Anti-tumor necrosis factor (TNF) therapy has emerged as a major advancement in the management of rheumatoid arthritis (RA). The anti-TNF agents adalimumab, etanercept, and infliximab are often used with conventional disease-modifying antirheumatic drugs (DMARDs), such as methotrexate, and have been shown in clinical trials to be effective in reducing the signs and symptoms of RA and in preventing the progression of joint damage.1-6
Significant limitations have been noted in the literature as to outcomes of clinical trials and observational studies of anti-TNF therapy.7,8 Clinical trial inclusion and exclusion criteria specify select patient populations, which may not necessarily reflect the diversity of patient and disease characteristics in real-world clinical scenarios.8 Therefore, real-world studies can offer unique insights over controlled clinical trials.9
Although there is greater diversity of patient and disease characteristics in real-world studies, there may be bias, because patients receive various treatments. Nevertheless, healthcare decision makers are increasingly relying on studies of real-world outcomes for decisions related to coverage and reimbursement.10 Many studies have evaluated the effectiveness of anti-TNF therapies in patients with RA,1-6 but few real-world studies have evaluated the clinical effectiveness of adalimumab, etanercept, and infliximab. The objective of this analysis was to assess the effectiveness of these agents in improving joint pain, joint swelling, joint stiffness, and fatigue among patients with RA.
We conducted a retrospective, observational chart review using data from 6 rheumatology clinics across the United States, and evaluated treatment effectiveness among a sample of patients with RA who were treated with anti-TNF therapy (ie, adalimumab, etanercept, or infliximab). The study period was from January 1, 2001, to November 30, 2006. For each patient, the observation period included a 1-year period before the index date, which was the date of the first anti-TNF prescription or administration, and a 2-year follow-up period after the index date (Figure). Baseline status was assessed during the year before the index date. Data were collected for a maximum of 2 years after the index date, or until discontinuation of anti- TNF therapy, whichever occurred first.
Patient selection was based on the following inclusion criteria: age ≥18 years; diagnosis of RA; and therapy initiated with adalimumab, etanercept, or infliximab between January 1, 2002, and November 30, 2004, with no documentation of previous anti-TNF therapy. Patients were required to be under the care of the participating physician or physician group, or have complete documentation (related to RA) from another provider for a minimum of 12 months before the index date and up to 24 months after the index date (no minimum follow-up period was required).
Patients with a diagnosis of psoriasis, ulcerative colitis, Crohn’s disease, or ankylosing spondylitis were excluded from the study, because of the differences in anti-TNF dosage recommendations. In addition, patients were excluded if they had a history of anti-TNF therapy or any of the following diagnoses any time before the index date—malignancy (eg, lymphoma, solid organ malignancy), infection (eg, tuberculosis, bacterial pneumonia), seizures, demyelinating disorders (eg, multiple sclerosis), heart failure, or drug-induced lupus.
Participating clinics identified and provided charts for all patients meeting the study inclusion criteria. Patient privacy was maintained through the use of unique alphanumeric values for each patient in compliance with the Health Insurance Portability and Accountability Act (HIPAA). The study protocol was approved by the New England Institutional Review Board.
Data were collected using a standardized chart review form with a graphical user interface developed in Microsoft Access. Patient demographics and concomitant DMARD use were collected. For this study, concomitant DMARD therapy included hydroxychloroquine, leflunomide, sulfasalazine, azathioprine, penicillamine, and auranofin.
Response to therapy was evaluated for joint pain, joint swelling, joint stiffness, and fatigue. Because of intra- and intervariability in documentation among physicians, a data abstraction form was used so that the chart reviewer could document both qualitative and quantitative data.
An algorithm was developed to translate qualitative data into quantitative values. Because there were no relevant studies available to guide the development of the algorithm, a clinical and psychometric rationale was applied to develop an algorithm that could be consistently applied to the varying levels of physician documentation.
A visual analog scale (VAS; 0-10 cm) was the preferred method for evaluating all 3 joint assessments (ie, pain, swelling, stiffness) and fatigue. If documented VAS was not available, the following algorithm was used:
A. Categorical documentation was converted to a numerical value using midpoints of a 0-to-10 scale (median [range]: mild = 1.65 [1-3.3], moderate = 4.95 [3.4-6.6], severe = 8.25 [6.7-9.9])
B. Dichotomous documentation was converted to a numerical scale using the midpoints of a 1-to-10 scale (yes = 5, no = 0).
If more than 1 form of documentation was present for a patient visit, the VAS was the preferred type of documentation, followed by categorical and dichotomous documentation.
Baseline scores were determined through available documentation on the index date or, if not available, documentation from the closest encounter before the index dates. The mean scores for joint pain, swelling, and stiffness, and fatigue were determined for the 2-year follow-up period. If a patient had multiple assessments, an average was derived.
The following criteria were used to define improvement from baseline:
A. Decrease in VAS of at least 1 point
B. Change in categorical documentation from “severe” to “moderate/mild” or from “moderate” to “mild”
C. Change in dichotomous documentation from yes to no.
Improvement was defined as any improvement during the 2-year follow-up period among any of the 3 documentation types. Overall improvement was based on improvement in any of the 4 effectiveness measures. Mean scores were based only on patients with documented information for each joint assessment at the given time point (baseline or follow-up period).
Basic descriptive statistics and univariate statistical testing were applied to this analysis. Analysis of variance or Kruskal-Wallis tests (as applicable) were applied to continuous variables to determine if differences existed among the 3 cohorts. If differences did exist (ie, P <.05), post-hoc multiple comparison tests (Tukey or Mann-Whitney) that controlled for the family- wise type 1 error rate were conducted. For the multiple comparison tests, P <.017 (Bonferroni correction) was considered statistically significant. Paired t-tests were conducted to determine if differences existed among the VAS scores at baseline and during the 2-year follow-up period within each cohort. Chi-square tests were conducted to assess differences between the 3 cohorts on categorical variables. All analyses were conducted using SAS version 9.1 (Cary, NC).
This study enrolled 496 patients, including 84 patients (16.9%) in the adalimumab group, 146 (29.4%) in the etanercept group, and 266 (53.6%) in the infliximab group (Table 1). Overall, 74.0% of the patients were female, and the mean age (standard deviation [SD]) was 56.1 (14.8) years. The mean age was 52.6 years in the etanercept group and 57.6 years in the infliximab group (P <.017).
Of patients with documented ethnicity (n = 217), whites comprised 80.2% of the overall population. The mean (SD) length of disease duration before initiation of anti-TNF therapy was 3.9 (3.5) years, and the mean (SD) length of follow-up after the index date was 614.8 (156.0) days. Approximately 76% of patients received concomitant DMARD therapy, which was more common (79.7%) in the infliximab cohort than in the etanercept (68.5%) or adalimumab (75.0%) cohorts (P <.040). The average time that patients remained on anti-TNF therapy ranged from 367 days to 423 days (etanercept, 423 days; adalimumab, 371 days; infliximab, 367 days; P = .126).
There was wide variation in the type of documentation used to indicate patient improvement (Table 2). For example, 99% of the patients had documentation for joint pain, using a dichotomous approach (yes, pain exists; no, pain does not exist), and 29% had VAS documentation for this assessment.
In total, 12.5% (n = 62) of all patients did not have joint assessment data to evaluate clinical improvement during the follow-up period. Of the 434 patients who had documented information for at least 1 of the 4 effectiveness measures, 46.5% (n = 202) reported a clinical benefit, including 48.7% (n = 115/236) in the infliximab group, 47.7% (n = 62/130) in the etanercept group, and 36.8% (n = 25/68) in the adalimumab group.
Overall, 150 of the 429 patients (35.0%) had an improvement from baseline in joint pain scores during the 2-year follow-up (Table 3). The adalimumab and infliximab groups reported significant improvements from baseline in joint pain scores (P = .004 and P <.001, respectively). The infliximab group had the highest percentage of patients with documented improvement injoint pain (36.9%) compared with the etanercept group (36.4%) and the adalimumab group (25.3%), although the difference was not significant.
Approximately 35% (122 of 348) of patients had an improvement from baseline in joint swelling scores. Among all patients, mean (SD) joint swelling scores significantly improved from baseline during the 2-year follow- up period (3.2 [1.9] vs 3.0 [1.8], P = .004). In the infliximab group, the mean (SD) score decreased from 3.2 (1.9) at baseline to 3.0 (1.8) during the follow-up period (P <.001). There were no significant differences between the mean scores at baseline and during the follow-up period in the adalimumab or the etanercept groups. The percentage of patients with improvement in joint swelling was highest in the infliximab group (38.0%), followed by the etanercept group (33.7%), and the adalimumab group (26.8%); however, these differences were not significant.
Among all patients, 107 of 380 patients (28.2%) had an improvement in joint stiffness score, and a significantdecrease from baseline was observed in mean scores (P = .022). There were no significant differences between the mean scores at baseline and during the follow-up period within the adalimumab group and the etanercept group (Table 3). In the infliximab group, the mean (SD) score decreased significantly from 3.8 (1.8) at baseline to 3.6 (1.8) during the follow-up period (P = .001).
However, the etanercept group had the highest percentage of patients who had an improvement in joint stiffness score (29.7%), followed by the infliximab group (28.2%), and the adalimumab group (25.0%); however, these differences were not significant.
Mean fatigue scores decreased from baseline for all patients and within each therapy group; however, these decreases were not significant. The percentage of patients with improvement in fatigue was greatest in the etanercept cohort (38.2%), followed by the infliximab group (27.4%), and the adalimumab group (26.3%); however, these differences were not significant.
When pairwise comparisons were conducted, the mean (SD) fatigue score for the infliximab group was significantly lower (4.2 [2.0]) than in the etanercept (5.2 [1.7]) cohort during follow-up (P <.017). How - ever, the scores at baseline were also lowest in the infliximab group.
The introduction of anti-TNF therapy is a valuable addition to the armamentarium of RA treatment options. The efficacy and safety of adalimumab, etanercept, and infliximab have been demonstrated in multiple randomized controlled trials with favorable results.3-5,11-13Anti-TNF therapy can improve symptoms, increase physical function, and slow or prevent radio - graphic joint damage in many patients with RA.2,14
Overall, the findings of this real-world analysis from 6 US rheumatology clinics support the benefit of anti- TNF therapy in patients with RA as reported in clinical trials.1,2,11 In general, the age and sex of the populations were similar to those typically reported in RA clinical trials with anti-TNF agents. Disease duration before anti-TNF agent intervention was shorter (ie, about 4 years) than that of patients who participated in previous clinical trials (ie, 9-11 years),1,2,11 which may be the result of recent evidence supporting earlier interventions or may reflect referral patterns within the participating rheumatology clinics.
Many patients in this analysis were treated with an anti-TNF therapy and concomitant DMARD therapy (75.6%). The data indicate that concomitant DMARD use was more common with infliximab treatment than with the other anti-TNF agents. This combination therapy is supported by clinical trials showing that anti-TNF therapy concomitant with methotrexate has improved benefit over methotrexate monotherapy.2,4,13
Nearly half of the patients included in this study reported an improvement in joint pain, joint swelling, joint stiffness, or fatigue during the 2-year follow-up period. The infliximab group was the only cohort to demonstrate significant improvements from baseline in 3 (joint pain, joint swelling, and joint stiffness) of the 4 clinical effectiveness measures. None of the cohorts demonstrated a significant change from baseline in the fatigue score, although numerical improvements were noted.
There is no standardized measure to assess outcomes in routine clinical practice; therefore, physicians typically utilize several measures, including laboratory tests, radiographic scores, joint counts, measures of functional status, global measures, and patient self-reported questionnaires.15-17 Indeed, wide variability has been demonstrated regarding the frequency with which different outcome measures are documented by providers.17
The present analysis assessed variable outcome measures with the development of an algorithm to integrate intra- and intervariability among providers reporting clinical status in patients with RA. A number of tools are currently available (eg, Simplified Disease Activity Index; Clinical Disease Activity Index; Disease Activity Score, including 28 joints).18,19 However, further exploration of a validated practical tool that could be adopted for widespread use to monitor RA treatment response is warranted, followed by exploration of the incentives needed to improve such data collection.
An advantage of this study is that it provides realworld outcomes associated with anti-TNF therapy for drugs that have already demonstrated efficacy and safety in controlled clinical trials. The 2-year follow-up period allowed for the assessment of long-term outcomes. However, because this analysis was designed to evaluate anti-TNF agents, other biologic therapies used for RA—including anakinra (Kineret),20 abatacept (Orencia),21 and rituximab (Rituxan)22—were not assessed.
Randomized clinical trials and observational studies in routine clinical practice are important for the evaluation of therapies and have unique advantages and limitations. Although not as scientifically rigorous, real-world studies can complement the information gained from clinical trials. Specifically, one observational cohort study of the effectiveness of anti-TNF therapy in patients who would have been ineligible for a clinical trial demonstrated that the majority of such patients would benefit from these treatments.23 Well-designed observational studies can be invaluable in gaining further understanding of response to therapy.
Consistent with observational research, the limitations of this study include missing data and inconsistencies with documentation. Lack of standardized outcome measures in clinical practice was confirmed by the results and limited the quantitative and qualitative nature of this study, thereby requiring the development of an algorithm to capture variability in assessment of clinical outcomes among providers. As such, this approach has not been validated. Although the study was designed to evaluate patients who were newly initiated with anti- TNF therapy, it is possible that patients might have received anti-TNF treatments before the pre-index assessment period.
Notable limitations of the study are inherent in the nature of observational studies. There exists the possibility of confounding by indication associated with lack of randomization, because other unmeasured factors might have influenced the selection of one anti-TNF agent over another, and variable reasons for changes in therapy may influence the outcomes. Also, the retrospective nature of the analysis prevented the collection of patient-reported outcomes. In addition, quantification of tender or swollen joints and radiographs to assess joint damage were not available.
These factors may have also contributed to the lack of significant improvements among patients who received etanercept in this analysis. Although these patients did not show significant improvements in the clinical response measures used in this study, etanercept has been shown to be effective for treating RA in randomized, placebo-controlled clinical trials.4,6 This discrepancy may also be a result of the different methods used in measuring treatment response.
Data from this analysis demonstrate the 2-year clinical effectiveness of anti-TNF therapy in patients with RA. This real-world study shows improvements in joint pain, joint swelling, joint stiffness, and fatigue. Overall clinical improvement was reported in 36.8% of the adalimumab group, 47.7% of the etanercept group, and 48.7% of the infliximab group over the 2-year followup period. These results are beneficial in that they provide confirmation that benefits seen in clinical trials are being translated into actual practice. Future observational research, with more uniform patient and provider assessments during the course of anti-TNF therapy, is warranted.
The authors thank Rebecca E. Clemente, PhD, and Robert Achenbach of Centocor Ortho Biotech, Inc, for editorial support.
This study was supported by funding from Centocor Ortho Biotech, Inc.
1. Weinblatt ME, Keystone EC, Furst DE, et al. Adalimumab, a fully human antitumor necrosis factor alpha monoclonal antibody, for the treatment of rheumatoid arthritis in patients taking concomitant methotrexate: the ARMADA trial. Arthritis Rheum. 2003;48:35-45.
2. Lipsky PE, van der Heijde DM, St. Clair EW, et al, for the Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy Study Group. Infliximab and methotrexate in the treatment of rheumatoid arthritis. N Engl J Med. 2000;343:1594-1602.
3. Keystone EC, Kavanaugh AF, Sharp JT, et al. Radiographic, clinical, and functional outcomes of treatment with adalimumab (a human anti-tumor necrosis factor monoclonal antibody) in patients with active rheumatoid arthritis receiving concomitant methotrexate therapy: a randomized, placebo-controlled, 52-week trial. Arthritis Rheum. 2004;50:1400-1411.
4. van der Heijde D, Klareskog L, Rodriquez-Valverde V, et al. Comparison of etanercept and methotrexate, alone and combined, in the treatment of rheumatoid arthritis: two-year clinical and radiographic results from the TEMPO study, a doubleblind, randomized trial. Arthritis Rheum. 2006;54:1063-1074.
5.Maini RN, Breedveld FC, Kalden JR, et al. Sustained improvement over two years in physical function, structural damage, and signs and symptoms among patients with rheumatoid arthritis treated with infliximab and methotrexate. Arthritis Rheum. 2004;50:1051-1065.
6. Bathon JM, Martin RW, Fleischmann RM, et al. A comparison of etanercept and methotrexate in patients with early rheumatoid arthritis. N Engl J Med. 2000;343:1586-1593.
7. Wolfe F, Michaud K, DeWitt EM. Why results of clinical trials and observational studies of anti-tumour necrosis factor (anti-TNF) therapy differ: methodological and interpretive issues. Ann Rheum Dis. 2004;63(suppl 2):ii13-ii17.
8. Pincus T, Sokka T. Should contemporary rheumatoid arthritis clinical trials be more like standard patient care and vice versa? Ann Rheum Dis. 2004;63(suppl 2):ii32-ii39.
9. Hawley DJ, Wolfe F. Are the results of controlled clinical trials and observational studies of second line therapy in rheumatoid arthritis valid and generalizable as measures of rheumatoid arthritis outcome: analysis of 122 studies. J Rheumatol. 1991;18:1008-1014.
10. Garrison LP Jr, Neumann PJ, Marshall D, Mullins CD. Using real-world data for coverage and payment decisions: the ISPOR real-world data task force report. Value Health. 2007;10:326-335.
11. Maini R, St Clair EW, Breedveld F, et al, for the ATTRACT Study Group. Infliximab (chimeric anti-tumour necrosis factor alpha monoclonal antibody) versus placebo in rheumatoid arthritis patients receiving concomitant methotrexate: a randomised phase III trial. ATTRACT study group. Lancet. 1999;354:1932-1939.
12.Moreland LW, Schiff MH, Baumgartner SW, et al. Etanercept therapy in rheumatoid arthritis. A randomized, controlled trial. Ann Intern Med. 1999;130:478-486.
13. Breedveld FC, Weisman MH, Kavanaugh AF, et al, for the PREMIER Investigators. A multicenter, randomized, double-blind clinical trial of combination therapy with adalimumab plus methotrexate versus methotrexate alone or adali-mumab alone in patients with early, aggressive rheumatoid arthritis who had not had previous methotrexate treatment. Arthritis Rheum. 2006;54:26-37.
14. Taylor PC. Anti-TNF alpha therapy for rheumatoid arthritis: an update. Intern Med. 2003;42:15-20.
15. American College of Rheumatology Subcommittee on Rheumatoid Arthritis Guidelines. Guidelines for the management of rheumatoid arthritis: 2002 update. Arthritis Rheum. 2002;46:328-346.
16. Pincus T, Sokka T. Quantitative measures for assessing rheumatoid arthritis in clinical trials and clinical care. Best Pract Res Clin Rheumatol. 2003;17:753-781.
17. Furst DE, Halbert RJ, Bingham CO, et al. Evaluating the adequacy of disease control in patients with rheumatoid arthritis: a RAND appropriateness panel. Rheumatology (Oxford). 2008;47:194-199. Epub 2008 Jan 4.
18. Leeb BF, Andel I, Sautner J, et al. Disease activity measurement of rheumatoid arthritis: comparison of the simplified disease activity index (SDAI) and the disease activity score including 28 joints (DAS28) in daily routine. Arthritis Rheum. 2005;53:56-60.
19. Aletaha D, Smolen J. The simplified disease activity index (SDAI) and the clinical disease activity index (CDAI): a review of their usefulness and validity in rheumatoid arthritis. Clin Exp Rheumatol. 2005;23(5 suppl 39):S100-S108.
20.Kineret (anakinra). Prescribing information. Thousand Oaks, CA: Amgen Inc; 2001.
21. Orencia (abatacept). Prescribing information. Princeton, NJ: Bristol-Myers Squibb; 2008.
22. Rituxan (rituximab). South San Francisco, CA: Biogen, Idec Inc, and Genentech, Inc; 2008.
23. Zink A, Strangfeld A, Schneider M, et al. Effectiveness of tumor necrosis factor inhibitors in rheumatoid arthritis in an observational cohort study: comparison of patients according to their eligibility for major randomized clinical trials. Arthritis Rheum. 2006;54:3399-3407.