Background/Purpose: Symptom clusters are stable groups of 2+ symptoms that are related to each other and frequently co-occur. Identifying symptom clusters in early RA may point to underlying mechanisms and RA subtypes, help predict disease trajectories and personalize symptom management to improve outcomes. We used scores from PROMIS-29 to identify symptom clusters at diagnosis in MTX-naïve patients starting MTX therapy. We evaluated the stability of these clusters and likelihood of transitioning between clusters over the first 6 months of treatment.

Methods: Data were from new onset RA patients (sx < 1 year) enrolled in the Canadian Early Arthritis Cohort (CATCH) who started MTX and had clinical and PROM measures available at 0, 3, and 6-months. Based on empirical and qualitative data, we used latent class analyses to identify symptom clusters from PROMIS-29 physical (pain, fatigue, sleep) and emotional (depression, anxiety) scales (levels: minimal, mild, moderate, severe). Models were compared using AIC, BIC, G-square and log-likelihoods. We estimated transitioning between clusters at 3- and 6-months using latent transition analyses.

Results: Of 310 adults with new RA, followed for 6 months and starting MTX, 67% were women and 78%White. Participants had a mean age of 56 yrs, CDAI 29.3, and RA symptom durationof 5 months. Optimal clusters included pain, fatigue, anxiety and depression and three intensity levels. We identified 4 discrete symptom clusters at diagnosis:MINIMAL (12%); Moderate-Severe Physical (M-S P: 40%); Mild Physical + Emotional (MILD P+E: 11%); Moderate-Severe Physical + Emotional (M-S P+E 37%). Clusters had similar sociodemographics except the MINIMALclass were slightly older and fewer were women; M-S P+E had a greater likelihood of depression history (Table). SJC and TJC were similar among classes; worse symptoms were associated with higher CDAI. Classes with emotional sx had significantly worse mood and sleep, and the M-S P class were more likely to report a history of depression. More patients with moderate-severe symptoms were on parenteral steroids. Most transitions occurred over the first 3 months. The best prognosis was for the MINIMAL class; 95% were classified similarly at 3 and 6 months. Among those with MILD P+E , almost all were classified similarly at 3 (82%) and 6 (96%) months. Participants with M-S P at diagnosis were also likely to transition to MINIMAL (64%) or MILD P+E (7%). The worse prognosis was among those with M-S P+E; at 3 months, a minority transitioned to MINIMAL (13%) or MILD P+E (33%) classes. Patients with either mild or moderate-severe emotional symptoms were the least likely to transition to milder symptom classes.

Conclusion: We identified 4 ERA patient subsets with varying physical (pain, fatigue) and emotional (depression, anxiety) symptoms and evaluated transitions over 6-months of initial MTX treatment. At baseline, most participants had moderate-severe Physical +/- Emotional symptoms. Patient subsets with emotional symptoms at RA-diagnosis were less likely to transition to better controlled symptom classes. Our results suggest that co-occurring emotional symptoms at diagnosis may reflect important subsets and signal a more guarded RA prognosis over the first 6 months.

« Back to ACR Convergence 2023

ACR Meeting Abstracts - https://acrabstracts.org/abstract/specific-symptom-clusters-at-diagnosis-signal-a-poorer-early-ra-prognosis-data-from-the-canadian-early-arthritis-cohort-catch/