Background/Purpose: We previously reported abnormal cardiac MRI (CMR)-determined aortic stiffness in patients with early, treatment-naive RA^1,2. We now report on whether this vascular stiffness is modifiable with RA DMARD therapy, and explore whether TNFi confers additional advantage over MTX csDMARD +/- subsequent escalation to TNFi.

Methods: A sub-grp of patients without CVD from an early (symptoms<1 yr), DMARD-naïve, RA (ACR/EULAR criteria) RCT³ with DAS28≥3.2 were recruited for 3.0T CMR to determine aortic distensibility (AD) at baseline, yrs 1 and 2 at a dedicated cardiology-CMR unit. Patients were randomised to first-line etanercept (ETN)+MTX (grp 1) or MTX/treat-to-target escalation to triple csDMARD, and switch to ETN+MTX at week 24 if failed to achieve clinical remission (grp 2; non-escalated subgrp 2a). At week 48, ETN was stopped, standard of care treatment maintained with observation up to week 96. AD values were natural log (ln) transformed prior to analysis; results are expressed as ratios between groups. Patients with DAS28≥2.6 at wk 48 defined non-responders. Change in AD value from baseline was evaluated in combined grps 1 and 2; between group, and within group (response states) change in AD at 1 and 2 years was compared adjusting for baseline DAS28ESR .

Results: Eighty early RA patients of mean(SD) age 49.4(13.08) and systolic BP 123(16) were recruited. Median(IQR) ESR, CRP and mean(SD) DAS28 were 31(31)mm/hr, 8(23)mg/L and 5.6(1.5) respectively. 66(85%) and 59(76%) patients were ACPA and RF positive respectively. 17(21%) were current smokers.

Table 1 details the analyses, confirming significant reduction in mean AD from baseline to year 1 [3.59 (3.14,4.11) vs 2.99 (2.66,3.36) respectively; p<0.01]. This was maintained at year 2 [3.55(33.09,4.09) vs 2.99(2.66,3.36); p=0.04]. There appeared to be no numerical difference in change at years 1 (and 2, data not shown) when comparing grp 1 vs grp 2, all responders vs non-responders and grp 1 responders vs non-responders (all adjusted for baseline AD, age, sex). To clarify this apparent absence of effect of disease activity as represented by response status, correlation analyses between AUC disease activity and AD at year 1 in the combined grps 1&2, and between grps 1 & 2 also did not identify an association. Planned exploratory comparison of grp 1 ETN responders and grp 2a responders (no ETN exposure) suggested a 16% difference (0.84 (0.60, 1.18), p=0.30).

Conclusion: This first longitudinal CMR treatment-naive ERA RCT cohort demonstrates vascular function abnormalities are modifiable (reduced) upon introduction of RA DMARD therapy. Treatment strategy rather than disease activity appeared to influence AD change; if confirmed in a larger trial this would suggest ETN+MTX confers a greater benefit over standard initial MTX/csDMARD.

References

1.     Erhayiem B, et al. Trials 2014

2.     Erhayiem B, et al. Arthritis & Rheumatology 2015

3.     Dumitru et al. BMC Musculoskeletal Disorders 2016

Table 1