Background/Purpose: Recent research suggests that rheumatoid arthritis (RA), an autoimmune systemic inflammatory disease, increases the risk of venous thromboembolism (VTE) including pulmonary embolism and deep vein thrombosis. We compared the risk of VTE in RA patients initiating treatment with biologic disease-modifying antirheumatic drugs (bDMARD), methotrexate (MTX) or non-biologic DMARD (nbDMARD). 

 

Methods: We conducted a population-based cohort study combining three U.S. commercial insurance claims databases (2001-2012). Adult patients with a new diagnosis of RA were identified based on ≥2 RA diagnoses that were ≥7 days apart with a baseline period free of RA diagnosis or DMARD use for ≥1 year. Among these patients, initiators (starting or switching) of various DMARDs were identified.  Patients with a history of VTE, malignancy and use of anticoagulants at baseline were excluded. Drug regimens were categorized into three mutually exclusive hierarchical groups: (1) a biologic DMARD with or without nbDMARDs, (2) MTX without a biologic DMARD, or (3) nbDMARDs without a biologic DMARD or MTX. We calculated incidence rates (IR) with 95% confidence intervals (CI) of VTE identified by a previously validated algorithm with inpatient diagnosis codes (PPV 75-90%). Cox proportional hazards models stratified by propensity score (PS) deciles after asymmetric trimming were used for 3 pairwise comparisons, controlling for baseline demographic factors, comorbidities, medications, and health care utilization. Sensitivity analysis limited to 180 days of follow-up and PS-matched analyses were also performed.

 

Results:  A total of 29,481 RA patients with 39,647 treatment episodes were identified.  Mean (SD) age was 49 (12) years for bDMARD and nbDMARD groups and 51 (12) for MTX. The crude IR of VTE per 1,000 person-years was higher in the bDMARDs group (5.53, 95% CI 3.67-8.32) compared to nbDMARDs and MTX (Table). In the PS decile-stratified Cox regression, the hazard ratio (HR) of VTE associated with initiation of bDMARDs was 1.83 (95% CI 0.92-3.63) compared to nbDMARDs and 1.39 (95% CI 0.73-2.64) compared to MTX. The HR of VTE associated with initiation of MTX versus nbDMARDs was 0.78 (95% CI 0.50-1.21). In a sensitivity analysis limiting the follow-up up to 180 days, bDMARDs was associated with a significantly increased risk (HR 2.48, 95% CI 1.14-5.40). In the PS-matched analyses, the HR was 1.34 (95% CI 0.65-2.75) in bDMARDs vs. nbDMARDs and 1.73 (95% CI 0.90-3.32) in bDMARDs vs. MTX.

 

Conclusion: Among newly diagnosed RA patients, initiating a bDMARD was associated with a likely increase, but not statistically significant, in the risk of incident VTE compared to those initiating MTX or nbDMARDs, albeit low absolute risks of VTE. Initiation of bDMARDs was significantly associated with a 2.5-times elevated risk of VTE in the first 180 days.

 

Table.  Risk of venous thromboembolism associated with initiation of DMARDs: PS decile-stratified ‘as treated’ analysis
Exposure	Treatment episodes	VTE	Person-years (PY)	IR (95% CI) per 1,000 PY	HR (95% CI)	HR (95% CI) 0-180d only
bDMARDs	4488	23	4157.90	5.53 (3.67-8.32)	1.83 (0.92-3.63)	2.48 (1.14-5.40)
nbDMARDs	12859	32	7220.08	4.43 (3.13-6.26)	Ref	Ref
bDMARDs	4597	21	4368.17	4.81 (3.14-7.38)	1.39 (0.73-2.64)	1.80 (0.84-3.85)
MTX	13912	32	9258.12	3.46 (2.45-4.89)	Ref	Ref
MTX	16352	42	11075.60	3.79 (2.80-5.13)	0.78 (0.50-1.21)	0.86 (0.50-1.48)
nbDMARDs	14618	41	8249.97	4.97 (3.66-6.75)	Ref	Ref

 

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/risk-of-venous-thromboembolism-and-use-of-disease-modifying-antirheumatic-drugs-for-rheumatoid-arthritis/