Efficacy and safety of golimumab in patients with ankylosing spondylitis: results of a randomized, double-blind, placebo-controlled, phase iii trial

Vol. 58, No. 11, November 2008, pp 3402–3412 2008, American College of Rheumatology Efficacy and Safety of Golimumab in Patients With Results of a Randomized, Double-Blind, Placebo-Controlled, Phase III Trial Robert D. Inman,1 John C. Davis, Jr.,2 D´ Joachim Sieper,5 Sung Il Kim,6 Michael Mack,7 John Han,7 Sudha Visvanathan,7 Zhenhua Xu,7 Objective. To evaluate the efficacy and safety of
primary end point was the proportion of patients with at
golimumab in patients with ankylosing spondylitis (AS)
least 20% improvement in the ASsessment in AS
in the GO-RAISE study.
(ASAS20) criteria at week 14.
Methods. Patients with active AS, a Bath AS
Results. At randomization, 138, 140, and 78 pa-
Disease Activity Index (BASDAI) score >4, and a back
tients were assigned to the 50-mg, 100-mg, and placebo
pain score of >4 were randomly assigned in a 1.8:1.8:1
groups, respectively. After 14 weeks, 59.4%, 60.0%, and
ratio to receive subcutaneous injections of golimumab
21.8% of patients, respectively, were ASAS20 responders
(50 mg or 100 mg) or placebo every 4 weeks. The
(P < 0.001). A 40% improvement in the ASAS criteria at
week 24 occurred in 43.5%, 54.3%, and 15.4% of patients,
ClinicalTrials.gov identifier: NCT00265083.
respectively. Patients receiving golimumab also showed
Supported by Centocor Research and Development, Inc. and significant improvement in the physical and mental
the Schering-Plough Research Institute, Inc.
component summary scores of the Short Form 36
Robert D. Inman, MD: University of Toronto, Toronto, Ontario, Canada; 2John C. Davis, Jr., MD, MPH: University of Health Survey, the Jenkins Sleep Evaluation Question-
naire score, the BASDAI score, and the Bath AS Func-
University Medical Center, Leiden, The Netherlands; 4Laura Diek- tional Index score, but not the Bath AS Metrology Index
man, MS, MT(ASCP): University of Texas–Houston; 5Joachim Sieper,MD: Charite ´ Hospital, Berlin, Germany; 6Sung Il Kim, MD, PhD: score. Through week 24, 85.6% of golimumab-treated
Pusan National University Hospital, Pusan, South Korea; 7Michael patients and 76.6% of patients in the placebo group had
Mack, PhD, John Han, PhD, Sudha Visvanathan, PhD, Zhenhua Xu,PhD, FCP, Benjamin Hsu, MD, PhD, Anna Beutler, MD: Centocor >1 adverse event, and 5.4% and 6.5% of patients,
Research and Development, Malvern, Pennsylvania; 8Ju respectively, had >1 serious adverse event. Eight
MD: Rheumazentrum Ruhrgebiet, Herne, Germany.
golimumab-treated patients and 1 placebo-treated pa-
Dr. Inman has received consulting fees, speaking fees, and/or tient had markedly abnormal liver enzyme values
honoraria from Schering-Plough, Centocor, Amgen, and Abbott (lessthan $10,000 each). Dr. van der Heijde has received consulting fees, (>100% increase from baseline and a value >150
speaking fees, and/or honoraria from Abbott, Amgen, Centocor, UCB, IU/liter), which were transient.
Wyeth, Roche, Schering-Plough, Novartis, and Chugai (less than$10,000 each). Dr. Sieper has received consulting fees, speaking fees, Conclusion. Golimumab was effective and well
and/or honoraria from Schering-Plough, Abbott, and Wyeth (less tolerated in a large cohort of patients with AS during a
than $10,000 each). Drs. Mack, Han, Visvanathan, Xu, Beutler, and 24-week study period.
Hsu own stock and/or stock options in Johnson & Johnson, of whichCentocor is a subsidiary. Dr. Braun has received consulting fees,speaking fees, and/or honoraria from Centocor, Schering-Plough, Ankylosing spondylitis (AS) is a chronic inflam- Wyeth, Amgen, Abbott, Pfizer, and Bristol-Myers Squibb (less than matory disease of unknown etiology that involves the sacroiliac joints, axial skeleton, entheses, and peripheral Address correspondence and reprint requests to Robert D.
Inman, MD, University of Toronto, Toronto Western Hospital, 399 joints. Chronic inflammation of entheses potentially Bathurst Street, Toronto, Ontario M5T 2S8, Canada. E-mail: leads to new bone formation in the form of syndesmo- phytes and ankylosis of vertebrae and joints, primarily in Submitted for publication February 6, 2008; accepted in the axial skeleton. Patients may also have extraarticular manifestations and comorbidities, including acute ante- the study agent; or any previous use of anti-TNF therapy, rior uveitis, psoriasis, colitis, aortitis, and cardiac con- rituximab, natalizumab, or cytotoxic drugs.
The study was conducted at 57 centers in the US, Canada, Europe, and Asia. The protocol was reviewed and Tumor necrosis factor ␣ (TNF␣) is a major approved by the institutional review board or independent therapeutic target in AS (1), and several biologic agents ethics committee at each site. All patients provided written that target this proinflammatory cytokine have been informed consent. A list of the principal investigators and their shown to significantly improve the signs and symptoms Study agent. Golimumab is a human IgG1␬ monoclo-
of AS in patients with moderate-to-severe disease (2–4).
nal antibody that specifically binds to both the soluble and Golimumab is a new human anti-TNF␣ monoclonal membrane-bound forms of TNF␣. Both golimumab and pla- antibody with pharmacologic properties that allow for cebo were supplied as sterile liquid for subcutaneous injection monthly subcutaneous dosing. Golimumab has been in single-use, 2-ml vials. Each patient received 2 injections (0.5ml and 1.0 ml) every 4 weeks. To maintain blinding, patients in studied in patients with rheumatoid arthritis (5) and the 50-mg group received active golimumab in the 0.5-ml patients with psoriatic arthritis (6), and more studies are syringe and placebo in the 1.0-ml syringe; patients in the ongoing. In the GO-RAISE study, we evaluated the 100-mg group received placebo in the 0.5-ml syringe and active efficacy and safety of golimumab in reducing signs and golimumab in the 1.0-ml syringe; and patients in the placebo group received placebo in both syringes.
Study protocol. In this 24-week, double-blind, placebo-
controlled study, patients were randomly assigned in a 1:1.8:1.8 PATIENTS AND METHODS
ratio to receive placebo or golimumab at a dose of 50 mg or100 mg. An interactive voice-response system with adaptive Patients. Adult patients who had AS (diagnosed ac-
treatment allocation was used to assign patients to treatment.
cording to the modified New York Criteria [7]) for Ն3 months Randomization was stratified by investigational site and before the first administration of the study agent, a Bath AS C-reactive protein (CRP) level (Յ1.5 mg/dl or Ͼ1.5 mg/dl). At Disease Activity Index (BASDAI) score of Ն4 (0–10-point week 16, patients who achieved Ͻ20% improvement from scale), a spinal pain assessment score of Ն4 on a visual analog baseline in both the total back pain and morning stiffness scale (VAS; 0–10-cm scale), and an inadequate response to measures entered early escape in a double-blinded manner: current or previous nonsteroidal antiinflammatory drugs patients in the placebo group received golimumab 50 mg, (NSAIDs) or disease-modifying antirheumatic drugs patients in the golimumab 50-mg group had a dose escalation (DMARDs) were eligible for participation in the study. Pa- to 100 mg, and patients in the 100-mg group continued to tients who were receiving NSAIDs had to have received continuous therapy for 3 months at the highest recommended Evaluations. The primary end point was the propor-
doses or had to have been unable to receive a full 3-month tion of patients who achieved at least 20% improvement in the course of full-dose NSAID therapy because of intolerance, ASsessment in AS International Working Group criteria(ASAS20) (8) at week 14. Secondary end points included toxicity, or contraindications. Patients were also required to ASAS 40% improvement (ASAS40) (9), ASAS partial remis- have normal results of a chest radiograph within 3 months sion (8), and 20% improvement in 5 of 6 ASAS domains before randomization and to have undergone screening for (ASAS5/6) (9). Disease activity was evaluated using the BAS- latent tuberculosis (TB) using a purified protein derivative skin DAI (10), the back pain VAS, the night pain VAS, the test and the QuantiFERON TB Gold test. Patients in whom patient’s global assessment, and the CRP level. Physical func- latent TB was discovered were required to initiate therapy for tion was evaluated using the Bath AS Functional Index TB prior to or simultaneously with the first dose of the study (BASFI) (11). Range of motion was assessed using the Bath agent. Patients were excluded from the study if they had any of AS Metrology Index (BASMI) (3-point scale) (12) and chest the following: complete ankylosis of the spine, any other expansion. Health-related quality of life was measured using inflammatory rheumatic disease, a serious infection within 2 the Short Form 36 (SF-36) Health Survey (13). Sleep distur- months before randomization, active or latent TB or positive bance was assessed using the Jenkins Sleep Evaluation Ques- results of a tuberculin skin test before screening or recent tionnaire (JSEQ) (14), in which patients indicated the number contact with a person with active TB, an opportunistic infec- of days they experienced problems falling asleep, staying tion within 6 months of screening, hepatitis, human immuno- asleep, early awakening, and awakening tired in the previous deficiency virus, a transplanted organ, malignancy, multiple 30 days. The numbers of days were grouped according to sclerosis, or congestive heart failure.
predefined categories, and each category was given a score of Patients were allowed to continue concurrent treat- 0–5, for a possible total score of 0–20.
ment with methotrexate (MTX), sulfasalazine, hydroxychloro- Serum samples collected before each injection of study quine, corticosteroids, and NSAIDs at stable doses during the agent were used to measure the trough golimumab concentra- study. The following treatments were not permitted: systemic tion. Serum samples obtained at baseline and at week 24 were immunosuppressives, DMARDs (other than MTX, sulfasala- assessed for the presence of antibodies to golimumab, using a zine, or hydroxychloroquine), or leflunomide within 4 weeks before the first administration of the study agent; alefacept or Statistical analysis. When designing the study, we
efalizumab within 3 months before the first administration of assumed that 20% of patients in the placebo group (regardless Figure 1. Patient disposition through week 24.
of CRP level), 35% of patients in the combined-golimumab using the Cochrane-Mantel-Haenszel test with the same strat- group with a CRP level Յ1.5 mg/dl, and 60% of patients in the ification criteria as for the primary end point. Changes from combined-golimumab group with a CRP level Ͼ1.5 mg/dl baseline in continuous variables were compared between treat- would achieve the primary end point. Under these assump- ment groups using an analysis of variance on the normal van tions, a sample size of 135 patients in each golimumab group and 75 patients in the placebo group had Ͼ99% power to Patients in the placebo or 50-mg groups who met the detect a significant difference in the proportion of patients criteria for early escape at week 16 were considered to be achieving the primary end point between the golimumab- nonresponders at week 24. The actual observed week 24 values treated groups and the placebo-treated group at the 0.05 level were used for patients in the 100-mg group who met the of significance. There was also Ͼ99% power for the compari- criteria for early escape at week 16.
son between each individual golimumab treatment group and A logistic regression analysis of the ASAS20 response at week 14 was performed based on the following factors: In the primary efficacy analysis, data from all random- treatment group (placebo or golimumab at either dose), ized patients were analyzed according to their assigned treat- screening CRP level (Յ1.5 mg/dl or Ͼ1.5 mg/dl), DMARD use ment group. A last observation carried forward procedure was (yes or no), and the continuous variables body weight and used to impute any missing ASAS components for patients who had data for at least 1 ASAS component at week 14.
Patients without data for any of the ASAS components at week14 were considered not to have achieved the primary end point. In addition, patients who met any 1 of the followingcriteria before week 14 were considered not to have achieved Data for this analysis were collected between the primary end point: initiated new DMARDs, biologic December 2005 and May 2007. A total of 356 patients agents, or systemic immunosuppressives; increased the MTX, were randomly assigned to treatment (Figure 1). One sulfasalazine, or hydroxychloroquine dose above the baselinelevel; initiated treatment with or increased the dose of cortico- patient was assigned to golimumab 50 mg but never steroids; or discontinued study treatment due to an unsatisfac- received any study treatment before withdrawing from participation. One patient in the placebo group received The proportions of patients who achieved the primary a 50-mg dose of golimumab in error. As per the pre- end point were compared between treatment groups using theCochrane-Mantel-Haenszel test with stratification by the specified data-handling rules, this patient was included screening CRP level (Յ1.5 mg/dl or Ͼ1.5 mg/dl). For second- in the placebo group for all efficacy analyses and in the ary end points, the proportions of patients were compared golimumab 50 mg group for all safety analyses.
Summary of demographics and baseline disease characteristics* Years since inflammatory back pain first occurred Years since symptoms of SpA first occurred† Years since diagnosis of ankylosing spondylitis Patient’s global assessment of disease activity (0–10-cm VAS) Patient’s assessment of total back pain (0–10-cm VAS) Inflammation, overall morning stiffness (0–10-cm VAS) Physical component summary score (0–50 scale) Mental component summary score (0–50 scale) Jenkins Sleep Evaluation Questionnaire (0–20 scale) History of extraaxial involvement, no. (%) Patients taking hydroxychloroquine, no. (%) * Except where indicated otherwise, values are the median (interquartile range). CRP ϭ C-reactive protein; VAS ϭ visual analog scale; BASDAI ϭBath Ankylosing Spondylitis Disease Activity Index; BASFI ϭ Bath Ankylosing Spondylitis Functional Index; BASMI ϭ Bath AnkylosingSpondylitis Metrology Index; NSAIDs ϭ nonsteroidal antiinflammatory drugs.
† Symptoms of spondylarthritis (SpA) include back pain, uveitis, psoriasis, dactylitis, enthesitis, peripheral arthritis, or inflammatory bowel disease.
The demographic and baseline disease charac- Efficacy. The primary end point was achieved. Of
teristics were generally well balanced across treatment the patients who received golimumab, 59.4% in the groups, except for disease duration (Table 1), which was 50-mg group and 60.0% in the 100-mg group achieved shorter in the golimumab groups. The treatment groups an ASAS20 response at week 14 compared with 21.8% were not balanced for the proportions of patients with in the placebo group (P Ͻ 0.001) (Figure 2). Sensitivity some extraaxial manifestations. Baseline disease activity analyses indicated that the primary end point was robust values indicated that patients had a moderately high when patients who discontinued treatment because of level of pain and inflammation (Table 1).
adverse events were considered to be nonresponders, not shown). The response at week 14 in the combined-golimumab group was greater in patients with a CRPlevel Ͼ0.6 mg/dl (66.1%) compared with the response inthose with a CRP level Յ0.6 mg/dl (49.5%), and inpatients with a CRP level Ͼ1.5 mg/dl (70.3%) comparedwith those with a CRP level Յ1.5 mg/dl (51.9%).
Greater proportions of patients in the golimumab groups achieved an ASAS20 response at the first assess-ment, 4 weeks after the first injection (Figure 3A). Themean BASDAI and BASFI scores were lower in thegolimumab groups through week 24 compared with thescores in the placebo group (Figures 3B and C). Inaddition, 43.5%, 54.3%, and 15.4% of patients in the Figure 2. Proportions of patients achieving improvement in the
ASsessment in Ankylosing Spondylitis (ASAS) International Working
Group criteria (20% improvement [ASAS20], 40% improvement
[ASAS40], 20% improvement in 5 of 6 ASAS domains [ASAS5/6], and
ASAS partial remission) at weeks 14 and 24. ء ϭ P Ͻ 0.001; ءء ϭ P Ͻ
0.01, versus placebo.
when patients with insufficient data to determine theASAS20 response were considered to be nonresponders,and when the analysis was based on observed data only.
The logistic regression analysis indicated that effectsattributable to treatment group (P Ͻ 0.001), screeningCRP level (P ϭ 0.0062), and body weight (P ϭ 0.0140)were significantly associated with ASAS20 responses,while use of DMARDs and duration of AS were notsignificantly associated.
The benefit of golimumab treatment was also consistent across subgroups of sex, race, age, geographicregion, and body weight, except for patients in the 50-mggroup in the weight quartile Ͼ87 kg and those in the100-mg group in the weight quartile Ͼ75.15 kg to Յ87 Figure 3. A, Proportion of patients achieving a 20% improvement in
the ASsessment in Ankylosing Spondylitis International Working
kg; for both of these groups, the percent of patients with Group criteria through week 24. B, Mean Bath Ankylosing Spondylitis
an ASAS20 response was not statistically significantly Disease Activity Index (BASDAI) score through week 24. C, Mean Bath
different from that observed in the placebo group (data Ankylosing Spondylitis Functional Index (BASFI) score through week 24.
Summary of changes in clinical signs and symptoms from baseline to week 14 and week 24* Patient’s global assessment of disease activity (0–10-cm VAS) Ϫ0.8 (Ϫ2.3, 0.30) Ϫ2.8 (Ϫ5.0, Ϫ1.0)† Ϫ3.4 (Ϫ5.3, Ϫ0.6)† Ϫ3.0 (Ϫ5.2, Ϫ0.8)†Patient’s assessment of total back pain (0–10-cm VAS) Ϫ0.8 (Ϫ3.1, 0.3) Ϫ3.5 (Ϫ5.5, Ϫ0.8)† Ϫ3.6 (Ϫ5.9, Ϫ0.9)† Ϫ3.5 (Ϫ5.8, Ϫ0.9)† Inflammation (overall morning stiffness) (0–10-cm VAS) Ϫ0.5 (Ϫ1.9, 0.7) Ϫ3.2 (Ϫ5.4, Ϫ1.2)† Ϫ3.3 (Ϫ5.7, Ϫ0.8)† Ϫ3.3 (Ϫ5.6, Ϫ1.0)† Ϫ1.4 (Ϫ3.1, Ϫ0.1)† Ϫ1.5 (Ϫ3.0, Ϫ0.1)† Ϫ1.4 (Ϫ3.1, Ϫ0.1)† Ϫ0.3 (Ϫ2.9, 0.5) Ϫ3.0 (Ϫ5.3, Ϫ0.5)† Ϫ3.1 (Ϫ5.9, Ϫ0.8)† Ϫ3.0 (Ϫ5.7, Ϫ0.6)† Patients with Ն1-unit improvement in the BASMI, no. (%) Physical component summary score (0–50 scale) Mental component summary score (0–50 scale) Patient’s global assessment of disease activity (0–10-cm VAS) Ϫ0.2 (Ϫ2.2, 1.0) Ϫ2.6 (Ϫ5.2, Ϫ1.0)† Ϫ3.6 (Ϫ5.9, Ϫ0.6)† Ϫ3.3 (Ϫ5.6, Ϫ1.0)† Patient’s assessment of total back pain (0–10-cm VAS) Ϫ0.40 (Ϫ2.0, 1.0) Ϫ3.5 (Ϫ5.6, Ϫ0.8)† Ϫ3.9 (Ϫ6.4, Ϫ1.2)† Ϫ3.7 (Ϫ6.3, Ϫ1.0)† Inflammation, (overall morning stiffness) (0–10-cm VAS) Ϫ0.2 (Ϫ2.3, 0.8) Ϫ3.5 (Ϫ5.4, Ϫ1.1)† Ϫ3.7 (Ϫ6.2, Ϫ1.4)† Ϫ3.6 (Ϫ5.8, Ϫ1.2)† Ϫ1.6 (Ϫ3.5, Ϫ0.3)† Ϫ1.6 (Ϫ3.5, Ϫ0.2)† Ϫ0.4 (Ϫ1.9, 0.9) Ϫ3.1 (Ϫ5.6, Ϫ0.8)† Ϫ3.5 (Ϫ6.7, Ϫ0.8)† Ϫ3.3 (Ϫ6.2, Ϫ0.8)† Patients with Ն1-unit improvement in the BASMI, no. (%) Physical component summary score (0–50 scale) Mental component summary score (0–50 scale) * Except where indicated otherwise, values are the median (interquartile range). VAS ϭ visual analog scale; BASFI ϭ Bath Ankylosing SpondylitisFunctional Index; BASDAI50 ϭ at least 50% improvement from baseline in the baseline Bath Ankylosing Spondylitis Disease Activity Index score;BASMI ϭ Bath Ankylosing Spondylitis Metrology Index; JSEQ ϭ Jenkins Sleep Evaluation Questionnaire.
† P Ͻ 0.001 versus placebo.
‡ P Ͻ 0.05 versus placebo.
§ P Ͻ 0.01 versus placebo.
50-mg, 100-mg, and placebo groups, respectively, 24. Of the 25 patients who entered early escape at week achieved an ASAS40 response at week 24. The propor- 16, changing from golimumab at a dose of 50 mg to a tions of patients achieving an ASAS20 response, an 100-mg dose of golimumab, 16.0% showed an ASAS20 ASAS40 response, an ASAS5/6 response, or partial remission at week 24 were similar to those at week 14 At both week 14 and week 24, patients who received golimumab showed statistically significant im- Of the 41 patients who entered early escape at provements in all components of the ASAS20 (Table 2).
week 16, changing from placebo to golimumab at a dose In addition, significantly more golimumab-treated pa- of 50 mg, 50.0% showed an ASAS20 response at week tients achieved Ն50% improvement in the BASDAI score compared with patients who received placebo Antibody-positive patients generally had low serum goli- (Table 2). Significant improvements in night back pain and the CRP level were observed in the golimumab Safety. Through week 16, before patients had the
groups from baseline to week 14 and week 24.
possibility of entering early escape, 77.3% of those in the The median changes from baseline in BASMI combined-golimumab group and 74.0% in the placebo scores at weeks 14 and 24 were similar between the group had Ն1 adverse event, with similar proportions in treatment groups (Table 2). However, significantly more the 50-mg (79.0%) and 100-mg (75.7%) groups. Five patients in the 50-mg and 100-mg golimumab groups patients (3.6%) in the 50-mg group, 7 patients (5.0%) in showed Ն1 unit improvement from baseline in the the 100-mg group, and 4 patients (5.2%) in the placebo BASMI score at week 14. In addition, 3 of the 5 group had serious adverse events through week 16.
component assessments of the BASMI (lumbar flexion, Through week 24, the proportion of patients who lumbar side flexion, and intermalleolar distance) im- had at least 1 adverse event was 79.9% in the all- proved at week 14 or week 24 for golimumab-treated golimumab group (which includes patients in the pla- patients. Although no improvement in chest expansion cebo group who met the criteria for early escape) and was observed at week 14, the median improvement in 85.6% in the combined-golimumab group (which in- chest expansion at week 24 was significantly greater in cludes only patients who were originally assigned to the 50-mg and combined-golimumab groups compared golimumab at the time of randomization), with similar with the placebo group (P ϭ 0.013 and P ϭ 0.016, proportions in the 50-mg (84.8%) and 100-mg (85.7%) groups (Table 3). Of the adverse events that occurred inՆ5% of patients in the combined-golimumab group, The SF-36 physical component summary scores nasopharyngitis, upper respiratory tract infections, fa- and JSEQ scores improved significantly from baseline to tigue, headache, diarrhea, injection-site erythema, and weeks 14 and 24 in all golimumab groups compared with increases in alanine aminotransferase (ALT) or aspar- the placebo group (Table 2). The SF-36 mental compo- tate aminotransferase (AST) levels occurred more fre- nent summary scores improved from baseline to week 14 quently in the combined-golimumab group than in the in all golimumab groups compared with the placebo placebo group. Patients receiving DMARDs at baseline group, with sustained and significant improvement (MTX, sulfasalazine, or hydroxychloroquine) had lower through week 24 in the 100-mg and combined- incidences of adverse events compared with patients who were not receiving DMARDs at baseline.
Golimumab serum concentration. Golimumab-
The proportion of patients who had at least 1 treated patients demonstrated an approximately dose- serious adverse event through week 24 was 3.6% in the proportional increase in serum golimumab concentrations 50-mg group, 6.4% in the 100-mg group, 5.4% in the (data not shown). Serum golimumab concentrations gen- combined-golimumab group, and 6.5% in the placebo erally achieved steady state by week 12. Patients in the group (Table 3). One patient in the 50-mg group had a 50-mg group who met the criteria for early escape had myocardial infarction on day 67, despite a normal lower median serum golimumab concentrations (0.36 ␮g/ screening cardiac evaluation 4 months prior. One pa- ml) before week 16 than did those who did not require tient in the 100-mg group experienced severe fatigue, early escape (0.59 ␮g/ml). The median serum concen- depression, and hypertension. There were no deaths, tration for the 21 patients who did not achieve an opportunistic infections, or cases of TB.
ASAS20 response after early escape (1.04 ␮g/ml) was Nine patients (2.9% of the 50-mg group, 2.9% of similar to that for the 4 patients who did achieve an the 100-mg group, and 1.3% of the placebo group) ASAS20 response (0.90 ␮g/ml). There was large vari- discontinued study treatment because of an adverse ability in serum golimumab concentrations. However, event, including 1 patient in the 50-mg group and 2 patients with a heavier body weight tended to have lower patients in the 100-mg group who had increases in liver transaminase levels; 1 patient in the 50-mg group who Antibodies to golimumab. The incidence of anti-
had alcohol withdrawal syndrome, hallucination, and a bodies to golimumab was low (4.1% of patients origi- suicide attempt; 1 patient in the 50-mg group with chest nally assigned to golimumab). None of the patients with pain; 1 patient in the 50-mg group with blepharitis, antibodies were receiving concomitant MTX. The high- nausea, and vomiting; 1 patient in the 100-mg group with est titer (1:2,560) was measured in a patient in the 50-mg hepatitis (noninfectious); 1 patient in the 100-mg group group who entered early escape to the 100-mg dose.
with depression and hypertension; and 1 patient in the Summary of adverse events and antibodies to golimumab through week 24* Discontinued study agent because of an adverse Adverse event occurring in Ն5% of patients in * Except where indicated otherwise, values are the number (%) of patients. DMARDs ϭ disease-modifying antirheumatic drugs; ALT ϭ alanineaminotransferase; AST ϭ aspartate aminotransferase; NA ϭ not applicable.
placebo group who had headache, influenza-like illness, markedly abnormal ALT or AST values. The remain- pain in the extremities, and increased body weight.
ing patient had the markedly abnormal value at week Nine patients had markedly abnormal postbase- 24, 8 weeks after receiving golimumab 50 mg as early line ALT or AST values (Ն100% increase from baseline escape, and this value had normalized after followup.
and a value Ͼ150 IU/liter), including 1 patient in the Some of the patients with markedly abnormal ALT or placebo group, 2 patients who were originally in the AST values were receiving concomitant hepatotoxic placebo group and then received 50 mg of golimumab, 2 medications (e.g., isoniazid, indomethacin, MTX, or patients in the 50-mg golimumab group, and 5 patients sulfasalazine), were consuming alcohol, or had evi- in the 100-mg golimumab group (2 patients in the 100-mg group had markedly abnormal levels of both Prophylaxis for latent TB consisted of isoniazid ALT and AST). Three patients (all in the 100-mg monotherapy or a combination of isoniazid and ri- group) discontinued the study agent because of ele- fampin. The proportions of patients with abnormal ALT vations in their ALT or AST level. Some of the 6 or AST values were generally higher among those who patients who continued to receive treatment had received TB prophylaxis compared with those who did concomitant medication adjustments. None of the 9 not receive prophylaxis (data not shown).
patients with a markedly abnormal ALT or AST value, A greater proportion of patients in the goli- including the patient with hepatitis, demonstrated mumab groups (46.4% and 48.6% in the 50-mg and symptoms of liver toxicity, and none had a concurrent 100-mg groups, respectively) had at least 1 infection markedly abnormal bilirubin value. Transaminase lev- through week 24 compared with the placebo group els decreased to normal or to Ͻ3-fold the upper limit (36.4%). Three patients had serious infections through of normal by week 24 in 8 of the 9 patients with week 24: 1 patient in the 100-mg group had infectious mononucleosis, 1 patient in the 100-mg group had The efficacy results of this study were similar in chronic otitis media, and 1 patient in the placebo group magnitude to those of previous studies of anti-TNF␣ had gastrointestinal inflammation. Through week 24, 2 agents in AS (2–4), although no data are available from patients (1 in the placebo group and 1 in the 100-mg direct comparisons in a single head-to-head study. How- golimumab group) had a malignancy; both had basal cell ever, currently available subcutaneous anti-TNF␣ agents are administered either twice weekly or once every 2 Through week 24, 8.7% and 6.4% of patients in weeks. Thus, monthly dosing with golimumab would the 50-mg and 100-mg groups, respectively, had at least provide a more convenient dosing schedule compared 1 injection-site reaction compared with 2.6% of patients with these agents. In addition, recent observational in the placebo group. The most common injection-site studies have shown that patients with spondylarthritis reaction was erythema. No injection-site reactions were respond well to treatment with a second anti-TNF␣ serious. One patient in the early escape group who agent (15,16), although golimumab was not evaluated in began treatment with 50 mg and then received 100 mg of golimumab experienced a severe injection-site reaction A recent study showed that patients with elevated (erythema over a 7.5 ϫ 6.5–inch area on the arm). No CRP levels and extensive spinal inflammation detected patients discontinued the study agent because of by magnetic resonance imaging were likely to respond to anti-TNF therapy (17). In our study, a greater propor- Through week 14, 10 (10.1%) of 99 patients in tion of patients with high screening CRP values achieved the 50-mg group, 16 (14.8%) of 108 patients in the ASAS20 responses compared with those with lower 100-mg group, and 8 (12.9%) of 62 patients in the screening CRP levels. However, significantly more placebo group had newly positive test results for antinu- golimumab-treated patients achieved an ASAS20 re- clear antibodies. Of these, 1 patient (6.3%) in the sponse compared with patients in the placebo group, 100-mg group also became positive for anti–double- regardless of the baseline CRP level.
stranded DNA antibodies. No patient had clinical fea- No statistically significant differences in median tures suggestive of drug-induced lupus.
BASMI scores at week 14 or week 24 were observedbetween the treatment groups. However, improvementsin individual BASMI components (lumbar flexion, lum- DISCUSSION
bar side flexion, intermalleolar distance), the propor- This phase III study was the first to evaluate the tions of patients with Ն1-unit improvement in the efficacy and safety of subcutaneous injections of goli- BASMI, and the improvement in chest expansion sug- mumab in patients with AS. The study population, gest that golimumab may have provided some improve- consisting of patients with moderate-to-severe symptoms ment in range of motion to patients in this study. In of back pain, was similar to that in previous studies of addition, we used a 3-point scale for the BASMI, which anti-TNF␣ agents in patients with AS, except that this was recently shown to be less sensitive to change than was the first study to include a sizable proportion of Fatigue is a well-described symptom in patients Injections of golimumab (50 mg or 100 mg) every with AS (19,20); however, few studies have systemati- 4 weeks resulted in rapid, significant, and sustained cally evaluated sleep. This was the first study to evaluate improvement in the signs and symptoms of AS through the effect of an anti-TNF agent on sleep in patients with week 24. No clear difference in efficacy was evident AS. The results show that golimumab-treated patients between the 50-mg and 100-mg dose groups through had a significantly greater reduction in sleep distur- week 24. The primary efficacy end point was robust to bance, as measured by the JSEQ, than patients in the sensitivity analyses, even when the longer disease dura- placebo group. Patients receiving golimumab also dem- tion in the placebo group was accounted for in a logistic onstrated improved health-related quality of life as regression analysis. Results of other outcomes, including measured by the physical and mental component sum- the ASAS40 response, ASAS partial remission, the mary scores of the SF-36 Health Survey.
ASAS5/6 response, back pain, inflammation, 50% im- provement in the BASDAI, and the BASFI, provided achieved steady state at week 12 and generally increased further evidence that golimumab-treated patients in a dose-proportional manner through week 24. Al- showed significant and clinically meaningful improve- though there was not a clear relationship between the trough serum golimumab concentration and the ASAS20 response, it is possible that the lower serum fective blockade of proinflammatory cytokines via an golimumab concentration observed in heavier patients acceptable route of administration is a priority for the (e.g., patients in the 50-mg group weighing Ͼ87 kg) management of AS. Golimumab doses of 50 mg or 100 might partially account for the lower ASAS20 response mg administered subcutaneously every 4 weeks were in these patients. Patients who met the criteria for early effective and well tolerated. Additional studies are escape from 50 mg to 100 mg of golimumab at week 16 planned to determine the long-term efficacy and safety generally had lower trough serum concentrations than those who did not require early escape. However, con-sidering the large interpatient variability, it is not clear ACKNOWLEDGMENTS
whether this difference between the groups was clinicallymeaningful. Only 4 of 25 patients (16%) who had dose We thank the patients, investigators, and study person- escalations from 50 mg to 100 mg were ASAS20 re- nel who made this study possible. We also thank Professors sponders after 8 weeks. The median serum trough Kurt de Vlam, Maxime Dougados, J. S. Hill Gaston, andMarjatta Leirisalo-Repo for their work as national coordina- concentrations at week 24 were similar for responders and nonresponders in this group. Thus, the inadequateresponse in most of these patients at week 16 was likelyattributable to factors other than trough serum concen- AUTHOR CONTRIBUTIONS
trations. Serum golimumab concentrations were gener- Dr. Inman had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the ally low in patients who were positive for antibodies to golimumab. The incidence of antibodies to golimumab Study design. Inman, Davis, van der Heijde, Mack, Han, Hsu, Beutler,
was low (4.1%), and thus there were too few antibody- Braun.
Acquisition of data. Inman, Davis Diekman, Sieper, Kim, Hsu, Braun.
positive patients to properly discern whether antibodies Analysis and interpretation of data. Inman, Davis van der Heijde,
to golimumab affected efficacy and safety.
Mack, Han, Visvanathan, Xu, Hsu, Beutler, Braun.
Golimumab was generally well tolerated by pa- Manuscript preparation. Inman, Davis van der Heijde, Sieper, Mack,
Han, Xu, Hsu, Beutler, Braun, and Scott Newcomer (nonauthor;
tients in this study, and the pattern of adverse events was consistent with the known safety profile of TNF␣ inhib- Statistical analysis. Mack, Han.
itors. Patients receiving golimumab had a slightly greater Trial coordination. Mischa Engel, Renato Gusinu, Andrea Bevis, Sean
Murphy, Cecile Spiertz, Angela Rommens (nonauthors; Centocor).
proportion of infections than those in the placebo group.
The most common infections were nasopharyngitis andupper respiratory tract infection. Serious infections were ROLE OF THE STUDY SPONSOR
uncommon, occurring in 1 patient in the placebo group The conduct of the study was managed by a steering commit- and 2 patients in the golimumab group. Serious adverse tee consisting of Drs. Inman, Davis, Braun, van der Heijde, and Hsu.
This committee also designed the study, with input from the Centocor events were also infrequent and occurred in similar clinical trial team. Clinical data were collected by the investigators proportions in the golimumab and placebo groups and/or study-site personnel and analyzed by Centocor statisticians and through weeks 16 and 24. Overall, adverse events oc- programmers, led by Drs. Mack and Han. An independent data-monitoring committee, paid for by Centocor, conducted periodic curred more frequently in patients who were not receiv- safety reviews. The study data were interpreted primarily by the ing DMARDs at baseline compared with those who steering committee, with contributions from the other authors. All were receiving concomitant DMARDs at baseline, re- authors reviewed the manuscript during its development, agreed tosubmit the manuscript, and approved the content of the final manu- gardless of treatment group assignment. There were no deaths, cases of TB, or opportunistic infections.
Injection-site reactions were infrequent and mostly mild, REFERENCES
consisting primarily of injection-site erythema.
Nine patients had transient, markedly abnormal 1. Braun J, Bollow M, Neure L, Seipelt E, Seyrekbasan F, Herbst H, et al. Use of immunohistologic and in situ hybridization techniques liver enzyme values (Ն100% increase from baseline and in the examination of sacroiliac joint biopsy specimens from a value Ͼ150 IU/liter), which have been previously patients with ankylosing spondylitis. Arthritis Rheum 1995;38: reported in patients with AS who receive anti-TNF␣ 2. Davis JC Jr, van der Heijde D, Braun J, Dougados M, Cush J, agents (3,4). Levels subsequently decreased in all pa- Clegg DO, et al, for the Enbrel Ankylosing Spondylitis Study tients, including those who continued to receive the Group. Recombinant human tumor necrosis factor receptor (et- study agent. None of these patients had symptoms of anercept) for treating ankylosing spondylitis: a randomized, con-trolled trial. Arthritis Rheum 2003;48:3230–6.
3. Van der Heijde D, Dijkmans B, Geusens P, Sieper J, DeWoody K, Identifying treatments that provide safe and ef- Williamson P, et al, and the Ankylosing Spondylitis Study for the Evaluation of Recombinant Infliximab Therapy Study Group.
Spadaro A, et al. Switching tumour necrosis factor ␣ antagonists in Efficacy and safety of infliximab in patients with ankylosing patients with ankylosing spondylitis and psoriatic arthritis: an spondylitis: results of a randomized, placebo-controlled trial (AS- observational study over a 5-year period. Ann Rheum Dis 2007; SERT). Arthritis Rheum 2005;52:582–91.
4. Van der Heijde D, Kivitz A, Schiff MH, Sieper J, Dijkmans BA, 17. Rudwaleit M, Schwarzlose S, Hilgert ES, Listing J, Braun J, Sieper Braun J, et al, for the ATLAS Study Group. Efficacy and safety of J. Magnetic resonance imaging (MRI) in predicting a major adalimumab in patients with ankylosing spondylitis: results of a clinical response to anti-TNF-treatment in ankylosing spondylitis.
multicenter, randomized, double-blind, placebo-controlled trial.
Ann Rheum Dis 2007. E-pub ahead of print.
18. Van der Heijde D, Landewe R, Feldtkeller E. Proposal of a linear 5. Kay J, Matteson EL, Dasgupta B, Nash P, Durez P, Hall S, et al.
definition of the Bath Ankylosing Spondylitis Metrology Index Golimumab in patients with active rheumatoid arthritis despite (BASMI) and comparison with the 2-step and 10-step definitions.
treatment with methotrexate: a randomized, double-blind, place- bo-controlled, dose-ranging study. Arthritis Rheum 2008;58: 19. Calin A, Edmunds L, Kennedy LG. Fatigue in ankylosing spondy- litis: why is it ignored? J Rheumatol 1993;20:991–5.
6. Kavanaugh A, McInnes I, Mease P, Krueger GG, Gladman D, 20. Van Tubergen A, Coenen J, Landewe R, Spoorenberg A, Chorus Gomez-Reino JJ, et al. Golimumab, a new, human, TNF-␣ A, Boonen A, et al. Assessment of fatigue in patients withankylosing spondylitis: a psychometric analysis. Arthritis Rheum antibody administered as a monthly subcutaneous injection in psoriatic arthritis: 24-week efficacy and safety results of therandomized, placebo-controlled GO-REVEAL study [abstract].
Arthritis Rheum 2007;56:4308–9.
APPENDIX A: PRINCIPAL INVESTIGATORS AND
7. Van der Linden S, Valkenburg HA, Cats A. Evaluation of STUDY CENTERS
diagnostic criteria for ankylosing spondylitis: a proposal for mod-ification of the New York criteria. Arthritis Rheum 1984;27:361–8.
The investigators who participated in this trial are as follows: 8. Anderson JJ, Baron G, van der Heijde D, Felson DT, Dougados Piet Geusens (Diepenbeek, Belgium), Frank Raeman (Merksem, M. Ankylosing Spondylitis Assessment Group preliminary defini- Belgium), Serge Steinfeld (Brussels, Belgium), Thierry Appelboom tion of short-term improvement in ankylosing spondylitis. Arthritis (Brussels, Belgium), Kurt de Vlam (Leuven, Belgium), Christopher Atkins (Victoria, British Columbia, Canada), Andre Beaulieu (Sainte 9. Brandt J, Listing J, Sieper J, Rudwaleit M, van der Heijde D, Foy, Quebec, Canada), Mary Bell (Toronto, Ontario, Canada), Robert Braun J. Development and preselection of criteria for short term Inman (Toronto, Ontario, Canada), Majed Khraishi (St. John’s, New- improvement after anti-TNF␣ treatment in ankylosing spondylitis.
foundland, Canada), Walter Maksymowych (Edmonton, Alberta, Can- ada), Proton Rahman (St. John’s, Newfoundland, Canada), Glen 10. Garrett S, Jenkinson T, Kennedy LG, Whitelock H, Gaisford P, Thomson (Winnipeg, Manitoba, Canada), Kamran Shojania (Vancou- Calin A. A new approach to defining disease status in ankylosing ver, British Columbia, Canada), Marjatta Leirisalo-Repo (Helsinki, spondylitis: the Bath Ankylosing Spondylitis Disease Activity Finland), Eero Jukka (Rauma, Finland), Yves Maugars (Nantes, Index. J Rheumatol 1994;21:2286–91.
¨rgen Braun (Herne, Germany), Christoph Fiehn (Baden- 11. Calin A, Garrett S, Whitelock H, Kennedy LG, O’Hea J, Mallorie Baden, Germany), Mathias Gruenke (Erlangen, Germany), Michael P, et al. A new approach to defining functional ability in ankylos- ¨uerle (Erlangen, Germany), Herbert Kellner (Munich, Germany), ing spondylitis: the development of the Bath Ankylosing Spondy- Stefan Schewe (Munich, Germany), Joachim Sieper (Berlin, Ger- litis Functional Index. J Rheumatol 1994;21:2281–5.
many), Henning Zeidler (Hannover, Germany), Reinhold Schmidt 12. Jenkinson TR, Mallorie PA, Whitelock HC, Kennedy LG, Garrett SL, Calin A. Defining spinal mobility in ankylosing spondylitis Piet van Riel (Nijmegen, The Netherlands), De (AS): the Bath AS Metrology Index. J Rheumatol 1994;21:1694–8.
(Maastricht, The Netherlands), Sjef van der Linden (Maastricht, The 13. Ware JE Jr, Sherbourne CD. The MOS 36-item Short-Form Netherlands), Won Tae Chung (Busan, South Korea), Sung Il Kim health survey (SF-36). I. Conceptual framework and item selec- (Busan, South Korea), Tae-Hwan Kim (Seoul, South Korea), Sung Hwan Park (Seoul, South Korea), Gwan Gyu Song (Seoul, South 14. Jenkins CD, Stanton BA, Niemcryk SJ, Rose RM. A scale for the Korea), Yeong Wook Song (Seoul, South Korea), Chung-Ming Huang estimation of sleep problems in clinical research. J Clin Epidemiol (Taichung, Taiwan), Joung-Liang Lan (Taichung, Taiwan), Wen-Chan Tsai (Kaohsiung City, Taiwan), John Davis, Jr. (San Francisco, CA), 15. Coates LC, Cawkwell LS, Ng NW, Bennett AN, Bryer DJ, Fraser Atul Deodhar (Portland, OR), Guy Fiocco (La Crosse, WI), Richard AD, et al. Real life experience confirms sustained response to Jimenez (Edmonds, WA), Philip Mease (Seattle, WA), Frederick long-term biologics and switching in ankylosing spondylitis. Rheu- Murphy (North Duncansville, PA), Mark Pearson (Brookfield, WI), matology (Oxford) 2008;47:897–900.
Eric Peters (Paradise Valley, AZ), John Reveille (Houston, TX), Yong 16. Conti F, Ceccarelli F, Marocchi E, Magrini L, Spinelli FR, Tsai (Ormond Beach, FL), Michael Weisman (Los Angeles, CA).

Source: http://simponi.se/Documents/simponi/Inman_RD_et_al._Arthritis_Rheum._2008%3B58(11)3402-12.pdf