Background/Purpose:

The HLA-DRB1 shared epitope (SE) is associated with joint destruction in ACPA+ patients (pts) with RA.¹ In the Early AMPLE trial, among ACPA+ pts with early RA, efficacy responses with abatacept (ABA) but not adalimumab were numerically higher in pts who were SE+ vs SE–.² This real-world analysis compared the mean change (∆) from baseline to the 6-mth follow-up visit in CDAI score among ACPA+/SE+ pts treated with ABA or TNF inhibitors (TNFi).

Methods:

Using data from CERTAIN (Comparative Effectiveness Registry to study Therapies for Arthritis and Inflammatory CoNditions; NCT01625650)³, this study included pts with RA who initiated ABA or a TNFi, were CCP3+ ( >20 U/mL), SE+ (presence of SE on HLA-DRB1 allele) and had moderate or high CDAI score ( >10) at initiation and 6-mth follow-up (+/− 3 mths). The primary outcome was mean ∆ in CDAI score over 6 mths; secondary outcomes included achievement of remission (CDAI ≤2.8) and low disease activity (CDAI ≤10) and mean ∆ in pain, fatigue and pt global assessment. Analyses were conducted using both propensity score (PS)-trimmed and -matched populations. PS was calculated within each line of therapy. The PS-trimmed cohort included pts with scores overlapping both populations. After PS trimming, TNFi initiators were PS matched 1:1 to ABA initiators within each line of therapy. A sensitivity analysis was conducted in biologic-experienced pts only. Mixed-effects models were used for primary and secondary outcomes and included baseline characteristics that remained unbalanced (standardized difference [sDiff] >0.1) between treatment groups.

Results:

In the overall PS-trimmed cohort (ABA, n=170; TNFi, n=157), baseline characteristics that were imbalanced (sDiff >0.1) between treatments included sex, RF+, physician and pt global assessments, pain, fatigue and prior biologic use (Table 1). Baseline characteristics in the overall PS-matched cohort (ABA, n=111; TNFi, n=111) were generally well balanced. Similar trends for imbalances in baseline characteristics were seen in biologic-experienced cohorts. In the overall population for both the PS-trimmed and -matched cohorts, there were numerically greater improvements in mean ∆ in CDAI  and nearly all secondary efficacy outcomes between ABA and TNFi (Table 2). There was no significant difference in the percentage of pts who remained on study medication at 6 mths. Similar trends for efficacy outcomes were seen for biologic-experienced pts; statistical significance was reached for mean ∆ in CDAI in the PS-trimmed cohort (Table 3).

Conclusion: There was a consistent numerical improvement in efficacy outcomes with abatacept over TNFi in pts with long-standing RA who were SE+ and ACPA+ over 6 mths in all four study cohorts. After adjusting for covariates that remained imbalanced between groups in the biologic-experienced PS-trimmed cohort, the improvement in CDAI with abatacept vs TNFi was significant. Additional studies to explore these findings in a larger population would be of interest.

References:

1. Huizinga TWJ, et al. Arthritis Rheum 2005;52:3433–3438.
2. Rigby W, et al. Presented at EULAR 2019. Poster LB008.
3. Pappas DA, et al. BMC Musculoskelet Disord 2014;15:113.

 Medical writing: Claire Line, PhD (Caudex)

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/the-comparative-effectiveness-of-abatacept-versus-tnf-inhibitors-in-patients-who-are-acpa-positive-and-have-the-shared-epitope-results-from-a-us-national-observational-study/