ACR Meeting Abstracts

ACR Meeting Abstracts

Abstract Number: 2735

Is Disease Duration an Independent Predictor of Treatment Response Among Patients with Rheumatoid Arthritis Initiating Abatacept?

Leslie Harrold1,2, KK Gandhi3, H Litman4, S Kelly3, YF Li2, E Alemao3, S Deveikis1 and J Kremer5, 1Corrona, LLC, Southborough, MA, 2University of Massachusetts Medical School, Worcester, MA, 3Bristol-Myers Squibb, Princeton, NJ, 4Corrona, LLC., Southborough, MA, 5Albany Medical College and The Center for Rheumatology, Albany, NY

Meeting: ACR/ARHP Annual Meeting 2015

Date of first publication: September 29, 2015

Keywords: ,

Session Information

Date: Tuesday, November 10, 2015

Title: Rheumatoid Arthritis - Small Molecules, Biologics and Gene Therapy Poster III

Session Type: ACR Poster Session C

Session Time: 9:00AM-11:00AM

Background/Purpose:
It has been postulated that patients with
longstanding RA have more treatment-resistant disease. We propose to examine
whether disease duration is an independent predictor of treatment response
among patients with RA initiating abatacept in a US national observational cohort. Methods: Using the Corrona RA registry, we identified new initiators of
abatacept between February 2006 and January 2014, with a follow-up visit at 1
year (±3 months), and assessment of disease activity using CDAI at both the
time of initiation and the 1-year follow-up, stratified by disease duration (0–2,
3–5, 6–10, >10 years). The primary outcome was mean Δ in CDAI, with a
secondary outcome as achievement of low disease activity (LDA; CDAI ≤10)
among those who initiated abatacept in moderate or high disease activity, or
achievement of remission (CDAI ≤2.8) in those who initiated abatacept in
low, moderate or high disease activity. For patients who switched agents before
1 year, the last observation before the switch was used to calculate mean
Δ in CDAI. Switchers were imputed as non-responders for calculation of LDA
and remission. Results: We identified 1746 abatacept initiators who met inclusion criteria
(disease duration 0–2 years: 243; 3–5 years: 313; 6–10 years: 431; >10 years:
759). Patients with longer disease duration were more likely to be older, not
working (including disabled and retired), more impaired (based on modified
Health Assessment Questionnaire), and with a greater number of prior biologics
used (Table 1). Unadjusted mean Δ in CDAI at 1 year ranged from –9.8 in
those with 0–2 years’ disease duration to –5.9 in those with >10 years
(p<0.001; Table 2). Unadjusted rates of LDA occurred in 44.7% of those with
0–2 years’ disease duration but diminished to 30.1% of those with >10 years
(p=0.002). Similarly, unadjusted rates of remission were observed in 20.8% of
those with 0–2 years’ disease duration but diminished to 11.7% of those with
>10 yrs (p=0.005). These differences were attenuated in multivariable models
and lost significance. The proportion of patients remaining on abatacept at 1
year among those with 0–2 years’ disease duration was 70% compared with 63.9% of
those with >10 years’ disease duration (p=0.07). Conclusion: Treatment with abatacept was associated with significantly greater
improvement in those with shorter disease duration in unadjusted analyses. This
significance was lost in adjusted models, but the trend remained. Further
exploration is needed to assess whether earlier treatment with abatacept is independently
associated with better outcomes.  

Table 1. Baseline characteristics
  Disease duration, years
  0–2 3–5 6–10 >10 p-value
Baseline characteristics N=243 N=313 N=431 N=759  
Mean age (SD), years 56 (13.9) 57 (13.3) 57.2 (12.6) 61.7 (11.5) 0.001
Female, n (%) 195 (80.2) 255 (81.5) 349 (81.0) 632 (83.3) 0.640
Smoking status         0.020
       Current, n (% of m*) 37 (15.4) 68 (21.8) 77 (17.9) 103 (13.6)  
Mean BMI (SD), kg/m2 30.0 (6.9) 29.9 (6.9) 29.5 (6.7) 29.5 (7.1) 0.22
Work status, n (% of m*)†         0.001
       Full time 114 (46.9) 102 (33.2) 159 (37.1) 209 (27.7)  
       Part time 26 (10.7) 32 (10.4) 41 (9.6) 70 (9.3)  
       Not working outside home 29 (11.9) 38 (12.4) 48 (11.2) 68 (9.0)  
       Student 4 (1.6) 10 (3.3) 9 (2.1) 13 (1.7)  
       Disabled 20 (8.2) 47 (15.3) 72 (16.8) 158 (21)  
       Retired 50 (20.6) 78 (25.4) 99 (23.1) 236 (31.3)  
Medicare, n (%) 63 (25.9) 101 (32.3) 142 (32.9) 363 (47.8) 0.001
Median mHAQ (IQR) 0.38 (0.75) 0.38 (0.75) 0.38 (0.75) 0.5 (0.75)  
Median CDAI (IQR) 20.5 (18.2) 19.8 (19.5) 19 (19.6) 18.3 (19.2) 0.45
Number of prior biologics/small molecules, n (%)         0.001
       0 107 (44.0) 45 (14.4) 50 (11.6) 79 (10.4)  
       1 85 (35.0) 130 (41.5) 161 (37.4) 244 (32.1)  
       2 42 (17.3) 105 (33.5) 132 (30.6) 246 (32.4)  
       3+ 9 (3.7) 33 (10.5) 88 (20.4) 190 (25.0)  
IQR=interquartile range; mHAQ=modified Health Assessment Questionnaire *m=number of patients with non-missing data; †P-value for number disabled vs others.
 

Table 2. Outcomes
  Disease duration, years  
  0–2 3–5 6–10 >10  
  N=243 N=313 N=431 N=759  
Primary outcome Mean (SE) Mean (SE) Mean (SE) Mean (SE) p-value
Δ in CDAI –9.8 (0.93) –8.7 (0.82) –7.0 (0.70) –5.9 (0.52) <0.001
Adjusted Δ in CDAI Model 1* –8.2 (0.80) –7.9 (0.68) –7.2 (0.58) –6.6 (0.45) 0.30
Adjusted Δ in CDAI Model 2† –7.9 (0.81) –7.9 (0.68) –7.2 (0.58) –7.0 (0.45) 0.64
Secondary outcome n (%) n (%) n (%) n (%) p-value
  N=206 N=259 N=336 N=585  
Unadjusted achievement of LDA/remission‡ 92 (44.7) 89 (34.4) 111 (33.0) 176 (30.1) 0.002
  OR

(95% CI)

OR

(95% CI)

OR

(95% CI)

OR

(95% CI)

 
Unadjusted likelihood of LDA Reference 0.65 (0.45, 0.94) 0.61 (0.43, 0.87) 0.53 (0.38, 0.74) 0.003
Adjusted likelihood of LDA,

Model 1*

Reference 0.80 (0.54, 1.18) 0.78 (0.53, 1.13) 0.67 (0.46, 0.96) 0.18
Adjusted likelihood of LDA,

Model 2†

Reference 0.90 (0.60, 1.35) 0.83 (0.56, 1.22) 0.75 (0.51,1.09) 0.44
*Adjusted for age, sex, baseline CDAI and the number of prior biologic/targeted synthetic DMARDs used †Adjusted for age, sex, CDAI, number of prior biologics/targeted synthetic DMARDs used, smoking status, BMI, work status, Medicare, and mHAQ ‡Among those in moderate or high disease at initiation LDA=low disease activity (CDAI ≤10); mHAQ=modified Health Assessment Questionnaire; OR=odds ratio
 
Disclosure: L. Harrold, Corrona, LLC., 3,Genentech and Biogen IDEC Inc., 5,Pfizer, AstraZeneca, 2; K. Gandhi, Bristol-Myers Squibb, 1,Bristol-Myers Squibb, 3; H. Litman, Corrona LLC, 3; S. Kelly, Bristol-Myers Squibb, 1,Bristol-Myers Squibb, 3; Y. Li, None; E. Alemao, Bristol-Myers Squibb, 3,Bristol-Myers Squibb, 1; S. Deveikis, Corrona, LLC., 3; J. Kremer, Corrona, LLC, 1,Corrona, LLC, 3.

To cite this abstract in AMA style:

Harrold L, Gandhi K, Litman H, Kelly S, Li Y, Alemao E, Deveikis S, Kremer J. Is Disease Duration an Independent Predictor of Treatment Response Among Patients with Rheumatoid Arthritis Initiating Abatacept? [abstract]. Arthritis Rheumatol. 2015; 67 (suppl 10). https://acrabstracts.org/abstract/is-disease-duration-an-independent-predictor-of-treatment-response-among-patients-with-rheumatoid-arthritis-initiating-abatacept/. Accessed .

« Back to ACR/ARHP Annual Meeting 2015

ACR Meeting Abstracts - https://acrabstracts.org/abstract/is-disease-duration-an-independent-predictor-of-treatment-response-among-patients-with-rheumatoid-arthritis-initiating-abatacept/