06/01/22
Evaluation of physician global assessment in patients with b/tsDMARD therapy selection aligned with a molecular signature response classifier: an analysis from The Study to Accelerate Information of Molecular Signatures (AIMS) in Rheumatoid Arthritis

 Johanna Withers, PhD; Lixia Zhang, PhD; Sam Asgarian, MD

Presented at EULAR 2022 Copenhagen, Denmark – June 1-4, 2022

Background:

Despite drug therapies that improve lives, tens of millions of patients annually are subjected to therapy using a “trial-and-error” approach because, until recently, it was not possible to use a patient’s unique molecular profile to inform targeted therapy selection. 

A blood-based molecular signature response classifier (MSRC) was shown to predict non-response to tumor necrosis factor-alpha inhibitor (TNFi) therapies in patients with rheumatoid arthritis (RA). This MSRC integrates disease associated gene expression and clinical features (anti-CCP, sex, BMI, PtGA)1. 

A recent study demonstrated that patient response to treatment, as defined by ACR50 at 6 months, informed by an MSRC result was more than 3 times better when an MSRC informed therapy selection (predicted non-responders prescribed an alternative mechanism of action vs predicted non-responders prescribed a TNFi; PNR-altMOA=34.8% vs PNR-TNFi=10.3%, p-value=0.05)1 (see Figure 1). Furthermore, when patients with no prediction of non-response were prescribed a TNFi for treatment (NP-TNFi), their outcomes (CDAI improvement ≥MID) were improved by 5 times compared to the cohort receiving TNFi therapy despite a signal of non-response from an MSRC (NP-TNFi=45.8% vs PNR-TNFi=10.3%, p-value = 0.005)1. 

Methods:

 This analysis reports on data from patients enrolled in AIMS between Sep 2020 and Oct 2021 who initiated a new b/tsDMARD, or continued existing therapy, following MSRC testing. Patients were ≥18 years of age with a clinical diagnosis of RA. Each had a PGA score measured at 3 or 6 months (n = 560). The cohort is composed of 560 subjects from 49 unique sites: 511 subjects paid in-person visits; 16 subjects paid remote visits; 33 subjects were missed to follow-up at 3 months; and 281 subjects paid an in-person visit at 6 months. 

The cohort was divided into patients whose targeted therapy selection was Aligned (PNR-altMOA, NP-TNFi or no prediction of non-response prescribed a drug with an alternative mechanism of action, NP-altMOA) and Not Aligned (PNR-TNFi) with MSRC results. Improvement from baseline of PGA was evaluated at 3 and 6 months, and swollen joints was evaluated at 3 months. 

Objective:

The objective of the Study to Accelerate Information of Molecular Signatures (AIMS) in RA was to build a clinical/molecular database of longitudinal data from RA patients managed in real-world settings with a focus on utilization of an MSRC test. The objective of this analysis was to evaluate physician global assessment (PGA) scores (0-100 visual analog scale) at 3 and 6 months and swollen joint counts at 3 months in RA patients with moderate to severe disease activity at baseline who received MSRC testing prior to a treatment decision. 

Results:

The PNR-altMOA cohort had significant improvement in PGA scores at 3 & 6 months compared to the PNR-TNFi cohort (3m: PNR-altMOA=18.4, n=163 vs PNR-TNFi =7.2, n=140, p-value=0.01; 6m: PNR-altMOA=13.1, n=80 vs PNR-TNFi=-1.11, n=90, p-value=0.013). 

The patient cohort NP-TNFi showed a trend of greater improvement in PGA scores at 3 & 6 months compared to PNR-TNFi (3m: NP-TNFi=11.5, n=134 vs PNR-TNFi=7.2, n=140, p-value=0.791; 6m: NP-TNFi=13.3, n=76 vs PNR-TNFi =-1.11, n=90, p-value=0.075) (Figure 2). Moreover, predicted non-responders prescribed an altMOA showed a trend of improvement in swollen joints at 3 months compared to predicted non-responders prescribed a TNFi (PNR-altMOA= 3.15, n=163 vs PNR-TNFi=1.7, n=140, p-value=0.08). The patient cohort NP-TNFi showed significantly more improvement in swollen joints at 3 months (NP-TNFi= 3.71, n=134 vs PNR-TNFi=1.7, n=140, p-value=<0.001). 

Conclusion:

The incorporation of MSRC testing into the b/tsDMARD selection process can improve patient outcomes and can help identify which patients may experience less benefit from TNFi therapies. 

View Full Poster