Improvement in cardiovascular biomarkers sustained at 4 years following an initial treat-to-target strategy in early rheumatoid arthritis

Rheumatology key message

• Initial treat-to-target therapy in early rheumatoid arthritis has sustained cardiovascular risk benefits at 4 years.

Sir. It is well recognized that individuals with RA are at greater risk of cardiovascular disease (CVD), with EULAR guidance advising optimal control of disease activity to reduce this risk [1]. The cardiovascular substudy of the Infliximab as Induction Therapy in Early Rheumatoid Arthritis (IDEA) trial evaluated infliximab (IFX) + MTX vs MTX + methylprednisolone (MP) (n = 38 and 41 in each group, respectively) using a treat-to-target approach in early RA and reported improvements in soluble cardiovascular biomarkers in both groups at week 78, with a greater improvement in insulin resistance in the IFX + MTX group [2, 3]. At week 78 the patients were discharged back to routine clinical care. Four years after their initial baseline IDEA visit, they were invited to participate in a follow-up study (IACON REC 09/H1307/98) to determine any long-term cardiovascular benefits of treat-to-target management and to evaluate for differences between the initial IFX + MTX and MTX + MP treatment arms.

Following obtaining informed consent, patients were assessed for RA disease activity and medication use and any new diagnosis of ischaemic heart disease (IHD), cerebrovascular accidents (CVAs), peripheral arterial disease (PAD), hypertension, hypercholesterolaemia or diabetes mellitus (following medical notes review/patient questioning). Mirroring our previous IDEA substudy, blood samples were taken to quantify three commonly measured soluble cardiovascular biomarkers: N-terminal pro-brain natriuretic peptide (NT-proBNP), homeostasis model assessment–estimated insulin resistance (HOMA-IR) and total cholesterol/high-density lipoprotein cholesterol (TC/HDL-C).

Eighteen patients were lost to follow-up between week 78 and year 4 (IFX + MTX, n = 10; MTX + MP, n = 8); follow-up data were available for 28 in the IFX + MTX group and 33 in the MTX + MP group. Comparing IFX + MTX vs MTX + MP at the baseline visit of the IDEA study, 64 vs 70% were female, 54 vs 64% were RF positive and 63 vs 78% were ACPA positive, respectively.

At year 4, of those initially in the IFX + MTX group, the three-variable 28-joint DAS (DAS28) was 1.80 (95% CI 1.10, 2.50), none were receiving oral steroids, 75% were receiving conventional synthetic DMARDs (csDMARDs; 61% MTX monotherapy) alone and 25% were receiving biologic DMARDs [bDMARDs; 18% TNF inhibitor (TNFi) therapy; 12% IFX], compared with those initially in the MTX + MP group, where the three-variable DAS28 was 1.9 (95% CI 1.1, 2.35), 9% were receiving oral prednisolone (median dose 5 mg), 73% were receiving csDMARDs (36% MTX monotherapy) alone and 18% were receiving bDMARDs (15% TNFi therapy, 12% IFX).

Since week 78, there were five new diagnoses of cardiovascular disease (IFX + MTX: 2 IHD, 1 CVA; MTX + MP: 1 IHD, 1 PAD), no new diagnoses of diabetes mellitus and seven new diagnoses of hypercholesterolaemia (IFX + MTX, n = 4; MTX + MP, n = 3). One (4%) of the IFX + MTX cohort received a new diagnosis of hypertension compared with six (18%) of the MTX + MP cohort (P = 0.225).

Soluble biomarker data were available for 40 patients (IFX + MTX, n = 20; MTX + MP, n = 20). Results at weeks 26 and 78 were comparable to those of the original IDEA cardiovascular substudy [2]. Continued improvements in soluble biomarkers of cardiovascular risk 4 years from baseline were shown regardless of the drug regimen (see Table 1); NT-proBNP values were 53–69% of baseline, HOMA-IR were 45–57% of baseline and TC/HDL-C decreased by 1.28–1.61. There were no significant differences observed between the treatment groups at year 4.

Table 1.

Changes to biomarker and lipoprotein values over time, separated by treatment regimen

Change in variable	MTX + MP	IFX + MTX	Unadjusted difference (95% CI), P-value	Adjusted difference (95% CI), P-valueb
Week 26
NT-proBNP mean ratio	0.88 (n=30)	0.95 (n=26)(0.96)a	1.09 (0.77, 1.54), 0.636c1.10 (0.77, 1.57), 0.610a,c	1.11 (0.78, 1.59), 0.548c1.14 (0.80, 1.61), 0.456a,c
HOMA-IR mean ratio	0.67 (n=30)(0.74)a	0.67 (n=25)	1.00 (0.57, 1.77), 0.990c 0.91 (0.53, 1.56), 0.725a,c	0.77 (0.50, 1.20), 0.244c0.74 (0.50, 1.11), 0.145a,c
TC/HDL-C mean	−0.64 (n=30)	−0.82 (n=26)	−0.18 (−0.88, 0.52), 0.619	−0.21 (−0.78, 0.37), 0.472
Week 78
NT-proBNP mean ratio	0.79 (n=30)	0.91 (n=22)(0.91)a	1.15 (0.75, 1.77), 0.517c 1.14 (0.74, 1.79), 0.529a,c	1.17 (0.76, 1.82), 0.473c 1.13 (0.75, 1.72), 0.550a,c
HOMA-IR mean ratio	0.81 (n=30)(0.89)a	0.66 (n=22)	0.81 (0.44, 1.51), 0.506c 0.74 (0.41, 1.33), 0.303a,c	0.64 (0.39, 1.06), 0.08c 0.62 (0.38, 0.098), 0.042a,c
TC/HDL-C mean	−0.94 (n=29)	−1.13 (n=21)	−0.19 (−1.08, 0.69), 0.663	−0.12 (−0.75, 0.51), 0.701
Year 4
NT-proBNP mean ratio	0.53 (n=20)	0.69 (n=20)(0.69)a	1.31 (0.74, 2.31), 0.342c 1.32 (0.73, 2.36), 0.346a,c	1.32 (0.73, 2.37), 0.350c 1. 31 (0.72, 2.38), 0.368a,c
HOMA-IR mean ratio	0.45 (n=20)(0.52)a	0.57 (n=20)	1.26 (0.67, 2.37), 0.471c 1.09 (0.61, 1.95), 0.765a,c	0.91 (0.58, 1.44), 0.679c 0.83 (0.59, 1.17), 0.272a,c
TC/HDL-C mean	−1.28 (n=20)	−1.61 (n=20)	−0.33 (−1.45, 0.78), 0.551	−0.47(−1.061, 0.122), 0.116
HDL-C mean (s.d.), (mg/dl)	11.6 (14.0) (n=20)	18.0 (11.7) (n=20)	6.4 (−1.8, 14.7), 0.125	6.8 (−1.2, 14.7), 0.093
LDL-C mean (s.d.), (mg/dl)	22.2 (28.2) (n=20)	5.9 (41.5) (n=20)	−16.3 (−39.0, 6.4), 0.155	−12.3 (−31.0, 6.4), 0.190
ApoA mean (s.d.), (g/l)	−0.025 (0.365)(n=20)	−0.018 (0.192)(n=20)	0.007 (−0.180, 0.193), 0.944	0.032 (−0.131, 0.195), 0.694
ApoB mean (s.d.), (g/l)	0.255 (0.159)(n=20)	0.187 (0.251) (n=20)	−0.069 (−0.204, 0.067),0.310	−0.062 (−0.188, 0.063),0.320
LpA, geometric mean, (g/l)	0.889 (n=16)	0.938 (n=15)	1.05 (0.72, 1.55), 0.777c	1.03 (0.69, 1.53), 0.887c

^aMinus extreme outlier.

^bAdjusted for baseline values.

^cValues exponentiated to give the ratio of the difference of one group vs another with associated CIs.

*P < 0.05.

ApoA: apolipoprotein A; ApoB: apolipoprotein B; LDL: low-density lipoprotein cholesterol; LpA: lipoprotein A.

While the difference in the incidence of new hypertension did not reach statistical significance, and notably some patients in the MTX + MP group were taking oral prednisolone at year 4, it may be clinically important. While glucocorticoids have a known association with hypertensive disease [4], IFX has been linked with reduced systolic blood pressure in patients with RA. In a trial of 16 RA patients, new IFX exposure was associated with a reduction in systolic blood pressure, along with reductions in plasma norepinephrine and renin activity [5]. This could suggest that the mechanism extends beyond that of simply reducing disease activity and inflammation.

Our cross-sectional analysis may also suggest that the initial beneficial impact of IFX + MTX on insulin resistance shown at week 78 had been lost by 4 years. However, the findings are limited by the small sample size and cross-sectional nature of the analysis, with a lack of knowledge about fluctuating disease activity/inflammation over the last 30 months—a known confounder of HOMA-IR [6]. In addition to patients receiving oral steroids at year 4 in the MTX + MP group, not all patients remained on IFX after week 78 in the alternative arm. Perhaps, for sustained insulin resistance improvement, continuous exposure to the drug is required.

To conclude, we report an intensive 78 week treat-to-target programme in early RA is associated with sustained long-term benefit in the improvement of soluble biomarkers of CVD, suggesting the potential for a reduction of cardiovascular risk in the long term.

Funding: This study was funded internally and supported by the National Institute for Health Research (NIHR) Leeds Biomedical Research Centre, NIHR Leeds Clinical Research Facility and Diagnostic Evaluation Co-operative and a research grant from the investigator-initiated studies program of Merck Sharp & Dohme. The views expressed are those of the authors and not necessarily those of the National Health Service, NIHR, the Department of Health or Merck Sharp & Dohme.

Disclosure statement: P.G.C. has participated in speakers bureaus or consulted for AbbVie, Bristol-Myers Squibb, GlaxoSmithKline, Novartis, Pfizer and Roche. M.H.B. has received research grant support and/or honoraria/consultancy fees from Pfizer, Roche-Chugai and UCB; has received honoraria/consultancy fees from AbbVie, Aurora, AstraZeneca, Bristol-Myers Squibb, Eli Lilly, Sandoz and Sanofi. P.E. has undertaken clinical trials and provided expert advice to Pfizer, Merck Sharp & Dohme, AbbVie, Bristol-Myers Squibb, UCB, Roche, Novartis, Samsung, Sandoz and Ely Lilly. E.V. has received consulting fees from AbbVie, Amgen, Bristol-Myers Squibb, Janssen and UCB and research support from AbbVie. H.K. has received honoraria from Janssen and research support from Centocor and Schering-Plough. A.W.M. has received research grant support and/or honoraria/consultancy fees from Merck Sharp & Dohme, Roche-Chugai, GlaxoSmithKline and Sanofi. L.A.B. has received honoraria from UCB. The other authors have declared no conflicts of interest.