Background/Purpose: Rheumatoid arthritis (RA) is associated with an increased risk of congestive heart failure, possibly via shared mechanisms of inflammation. This study was undertaken to test the hypothesis that the powerful anti-inflammatory effect of anti-interleukin 6 (tocilizumab: TCZ) therapy might lead to a reduction in left ventricular (LV) dysfunction in patients (pts) with RA. We sought to measure LV regional function by using a cardiac magnetic resonance imaging (CMR) approach in RA pts without cardiac symptoms, and also to evaluate the changes in these measurements at 52 wks of TCZ treatment.

Methods: This was an open-label prospective pilot study to directly evaluate the effect of TCZ on LV function in RA pts without a clinical diagnosis of cardiovascular disease. Consecutive RA pts with active disease and healthy control subjects were enrolled. The RA pts each had inadequate clinical response to non-biologic DMARDs or non-TNF directed therapy, and were prescribed TCZ therapy (8 mg/kg IV every 4 wks). All subjects underwent baseline evaluation of LV function, as measured by non-contrast CMR on a 1.5 T scanner. Peak systolic regional radial strain (Err, %) was calculated by feature tracking of cine MRI in the six segments of the mid-slice of LV. After the baseline (BL) CMR, treatment with the prescribed TCZ was initiated and pts were followed for 52 wks. Pts underwent follow-up CMR evaluation at 52 wks of treatment with TCZ. We examined the differences in peak Err between the control subjects and RA pts. We compared peak Err of RA pts at BL and at 52 wks, and determined the association of peak Err with disease activity and severity measures.

Results: All RA pts received TCZ treatment for 52 wks. We compared 11 RA pts (100% female; mean age 52.6±10.4 y) at BL and at 52 wks, with 10 non-RA controls (100% female; mean age 55.7±4.6 y). DAS28-ESR, SDAI, and Swollen Joint Count (SJC) were significantly lower in RA pts at 52 wks than at BL (p<0.0001, p<0.0001, p=0.003, respectively). Mean peak Err in RA pts at BL was significantly lower than in normal subjects (0.58±0.13 vs. 0.69±0.10; p=0.04); mean peak Err in RA pts at 52 wks was not significantly different from those of normal subjects (0.65±0.12 vs. 0.69±0.10; p=0.56) (figure). Mean peak Err tended to be lower at BL than at 52 wks, but not significantly (p=0.26). Percentage change of mean peak Err in RA pts was significantly associated with percentage change of SJC (r=0.74, p=0.009) and DAS28-ESR (r=0.70, p=0.017).

Conclusion: Our findings suggested sub-clinical LV regional dysfunction in RA pts without cardiac symptoms. Furthermore, we demonstrated that improvements in LV regional function by TCZ treatment correlated with reduction in measures of systemic inflammation and RA disease activity. This suggests that systemic inflammation contributes to myocardial dysfunction in RA.