Tender Joint Count May Not Reflect Inflammation Activity in Rheumatoid Arthritis

In patients with long-standing rheumatoid arthritis (RA), tender joint count was strongly associated with subjective measures such as Composite Disease Activity Scores (CDAS) and patient-reported outcome measures (PROMs), but was weakly associated with objective assessments of inflammatory activity such as ultrasound, according to study findings published in Arthritis Care & Research.

While swollen joint count is recognized as a useful measure of RA disease activity, the utility of TJC is less clear. Because TJC is based on multiple factors, it may not accurately represent current inflammation. Investigators sought to explore the relationship between TJC and various PROMs, as well as clinical assessment and ultrasound findings.

A secondary analysis of a longitudinal observational study enrolled 209 patients with RA (mean age, 53.3 years; 81% women; mean disease duration, 10.0 years) taking biologic disease-modifying antirheumatic drugs and assessed them at baseline and 1, 2, 3, 6, and 12 months. There were 152 (72.7%) individuals still enrolled at the end of the 12-month study.

The difference between TJC and SJC was used to evaluate the effects of tender joints on disease burden and inflammation, with a difference >0 representing patients with mainly tender joints (n=84; mean difference, 4.9) and a difference ≤0 indicating predominantly swollen joints (n=125; mean difference, −4.3). On ultrasound, grey scale changes were used to score the level of synovitis, while power Doppler was used to assess the degree of vascularization.

Pearson correlations and regression analysis were employed to evaluate the relationships between TSJD, inflammatory activity by ultrasound, and PROMs and CDAS. All clinical and radiographic evaluators were blinded to relevant results and outcomes.

Calculations revealed a strong association between TJC and PROMs (P <.001), with these outcomes and CDAS significantly higher in participants with a difference between TJC and SJC >0 compared with participants with a difference ≤0 at all follow-up visits (P <.001). In contrast, SJC was weakly associated with PROMs. Assessor’s global scores and laboratory markers were equivocal, without significant differences between groups. An analysis of ultrasound findings demonstrated significantly lower sum scores in participants with predominantly tender joints vs people with predominantly swollen joints (P <.001 to .03).

Overall, baseline difference between TJC and SJC was a positive predictor of all 6-month CDAS (P <.001-.019) and PROMs at all time points (P =.03 to <.001) but negatively predicted sum scores (grey scale and power Doppler) on 6- and 12-month ultrasound readings (P <.001). All variables and scores were significantly improved across 12 months in all patients, regardless of difference between TJC and SJC grouping (P <.001). There was no significant relationship between baseline difference between TJC and SJC and remission by CDAS, but the difference was significantly associated with 6- and 12-month remission according to power Doppler sum scores (odds ratios, 1.07-1.10; 95% CI, 1.01-1.18; P =.002 to .02).

Study strengths included a longitudinal design with comprehensive ultrasound and clinical evaluations, joint assessments carried out by 2 trained nurses, and sonographer expertise. Study limitations included possible limited generalizability owing to the use of a single center, analysis of only known patients with RA, and use of “last observation carried forward” to account for missing data.

“These results indicate that tender joints do not reflect the same pathology as found with [ultrasound],” concluded the authors, adding, “The novelty of our study is the inclusion of a comprehensive [ultrasound] assessment as an objective assessment of inflammation.”

This work was supported by AbbVie, Pfizer and Roche in form. Please see original article for conflicts of interest statement.

follow @RheumAdvisor

Reference

Hammer HB, Michelsen B, Provan SA, et al. Tender joint count may not reflect inflammatory activity in established rheumatoid arthritis patients – results from a longitudinal study [published online November 26, 2018]. Arthritis Care Res. doi:10.1002/acr.23815