Session Information
Title: Rheumatoid Arthritis - Clinical Aspects (ACR): Comorbidities, Treatment Outcomes and Mortality
Session Type: Abstract Submissions (ACR)
Background/Purpose: Group-based trajectory modeling defines clusters of individuals in observed data with similar disease trajectories. We have previously defined five distinct trajectories in early rheumatoid arthritis based on DAS28 scores. Our objective was to examine radiographic progression in these defined clusters.
Methods: Patients were assigned to mutually exclusive trajectories by their DAS28 scores over 24 months. Patients with baseline and follow-up radiographs (n=601) were included in this analysis and baseline demographics did not differ from the larger cohort (n=1,568). Radiographs were scored with the van der Heijde modification of the Sharp score (vdHSS). ANOVA with the Bonferroni correction was applied to test for mean differences in baseline and follow-up vdHSS between clusters. Paired t-tests were applied to test differences in mean total vdHSS between baseline and last follow-up by cluster group. Pearson’s chi-squared test was used to identify differences between clusters in the proportion of patients with vdHSS progression (³3.5 units/year) and rapid progression (³5 units/year).
Results: Clusters (Cl) were characterized as: Cl-1 (n=163, 27%) began in high disease activity and achieved remission; Cl-2 (n=121, 20%) began in moderate or low disease activity and achieved remission; Cl-3 (n=158, 26%) began in moderate disease activity and achieved low disease activity; Cl-4 (n=132, 22%) began in high disease activity and achieved moderate disease activity; and Cl-5 (n=27, 5%) began and remained in high disease activity. At 24 months, patients in Cl-5 were more frequently on biologics (35%) and steroids (38%) but were less frequently on methotrexate monotherapy (18%), despite earlier use of DMARD combination therapy (43% at 3 months) relative to other clusters. The overall mean baseline vdHSS score was 5.3 units (SD 8.4) with progression of 2.8 units annually (95%CI 2.4-3.2, p<0.001); 43% of the cohort had no change in radiographic scores in follow-up. Clusters differed in baseline and follow-up scores for joint space and total vdHSS but not erosions, with Cl-4 and Cl-5 experiencing the worst radiographic progression by 3.6 units annually (95%CI 2.5-4.6) and 4.4 units annually (95%CI 1.1-7.7) respectively (Table 1).
|
Table 1. Baseline and Mean Change in van der Heijde Sharp Scores over 24 months in Early Rheumatoid Arthritis, Overall and by Trajectory Cluster |
||||||
|
Erosion Score |
Joint Space Score |
Total vdHSS
|
||||
|
Baseline |
Change |
Baseline |
Change |
Baseline |
Change |
|
|
Mean (SD) |
Difference (95%CI, p value) |
Mean (SD) |
Difference (95%CI, p value) |
Mean (SD) |
Difference (95%CI, p value) |
|
|
OVERALL n=601 |
2.1 (3.8) |
1.4 (1.2-1.7, p<0.001) |
3.3 (6.1) |
1.4 (1.1-1.6, p<0.001) |
5.3 (8.4) |
2.8 (2.4-3.2, p<0.001) |
|
Cluster 1 (HDA to REM) n=163 |
1.8 (4.2) |
1.0 (0.6-1.4, p<0.001) |
2.1 (3.9) |
1.1 (0.7-1.6, p<0.001) |
3.9 (6.8) |
2.1 (1.4-2.9, p<0.001) |
|
Cluster 2 (LDA to REM) n=121 |
2.0 (3.4) |
1.5 (1.0-2.0, p<0.001) |
3.0 (4.7) |
1.3 (0.9-1.7, p<0.001) |
5.0 (6.7) |
2.7 (2.0-3.5, p<0.001) |
|
Cluster 3 (MDA to LDA) n=158 |
2.2 (3.9) |
1.5 (1.0-2.0, p<0.001) |
3.7 (6.1) |
1.1 (0.7-1.5, p<0.001) |
5.9 (8.8) |
2.6 (1.8-3.4, p<0.001) |
|
Cluster 4 (HDA to MDA) n=132 |
2.4 (3.6) |
1.8 (1.2-3.1, p<0.001) |
4.6 (8.8) |
1.8 (1.2-2.3, p<0.001) |
7.0 (11.0) |
3.6 (2.5-4.6, p<0.001) |
|
Cluster 5 (HDA) n=27 |
1.2 (2.0) |
1.7 (-0.1-3.6, p=0.065) |
2.3 (4.7) |
2.7 (0.9-4.5, p=0.0054) |
3.5 (5.4) |
4.4 (1.1-7.7, p=0.0107) |
|
Legend: HDA high disease activity; REM remission; LDA low disease activity |
||||||
Cl-4 and Cl-5 had the highest proportion of progressors (24 and 26% respectively) compared to the other clusters (range 14-19%) although not statistically significant (p=0.204) and Cl-5 was characterized by all patients being rapid progressors.
Conclusion: We have defined five clinical disease trajectories in early rheumatoid arthritis. The cluster defined by patients beginning and remaining in high disease activity have the lowest baseline radiographic scores but the highest propensity for radiographic progression despite aggressive therapy. This highlights the importance of defining patient prognosis at baseline and immediately optimizing therapy to prevent functional decline.
Disclosure:
C. Barnabe,
None;
Y. Sun,
None;
G. Boire,
None;
C. Hitchon,
None;
E. C. Keystone,
Abbott Laboratories,
2,
Amgen Canada,
2,
Astrazeneca Pharmaceuticals LP,
2,
Bristo-Myers Squibb,
2,
F. Hoffman La-Roche Inc.,
2,
Janssen Pharmaceutica Product, L.P.,
2,
Eli Lilly and Company,
2,
Novartis Pharmaceutical Corporation,
2,
Pfizer Inc,
2,
Sanofi-Aventis Pharmaceutical,
2,
Abbott Laboratories,
5,
AstraZeneca,
5,
Biotest,
5,
Bristol-Myers Squibb,
5,
F. Hoffman-La Roche Inc.,
5,
Genentech and Biogen IDEC Inc.,
5,
Janssen Pharmaceutica Product, L.P.,
5,
Eli Lilly and Company,
5,
Merck Pharmaceuticals,
5,
Pfizer Inc,
5,
Abbott Laboratories,
8,
AstraZeneca,
8,
Bristol-Myers Squibb,
8,
F. Hoffman La-Roche Inc.,
8,
Janssen Pharmaceutica Product, L.P.,
8,
Pfizer Inc,
8,
UCB,
8,
Amgen,
8;
J. C. Thorne,
None;
B. Haraoui,
AbbVie,
2,
AbbVie,
5,
Amgen,
2,
Amgen,
5,
Bristol-Myers Squibb,
2,
Bristol-Myers Squibb,
5,
Janssen Pharmaceutica Product, L.P.,
2,
Janssen Pharmaceutica Product, L.P.,
5,
Pfizer Inc,
2,
Pfizer Inc,
5,
Roche Pharmaceuticals,
2,
Roche Pharmaceuticals,
5,
UCB,
2,
UCB,
5;
J. R. Curtis,
Roche, Genentech, UCB Pharma, Janssen, CORRONA, Amgen, Pfizer, BMS, Crescendo, AbbVie,
2,
Roche, Genentech, UCB Pharma, Janssen, CORRONA, Amgen, Pfizer, BMS, Crescendo, AbbVie,
5;
D. van der Heijde,
Director Imaging Rheumtology BV,
4;
D. Tin,
None;
J. E. Pope,
Amgen,
2,
Amgen Inc.,
5;
V. P. Bykerk,
Amgen,
5,
Bristol-Myers Squibb,
5,
Pfizer Inc,
5,
UCB ,
5.
« Back to 2014 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/radiographic-progression-differs-between-trajectory-clusters-defined-by-das28-scores-in-early-rheumatoid-arthritis/