Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

ABSTRACTS in PERSPECTIVE based on presentations from the 72nd Annual Meeting of the American College of Rheumatology

San Francisco, California / October 24-29, 2008

EDITORIAL REVIEW:

Majed M. Khraishi, MD, FRCPC

Medical Director (Rheumatology), Nexus Clinical Research, Professor of Medicine, Memorial University of Newfoundland, St. John’s, Newfoundland

SEARCHING FOR OPTIMAL ANTI-TNF THERAPY FOR RA

Inhibitors of tumour necrosis factor-alpha (TNF-a) have profoundly affected the treatment of rheumatoid arthritis (RA). More than a decade after their introduction, TNF-a inhibitors have evolved from salvage therapy to a potential option for first-line treatment of early-stage disease.

The clinical success of TNF-a inhibitors has helped fuel continued research and development in the field, as researchers seek to optimize use of currently available agents and produce new agents that offer potential advantages over existing treatment options. As the newest member of the TNF-a inhibitor class, golimumab is a fully human IgG1 monoclonal antibody and has simplified the treatment process to a once-a-month subcutaneous (s.c.) injection. It also is the first agent in a randomized, controlled clinical trial to have proven that RA patients with prior exposure to anti-TNF therapy can be treated successfully with another agent in the same class. Studies presented at this year’s 2008 Annual Meeting of the American College of Rheumatology demonstrated its efficacy in a wide-ranging RA patient population, including methotrexate (MTX)-naive patients and in those whose RA remains active despite MTX treatment. The studies reviewed also demonstrated good tolerability with no new safety signals related to anti-TNF therapy.

The reflected ongoing research efforts to identify potential biomarkers of treatment success with TNF-a inhibitors improved our understanding of currently available TNF-a inhibitors and provided additional support for the use of TNF-a inhibitors in RA.

Anti-TNF Therapy After Discontinuing Prior Anti-TNF Therapy

In clinical practice, physicians will often prescribe a TNF-a inhibitor for RA patients who have had previous exposure to a different one. However, the clinical wisdom of the strategy had never been fully examined in a randomized, controlled clinical trial. Dr. Josef Smolen, Medical University in Vienna, Austria, presented data from a study of 461 patients randomized to placebo or to golimumab 50 mg or 100 mg every 4 weeks in patients with previous history of TNF-a inhibitor therapy. Patients on background treatment with nonbiologic disease-modifying antirheumatic drugs (DMARDs) at enrolment continued those agents. About 60% of the patients had discontinued prior anti-TNF therapy because of lack of effectiveness. Two-thirds of the patients had been treated previously with a single TNF-a inhibitor, 25% had exposure to two TNF-a inhibitors and 9% had already been treated with three. The primary end point was the proportion of patients who achieved an ACR20 response at 14 weeks. Significantly more patients in the golimumab arms met the primary end point (35.3% with 50 mg and 37.9% with 100 mg vs. 18.1% with placebo, P<0.001). The improvement was maintained at 24 weeks (34% and 43.8% vs. 16.8%, P<0.001). The golimumab groups also demonstrated superiority with respect to disease activity score 28 (DAS28) improvement at weeks 14 and 24 and health assessment questionnaire (HAQ) improvement at 24 weeks (P<0.001 for all comparisons). The frequency and severity of adverse events did not differ substantially among treatment groups.

Methotrexate-refractory Disease

Before the advent of TNF-a inhibitors, MTX had few challengers as the most effective therapy for RA. Although effective, MTX does not achieve or maintain adequate responses in a substantial proportion of patients. Dr. Edward Keystone, University of Toronto, Ontario, reported findings from a study that demonstrated the superiority of combination therapy with MTX and golimumab vs. MTX monotherapy. The trial involved 444 patients who were randomized to four treatment groups: MTX alone, golimumab 100 mg alone, or MTX combined with either golimumab 50 mg or 100 mg. As early as 14 weeks, patients treated with the combinations demonstrated significantly greater improvement vs. MTX alone in ACR response (20, 50, and 70), HAQ, DAS28 response and remission.

Impact of Therapy on Fatigue and Physical Function

RA can result in substantial fatigue and profoundly and adversely affect physical function, both of which are important outcomes from the patient’s perspective. An analysis of another study of golimumab in patients with prior anti-TNF therapy showed that the new TNF-a inhibitor significantly improved patient-reported fatigue and physical function. Reported by Dr. Jonathan Kay, Harvard Medical School, Boston, Massachusetts, the study involved 459 patients with active RA who were randomized to monthly treatment with placebo, golimumab 50 mg or 100 mg. The primary end point was the change in physical function and fatigue at 24 weeks, as assessed by validated survey instruments. Scores on the health assessment questionnaire (HAQ) decreased by 0.05 in the placebo group vs. 0.25 in the 50-mg golimumab group and 0.38 in the 100-mg golimumab group (P<0.001). The proportion of patients who had a change in HAQ of at least 0.25 (minimal value for clinical significance) also was significantly greater with 50 mg and 100 mg doses (50.3% and 53.6%, respectively) compared with 34.2% in the placebo group. Both golimumab groups were significantly different from placebo, as was the combined results of the two groups (51.9%, P<0.001). The TNF-a inhibitor also demonstrated superiority on the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F). Placebo-treated patients had mean improvement on FACIT-F of 3 points at 24 weeks, vs. 6.0 with golimumab 50 mg (P<0.05) and 7.5 with golimumab 100 mg (P<0.001). The between-group differences were statistically significant at a 14-week interim analysis (P<0.001).

Biomarkers to Predict Treatment Outcomes

The ability to recognize positive and negative laboratory markers of treatment outcome would benefit clinicians and patients alike. Clinicians might be able to discontinue an ineffective therapy and switch a patient to a different drug without a prolonged clinical trial of the ineffective agent. Patients would benefit from avoiding the pain and disability that would persist in the face of ineffective therapy. Timely therapeutic switching would also save the expense associated with a trial of a drug that ultimately will not help the patient. Protein profiling was performed for more than 100 acute-phase, inflammatory and other biomarkers to determine differential expression of the proteins. The laboratory analysis involved serum samples from a randomized clinical trial of MTX plus placebo, golimumab plus placebo, or MTX plus two different doses of golimumab (50 mg and 100 mg) in patients with early RA. Levels of some markers changed significantly as early as 4 weeks after the start of treatment. The markers were compared with response data to identify potential predictors of response. More than a dozen markers had significant associations at 14 weeks with ACR20 response, ACR50 response and DAS28. Changes in vonWillebrand Factor and pyridinoline proved to have the strongest associations.

TNF Variant Predicts Treatment Response

In RA, considerable interest has centred on a single nucleotide polymorphism in the human TNF-a gene promoter: the -308 G/A variant. Unfortunately, studies conducted to date have been too small to provide conclusive evidence that the 308 genotype predicts response to anti-TNF therapy. A meta-analysis led by Dr. Darren D. O’Rielly, Memorial University of Newfoundland, St. John’s, provided the first comprehensive examination of the variant’s influence on response to anti-TNF therapy in patients with RA. Updated findings of nine published studies included a total of 692 patients, 179 of whom had either the GA or A/A genotype were presented. As compared with the homozygous G/G genotype, patients with the variant genotype were less likely to respond to anti-TNF therapy, as determined by ACR20 and DAS28 criteria. Patients with the G/G genotype had an overall response rate of 80%, whereas G/A and A/A combined were associated with an overall response rate of 66%. The odds ratio for response among carriers of the A allele was 0.43 (0.28, 0.68).

Efficacy by Disease Severity

Randomized clinical trials of TNF inhibitors invariably have applied exclusion criteria to enrolment of patients with RA. In clinical practice, however, TNF-a inhibitors are prescribed more broadly, and include many patients with low disease-activity states. The effectiveness of TNF-ainhibitors across a range of RA severity had not been carefully examined. Dr. Jeffrey Greenberg, New York University, New York, reported findings from a retrospective study that compared anti-TNF outcomes in patients with high vs. moderate disease activity. The analysis involved 437 patients from a large RA registry. The patients had been followed for five to 10 months after initiation of treatment. Disease activity was determined by Clinical Disease Activity Index (CDAI) criteria. The primary outcomes were the proportion of patients in remission, low disease activity state (LDAS), and modified ACR20 and ACR50 responses. Unadjusted outcomes were similar between patients with high and moderate disease activity, except for LDAS, which occurred in more patients with moderate disease activity. Sensitivity analysis involving a subgroup of 134 patients revealed similar results when employing different definitions for remission and other outcomes. The data provided evidence that outcomes achieved with TNF inhibitors in clinical practice are similar to those achieved in clinical trials and that patients with moderate and severe RA derive similar benefits from anti-TNF therapy.

Note: At press time, golimumab is not available in Canada.

PSORIATIC ARTHRITIS, ANKYLOSING SPONDYLITIS AND TNF INHIBITION

EDITORIAL REVIEW:

Edward C. Keystone, MD, FRCPC

Director, The Rebecca MacDonald Centre for Arthritis and Autoimmune Disease Director, Division of Advanced Therapeutics in Arthritis Chair, Canadian Rheumatology Research Consortium Consultant (Rheumatology) UHN-Mount Sinai Hospital, Professor of Medicine, University of Toronto Toronto, Ontario

The role of tumour necrosis factor alpha (TNF-a) inhibitors in rheumatic diseases has expanded since the drug class’ introduction more than a decade ago. The clinical expansion has included treatment of psoriatic arthritis (PsA) and ankylosing spondylitis (AS), conditions that have the potential to cause as much physical and emotional harm as rheumatoid arthritis (RA).

The estimated prevalence of PsA ranges from 6% to 39% of patients with psoriasis. The Psoriasis Foundation 2001 Benchmark Survey estimated that PsA may affect as many as 23% of the psoriasis patient population.

Differential diagnosis of PsA can pose a major clinical challenge, as the clinical features can overlap with those of a variety of other conditions, including RA, reactive arthritis, AS and inflammatory bowel disease (IBD). Diagnosis can be further complicated by coexistence of other rheumatologic conditions, such as osteoarthritis, soft-tissue rheumatism and septic arthritis. Moverover, studies have shown that patients with IBD and AS have a higher prevalence of psoriasis.

Historically, PsA has been treated with conventional DMARDs, particularly methotrexate (MTX) and cyclosporine. More recently, several TNF-a inhibitors have received an indication for the disease.

As the newest of the TNF-a inhibitors, golimumab has demonstrated efficacy in PsA. The agent has several potential advantages in clinical practice. It is administered as a once-monthly subcutaneous injection. It neutralizes both soluble and membrane-bound forms of TNF. Additionally, it is the only anti-TNF agent that has data from a randomized clinical trial demonstrating efficacy in patients with prior exposure to one or more TNF-a inhibitors.

AS is one of the major subtypes of the chronic inflammatory disorders known as spondyloarthropathies. AS primarily affects sacroiliac joints and the axial skeleton. Principal clinical features include back pain, inflammation at various sites in the axial skeleton, peripheral arthritis, enthesitis and anterior uveitis.

TNF-a inhibitors have demonstrated efficacy in the treatment of AS, particularly in patients with high disease activity. The rationale for treating PsA and AS with TNF-ainhibitors has emerged from recognition that levels of the cytokine are elevated in patients with those diseases. When released into circulation, TNF binds to naturally occurring receptors on cell surfaces. In a healthy state, TNF receptors appear to serve as a counterbalance to TNF. In disease states associated with elevated levels of TNF, however, receptors are thought to be overwhelmed by the increased volume of the cytokine and lose their regulatory capacity for TNF.

As with RA and PsA, AS is associated with elevated levels of TNF, making TNF-a inhibitors a rational and reasonable approach to therapy. The three existing TNF-a inhibitors are indicated for treatment of AS, and patients in phase III clinical trials of golimumab have experienced marked improvement in the disease.

Presentations at the American College of Rheumatology meeting reinforced the value of TNF-a inhibition in the management of these complex diseases.

Maintaining Quality of Life in AS

As with any chronic disease, preserving quality of life is a major goal of therapy for AS. Therefore, a treatment that improves disease status and preserves or improves quality of life would offer an attractive therapeutic option for patients with AS. To that end, Dr. Désirée van der Heijde, Leiden University Medical Center, The Netherlands, reported health-related quality-of-life data from a placebo-controlled clinical trial evaluating two different doses of golimumab. The analysis comprised both the physical and mental component summaries of the widely used SF-36 quality-of-life assessment instrument. The results showed significantly greater improvement in both summary scores among patients treated with golimumab.

For the physical component summary, the two doses of golimumab demonstrated superiority vs. placebo at week 14 (P<0.001), and the improvement was maintained with both doses at week 24 (P<0.001). Moreover, the degree of improvement was similar in the two golimumab groups. Actively-treated patients also experienced significantly greater improvement in the mental component summary of the SF-36 at week 14, with somewhat more improvement with the higher dose (P<0.001 vs. P=0.014 for 50 mg). The improvement was maintained at week 24 with the 100-mg dose but not the 50-mg dose. Collectively, the results showed that treatment of AS with the new TNF-a inhibitor golimumab significantly improves health-related quality of life which is maintained over time.

Improved Productivity in AS

Dr. Juergen Braun, Rheumazentrum Ruhrgebiet, Herne, Germany, and international collaborators examined the effects of golimumab on productivity in patients with active AS. Patients were randomized to placebo or to one of two doses of golimumab (50 or 100 mg once a month) and followed for 24 weeks. Patients with inadequate response on placebo at week 16 entered early escape with golimumab 50 mg for the remaining eight weeks, and patients in the 50-mg golimumab group could enter early escape with golimumab 100 mg. Patients in the 100-mg group continued treatment at the same dose. The primary end point was patient self-rated change in productivity as assessed by a visual analogue scale at 24 weeks. Both active treatment groups resulted in significantly greater improvement at 16 and 24 weeks (P<0.001). At the end of the study, patients originally assigned to placebo had mean improvement of 0.5 compared with 2.7 and 2.9 in the 50- and 100-mg groups.

Long-Term Efficacy in PsA

Dr. Arthur Kavanaugh, University of California, San Diego, reported data on behalf of colleagues in the US and Canada from a trial evaluating 52-week results with golimumab in patients with active PsA. Patients were randomized to placebo or to 50 mg or 100 mg and followed for a total of 52 weeks. Placebo patients with an inadequate response at 16 weeks were randomized to early escape with one of the two golimumab doses. Patients still on placebo at 20 weeks switched to 50 mg for weeks 24 to 52. The primary end point was the proportion of patients who achieved an ACR20 response and improvement of at least 75% on the Psoriasis Area and Severity Index (PASI75). At the end of 24 weeks, 12% of placebo patients had achieved an ACR20 response compared with 52% and 61% for golimumab 50 mg and 100 mg, respectively (P<0.001). Only 1% of placebo patients had a PASI75 at 24 weeks vs. 56% and 66% of the golimumab groups (P<0.001). Both actively-treated groups also had significantly greater improvement in the DAS28, Health Assessment Questionnaire (HAQ), and Nail Psoriasis Activity Index at weeks 14 and 24.

Effects on Nail Disease, Enthesitis and Dactylitis

Investigators in a multicentre North American study examined the efficacy of golimumab on the outcomes of nail disease, enthesitis and dactylitis in 405 patients with PsA. As in the previously described studies, patients were randomized to placebo or to one of two doses of golimumab and followed for 24 weeks. Nail changes were based on the single worst psoriasis-affected nail at baseline, as assessed by the Nail Psoriasis Severity Index (NAPSI) and by the Nail Physician’s Global Assessment. Dactylitis was assessed in all 20 digits and rated on a scale of 0 (none) to 3 (severe). The presence of enthesitis was evaluated at 15 body sites. At the 24-week evaluation, as reported by Dr. Dafna Gladman, University of Toronto, Ontario, patients treated with either dose of golimumab had significant improvement in the NAPSI, Nail Physician’s Global Assessment (PGA), and enthesitis (P<0.001 for all comparisons). Additionally, the 100-mg dose led to statistically significant improvement in dactylitis compared with placebo (P<0.001). The study is ongoing, and follow-up will continue for 52 weeks.

Predictors of Spondylitis in PsA

Dr. Vinod Chandran, Toronto Western Hospital, described the frequency of spondylitis and factors associated with the condition in 206 patients with PsA. During 10 years of follow-up, 15% of the cohort developed spondylitis. Factors associated with spondylitis were gender, family history of PsA, nail dystrophy, number of swollen joints, enthesitis, number of radiographically damaged joints, periostitis, calcaneal spurs, erythrocyte sedimentation rate (ESR), and hypertension. In a multivariate analysis, factors that independently predicted spondylitis were nail dystrophy, periostitis, number of radiolographically damaged joints and ESR. Somewhat paradoxically, the swollen joint count appeared to have a protective effect against sacroiliitis (RR 0.83, P<0.01). However, additional analyses exploring the robustness of the findings showed that swollen joint count was neither protective nor predictive.

Note: At press time, golimumab is not available in Canada.

Abstracts Relevant to

SEARCHING FOR OPTIMAL ANTI-TNF THERAPY FOR RA

ABSTRACT 968 - Golimumab, a Human Anti-TNF-a Monoclonal Antibody, Injected Every 4 Weeks: Efficacy & Safety in RA Patients Previously Treated with Anti-TNF-a Agents (GO-AFTER Study)

J. Smolen, J. Kay, M. K. Doyle, R. Landewé, E. L. Matteson, J. Wollenhaupt, N. Gaylis, F. Murphy, J. Neal, Y. Zhou, S. Visvanathan, E. C. Hsia, M. U. Rahman

Purpose: To evaluate the efficacy and safety of golimumab (GLM) in patients (pts) with active rheumatoid arthritis (RA) previously treated with anti-TNFa agent(s).

Methods: In this multicenter, randomized, double-blind, PBO-controlled study, pts with active RA (=4 tender joints and 4 swollen joints) were randomized to subcutaneous injections of PBO, GLM 50 mg or 100 mg q 4 wks. Pts continued to receive stable doses of methotrexate (MTX), sulfasalazine (SSZ), and/or hydroxychloroquine (HCQ) if receiving them at baseline. Pts could have received =1 anti-TNFa agent(s) and discontinued for any reason(s). The primary endpoint was the proportion of pts with =20% improvement in ACR 20 at wk 14. Improvement from baseline in the Health Assessment Questionnaire (HAQ) at wk 24 was also assessed.

Results: 461 pts were randomized. Mean disease duration ranged from 8.65-9.80 yrs; 97% of pts were RF or anti-CCP positive, 72% were positive for both. At baseline, 66% were receiving MTX; 5% and 7% SSZ and HCQ, respectively. All pts had received at least one anti-TNFa agent, 24.9% had received two, and 9.3% had received three. Prior anti-TNFa therapy had been discontinued due to lack of efficacy (58.4%) or non-efficacy related reasons (53.4%).

Among pts who discontinued previous anti-TNFa therapy due to lack of efficacy, 35.7% (p=0.006) and 42.7% (p<0.001) in the GLM 50 mg and 100 mg grps had an ACR 20 response at wk 14 compared with 17.7% of pts in the PBO grp. Through wk 24, 72.3%, 66.4%, and 78.3%, in the PBO, GLM 50 mg, and 100 mg grp had >1 adverse event, respectively. Among pts in the PBO, GLM 50 mg and 100 mg grps, serious adverse events occurred in 9.7%, 7.2%, and 4.6%, respectively. Serious infections were reported in 3.2%, 3.3%, 0.7%, and injection-site reactions (ISR) through wk 16 were reported in 2.6%, 3.9%, and 10.5%, of pts in the PBO, GLM 50 mg, and 100 mg grps, respectively. The most commonly reported ISR was erythema. No severe reactions were reported and none led to discontinuation.

Conclusion: GLM significantly reduced the signs and symptoms of RA and improved physical function in pts who had previously received anti-TNFa therapy and discontinued for any reason. GLM was generally well-tolerated.

Commentary on abstract 968

This study represented a systematic examination of a practice thought to be widespread in clinical management of RA: switching from one TNF-a inhibitor to a different agent in the same class. The trial was the first-ever randomized, placebo-controlled trial of anti-TNF therapy in patients with a history of treatment with a different anti-TNF agent. The study involved 461 patients who were randomized to placebo or to golimumab 50 mg or 100 mg every four weeks. Patients on background treatment with nonbiologic disease-modifying agents at enrolment continued those agents. Patients’ mean disease duration was nine to 10 years, and 97% tested positive for rheumatoid factor or anti-CCP (72% were positive for both). Two-thirds were on background MTX. Almost 60% of the patients had discontinued prior anti-TNF therapy due to lack of efficacy. Two-thirds of the patients had been treated previously with a single TNF-a inhibitor, 25% had exposure to two anti-TNF therapies, and 9% had already been treated with three. The primary end point was the proportion of patients who achieved an ACR20 response at 14 weeks. Significantly more patients in each of the golimumab arms met the primary end point (35.3% with 50 mg and 37.9% with 100 mg vs. 18.1% with placebo [P<0.001]). The improvement was maintained at 24 weeks (34% and 43.8% vs. 16.8%, P<0.001). A significantly higher proportion of actively-treated patients also achieved ACR50 and ACR70 responses at 14 and 24 weeks (P=0.011 to P<0.001). Both golimumab groups demonstrated superiority with respect to DAS28 improvement at weeks 14 and 24 and HAQ improvement at 24 weeks (P<0.001 for all comparisons). Aside from more injection-site reactions in the 100-mg group, adverse events and serious adverse events did not differ substantially among treatment groups.

Questions and answers with Dr. Josef Smolen, Medical University of Vienna, Austria

Q: Did response to golimumab differ according to the number of prior exposures to anti-TNF agents?

A: We saw a decline in response from patients with one prior anti-TNF therapy compared with those who had been treated with two agents. Differences between patients with two and three prior anti-TNF therapies were difficult to assess because of the small numbers, but they did not appear to differ substantially.

Q: This was not a direct comparison of two anti-TNF therapies. Nonetheless, how did the responses with golimumab compare with what has been reported for other agents in the class?

A: The efficacy is on the order of what is seen with other biologic agents in a similar population of patients. This suggests that it is viable to use a TNF-a inhibitor after another TNF-a inhibitor, and it is not absolutely necessary to swap biologic agents because the overall response rate will be similar.

Q: How does the activity compare with what would be expected in patients with no history of anti-TNF therapy?

A: Responses are lower, which is what would be expected. The ACR20 response rate is 10 to 15% lower than what is seen across the class of TNF-a inhibitors in patients who have not yet been treated with a TNF inhibitor.

ABSTRACT 1211 - Golimumab Administered Subcutaneously Every 4 Wks in Patients with Active Rheumatoid Arthritis Despite Methotrexate: Wk 24 Results of the Randomized, Double-Blind, Placebo-Controlled, GO-FORWARD Study

E. Keystone, M. C. Genovese, L. Klareskog, E. C. Hsia, J. Livingston, M. Wiekowski, S. T. Hall, P. Miranda, J. Pazdur, S. C. Bae, W. Palmer, Z. Wu, M. U. Rahman

Purpose: To assess the efficacy and safety of golimumab (GLM) in patients (pts) with active RA despite methotrexate (MTX).

Methods: In this multicenter, randomized, DB, placebo (PBO)-controlled study, adult pts with active RA (=4 tender & 4 swollen joints) despite MTX were randomized (3:3:2:2 ratio) to PBO injections + MTX, GLM 100 mg injections + PBO capsules, GLM 50 mg injections + MTX, or GLM 100 mg injections + MTX. Injections were administered q4 wks. Coprimary endpts were the proportion of pts achieving ACR 20 at wk 14 and improvement from baseline (BL) in HAQ at wk 24. Data through wk 24 are presented.

Results: 444 pts with active RA were enrolled. Treatment groups were balanced for BL demographic and disease characteristics. GLM + MTX was significantly better than PBO + MTX in improving signs and symptoms of RA, as well as in improving physical function (Table). The 2 GLM doses + MTX were comparable in efficacy. While the GLM alone group had numerically better efficacy parameters than MTX alone, they generally did not reach statistical significance. GLM was generally well-tolerated. Through wk 16, the proportions of patients who had >1 adverse event was 60.9%, 63.2%, 68.5%, and 69.7% in the PBO + MTX, GLM 100 mg + PBO, GLM 50 mg + MTX, and GLM 100 mg + MTX, respectively. Serious adverse events occurred in 2.3%, 3.8%, 5.6%, and 9.0% of pts, respectively. Serious infections occurred in 0.8%, 2.3%, 2.2%, and 5.6% of pts, respectively. Injection site reactions occurred in 0.0%, 5.3%, 3.4%, and 4.5% of pts, respectively. One patient in the GLM 100 mg alone group died due to diarrhea, colitis, and sepsis. Three pts had malignancies: 1 breast cancer (GLM 100mg + MTX), 1 Bowen’s disease and squamous cell skin cancer (GLM 100 mg + PBO),and 1 basal cell carcinoma (PBO + MTX). There were no reports of
unistic infections.

<img2729|center>

Conclusion: In patients with active RA despite MTX, the addition of GLM 50 mg or 100 mg SC injections q4 wks to MTX significantly reduced the signs and symptoms of RA and improved physical function through wk 24. GLM was generally well-tolerated.

Commentary on abstract 1211

In this randomized clinical trial, golimumab, the newest member of the anti-TNF class of agents, demonstrated superiority over MTX monotherapy. The trial involved 444 patients with active RA despite treatment with MTX. Patients were randomized to four treatment groups: MTX alone, golimumab 100 mgalone, or golimumab either 50 mg or 100 mg plus MTX. As early as 14 weeks, patients treated with either combination therapy demonstrated significantly greater improvement vs. MTX alone in ACR response (20, 50 and 70), HAQ, DAS28 response and remission. The improvement was maintained out to 24 weeks. More than half of patients treated with the treatment combinations had achieved an ACR20 response by week 24 vs. a third of MTX/placebo patients (P<0.001). As monotherapy, it demonstrated non-inferiority even a trend toward superiority over MTX alone (44.4%, P=0.059). At 24 weeks, almost 60% of patients randomized to the two TNF inhibitor combinations met criteria for an ACR20 response (59.6% vs. 27.8% for MTX alone, P<0.001). Additionally, three-fourths of golimumab combination patients had a good or moderate DAS28 response at week 24 vs. 42% for MTX monotherapy (P<0.001), and 21 to 22% of the golimumab combination groups achieved DAS28 remission vs. 6% of the MTX monotherapy group. Rates of adverse events did not differ among treatment groups.

Questions and answers with Dr. Edward Keystone, University of Toronto, Ontario

Q: Does golimumab have any advantages over other TNF inhibitors?

A: The most obvious advantage is convenience. Golimumab is administered once a month, compared with weekly or biweekly with the other agents in the class.

Q: What is its tolerability?

A: It has a safety profile similar to that of other anti-TNF agents.

Q: How did the different golimumab regimens compare with each other and with MTX alone?

A: The 50-mg or 100-mg dose of golimumab combined with MTX was superior to MTX alone. The two combinations were comparable to each other. As a single agent, it generally was no better than MTX alone.

ABSTRACT 1035 - Golimumab Significantly Improves Physical Function & Fatigue in RA Patients Previously Treated with Anti-TNFa Agents: Results from the GO-AFTER Study

J. Kay, M. K. Doyle, J. Smolen, J. Buchanan, E. L. Matteson, E. C. Hsia, R. Landewe, Y. Zhou, S. Parasuraman, M. U. Rahman

Purpose: To evaluate the effect of golimumab (GLM) on physical function & fatigue in patients (pts) with active rheumatoid arthritis (RA) previously treated with anti-TNFa agent(s).

Methods: In this multicenter, double-blind study, pts with active RA (=4 tender & =4 swollen joints) were randomized to subcutaneous placebo (PBO) or GLM 50 mg or 100 mg every 4 wks. Randomization was stratified by investigational site & baseline (BL) methotrexate (MTX) use (yes/no). Pts continued to receive stable doses of MTX, SSZ, &/or HCQ if receiving them at BL. Pts could have received >1 anti-TNF a agent(s) & may have discontinued them for any reason(s). Impact on physical function was assessed using the disability index of the Health Assessment Questionnaire (HAQ). Fatigue was measured using the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F) questionnaire. The total FACIT-F score ranges from 0 to 52, with a higher score indicating less fatigue. An ANOVA on van der Waerden normal scores was performed for between-group comparisons.

Results: Prior anti-TNFa therapy had been discontinued due to lack of efficacy & non-efficacy related reasons (53.4%). At BL, all pts had similar mean ± SD HAQ scores, indicating moderate functional impairment. GLM 50 mg & 100 mg significantly improved HAQ compared with PBO. Pts who had discontinued anti-TNFa therapies due to lack of efficacy had significant improvements from BL in HAQ score compared with PBO at wk 24 (GLM 50 mg, 0.20 ± 0.50; GLM 100 mg 0.26 ± 0.50; Combined 0.23 ± 0.50; PBO 0.06 ± 0.51; p <0.05 for all comparisons). Pooled analysis of pts who had discontinued anti-TNFa therapies due to non-efficacy related reasons showed significant improvements from BL in HAQ score compared with PBO at wk 24 (GLM 50 mg, 0.36 ± 0.58; GLM 100 mg 0.31 ± 0.44; Combined 0.33 ± 0.52; PBO 0.08 ± 0.52; p<0.01 for all comparisons). The proportions of pts achieving >0.25 improvement from BL in HAQ was significantly greater for GLM groups compared with PBO at wk 24. The mean change from BL in FACIT
reater for the GLM groups compared with PBO at wk 14 & wk 24.

<img2730|center>

Conclusion: In RA pts who had received prior anti-TNFa therapy & discontinued for any reason, GLM significantly improved physical function & fatigue through wk 24. Significantly higher proportions of pts in the GLM groups achieved >0.25 improvement in HAQ scores, the minimal clinically important difference.

Commentary on abstract 1035

This study examined the effect of golimumab treatment on physical function and fatigue in patients with active RA and a history of anti-TNF therapy. Investigators randomized 459 patients to monthly treatment with placebo, golimumab 50 mg or 100 mg. The primary end point was the change in physical function and fatigue at 24 weeks, as assessed by validated survey instruments. Scores on the HAQ decreased by 0.05 in the placebo group vs. 0.25 in the 50-mg golimumab group and 0.38 in the 100-mg golimumab group (P<0.001). The proportion of patients who had a change in HAQ of at least 0.25 (minimal value for clinical significance) also was significantly greater with 50 mg and 100 mg (50.3% and 53.6%, respectively) compared with 34.2% in the placebo group. Both golimumab groups were significantly different from placebo, as was the combined results of the two groups (51.9%, P<0.001). It also demonstrated superiority on the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F). Placebo-treated patients had mean improvement on FACIT-F of 3 points at 24 weeks vs. 6 points with 50 mg (P<0.05) and 7.5 points with 100 mg (P<0.001). The between-group differences were statistically significant at a 14-week interim analysis (P<0.001). Overall, the data showed that this agent favourably affects two key issues that significantly influence quality of life in patients with RA.

ABSTRACT 1652 - Serum Markers Associated with Improvement in Clinical Response Measures in RA Patients Treated with Golimumab, A New Human Anti-TNFa Monoclonal Antibody

S. Visvanathan, M. U. Rahman, E. C. Keystone, M. C. Genovese, L. Klareskog, E. C. Hsia, M. Mack, M. Elashoff, C. L. Wagner

Purpose: To identify serum markers that are associated with clinical response in rheumatoid arthritis (RA) patients treated with golimumab (GLM) with and without methotrexate (MTX).

Methods: Sera were collected at wks 0, 4 and 14 from a subset of 100 - 150 patients from the GO-FORWARD study of patients with active RA despite MTX therapy. Patients were randomly assigned in a 3:3:2:2 ratio to receive placebo (PBO) + MTX: GLM 100 mg+ PBO; GLM 50 + MTX; GLM 100 mg + MTX. GLM/PBO was administered SC every 4 wks and MTX/PBO was administered orally every wk. Samples were tested for select inflammatory and bone and cartilage markers and protein profiling using multi-analyte profiles. The % change from baseline in markers was compared between the PBO+MTX and combined GLM + MTX group (50 mg + 100 mg ) and the GLM + PBO group using an ANOVA on the van der Waerden normal scores and t tests. Logistic regression models were used to test for the association of serum markers with clinical endpoints.

Results: Levels of acute phase (CRP, IL-6, a1 anti-trypsin, serum amyloid P, complement 3, haptoglobin), inflammatory (IL-6, VEGF, ICAM-1, MMP-3, TIMP-1, ENRAGE) and other (thyroxin binding globulin, von willebrand factor) markers were significantly decreased at wks 4 and 14 after treatment with GLM + MTX compared with PBO + MTX (p<0.01). Similar markers were modulated by GLM + PBO treatment. Further, baseline levels of pyridinoline and von willebrand factor were significantly associated (OR from 0.101 to 0.444, p<0.05) with select clinical measures (ACR 20, ACR 50, DAS 28) at wk 14 in the combined GLM + MTX group. Changes from baseline to wk 4 in these and additional markers (complement 3, thyroxin binding globulin, plasminogen activator inhibitor 1, IgA, a1 anti-trypsin, hyaluronic acid, IL-16, serum amyloid P, haptoglobin, MMP-3, ENRAGE) were also significantly associated (OR from 0.007 to 3.744, p<0.05) with the same select clinical measures at wk 14 in the combined GLM + MTX group. In addition, GLM + MTX treatment resulted in significant increases at wk 4 in the bone formation markers osteocalcin and PINP, and decreases in the bone degradation marker deoxypyridinoline as compared to PBO + MTX (p<0.05).

Conclusion: Baseline levels and changes at wk 4 in levels of select markers were associated with clinical response to GLM + MTX or GLM + PBO treatment at wk 14 in patients with active RA despite MTX therapy. These results suggest that GLM treatment impacts multiple proteins associated with the TNFa pathway and disease processes of RA.

Commentary on abstract 1652

The ability to identify biomarkers predictive of response to anti-TNF therapy would provide a valuable clinical tool to guide therapy. This study examined changes in serum levels of pro-inflammatory cytokines during treatment with golimumab. Elevated levels of multiple markers of inflammation have been associated with RA. Investigators used serum samples from a randomized clinical trial that evaluated MTX plus placebo, golimumab plus placebo or MTX plus two different doses of golimumab (50 mg and 100 mg).Protein profiling was performed for more than 100 acute-phase, inflammatory and other biomarkers to determine differential expression of the proteins, and 78 met criteria for inclusion in the data analysis. Baseline levels of most of the markers differed minimally among the treatment groups. The analysis revealed a larger number of markers that varied significantly with MTX/golimumab therapy vs. MTX plus placebo. Levels of acute-phase, inflammatory, and select lipid-related markers had declined significantly as early as week 4 and week 14 (P<0.01). Additionally, significant decreases occurred as early as week 4 in markers of bone formation, whereas a marker of bone degradation decreased significantly (P<0.05). The markers were then compared with response data to identify potential predictors of response. Changes in more than a dozen markers had significant associations at 14 weeks with ACR20 response, ACR50 response and DAS28. Changes in two markers, von Willebrand Factor and pyridinoline, had significant associations with multiple clinical end points. It appears that golimumab affects multiple proteins associated with the TNF-a pathway and with disease processes of RA.

Questions and answers with Dr. Sudha Visvanathan, Malvern, Pennsylvania

Q: What is the ultimate goal for evaluation of serum proteins?

A: We hope that individual markers may be used to monitor response to therapy. Changes in the expression of markers early in the course of treatment might identify patients who are likely to respond to therapy. Patients who appear unlikely to have a good response can then be switched to another therapy that might lead to better results.

Q: Do you have any thoughts about why von Willebrand Factor and pyridinoline emerged as the markers most closely associated with clinical end points?

A: This was just a preliminary study. Samples from many more patients and different types of patients need to be analyzed to determine which biomarkers correlate most strongly with clinical outcomes. Not much can be made of associations until there is some evidence of consistency in the associations.

ABSTRACT 1023 - TNF-a -308(a) Variant Predicts Poor Response to TNFa Inhibitors in Rheumatoid Arthritis Patients: A Meta-Analysis

D. D. O’Rielly, N. Roslin, J. Beyene, P. Rahman

Background: TNF-a blockade is a very effective therapy for rheumatoid arthritis (RA). However, a proportion of patients experience either primary or secondary failure with anti-TNF agents in RA.

Objective: To determine if the common TNF-a variant -308(A) can predict poor response to TNF-a inhibitors in RA patients by performing a meta-analysis.

Methods: Studies examining the association of TNF-a-308 variant and responsiveness to TNF-a inhibitors were identified using MEDLINE and abstracts in rheumatology journals. Studies were limited to RA patients, irrespective of ethnicity, receiving any TNF-a inhibitor. Data were extracted for the number of responders and non-responders to TNF-a inhibitors based on DAS28 (response being improvement of at least 1.2 in the DAS 28 score) or achieving at least ACR20, at least 12 weeks after initiation of therapy. Meta-analysis was performed for the presence of the A allele (A/A + A/G) between responders to TNF-a inhibitors and non-responders using a random effects model.

Results: A total of 8 studies met the inclusion criteria representing a total of 498 RA patients, including 130 patients with the -308(A) variant. There was no heterogeneity between the studies (p=0.78). The overall OR for the A allele carrier status was significantly decreased in the responder group (OR=0.32, 95% CI=0.19-0.82, p=0.000017). The frequency of the A allele was 88/402 (21.9%) in responders and 42/96 (43.8%) in non-responders.

Conclusion: This meta-analysis indicates that the -308(A) variant significantly predicts poor response to TNF-a inhibitors and this may have therapeutic implications. The c
spectively genotyping for TNF-a -308 variant when instituting anti-TNF agents should now be formally assessed.

<img2731|center>

Commentary on abstract 1023

This study reflects the ongoing interest in use of pharmacogenetics to individualize therapy. In RA, considerable interest has centered on the -308 G/A variant in the TNF-a gene promoter. This meta-analysis provided the first comprehensive examination of the variant’s association with response to anti-TNF therapy in patients with RA. The analysis included nine published studies involving 692 patients, 179 of whom had either the GA or A/A genotype. The studies defined response by ACR20 or DAS28 criteria. As compared with patients who were homozygous G/G, patients with the variant genotype were less likely to respond to anti-TNF therapy. For all studies combined, patients with the G/G genotype had an overall response rate of 80%, whereas G/A and A/A combined were associated with an overall response rate of 66%. The odds ratio for response among carriers of the A allele was 0.43 (0.28, 0.68). In six of the nine studies, authors quantified the decrease in DAS28 score. As compared with the G/G genotype, the A allele carrier state was associated with a significantly smaller decrease in DAS28 (P=0.00527).

ABSTRACT 1656 - Effectiveness of TNF Inhibitors in RA Patients with Moderate vs. High Disease Activity

E. Keystone, G. Reed, M. Hooper, J. Greenberg

Background: Randomized controlled trials (RCTs) of TNF inhibitors (TNF-Is) have generally applied disease activity inclusion criteria for RA patients (pts). In the U.S., TNF-Is are prescribed more broadly to patients with lower disease activity. A previous post-hoc subanalysis of etanercept RCTs has demonstrated better outcomes in pts with more moderate disease activity.

Purpose: To examine outcomes in pts with high versus moderate baseline disease activity initiating a TNF-I using RA pts from a U.S.-based registry.

Methods: Biologic naïve pts with RA prescribed TNF-Is in the Consortium of Rheumatology Researchers of North America (CORRONA) registry were included. Patients with follow up 5-10 months after TNFi initiation were selected and stratified into high disease activity (HDA, n=172) vs moderate (MDA, n=124) disease activity at baseline. Previously published CDAI cut-points for MDA were 10-22 and HDA >22. Pts with baseline values of zero for any component of CDAI and ACR outcomes were excluded to minimize bias due to floor effects, as well pts who discontinued due to toxicity. Outcome measures evaluated were remission (CDAI =2.8), low disease activity state (LDAS) defined as CDAI =10, as well as modified (m)ACR 20 and 50 outcomes requiring 4 of 6 components including tender and swollen jt counts. Sensitivity analyses using DAS28 definitions of MDA vs HDA cohorts for pts with ESR values (n=75 HDA, n=57 MDA), as well as DAS28-derived remission (=2.6) and LDAS <3.2) outcomes were also performed.

Results: Baseline demographics were similar for pts with MDA vs HDA. The proportion of pts achieving CDAI remission were comparable in MDA (13.6%) vs HDA (14.9%) pts (P=0.86). For LDAS outcome, a higher proportion of MDA vs HDA pts achieved LDAS based on CDAI (45.8% vs 32.9%) (P=0.03). Comparable response rates were observed for pts with MDA vs HDA: mACR20 of 34.5% vs 41.0% (P=0.31) and mACR 50 of 26.7% vs 30.8% (P=0.50). Similar results were achieved using DAS-28 cohort and outcome definitions. Of note, DAS-28 remission rates for MDA (33.3%) were numerically higher than HDA pts (22.7%), but not significantly different (P=0.33). Similar results were obtained without excluding pts with floor effects as well.

Conclusion: These results from a registry of U.S. clinical practices demonstrate comparable or superior outcomes in pts with MDA vs HDA prescribed TNF-I therapy. These data support treatment strategies that include TNFI-s in pts with MDA, and suggest that broadening RCT entry criteria may improve generalizability without compromising the ability to reach remission and response endpoints.

Commentary on abstract 1656

This study examined an issue not previously evaluated in the RA field: outcomes of TNF inhibitor therapy inpatients with high vs. moderate baseline disease activity. From a large RA registry, investigators identified 437 patients who had five to 10 months of follow-up after initiation of anti-TNF therapy. The patients were stratified into groups representing high disease activity (n=230) and moderate disease activity (n=207). Disease activity was determined by CDAI criteria (moderate=CDAI 10 to 22, high=CDAI >22). The primary outcomes were the proportion of patients in remission (CDAI =2.8); LDAS defined as a CDAI =10; and modified ACR20 and ACR50 outcomes requiring four of six components, including tender and swollen-joint counts. Unadjusted outcomes were similar between patients with high and moderate disease activity, except for LDAS, which occurred in more patients with moderate disease activity (42.42% vs. 31.78%, P=0.032). Investigators performed a sensitivity analysis in a subgroup of 134 patients with acute-phase reactants using DAS28 =2.6 to define remission, LDAS =3.2, and standard ACR20, 50 and 70 outcomes. The analysis yielded similar results except for ACR20 outcomes, which were significantly lower in patients with moderate baseline disease activity (21.79% vs. 41.07%, P=0.022). The data provide evidence that outcomes achieved with TNF inhibitors in clinical practice are similar to those achieved in clinical trials and that patients with moderate and severe RA derive similar benefits from anti-TNF therapy.

Questions and answers with Dr. Jeffrey Greenberg, New York University, New York

Q: What did you hope to accomplish with this study?

A: Randomized clinical trials of TNF inhibitors have generally involved patients with high baseline disease activity as compared with patients seen in clinical practice. To our knowledge, no one had compared outcomes of anti-TNF therapy in the two types of patients. These results confirm that the benefits that clinicians and patients can expect in clinical practice are similar to those reported from clinical trials.

Q: What are the principal implications of the results?

A: First, they support the use of treatment strategies that include TNF inhibitors in patients with moderate disease activity. Second, they suggest entry criteria for clinical trials can be expanded without compromising the ability to achieve clinically meaningful disease activity states and responses.

Abstracts Relevant to

PSORIATIC ARTHRITIS, ANKYLOSING SPONDYLITIS AND TNF INHIBITION

ABSTRACT 1114 - Golimumab Significantly Improves Health-Related Quality of Life in Patients with Active Ankylosing Spondylitis: Results from the Phase III GO-RAISE Study

D. van der Heijde, A. Deodhar, J. Braun, M. Mack, S. Parasuraman, J. Buchanan, B. Hsu, A. Beutler, C. Han, R. D. Inman

Purpose: To evaluate the impact of golimumab (GLM) on health-related quality of life (HRQoL) in patients with ankylosing spondylitis (AS).

Methods: GLM, a new human, TNF-a antibody, was studied in a multicenter, randomized, placebo (PBO)-controlled study (GO-RAISE). A total of 356 patients were randomly assigned in a 1.8:1.8:1 ratio to receive subcutaneous injections of GLM 50 or 100 mg or PBO every 4 wks. Patients with definite AS according to the modified NY criteria, a BASDAI score =4, and a back pain score =4 were eligible for inclusion in the study. Impact on HRQoL was assessed using the Physical Component Summary (PCS) and Mental Component Summary (MCS) scores of the SF-36. Primary efficacy analyses were performed at wk 14. At wk 16, patients in the PBO or GLM 50 mg groups who had <20% improvement from baseline in both total back pain and morning stiffness measures entered early escape in a double-blind fashion. All other patients remained on their previous medication until wk 24. For those patients in the GLM 50 mg or PBO groups who entered early escape, their last observation prior to change in treatment was carried forward for the wk 24 analyses; the observed values at wk 24 were used for GLM 100 mg pts. An ANOVA on van der Waerden normal scores was performed for between-group comparisons.

Results: GLM 50 mg and 100 mg significantly improved HRQoL. The mean ±SD baseline PCS scores of 29.3 ±8.3, 30.7 ±8.1, and 29.2 ±7.1 for the GLM 50 mg, 100 mg, and PBO groups, respectively, were significantly lower than the norm of 50 ±10 reported for the US population[1]. Patients treated with GLM 50 mg and 100 mg showed significantly greater improvement from baseline in SF-36 PCS scores compared with PBO at wk 14 (8.8 ±9.6 and 8.9 ±9.8 vs. 3.0 ±7.2; p<0.001, for both comparisons) and wk 24 (9.6 ±10.6 and 9.2 ±10.3 vs. 2.8 ±8.1; p<0.001 for both comparisons). The improvements observed at wk 14 for the PCS of the SF-36 were maintained through wk 24. The improvements in SF-36 PCS scores were similar in the GLM 50 mg and GLM 100 mg groups at wk 14 and wk 24. Patients treated with GLM 50 mg showed significantly greater improvement from baseline in SF-36 MCS scores when compared with PBO at wk 14 (2.6 ±8.5 vs. -0.4 ±8.3; p=0.014). GLM 100 mg also showed a significant improvement in MCS scores when compared with PBO at wk 14 (4.8 ±10.4 vs. 0.4 ±8.3; p<0.001). The improvement in the MCS score was maintained through wk 24 for GLM 100 mg, but not for GLM 50 mg.

Conclusions: HRQoL was significantly improved with both GLM 50 mg and 100 mg. Improvements in PCS scores were observed as early as wk 14 and maintained through wk 24.

Reference: [1] Ware JE Jr. SF-36 Health Survey Update. Spine 2000;25(24):3130-9.

Commentary for abstract 1114

This placebo-controlled study provided objective evidence of improved quality of life in patients with AS treated with golimumab. Eligibility criteria included a disease activity index score and a back pain score =4. Investigators randomized 356 patients with AS to placebo or to one of two doses of the new anti-TNF-a agent: 50 mg or 100 mg, administered s.c. once every 4 weeks. Impact on health-related quality of life was assessed by the physical component and mental component summaries of the SF-36 form. The primary end point was improvement in each of the two component summaries, as assessed at 14 and 24 weeks. Patients in the placebo and 50-mg golimumab groups could be switched to 100 mg golimumab at week 16 if they had <20% improvement from baseline in both summary scores (early escape). Among patients who entered early escape, the score at last observation prior to treatment change was carried forward to week 24. The results demonstrated significant improvement vs. placebo in the physical component summary at week 14 and week 24 for the 50-mg dose (P<0.001) and the 100-mg dose (P<0.001). Improvement observed at week 14 was observed through week 24. With respect to the mental component summary, both doses of golimumab were superior to placebo at week 14: P<0.014 for the 50-mg dose and P<0.001 for the 100-mg dose. The improvement was maintained at week 24 with the 100-mg dose but not the 50-mg dose.

ABSTRACT 1113 - Golimumab Significantly Improves Productivity in Patients with Active Ankylosing Spondylitis: Results from the Phase III GO-RAISE Study

J. Braun, R. D. Inman, D. van der Heijde, M. Mack, S. Parasuraman, J. Buchanan, B. Hsu, A. Beutler, C. Han, A. Deodhar

Purpose: To evaluate the impact of golimumab on productivity in patients with active ankylosing spondylitis (AS).

Methods: Golimumab, a new, human, TNF-a antibody was studied in a multicenter, randomized, placebo (PBO)-controlled study (GO-RAISE). A total of 356 patients were randomly assigned in a 1.8:1.8:1 ratio to receive subcutaneous injections of golimumab 50 or 100 mg or PBO every 4 wks. Patients with definite AS according to the modified NY criteria, a BASDAI score =4, and a back pain score of =4 were eligible for inclusion in the study. Productivity was measured on a VAS scale (0 cm = not at all affected, 10 cm = affected very much). The change in productivity from baseline to wk 16 and wk 24 was summarized and compared between groups. At wk 16, patients in the PBO or golimumab 50 mg group who had <20% improvement from baseline in both total back pain and morning stiffness measures entered early escape in a double-blind fashion. All other patients remained on their previous medication until wk 24. For those patients in the golimumab 50 mg or PBO groups who entered early escape, their last observation prior to change in treatment was carried forward for the wk 24 analyses. The observed values at wk 24 were used for golimumab 100 mg patients. An ANOVA on van der Waerden normal scores was performed for between-group differences.

Results: At baseline, patients in the golimumab 50 mg, 100 mg group, and the PBO group had similar mean ±SD baseline scores of 6.6 ±2.5, 6.8 ±2.3 and 6.3 ±2.5, respectively. Mean improvement in self-reported productivity was significantly greater in the golimumab 50 mg group compared with the PBO group at wk 16 (-2.8 ±3.0 vs. -0.4 ±2.7; p<0.001) and wk 24 (-2.7 ±3.1 vs. -0.4 ±2.7; p<0.001). Mean improvement in self-reported productivity was also significantly greater in the golimumab 100 mg group compared with the PBO group at wk 16 (-2.9 ±2.9 vs. -0.4 ±2.7; p<0.001) and wk 24 (-2.9 ±3.0 vs. -0.4 ±2.7; p<0.001). The change from baseline in self-reported productivity was similar in the golimumab 50 mg and golimumab 100 mg groups at wk 16 and wk 24.

Conclusions: Patients with AS treated with golimumab 50 mg and 100 mg had significant improvement in self-reported productivity. Improvements in productivity at wk 16 were maintained through wk 24.

Commentary on abstract 1113

The principal finding of this study was that treatment with either 50 or 100 mg of golimumab significantly improved productivity compared with placebo in patients with active AS. Patients self-rated productivity by means of a visual analogue scale (0 to 10 cm) at baseline and at weeks 16 and 24. At week 16, patients who reported <20% improvement in total back pain and morning stiffness entered early escape for the remaining eight weeks. Patients originally randomized to placebo then received 50 mg of golimumab. Those initially treated with 50 mg received 100 mg, and patients randomized to 100 mg continued treatment at that dose. For patients who entered early escape, their productivity scores at 16 weeks were carried forward to week 24. Baseline productivity scores ranged from 6.3 to 6.8. Patients in both active treatment groups had significant improvement in productivity at 16 and 24 weeks compared with placebo. At week 16, productivity had improved by an average of 2.8 and 2.9 in the 50-mg and 100-mg groups, respectively, vs. 0.4 in the placebo group. At the end of the study, patients originally assigned to placebo had mean improvement of 0.5 compared with 2.7 and 2.9 in the 50-mg and 100-mg groups (P<0.001 for weeks 16 and 24).

ABSTRACT 1097 - Golimumab Administered Every Four Weeks as a Subcutaneous Injection in Psoriatic Arthritis: Nail, Enthesitis, and Dactylitis Response in the Randomized, Placebo-Controlled, GO-REVEAL Study

D. Gladman, A. Kavanaugh, I. McInnes, P. Mease, J. J. Gómez-Reino, K. Papp, J. Livingston, S. Mudivarthy, M. Mack, A. Beutler, G. G. Krueger

Purpose: To assess the effect of golimumab treatment on nail disease, dactylitis, and enthesitis in patients with active psoriatic arthritis (PsA).

Methods: Adult PsA patients with =3 swollen and 3 tender joints and active plaque psoriasis (=1 lesion =2 cm diameter), were randomized to subcutaneous (SC) placebo (n=113), golimumab 50 mg (n=146), or golimumab 100 mg (n=146) every 4 weeks (q4wk). Week 24 comparisons are presented for placebo compared with each golimumab dose group. At baseline, the single worst psoriasis-affected nail (target) was assessed using the technique of the Nail Psoriasis Severity Index (NAPSI). The Nail Physician’s Global Assessment (PGA; 1 to 5; 1=absence of; 5=very severe) evaluated all fingernails. Dactylitis was assessed in each of the 20 digits of the hands and feet (0=no dactylitis to 3=severe dactylitis). Entheses tenderness was scored in 15 body sites (0=absent; 1=present) using the PsA-modified (plantar fascia added) Maastricht Ankylosing Spondylitis Enthesitis Score (MASES).

Results: The target nail NAPSI median percent change from baseline at week 24 was significantly greater for both golimumab 50 mg (33) and golimumab 100 mg (54) vs. placebo (0; p<0.001 for both comparisons). Nail PGA improved in a significantly greater proportion of patients in the golimumab 50 mg (60%) and golimumab 100 mg (63%) group than in the placebo group (18%; p<0.001 for both comparisons). Dactylitis score median percent change from baseline at week 24 was 100 for both golimumab 50 mg and 100 mg versus 42 for placebo with a statistically significant difference observed for only the golimumab 100 mg group (p<0.001). Enthesitis score median percent change from baseline at week 24 was 60 and 67 for the golimumab 50 mg and 100 mg groups, respectively, compared with 12 for the placebo group (p<0.001).

Conclusion: Golimumab, administered as a monthly subcutaneous injection at doses of 50 and 100 mg, significantly improved psoriatic nail changes and enthesitis, in patients with active PsA. Both golimumab doses also improved dactylitis, but statistical significance was achieved only with the golimumab 100 mg dose for this variable.

Commentary on abstract 1097

This randomized double-blind clinical trial of patients with active PsA examined the clinical impact of two doses of golimumab on nail disease, enthesitis and dactylitis. Patients were randomized to once-monthly s.c. injections of placebo or golimumab 50 mg or 100 mg. Follow-up continued for 24 weeks. At the 24-week evaluation, patients treated with either active dose had significant improvement in the Nail Psoriasis Severity Index, Nail PGA, and enthesitis (P<0.001 for all comparisons). Additionally, the 100-mg dose led to statistically significant improvement in dactylitis compared with placebo (P<0.001). Both golimumab doses led to >50% improvement in nail score, whereas the improvement with placebo averaged 0%. Nail PGA improved by more than 60% in both golimumab groups vs. 13% with placebo. Both active doses resulted in 100% improvement in dactylitis vs. 42% for placebo. Median per cent improvement in enthesitis was fivefold greater with golimumab compared with placebo. This study is ongoing, and 52-week results will be reported in the near future.

Questions and answers with Dr. Dafna Gladman, University of Toronto, Ontario

Q: Why were dactylitis, nail changes and enthesitis chosen as the principal outcomes of the study?

A: These are important outcomes in PsA. They were the principal outcomes for this analysis since they are important, but they are not included in the core set of outcomes which are based on joint count, patient and physician global assessments, assessment of pain and HAQ, as well as acute-phase reactants.

Q: Does golimumab or other anti-TNF therapies affect the skin manifestations of PsA?

A: Yes. There was an oral presentation at the spondyloarthritis treatment section at the ACR meeting which demonstrated the improvement in skin manifestations by golimumab. The same results have been published for etanercept, adalimumab and infliximab.

Q: Even though this study did not involve direct comparison of different therapies, can you comment on how the improvement with golimumab compared with what is usually seen with standard therapy?

A: The only other drug which has shown statisticially significant improvement in dactylitis and enthesitis was infliximab. The outcome was not studied with etanercept, and in the adalimumab trial the results were encouraging but not statistically significant. Improvement of dactylitis and enthesitis has not been shown with traditional DMARDs such as MTX, leflunomide, sulfasalazine and cyclosporine. The effect on the nails was studied in the golimumab study for the first time.

ABSTRACT 671 - Golimumab Administered Subcutaneously Every 4 Weeks in Psoriatic Arthritis Patients: 52-Week Efficacy and Safety Results of the Randomized, Placebo-Controlled GO-REVEAL Study

A. Kavanaugh, P. Mease, G. G. Krueger, D. Gladman

Purpose: The efficacy and safety of golimumab (GLM) in reducing signs and symptoms of active psoriatic arthritis (PsA) were assessed.

Methods: Adult PsA patients (pts) with =3 swollen & =3 tender joints were randomized to receive SC injections of placebo (PBO) or GLM (50 or 100 mg) every 4 weeks (wk). At wk16, pts with inadequate arthritis response entered early escape (EE) in a blinded manner to GLM 50 mg (PBO pts) or GLM 100 mg (GLM 50 mg pts). All pts on PBO at wk 20 received GLM 50 mg from wk24 through wk52. Wk24 comparisons were between GLM groups and the PBO group. The proportion of pts with ACR20 response and PASI75 improvement at wk 24 was analyzed by the Cochran-Mantel-Haenszel test with stratification by baseline MTX. At wk52, efficacy comparisons by dose were not performed; response for pts (including those who stopped treatment but had wk 52 data; excluding those whose dose changed) treated with either GLM 50 mg or 100 mg beginning at wk 0 was determined using observed data.

Results: 405 pts enrolled. Groups were balanced for baseline demographic and disease characteristics; mean age was 46-48 years, median swollen/tender joint counts were 12-14/22-24, HAQ scores were 1.0-1.1, and PASI scores were 8.4-11.1. GLM was significantly better than PBO in improving signs and symptoms of PsA at wk24. Efficacy was maintained through wk 52. 2.4% of GLM pts reported SAEs through wk24 vs. 6.2% of PBO pts. Through wk52, 4.6% of pts reported SAEs. Injection site reactions (ISR) occurred in 4.8% of GLM and 2.7% of PBO pts through wk 24 and 7.6% of GLM pts through wk 52. Tuberculosis or opportunistic infections were not reported. Through wk24, there were 3 malignancies in GLM-treated pts (1 prostate and 2 basal cell skin cancers). Between wk24 and wk52, 1 colon and 1 small cell lung cancer were reported. The small cell lung cancer pt died; another pt died in a climbing accident.

Conclusions: In pts with active PsA, GLM 50 and 100 mg administered SC every 4 weeks i
psoriatic skin disease through wk52. GLM was generally well-tolerated with a safety profile similar to that observed in the first 24 wks of treatment.

Figure<img2732|center>

Commentary for abstract 671

This phase III, randomized, placebo-controlled trial examined the efficacy of golimumab 50 and 100 mg vs. placebo in 405 patients with active PsA and qualifying target lesions >2 cm. Treatment and follow-up continued for a year, at which time patients entered a long-term extension with open-label golimumab. Patients who achieved an inadequate response at 16 weeks entered randomized early escape to 50- or 100-mg doses. All patients still on placebo at 20 weeks received 50 mg from weeks 24 to 52. The primary end point was the proportion of patients achieving an ACR20 response and PASI75 at 24 weeks. Both doses had significantly higher ACR responses (ACR20, ACR50, and ACR70) than placebo at weeks 14, 24 and 48. Response rates were similar at all time points between the two active treatment groups. Additionally, a significantly higher proportion of patients randomized to either active dose achieved PASI75 at weeks 14 and 24, but the two groups did not differ significantly from each other. Regarding the primary end point, 12% of placebo patients achieved an ACR20 response by week 24 compared with 52% and 61% for golimumab 50-mg and 100-mg, respectively (P<0.001). Only 1% of placebo patients had a PASI75 at 24 weeks vs. 56% and 66% of the golimumab groups (P<0.001). Patients treated with either active dose also had significantly more improvement in the DAS28, HAQ and NAPSI at weeks 14 and 24.

Questions and Answers with Dr. Arthur Kavanaugh, University of California, San Diego

Q: What was the tolerability of golimumab compared with placebo?

A: The proportion of patients who had at least one serious adverse event was about 4% in the placebo group and about 4.5% in the two golimumab groups combined.

Q: Did the two golimumab groups differ significantly with respect to any of the primary or secondary outcomes?

A: No, the outcomes in the two groups were similar and did not differ significantly in a combined analysis at 52 weeks.

ABSTRACT 536 - Risk Factors for the Development of Spondylitis in Patients with Psoriatic Arthritis (PsA)

V. Chandran, D. C. Tolusso, C. T. Schentag, R. J. Cook, D. D. Gladman

Purpose: To determine the incidence and risk factors for the development of spondylitis in patients with PsA.

Methods: Patients were identified from a prospective open dynamic cohort of patients with PsA established in 1978. Clinical evaluation is done every 6-12 months, and radiographic evaluation of the peripheral joints and spine is done in all patients every 2 years. Data are tracked in a computerized database. Patients who did not have spondylitis at first clinic visit were identified and the incidence and risk factors for the development of spondylitis (bilateral sacroiliitis = grade 2 or unilateral sacroiliitis =3 and inflammatory neck or back pain or limitation in spinal mobility) were determined. Under a multistate framework, the proportion of PsA patients with spondylitis was estimated robustly using marginal methods and based on a Markov model. Risk factors assessed included sex, age, race, family history of psoriasis/PsA, duration of psoriasis and PsA, number of actively inflamed, swollen, clinically damaged and radiographically damaged joints, dactylitis, nail dystrophy, periostitis, enthesitis, calcaneal spurs, prior NSAID use, DMARD use, smoking, hypertension, ESR and HLA-B27 status. Risk factors identified in univariate analyses at the 0.25 level were included in a multivariate model to determine risk factors associated with development of spondylitis.

Results: 206 patients with complete covariate information were included in the present study. There were 119 males and 87 females with a mean age at presentation of 42.4 years and mean disease duration of 6.6 years. After 10 years of follow up, 15% (95% CI 0.06, 0.22) of patients with PsA were observed to develop spondylitis. Sex, family history of PsA, nail dystrophy, number of swollen joints, enthesitis, number of radiographically damaged joints, periostitis, calcaneal spurs, ESR and presence of hypertension was associated with the development of spondylitis in univariate analyses (p<0.25). Nail dystrophy (RR=7.07, 95% CI 1.56, 32.04), number of swollen joints (RR=0.83, 95% CI 0.73, 0.95), number of radiographically damaged joints (RR=1.11, 95% CI 1.05, 1.17), periostitis (RR=5.67, 95% CI 2.11:15.27) and ESR (RR=1.03, 95% CI 1.01, 1.05 for each 1mm increase) were associated with risk of spondylitis in multivariate analysis.

Conclusions: 15% of patients with peripheral PsA without axial involvement develop spondylitis over 10 years of follow up in a tertiary care clinic. Nail dystrophy, number of radiographically damaged joints, periostitis and increased ESR is associated with increased risk of developing spondylitis, whereas number of swollen joints is “protective.”

Commentary on abstract 536

This study explored factors associated with the development of spondylitis in patients with PsA and no axial involvement. The study group involved 206 patients whose mean age was 42.4 years and whose mean disease duration was 6.6 years. During approximately 10 years of follow-up, 15% of the patients developed spondylitis. Factors associated with spondylitis were gender, family history of PsA, nail dystrophy, number of swollen joints, enthesitis, number of radiographically damaged joints, periostitis, calcaneal spurs, ESR, and hypertension. Multivariate analysis pared the list of factors down to five: nail dystrophy (RR 7.07, P=0.01); periostitis (RR 5.67, P<0.0001); number of swollen joints (RR 0.73, P<0.01); number of radiographically damaged joints (RR 1.11, P<0.0001); and ESR (RR 1.03, P=0.001).

Questions and Answers with Dr. Vinod Chandran, Toronto Western Hospital, Ontario

Q: Can you speculate about the reason for an apparent protective effect of swollen joint count?

A: Since swollen joints predict future damage in that joint, one would expect swollen joints to be a risk factor for spondylitis, if damaged joints are. It is possible that there is an inverse relationship between the degree and severity of inflammation in the peripheral joints and the spine. It is well known that patients with classical AS have relatively mild peripheral arthritis, and those with PsA have relatively mild spinal disease. Moreover, there may be different pathogenic mechanisms for these inflammatory processes. In another poster at ACR, we examined different criteria to determine the robustness of the results. In that analysis, swollen joints were not protective or predictive. Therefore, although radiographic damage, periostitis and nail dystrophy are associated with development of psoriatic spondylitis, the question of whether swollen joints are protective is unresolved.

Q: How might this information about risk factors for spondylitis be used in clinical practice?

A: This information will help focus attention of clinicians to investigate spinal disease in patients with these risk factors so that spinal involvement can be detected early, and appropriate intervention to prevent restricted mobility be initiated.

Q: How does the clinical course of patients with psoriatic spondylitis differ from that of patients with PsA?

A: Psoriatic spondylitis is a subgroup of PsA. Although only 2% of patients with PsA have spondylitis alone, 30 to 50% of patients with PsA have evidence of both spondylitis and peripheral arthritis. On long-term follow-up, there is progressive restriction of lateral spinal flexion and cervical rotation. However, only 45% of patients with radiographic evidence of psoriatic spondylitis have axial symptoms. Thus, a significant proportion of patients with PsA have peripheral and axial arthritis and have progressive damage to peripheral joints, restriction of lateral spinal flexion, and cervical rotation.