Background/Purpose: The goal of treatment
for patients with rheumatoid arthritis (RA) is to achieve and maintain low
disease activity (LDA) or remission. Little information is available on the
likelihood and predictors of sustained response following LDA/remission among
patients receiving rituximab. The
objective of this analysis was to examine factors associated with sustained
response in patients with RA treated with rituximab in the Corrona registry.

Methods: Corrona is a US-based, prospective,
observational cohort of nearly 40,000 patients with RA. Patients with moderate
or high disease activity (Clinical Disease Activity Index [CDAI] > 10) who
initiated rituximab and who achieved LDA/remission (CDAI ≤ 10) within 12
months of their last rituximab infusion were included. Patients were followed
until loss of response (CDAI > 10), initiation of another biologic or small
molecule or exit from the registry. Covariates ascertained at the time of
rituximab initiation included functional status, CDAI and patient-reported
pain. Covariates examined at LDA/remission were demographic and clinical
characteristics, prior medications and concomitant steroid and conventional synthetic
disease-modifying antirheumatic drug (csDMARD) use. Time from rituximab
initiation to LDA/remission was estimated. Time-varying covariates included rate
of rituximab retreatment, number of prednisone dose increases and csDMARD
initiations over time. Survival analyses and Cox proportional hazard regression
models estimated the likelihood of sustained response following rituximab
treatment and associations between covariates of interest and sustained
response.

Results:
Of the
1184 patients who initiated rituximab and had ≥ 1 follow-up appointment,
306 patients achieved LDA/remission within 12 months of their last rituximab infusion
(median [IQR] time to LDA/remission, 5 [3-10] months). Most patients were female
(76%) with a median (IQR) age of 60 (52-68) years; over 84% of patients had
received ≥ 2 prior biologics. Of these patients, 34% received rituximab
retreatments at a median (IQR) rate of 1.01 (0.71-1.36) retreatments per year.
Twelve months after achievement of LDA/remission, 49% of patients maintained their
response. In adjusted models (Figure), rituximab retreatment was
associated with a significantly increased likelihood of sustained response,
while prednisone dose increases ≥ 7.5 mg and use of 1 prior non–tumor
necrosis factor inhibitor (TNFi) biologic were associated with increased
likelihood of loss of response.

Conclusion: In this cohort of patients
with refractory disease and use of several prior biologics who achieved
LDA/remission following rituximab treatment, rituximab retreatment was
associated with significantly greater sustained response. Increased prednisone
doses and prior non-TNFi biologic use, factors that are likely surrogate
markers of more severe disease, were associated with shorter sustained
response.