Rheumatoid Arthritis - Clinical - Predictors and Prognosis
- OP0114: HLA Antigens and Anti-CCP Association with Joint Damage
- OP0116: Overweight is Associated with Less Joint Damage in Seropositive RA Patients
- THU0137: Longitudinal Study: Are RANKL and OPG Levels Predictive of Radiographic Progression?
- OP0138: Barriers to Medication Adherence in RA Patients
OP0114 THE ASSOCIATION OF HLA CLASS II ANTIGENS AND ANTI-CCP ANTIBODIES WITH PROGRESSION OF JOINT DAMAGE IS AFFECTED BY EARLY AND TARGETED TREATMENT OF RHEUMATOID ARTHRITIS
J.K. De Vries-Bouwstra, Y.P.M. Goekoop-Ruitermano, K.N. Verpoort, G.M.T. Schreuder, M.H.W. De Bois, C. Mallee, R.E.M. Toes, R.R.P. De Vries, F.C. Breedveld, B.A.C. Dijkmans, T.W.J. Huizinga, C.F. Allaart.
The interaction of genetic predisposition, autoimmunity, and therapy on disease outcomes in RA has received little systematic exploration in clinical studies. Here, De Vries-Bouwstra et al examine the association of the HLA-DR4 shared epitope and autoantibody status on radiographic progression in early RA patients receiving different treatment strategies in the BeSt (Dutch for Treatment Strategies) trial.
Methods: The study design of the BeSt has been described here previously (see design). For this post-hoc analysis, the effect of treatment strategy according to patient serologic status for rheumatoid factor (RF), anti-CCP antibodies (CCP), and shared epitope (SE) on 2-year radiographic progression, measured using the Van der Heijde modification of the Sharp scoring method (Sh/VdH), was determined. Odds ratios for radiographic progression > 4.6 Sh/VdH units over 2 years were calculated according to serologic status.
Results: Of the 508 subjects enrolled in BeSt, RF was available in all subjects. Approximately 80% of subjects were assessed for CCP and HLA-DR4. Precise determination of SE status was available on 315 subjects. Sixty-five percent of enrolled subjects with available results were seropositive for RF, 61% for CCP, and 63% for SE. Serologies were balanced between the 4 treatment strategy allocations.
Neither HLA DR4 nor SE positivity was significantly associated with radiographic progression in any of the 4 treatment strategy allocations. RF positivity was only significantly predictive of radiographic progression in subjects allocated to the sequential monotherapy group, with RF seropositive subjects in this group demonstrating a greater than 4-fold higher risk of radiographic progression compared to RF seronegative subjects within that group (95% CI 1.7 11.0). CCP positivity was only significantly predictive of radiographic progression in subjects allocated to the sequential monotherapy and step-up therapy groups. CCP seropositive subjects treated with sequential monotherapy were at a 6.6 times greater risk of radiographic progression compared with CCP seronegative patients within this treatment group (95% CI 2.4 18.4). CCP seropositive subjects treated with step-up therapy were at a 2.5 times greater risk of radiographic progression compared with CCP seronegative subjects within this treatment group (95% CI 1.1 5.9).
Conclusions: HLA DR4 and SE do not convey an increased risk of radiographic progression in early RA subjects receiving treatment. Increasing the aggressiveness of early treatment can abrogate the adverse influence of RF or anti-CCP antibodies on 2-year radiographic progression.
Editorial Comment: These findings are both novel and compelling for several reasons. For one, they illustrate for the first time in a direct and elegant way that choice of treatment strategy can counteract the harmful effects of risk factors for radiographic progression. In addition, contrasting with prior research, there appears to be a distinct disconnect in risk associated with genetic RA-associated risk factors and auto-antibody status. It is possible that the incomplete capture of genotyping within the cohorts could have reduced the ability to detect statistical differences within the treatment groups. Thus, confirmation of these findings is necessary.
In the original BeSt publication of clinical outcomes, there were no differences in outcomes between the sequential monotherapy and step-up combination therapy groups, largely because most of the subjects in both groups were treated with methotrexate monotherapy for the study duration. Therefore, it is interesting here to see a separation between these two treatment allocations favoring the step-up combination strategy in RF and CCP seropositive subjects.
As for clinical outcomes, these findings support aggressive treatment, either with initial combination non-biologic DMARD therapy (here including supplemental glucocorticoids) or with combinations including biologics. In particular, these findings would suggest that less aggressive treatment should not be contemplated in CCP seropositive subjects.
OP0116 OVERWEIGHT IS ASSOCIATED WITH SIGNIFICANTLY LESS RADIOGRAPHIC JOINT DAMAGE IN SEROPOSITIVE BUT NOT IN SERONEGATIVE PATIENTS WITH RHEUMATOID ARTHRITIS
G. Westhoff, Z. Zhivkov
A paradoxical protective association of increasing body mass index (BMI) has been described in RA for both cardiovascular and all cause mortality (Links). The mechanisms of these associations are not understood, but may involve RA and non-RA disease characteristics. Here, Westhoff et al report an negative association between overweight and obesity and radiographic damage in RA.
Methods: Subjects derived from a German multicenter prospective inception cohort of early RA patients enrolled within 2 years of disease onset. Demographic and disease specific data (including BMI) were collected at 6 month intervals. Plain radiographs of the hands and feet were obtained at baseline and 3 years and radiographic progression of RA scored using the Ratingen method. The risk of significant radiographic progression at 3 years (Ratingen score > 7 units) was compared according to baseline BMI, grouped by WHO category (BMI < 25, normal; BMI 25 - < 30, overweight; BMI > 30, obese).
Results: Among the 1031 subjects enrolled in the inception cohort, 916 were followed to 3 years. Among these, 767 had complete radiographic data. The proportion of subjects of normal weight, overweight, and obese was 40%, 41%, and 19%, respectively. Compared to normal weight subjects, obese subjects tended to be female (82%), older (59 vs. 55 years), higher baseline HAQ (1.29) and DAS28 (4.9 vs. 4.6), with less diagnosed osteoporosis. Normal weight and obese subjects had similar treatment characteristics for both DMARDs and glucocorticoids.
Normal weight subjects had significantly higher baseline radiographic damage compared to obese subjects (Ratingen score 4.5 vs. 2.4; p = 0.006) and demonstrated a significantly higher rate of 3-year radiographic progression (Ratingen score 3.6 vs. 1.3; p = 0.011), despite comparable disease activity. In analyses stratified by RF status, the BMI associated difference in radiographic progression was observed only in RF seropositive subjects. In multivariable regression models, the risk of significant radiographic progression was greater than 3-fold higher in RF seropositive normal weight subjects compared with RF seropositive obese subjects (OR 3.2 (95% CI 1.7 6.2). In contrast, no significant difference in the risk for significant radiographic progression was observed when RF seronegative normal weight subjects were compared to RF seronegative obese subjects.
Conclusions: Obesity was associated with lower baseline and progressive radiographic damage compared to normal weight in early RA patients, despite comparable baseline disease activity and treatment.
Editorial Comment: Even though obesity has been identified as a risk factor for the development of RA, studies like the one presented here add to a short list of others suggesting a protective effect of increasing BMI in patients who have developed the disease. The underlying basis for this protective effect is not clear and is somewhat paradoxical, as adipose tissue is a known source of inflammatory cytokines. Although no studies have directly addressed the issue, one might speculate that these adipose derived cytokines would be associated with more, rather than less, severe RA outcomes. Additional work delineating fat and muscle composition is required to clarify the underlying mechanisms contributing to the effects seen here.
OP0114 Serum Levels of RANKL and OPG in Rheumatoid Arthritis A 10-Year Longitudinal Study with Associations to Radiographic Progression
Syvesen, Odegard, Gaarder, van der Heijde, et al.
Levels of RANKL, an osteoclast activator, and OPG, an antagonist of RANKL, may be predictive of radiographic progression in RA given the role of osteoclasts in joint erosions in patients with RA. Syversen et al assess the association of serum levels of OPG and RANKL with radiographic progression over 10-years in an RA cohort.
Methods: 238 patients with RA (<4 years duration at onset, mean 2.3 years) were followed longitudinally for 10 years. Radiographs and serum collections were done at baseline, 5 years, and 10 years. Using the van der Heijde modified Sharp score, one reader read the radiographs. The time order was not blinded. Associations of RANKL, OPG and RANKL/OPG ratio with radiographic progression were examined with correlation analysis (Spearman) and ANOVA.
Results: 125 patients were included in this analysis. The levels of RANKL and OPG remained stable in individual patients over the 10-year period. Baseline levels of OPG, RANKL and RANKL/OPG ratio were not associated with radiographic progression at 10 years (see table below).
| OPG | R=-0.07, p=0.40 |
| RANKL | R=0.03, P=0.74 |
| RANKL/OPG ratio | R=0.04, P=0.60 |
No associations between OPG, RANKL, and RANKL/OPG ratios were observed at any of the 3 time points.
Conclusion: OPG and RANKL levels remained stable over time, and appeared to have no predictive value in determining radiographic outcomes in patients with RA over time.
Editorial Comment: These results are surprising since some previously published studies have shown the opposite. Although the number of patients studied is relatively small, the long duration of the study with three time points per study participant lends considerable power to the analysis. The hunt for a biomarker(s) that will prospectively identify those patients with the most destructive, erosive RA continues. RANKL and OPG dont appear to satisfy the requirement.
OP0138 Barriers to Medication Adherence in people with Rheumatoid Arthritis Using DMARDs
Van den Bemt, Van Lankveld, Van den Hoogen, et al.
Lack of efficacy of a prescribed medication raises questions about the level of patient adherence. Non-adherence to medications in other chronic illnesses has been shown to be as high as 50%. This study investigates factors associated with medication adherence in patients with rheumatoid arthritis (RA) currently using DMARDs.
Methods: 228 patients were interviewed about when and how they take their medication, and self-perceived efficacy, side effects, and adherence. After the interview, standardized self report questionnaires were administered to assess adherence (CQR), coping, beliefs about medicines, satisfaction about medicine information, and functional score (HAQ). Adherence was defined as compliance > 80%, and poor adherence as compliance < 40%.
Results: Average drug use was 5.4 drugs/day (2-19 drugs/day). In the interviews, more than 95% of patients reported being adherent, while the CQR suggested only 68% were in fact adherent. In this study population, only 2 factors significantly correlated with non-adherence; number of adverse events (r=0.17, p<0.01) and active coping style (r=0.13, <0.01). No correlation was found for with age, number of drugs taken per day, social support, HAQ, most coping styles, trust in rheumatology, or information satisfaction.
Conclusion: Not surprising, a higher percentage of patients believe they are adherent with their medication than actually are. However, of the variables examined, only adverse events and active coping style appear to correlate with adherence in this group of RA patients.
Editorial Comment: This study addresses an important question as some of the loss of efficacy of RA medications reported in real world experience compared to clinical trials could be explained by a decrease in adherence to medication. It is disturbing that few predictors of non-adherence could be identified in this rather large study. This makes it difficult to devise a systematic approach to improve adherence in patients with chronic illnesses such as RA. Also, a weakness in this study is the reliance on self-report. The use of an electronic cap that measures and records each opening of a medication bottle is a more objective measure of adherence.