VAS (mm) –38·3 –41·6 –32·7 week 4 (figure 2). An increase from baseline in HAQ-DI
Difference from placebo group –5·6 (–10·5 to –0·8) –9·0 (–13·8 to –4·2) NA of 0·3 points or more was achieved by 77 (61%) patients
p=0·0229 p=0·0002 in the 4 mg/kg group, 83 (59%) in the 8 mg/kg group,
CRP (mg/L) –16·6 –25·1 –3·5 and 47 (47%) in the placebo group at week 24. There were
Difference from placebo group –13·0 (–20·1 to –5·9) –21·6 (–28·6 to –14·6) NA greater improvements from baseline in FACIT-Fatigue
p=0·0004 p<0·0001
scores over time with both doses of tocilizumab than
ESR (mm/h) –25·5 –39·5 –7·1
with placebo (figure 2 and table 4). By week 24, there were
Difference from placebo group –18·3 (–24·3 to –12·4) –32·3 (–38·2 to –26·5) NA
p<0·0001 p<0·0001
greater improvements from baseline in both physical
HAQ –0·52 –0·55 –0·34
and mental SF36 scores with both doses of tocilizumab
Difference from placebo group –0·18 (–0·34 to –0·02) –0·21 (–0·37 to –0·05) NA
than with placebo (table 4).
p=0·0296 p=0·0082 More patients receiving tocilizumab reported at least
one adverse event than did those receiving placebo.
Data are mean change from baseline, or mean difference from placebo (95% CI).
Adverse events with a cumulative incidence by week 24
Table 3: Mean changes in core variables from baseline at week 24 (adjusted means) of 5% or more are summarised in table 5. More infections
were seen in the tocilizumab groups than in the placebo
group (table 5); the rate of all infections was 98·7 per
Tocilizumab 4 mg/kg Tocilizumab 8 mg/kg Placebo 100 patient-years of treatment in the 4 mg/kg group,
Serum amyloid A concentration (µg/mL)* –43·2 –64·3 –2·4
101·9 per 100 patient-years in the 8 mg/kg group, and
96·1 per 100 patient-years in the placebo group.
Haemoglobin concentration (g/L) 9·2 12·4 –0·3
Cutaneous adverse events were more frequent during
Difference from placebo group 9·5 (6·3 to 12·7) 12·7 (9·6 to 15·9) NA
p<0·0001 p<0·0001 treatment with tocilizumab than with placebo. Such
RF concentration (U/mL) –8·5 –65·1 17·1 events were characterised as mild, self-limited, and of
Difference from placebo group –25·6 (–116·8 to 65·6) –82·2 (–172·9 to 8·5) NA short duration and included localised rashes and
p=0·58 p=0·08 dermatitis with or without pruritus; most occurred only
FACIT-Fatigue score 7·3 8·6 4·0 once, and did not prevent further treatment. There was
Difference from placebo group 3·3 (0·9 to 5·6) 4·6 (2·3 to 6·9) NA no apparent association between the rashes and the few
p=0·0063 p<0·0001 hypersensitivity events recorded.
SF36 score (physical) 9·7 9·5 5·0 Abnormal laboratory results reported as adverse
Difference from placebo group 4·8 (2·5 to 7·0) 4·5 (2·3 to 6·7) NA events were more frequent in the tocilizumab groups
p<0·0001 p<0·0001 than in the placebo group; most such events were
SF36 score (mental) 5·7 7·3 2·7 transient increases in the concentrations of hepatic
Difference from placebo group 3·0 (0·1 to 5·8) 4·7 (1·8 to 7·5) NA aminotransferases, increases in lipid concentrations, or
p=0·0394 p=0·0012
transient decreases in absolute neutrophil counts.
RF=rheumatoid factor. *Exploratory analysis, difference from placebo not calculated. Notably, 12 (6%) patients in the 4 mg/kg group, 21 (10%)
in the 8 mg/kg, and eight (4%) in the placebo group had
Table 4: Mean change from baseline in other variables at week 24 (adjusted means)
increases in alanine aminotransferase concentrations

992 www.thelancet.com Vol 371 March 22, 2008

 Articles

of more than three times the upper limit of normal, and

Tocilizumab 4 mg/kg Tocilizumab 8 mg/kg Placebo
11 patients in the study exhibited concentrations of (n=212)* (n=206)* n=204
more than five times the upper limit of normal (three
At least one adverse event
in the 4 mg/kg group, seven in the 8 mg/kg group, and
At least one adverse event 151 (71%) 143 (69%) 129 (63%)
one in the placebo group). None of these patients had a
Difference from placebo group 8·0% (–1·0 to 17·0) 6·2% (–3·0 to 15·3) NA
concurrent increase in total bilirubin or alkaline
Adverse events occurring with a frequency of ≥5%†
phosphatase, nor any clinical signs of hepatitis or
Any infection or infestation 65 (31%) 66 (32%) 56 (27%)
hepatic dysfunction. In most cases, these increases
Difference from placebo group 3·2% (–5·5 to 11·9) 4·6% (–4·3 to 13·4) NA
declined or normalised spontaneously or after
Upper respiratory tract infection 12 (6%) 17 (8%) 13 (6%)
temporary interruption of study treatment, without
recurrent increases after resumption of treatment. One Nasopharyngitis 11 (5%) 12 (6%) 10 (5%)

patient in the 4 mg/kg group experienced an increase Any gastrointestinal disorder 46 (22%) 48 (23%) 44 (22%)

in alanine aminotransferase concentration of more Difference from placebo group 0·1% (–7·8 to 8·0) 1·7% (–6·3 to 9·8) NA
than five times the upper limit of normal only at Dyspepsia and abdominal pain 17 (8%) 20 (10%) 11 (5%)
week 24; the concentration fell to 2·3 times the upper Mouth ulcers 4 (2%) 5 (2%) 1 (0·5%)
limit of normal a week later. We found no evidence that Peptic ulcers 1 (0·5%) 0 (0%) 2 (2%)
the 5–7% of patients with raised hepatic aminotrans- Any musculoskeletal and connective 29 (14%) 25 (12%) 33 (16%)
tissue disorder
ferase concentrations before randomisation had a
Difference from placebo group –2·5% (–9·4 to 4·4) –4·0% (–10·8 to 2·7) NA
higher likelihood for abnormal liver functional tests
Aggravation of rheumatoid arthritis 7 (3%) 7 (3%) 13 (6%)
that affected treatment adherence during the study
(data not shown). Any nervous system disorder 25 (12%) 31 (15%) 26 (13%)

Mean plasma concentrations of total cholesterol, Difference from placebo group –1·0% (–7·3 to 5·4) 2·3% (–4·4 to 9·0) NA

HDL cholesterol, and LDL cholesterol were increased Headache 15 (7%) 14 (7%) 9 (4%)
from baseline in the tocilizumab groups at the first Any skin and subcutaneous disorder 27 (13%) 37 (18%) 15 (7%)
scheduled assessment at week 6 (table 6); by contrast, Difference from placebo group 5·4% (–0·4 to 11·1) 10·6 (4·3 to 17·0) NA
such levels were unchanged in the placebo group. Plasma Rash‡ 13 (6%) 11 (5%) 3 (1%)
cholesterol levels remained raised at weeks 14 and 24 in Any laboratory variable 26 (12%) 30 (15%) 11 (5%)
the two tocilizumab groups; these increases coincided Difference from placebo group 6·9% (1·5 to 12·3) 9·2% (3·4 to 14·9) NA
with moderate to large decreases in CRP (data not Raised alanine aminotransferase 12 (6%) 11 (5%) 3 (1%)
shown). An increase in total cholesterol to more than Raised aspartate aminotransferase 1 (0·5%) 2 (1%) 1 (0·5%)
6·2 mmol/L (an indication for intervention25) at the last Any respiratory, thoracic, or mediastinal 14 (7%) 16 (8%) 11 (5%)
observation was noted more often in the tocilizumab disorder)
groups than in the placebo group (55 [26%] patients in Difference from placebo group 1·2% (–3·4 to 5·8) 2·4% (–2·4 to 7·2) NA
the 4 mg/kg group; 43 [21%] in the 8 mg/kg group; seven Cough 7 (3%) 3 (1%) 3 (1%)
[3%] in the placebo group). Increases in the ratio of total Pharyngeal pain 2 (0·9%) 4 (2%) 3 (1%)
to HDL cholesterol of more than 30% above baseline Dyspnoea 3 (1%) 3 (1%) 0 (0%)
were recorded in 16 (8%) patients in the 4 mg/kg group, Any vascular disorder 17 (8%) 12 (6%) 10 (5%)
34 (17%) in the 8 mg/kg group, and ten (5%) in the Difference from placebo group 3·1% (–1·6, 7·8) 0·9% (–3·4 to 5·3) NA
placebo group. By contrast, the number of patients with Hypertension 7 (3%) 8 (4%) 9 (4%)
increases of more than 30% above baseline in the ratio of Any general disorder and administration 19 (9%) 22 (11%) 14 (7%)
apolipoprotein B to A was comparable across treatment site conditions

groups (data not shown). There was no excess in major Difference from placebo group 2·1% (–3·1 to 7·3) 3·8% (–1·6 to 9·3) NA
adverse cardiovascular events in the tocilizumab groups Fatigue 3 (1%) 6 (3%) 5 (2%)
compared with the placebo group during the 24 weeks of Peripheral oedema 5 (2%) 4 (2%) 2 (1%)
this study (table 5). Seven patients began therapy with a Pyrexia 2 (0·9%) 1 (0·5%) 5 (2%)
lipid-lowering agent according to guidelines provided in (Continues on next page)
the protocol (one patient in the placebo group, three in
each tocilizumab group). This therapy led to decreases in
lipid levels, without increases in hepatic aminotransferases decreases in neutrophil counts below the lower limit of
(data not shown). normal were seen more often in patients on tocilizumab
As expected with the active inflammation typical of than in those on placebo (37 patients in the 4 mg/kg
rheumatoid arthritis, neutrophil counts at baseline were group, 67 in the 8 mg/kg group, and four in the placebo
raised, with mean values close to the upper limit of group). There was no apparent association between
normal (table 1). Mean neutrophil counts decreased to neutropenia (<1·0×10⁹ neutrophils per L) and the
within the normal range by week 2 after each infusion in occurrence or intensity of infections (data not shown).
both tocilizumab treatment groups; the effect was more Other adverse events occurred with comparable
stable in the 8 mg/kg group (data not shown). Transient frequency in all groups. Infusions were generally well

www.thelancet.com Vol 371 March 22, 2008 993

 Articles

group, 96 [47%] in the 8 mg/kg group, and 61 [30%] in

Tocilizumab 4 mg/kg Tocilizumab 8 mg/kg Placebo
(n=212)* (n=206)* (n=204) the placebo group).
The incidence of serious adverse events was much the
(Continued from previous page)
same in all three groups (table 5). Most occurred only
Serious adverse events§
once, and only 11 of the 41 serious adverse events led to
Patients with at least one serious 13 (6%) 13 (6%) 12 (6%)
discontinuation of treatment. 15 serious adverse events
adverse event
(in 14 patients) were deemed to be possibly or probably
Difference from placebo group 0·3% (–4·3 to 4·8) 0·4% (–4·2 to 5·1) NA
related to study treatment. Three of these occurred in
Patients with at least one serious adverse 5 (2%) 6 (3%) 3 (1%)
event related to study treatment patients in the placebo group: urinary tract infection
Difference from placebo group 0·9% (–1·7 to 3·5) 1·4% (–1·4 to 4·3) NA (week 19), dysphasia (week 24), and squamous cell
Total number of serious adverse events 13 14 14 carcinoma of the skin (week 9). Five events occurred in
Infections and infestations 3 (1%) 6 (3%) 2 (1%) patients receiving tocilizumab 4 mg/kg: pneumonia
Musculoskeletal 1 (0·5%) 2 (1%) 1 (0·5%)
(week 4), Pneumocystis jirovecii pneumonia (week 9),
Gastrointestinal 0 2 (1%) 1 (0·5%)
interstitial lung disease (week 9), leukaemoid reaction
Nervous system 1 (0·5%) 1 (0·5%) 1 (0·5%)
(week 8), and anaphylactic reaction (week 9). Seven events
occurred in six patients receiving tocilizumab 8 mg/kg:
Blood and lymphatic system 2 (0·9%) 0 0
cellulitis (week 8, in two patients), pneumonia (week 12),
Cardiac 0 1 (0·5%) 1 (0·5%)
peridiverticular abscess (week 1), upper respiratory tract
Ear and labyrinth disorders 0 0 2 (1%)
infection (week 15), cerebrovascular accident (week 20, in
General disorders and administration 0 0 2 (1%)
site conditions one of the patients with cellulitis), and idiopathic
Injury, poisoning, and procedural 1 (0·5%) 0 1 (0·5%) pulmonary fibrosis (week 13). The patient who developed
complications P jirovecii pneumonia had a history of chronic obstructive
Neoplasms 0 0 2 (1%) pulmonary disease. Treatment was discontinued and,
Reproductive 0 1 (0·5%) 1 (0·5%) with appropriate treatment for the pneumonia, the patient
Respiratory 1 (0·5%) 1 (0·5%) 0 recovered, with a normal chest radiograph within 11 days.
Vascular 2 (0·9%) 0 0 Study treatment was not resumed in this patient. The
Immune system 1 (0·5%) 0 0 patients with presumed interstitial lung disease and
Surgical and medical procedures 1 (0·5%) 0 0 idiopathic pulmonary fibrosis recovered rapidly on
appropriate therapy and one continued study treatment.
Data are n (%) or % difference from placebo (95% CI). *One patient was randomised to tocilizumab 4 mg/kg but 11 serious adverse events were serious infections
received tocilizumab 8 mg/kg throughout the study; data for this patient are presented in the 4 mg/kg group in
summaries of the ITT population and in the 8 mg/kg group in summaries of the safety population. †Cumulative (table 5). Eight of these were classified as possibly or
incidence by week 24 of most common adverse events and of serious adverse events; adverse events occurred probably related to study drug. The three unrelated
sporadically throughout the course of the study without apparent accumulation at any particular week. ‡Including events included ovarian abscess (week 10) in the placebo
rash, popular rash, urticaria, erythematous rash, maculopapular rash, pruritic rash, generalised rash, macular rash.
§SAEs were reported within 1–3 days of occurrence, in accordance with regulatory guidelines.
group, gastroenteritis (week 10) in the tocilizumab
4 mg/kg group, and empyema (week 14) in the
Table 5: Summary of adverse events (safety population) tocilizumab 8 mg/kg group. Overall, the rates of serious
infections were 3·05 per 100 patient-years in the 4 mg/kg
Tocilizumab Tocilizumab Placebo
group, 6·05 per 100 patient-years in the 8 mg/kg group,
4 mg/kg 8 mg/kg and 2·29 per 100 patient-years in the control group.
Total cholesterol (mmol/L) 0·9 (0·7) 0·9 (0·8) 0·0 (0·6)
The main adverse events that led to withdrawal from
the study were serious infections (one patient in the
HDL cholesterol (mmol/L) 0·1 (0·3) 0·1 (0·3) –0·01 (0·23)
4 mg/kg group, two in the 8 mg/kg group, and one in
LDL cholesterol (mmol/L) 0·5 (0·6) 0·6 (0·7) 0·03 (0·48)
the placebo group) and abnormal liver function tests
Data are mean change from baseline (SD). (six in the 4 mg/kg group, seven in the 8 mg/kg group,
and one in the placebo group). Of the 11 patients in the
Table 6: Mean change from baseline in plasma lipid concentrations
at week 6 tocilizumab groups who withdrew from study treatment
due to raised concentrations of alanine aminotransferase,
four had concentrations raised to three to five times the
tolerated, with minor incidences of nausea (one patient upper limit of normal and seven had increases of more
in each group), rash (one patient in tocilizumab 4 mg/kg than five times the upper limit of normal. One patient
group), or hypertension (three cases in the 4 mg/kg in the control group withdrew because of an increase in
group, one in the 8 mg/kg group, and one in the placebo alanine aminotransferase concentration more than five
group) occurring during or within 24 h of infusion. times the upper limit of normal. Two patients in the
Frequencies of adverse events over the entire study period tocilizumab 4 mg/kg group discontinued because of
are shown in table 5. Adverse events deemed to be related increases in total bilirubin concentration of 2·8 times
to study treatment were more frequent in the tocilizumab the upper limit of normal and 1·8 times the upper limit
groups than in the placebo group (90 [43%] in the 4 mg/kg of normal without concomitant increases in amino-

994 www.thelancet.com Vol 371 March 22, 2008


transferase concentrations. No patient developed signs
or symptoms suggestive of hepatitis.
Anti-tocilizumab antibodies were detected in five
patients: one in the 4 mg/kg group and four in the
8 mg/kg group. Two hypersensitivity reactions leading to
withdrawal were recorded, one in each tocilizumab
group.
Discussion
These data provide evidence that inhibition of
interleukin-6-mediated proinflammatory effects signifi-
cantly and rapidly improves the signs and symptoms of
rheumatoid arthritis. Thus, tocilizumab could be an
effective agent for the treatment of patients with moderate
to severe rheumatoid arthritis.
In this study, tocilizumab produced a marked
improvement from baseline in all ACR core set variables
and significantly more patients on the drug achieved
ACR20, ACR50, and ACR70 responses than did those in
the placebo group. About a quarter of patients treated
with tocilizumab 8 mg/kg—compared with less than 1%
on placebo—achieved DAS28 remission. Impairment of
physical function, as indicated by a mean baseline
HAQ-DI of about 1·6 across all treatment groups, was
improved by tocilizumab more than it was by placebo,
reflecting substantial functional benefits for the patients.
Furthermore, health-related quality of life was improved
more with tocilizumab than with placebo.
Rapid and sustained improvements in the acute phase
response markers CRP and ESR, as well as in
haemoglobin, were seen, especially with tocilizumab
8 mg/kg. Low haemoglobin concentrations are
indicative of severe chronic inflammation and are
reported in 30–40% of patients with rheumatoid
arthritis;27 the increases observed here suggest a
reduction in the severity of systemic inflammation with
tocilizumab. Although tocilizumab affects the acute
phase response by liver cells directly, the improvement
of all signs and symptoms of rheumatoid arthritis, as
evidenced in particular by the consistent decrease in
swollen and tender joint counts, suggests that the
response to treatment involved an improvement in
synovitis. Lastly, fatigue—a major debilitating factor in
rheumatoid arthritis—was reduced from baseline
significantly more by tocilizumab than by placebo.
Improvements with tocilizumab were apparent at the
first assessments—ie, 2–4 weeks after treatment
initiation—for all efficacy variables studied and were
maintained or further improved over the treatment
period. Although comparison of the two doses was not
an objective of this study, tocilizumab 8 mg/kg provided
a more frequent and robust therapeutic effect.
As expected with the involvement of interleukin 6 in
the immune response, previous experience with
tocilizumab treatment, and knowledge of agents that
interfere with the immune response,3,14,15,28–30 upper
respiratory tract infections were the most common
www.thelancet.com Vol 371 March 22, 2008
Articles
adverse events seen, occurring more frequently in those
receiving tocilizumab than in those on placebo. The
number of patients experiencing serious infections was
higher with tocilizumab 8 mg/kg compared with control.
Notably, no cases of tuberculosis occurred during the
study period. Since Mycobacterium tuberculosis-induced
interleukin 6 has been shown to inhibit the ability of the
cellular immune response to eradicate infection,31
inhibition of interleukin-6 receptor signalling would not
be anticipated to increase the risk of reactivation of latent
tuberculosis infections.
Tocilizumab therapy was associated with increases in
mean hepatic aminotransferase concentrations, con-
sistent with previous experience.14,15 These increases were
typically single events detected at various visits, without
concomitant increases in bilirubin; no patient experienced
any clinical symptoms of hepatic disease. Since patients
with baseline aminotransferase concentrations of more
than 1·5 times the upper limit of normal were excluded
from this study, the risk of hepatic adverse events in such
patients requires further examination.
In association with the inflammatory process, patients
with active rheumatoid arthritis often have lower lipid
concentrations than the general population32 and increases
have been seen with improvement in chronic
inflammation.33 At 6 weeks, mean fasting plasma lipid
concentrations were raised in the tocilizumab groups,
compared with baseline levels; by contrast, no substantial
changes were seen in the placebo group. These increases
coincided with decreases in CRP levels and were stable
over the treatment period. Increases in atherogenic
indices (ie, the ratio of total to HDL cholesterol and the
ratio of apolipoprotein B to A) were seen in a minority of
patients and in the short term there was no indication of
an increase in major adverse cardiac events. Further
studies are needed to fully assess the effects of tocilizumab
on cardiovascular risk markers. Interestingly, increases
from lower than normal baseline lipid concentrations
have been observed with other efficacious therapies,
including TNF inhibitors.34–37 Studies in which the
inflammatory response in rheumatoid arthritis was
reduced, as with TNF inhibitors, have shown decreased
rates of cardiovascular events, despite the increase in lipid
concentrations.38–42
There are several limitations to our study. First,
although the 6-month trial time is sufficient to judge
efficacy, persistence of clinical and functional
improvement will need long-term follow-up. Second,
inhibition of structural damage is an important aspect of
treatment for rheumatoid arthritis, but was not assessed
here. Although analysis of radiographic changes with
tocilizumab monotherapy is suggestive of a beneficial
effect on progression of joint damage,30 these data came
from an open label trial and the combination of
tocilizumab with methotrexate was not assessed. Third,
we did not assess long-term safety; this information will
only be available from long-term extension studies.
995
 Articles
Nevertheless, no new types of adverse events were
observed when compared with early phase trials.14,15,30
Fourth, although the power of this study is over 90%, the
actual number of patients enrolled was slightly less than
planned. Due to the conservative nature of the
assumptions of the initial sample size and the lower than
expected dropout rate,15 we believe that there was still
more than adequate power within this trial.
Contributors
JSS participated in the development of the study design and was the
principal investigator. AB, ARR, CRR, JR, and RA participated in the
study as investigators; EA and TW analysed the data. JS and TW wrote
the manuscript with input from all the authors. All authors saw and
approved the final version of the manuscript.
OPTION Investigators
Argentina—G Tate (Organizacion Medica En Investigacion,
Buenos Aires; 14 patients); J A Maldonado-Cocco (Instituto de
Rehabilitacion Psicofisica, Bueno Aires; ten patients); J Scali (Hospital
Durand, Buenos Aires; 12 patients). Australia—A Taylor, P Hanrahan
(Sir Charles Gairdner Hospital, Nedlands; ten patients); P Nash
(Rheumatology Research Unit Sunshine Coast, Maroochydore;
nine patients); M Smith (Repatriation General Hospital, Daw Park;
one patient). Austria—J Smolen, M Koeller (Division of Rheumatology,
Medical University of Vienna; four patients); J Smolen, G Eberl
(Krankenhaus der Stadt Wien—Hietzing, Wien; six patients); A Dunky
(Wilhelminenspital, Wien; eight patients); O Zamani (Kaiser Franz Josef
Spital, Wien; four patients). Brazil—J C Simon (Hospital Nossa Senhora
Da Conceicao, Porto Alegre; 20 patients); M A Scheinberg (Hospital
Israelita Albert Einstein, Sao Paulo; eight patients). Bulgaria—D Yaneva
(Umhat St Anna, Sofia; 17 patients); B Oparanov (Military Medical
Academy, Sofia; 16 patients); D Karastatev (Mhar St Marina, Varna;
15 patients). Canada—C Atkins (Percuro Clinical Research Ltd, Victoria;
six patients); A Beaulieu (Centre de Rhumatologie St-Louis, Sainte-Foy,
QC; 18 patients); M Bell (Sunnybrook and Women’s College Health
Service, Toronto; three patients); B Haraoui (Institut de Rhumatologie
de Montreal, Montreal; four patients); L Marin (University of Calgary,
Calgary; three patients); J C Thorne (Arthritis Program Research Group,
Newmarket, ON; one patient); M Zummer (Hôpital
Maisonneuve-Rosemont, Montreal; two patients); M Khraishi (Nexus
Clinical Research Centre, St John’s; six patients); R J R McKendry
(Rheumatology Reseach Associates, Ottawa; four patients); V Pandith
(JBN Medical Diagnostic Services Inc, Burlington; three patients);
T McCarthy (Manitoba Clinic, Winnipeg; five patients). China—C-S Lau
(Queen Mary Hospital, Hong Kong; eight patients); E Li (Prince of Wales
Hospital, Chinese University, Hong Kong; eight patients); C-C Mok
(Tuen Mun Hospital, Tuen Mun; three patients). France—A Kahan
(Hôpital Cochin, Paris; seven patients); D Wendling (Hôpital Jean
Minjoz, Bescancon; three patients); T Bardin (Hôpital Lariboisiere, Paris;
one patient); M Nguyen (Hôpital Cochin, Paris; nine patients);
P Claudepierre (Hôpital Henri Mondor, Creteil; one patient);
F Berenbaum (Hôpital Saint Antoine, Paris; two patients); X Puechal
(Centre Hospitalier du Mans, Le Mans; five patients). Germany—R Alten
(Schlosspark Klinik, Berlin; 30 patients); C Fiehn (Rheuma-Zentrum
Baden-Baden GmbH; five patients); B Heilig (Rheumapraxis Heidelberg,
Heidelberg; 14 patients); B Hellmich (Rheumaklinik Bad Bramstedt, Bad
Bramstedt; four patients); U Lange (Kerckhoff-Klinik GmbH, Bad
Nauheim; five patients); H-M Lorenz (Medizinische Klinik und
Poliklinik V, Heidelberg; three patients); A Rubbert-Roth (Klinikum der
Universität zu Köln, Köln; 16 patients); J Wendler (Praxis für
Rheumatologie, Erlangen; five patients). Hungary—L Czirijak
(Magyarorszagi Irgalmasrend Es A Pesci, Pecs; five patients); L Hodinka
(Orszago Reumatologiai Es Fizioterapi, Budapest; 15 patients);
Z Szekanecz (Debrenceni Egyetem Orvos-Es Egeszsegtu, Debrecen;
15 patients). Israel—Y Molad (Rabin M C-Beilinson Campus,
Petach-Tikva; two patients); M Nahir (Rambam Medical Center, Haifa;
13 patients); I Rosner (Bnei Zion Medical Center, Haifa; seven patients);
A Rubinow (Hadassah Hospital, Jerusalem; one patient); M Abu Shakra
(Soroka, Beer Sheva; four patients); O Elkayam (Sourasky/Ichilov
996
Hospital; two patients). Italy—R Marcolongo (Universita di Siena—
Azienda Ospedal, Siena; four patients); G Bagnato (A U O G Martino—
Policlinico Universita, Gazzi; five patients); G Triolo (A U Policlinico
P Giaccone, Palermo; four patients); F Trotta (Arcispedale S Anna,
Ferrara; six patients); S de Vita (Policlinico Universitario, Udine; six
patients). Mexico—C Ramos-Remus (Unidad de Enfermedades Cronicas
Degenerativas, Guadalajara; 22 patients); G E Lugo (Hospital Juarez
de Mexico, Mexico City; 23 patients); C Abud-Mendoza (Hospital
Central—Dr Ignacio Morone, San Luis Potosi; 30 patients); C Pineca
(Instituto Nacional de Cardiologia, Mexico City; 14 patients);
I G de la Torre (Centro de Estudios de Investigacion Básica y Clínica,
Guadalajara; 15 patients); C Pacheco (Hospital Clinica Del Parque,
Chihuahua; seven patients). Singapore—K H Leong (Gleneagles Medical
Centre, Singapore; four patients); D R Koh (National University Hospital,
Singapore; three patients); Slovakia—J Rovensky (Narodny Ustav
Reumatickych Chorob—3, Piestany; 28 patients). Switzerland—J Dudler
(Hôpital Nestlé, Lausanne; five patients); P Villiger (Inselspital, Bern;
two patients). Thailand—W Lothrenoo (Maharaj Nakorn Chiang Mai
Hospital, Chang Mai; nine patients); P Asavatanabodee (Pramongkutklao
Hospital, Bangkok; 12 patients); S Nilganuwong (Siriraj Hospital,
Mahidol University, Bangkok; 12 patients); K Totemchokchaiyakarn
(Ramathibodi Hospital, Mahidol University, Bangkok; ten patients).
Conflict of interest statement
JSS has consulted for, or received lecture fees from Roche,
Centocor/Schering-Plough, Amgen, Wyeth, Sanofi-Aventis, Abbott,
Bristol-Myers Squibb, UCB, AstraZeneca, and Novartis, and grant support
from Centocor/Schering-Plough, Abbott, Roche, UCB, and Wyeth. AB has
received support for clinical trials from Roche, Novartis, Amgen, Abbott,
and Wyeth. ARR has received consulting and lecture fees from Roche,
Bristol-Myers Squibb, Wyeth, Abbott, Essex, UCB, Merck Sharp &
Dohme, and Merck. CRR has received consulting fees from Roche,
Wyeth, Novartis, Merck, Abbott, and Schering-Plough, as well as clinical
trial support from Roche, Wyeth, and Novartis. RA has received
consulting fees from Abbott, Bristol-Myers Squibb, Merck Pharma
GmbH, Schering-Plough, Pfizer, and Roche. JR has received lecture fees
from Roche, Schering-Plough, Wyeth, Abbott, Pfizer, Sanofi-Aventis, and
Novartis and clinical trial support from Roche, Abbott. EA and TW are
employees of Roche.
Acknowledgments
This trial was supported by F Hoffmann-La Roche and Chugai
Pharmaceutical. Preliminary data were presented at the EULAR 2007
Congress, June 13–16, 2007 (Barcelona, Spain). We acknowledge the
collaboration and commitment of all investigators and their staff,
without whom the present study would not have been possible. Editorial
support was provided by Emiliana Jelezarova (Phocus Services, Basel,
Switzerland), supported by F Hoffmann-La Roche.
References
1 Smolen JS, Aletaha D, Koeller M, Weisman MH, Emery P. New
therapies for treatment of rheumatoid arthritis. Lancet 2007;
370: 1861–74.
2 Firestein GS. Evolving concepts of rheumatoid arthritis. Nature
2003; 423: 356–61.
3 Emery P, Fleischmann R, Filipowicz-Sosnowska A, et al. The
efficacy and safety of rituximab in patients with active rheumatoid
arthritis despite methotrexate treatment: results of a phase IIB
randomized, double-blind, placebo-controlled, dose-ranging trial.
Arthritis Rheum 2006; 54: 1390–400.
4 Jiang Y, Genant HK, Watt I, et al. A multicenter, double-blind,
dose-ranging, randomized, placebo-controlled study of
recombinant human interleukin-1 receptor antagonist in patients
with rheumatoid arthritis: radiologic progression and correlation
of Genant and Larsen scores. Arthritis Rheum 2000; 43: 1001–09.
5 Kremer JM, Genant HK, Moreland LW, et al. Effects of abatacept in
patients with methotrexate-resistant active rheumatoid arthritis:
a randomized trial. Ann Intern Med 2006; 144: 865–76.
6 Lipsky PE, van der Heijde DM, St Clair EW, et al. Infliximab and
methotrexate in the treatment of rheumatoid arthritis. Anti-Tumor
Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant
Therapy Study Group. N Engl J Med 2000; 343: 1594–602.
7 Emery P, Salmon M. Early rheumatoid arthritis: time to aim for
remission? Ann Rheum Dis 1995; 54: 944–47.
www.thelancet.com Vol 371 March 22, 2008

8 Mierau M, Schoels M, Gonda G, Fuchs J, Aletaha D, Smolen JS.
Assessing remission in clinical practice. Rheumatology (Oxford)
2007; 46: 975–79.
9 Firestein GS, Alvaro-Gracia JM, Maki R. Quantitative analysis of
cytokine gene expression in rheumatoid arthritis. J Immunol 1990;
144: 3347–53.
10 Steiner G, Tohidast-Akrad M, Witzmann G, et al. Cytokine
production by synovial T cells in rheumatoid arthritis.
Rheumatology (Oxford) 1999; 38: 202–13.
11 Naka T, Nishimoto N, Kishimoto T. The paradigm of IL-6: from basic
science to medicine. Arthritis Res 2002; 4 (suppl 3): S233–42.
12 Aletaha D, Nell VP, Stamm T, et al. Acute phase reactants add little
to composite disease activity indices for rheumatoid arthritis:
validation of a clinical activity score. Arthritis Res Ther 2005;
7: R796–806.
13 van Leeuwen MA, van der Heijde DM, van Rijswijk MH, et al.
Interrelationship of outcome measures and process variables in
early rheumatoid arthritis. A comparison of radiologic damage,
physical disability, joint counts, and acute phase reactants.
J Rheumatol 1994; 21: 425–29.
14 Nishimoto N, Yoshizaki K, Miyasaka N, et al. Treatment of
rheumatoid arthritis with humanized anti-interleukin-6 receptor
antibody: a multicenter, double-blind, placebo-controlled trial.
Arthritis Rheum 2004; 50: 1761–69.
15 Maini RN, Taylor PC, Szechinski J, et al. Double-blind randomized
controlled clinical trial of the interleukin-6 receptor antagonist,
tocilizumab, in European patients with rheumatoid arthritis who
had an incomplete response to methotrexate. Arthritis Rheum 2006;
54: 2817–29.
16 Arnett FC, Edworthy SM, Bloch DA, et al. The American
Rheumatism Association 1987 revised criteria for the classification
of rheumatoid arthritis. Arthritis Rheum 1988; 31: 315–24.
17 Felson DT, Anderson JJ, Boers M, et al. Preliminary definition of
improvement in rheumatoid arthritis. Arthritis Rheum 1995;
38: 727–35.
18 Prevoo ML, van ‘t Hof MA, Kuper HH, van Leeuwen MA,
van de Putte LB, van Riel PL. Modified disease activity scores that
include twenty-eight-joint counts. Development and validation in
a prospective longitudinal study of patients with rheumatoid
arthritis. Arthritis Rheum 1995; 38: 44–48.
19 van Gestel AM, Prevoo ML, van ‘t Hof MA, van Rijswijk MH,
van de Putte LB, van Riel PL. Development and validation of the
European League Against Rheumatism response criteria for
rheumatoid arthritis. Comparison with the preliminary American
College of Rheumatology and the World Health
Organization/International League Against Rheumatism Criteria.
Arthritis Rheum 1996; 39: 34–40.
20 Fransen J, Creemers MC, van Riel PL. Remission in rheumatoid
arthritis: agreement of the disease activity score (DAS28) with the
ARA preliminary remission criteria. Rheumatology (Oxford) 2004;
43: 1252–55.
21 Bruce B, Fries JF. The Stanford Health Assessment Questionnaire:
a review of its history, issues, progress, and documentation.
J Rheumatol 2003; 30: 167–78.
22 Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health
survey (SF-36). I. Conceptual framework and item selection.
Med Care 1992; 30: 473–83.
23 Cella D, Lai JS, Chang CH, Peterman A, Slavin M. Fatigue in cancer
patients compared with fatigue in the general United States
population. Cancer 2002; 94: 528–38.
24 Cella D, Yount S, Sorensen M, Chartash E, Sengupta N, Grober J.
Validation of the Functional Assessment of Chronic Illness Therapy
Fatigue Scale relative to other instrumentation in patients with
rheumatoid arthritis. J Rheumatol 2005; 32: 811–19.
25 Wells GA, Tugwell P, Kraag GR, Baker PR, Groh J, Redelmeier DA.
Minimum important difference between patients with rheumatoid
arthritis: the patient’s perspective. J Rheumatol 1993; 20: 557–60.
www.thelancet.com Vol 371 March 22, 2008
Articles
26 National Cholesterol Education Program, National Heart, Lung and
Blood Institute, and National Institutes of Health. Detection,
evaluation, and treatment of high blood cholesterol in adults (Adult
Treatment Panel III). Final report. Bethesda: National Institutes of
Health, 2002.
27 Wolfe F, Michaud K. Anemia and renal function in patients with
rheumatoid arthritis. J Rheumatol 2006; 33: 1516–22.
28 Keystone EC, Kavanaugh AF, Sharp JT, et al. Radiographic, clinical,
and functional outcomes of treatment with adalimumab (a human
anti-tumor necrosis factor monoclonal antibody) in patients with
active rheumatoid arthritis receiving concomitant methotrexate
therapy: a randomized, placebo-controlled, 52-week trial.
Arthritis Rheum 2004; 50: 1400–11.
29 Maini R, St Clair EW, Breedveld F, et al. Infliximab (chimeric
anti-tumour necrosis factor alpha monoclonal antibody) versus
placebo in rheumatoid arthritis patients receiving concomitant
methotrexate: a randomised phase III trial. ATTRACT Study Group.
Lancet 1999; 354: 1932–39.
30 Nishimoto N, Hashimoto J, Miyasaka N, et al. Study of active
controlled monotherapy used for rheumatoid arthritis, an IL-6
inhibitor (SAMURAI): evidence of clinical and radiographic benefit
from an x-ray reader-blinded randomized controlled trial of
tocilizumab. Ann Rheum Dis 2007; 66: 1162–67.
31 Nagabhushanam V, Solache A, Ting LM, Escaron CJ, Zhang JY,
Ernst JD. Innate inhibition of adaptive immunity:
Mycobacterium tuberculosis-induced IL-6 inhibits macrophage
responses to IFN-gamma. J Immunol 2003; 171: 4750–57.
32 Sattar N, McCarey DW, Capell H, McInnes IB. Explaining how
“high-grade” systemic inflammation accelerates vascular risk in
rheumatoid arthritis. Circulation 2003; 108: 2957–63.
33 Dursunoglu D, Evrengul H, Polat B, et al. Lp(a) lipoprotein and
lipids in patients with rheumatoid arthritis: serum levels and
relationship to inflammation. Rheumatol Int 2005; 25: 241–45.
34 Sattar N, Crompton P, Cherry L, Kane D, Lowe G, McInnes IB.
Effects of tumor necrosis factor blockade on cardiovascular risk
factors in psoriatic arthritis: a double-blind, placebo-controlled
study. Arthritis Rheum 2007; 56: 831–39.
35 Vis M, Nurmohamed MT, Wolbink G, et al. Short term effects of
infliximab on the lipid profile in patients with rheumatoid arthritis.
J Rheumatol 2005; 32: 252–55.
36 Dahlqvist SR, Engstrand S, Berglin E, Johnson O. Conversion
towards an atherogenic lipid profile in rheumatoid arthritis patients
during long-term infliximab therapy. Scand J Rheumatol 2006;
35: 107–11.
37 Allanore Y, Kahan A, Sellam J, Ekindjian OG, Borderie D. Effects of
repeated infliximab therapy on serum lipid profile in patients with
refractory rheumatoid arthritis. Clin Chim Acta 2006; 365: 143–48.
38 Choi HK, Hernan MA, Seeger JD, Robins JM, Wolfe F.
Methotrexate and mortality in patients with rheumatoid arthritis:
a prospective study. Lancet 2002; 359: 1173–77.
39 Jacobsson LT, Turesson C, Hanson RL, et al. Joint swelling as a
predictor of death from cardiovascular disease in a population study
of Pima Indians. Arthritis Rheum 2001; 44: 1170–76.
40 Turesson C, Matteson EL. Cardiovascular risk factors, fitness and
physical activity in rheumatic diseases. Curr Opin Rheumatol 2007;
19: 190–96.
41 Wolfe F, Freundlich B, Straus WL. Increase in cardiovascular and
cerebrovascular disease prevalence in rheumatoid arthritis.
J Rheumatol 2003; 30: 36–40.
42 Wolfe F, Michaud K. Heart failure in rheumatoid arthritis: rates,
predictors, and the effect of anti-tumor necrosis factor therapy.
Am J Med 2004; 116: 305–11.
997