Background/Purpose: Venous thromboembolism (VTE), the combined endpoint of deep venous
thrombosis (DVT) and pulmonary embolism (PE), is a common and potentially
deadly medical problem with an annual incidence between 1-3/1,000.  While rheumatoid arthritis (RA) has been
linked to an increased incidence of VTE, few studies have examined the risk associated
with psoriasis and psoriatic arthritis (PsA).  The objective of this study was to
determine the risk of VTE among patients with PsA, psoriasis and RA compared with
the general population.

Methods: A
longitudinal cohort study was conducted in The Health Improvement Network
(THIN), a primary care medical record database in the United Kingdom, to
determine the risk of VTE, DVT only and PE only using data from 1994-2014.
Patients aged 18-89 were selected if they had a diagnosis of PsA, RA, or
psoriasis.  Diagnosis codes for PsA,
RA, psoriasis, DVT and PE have been validated in THIN.  Up to 5 unexposed controls matched on
practice and start date within the practice were selected for each exposed patient.  Cox proportional hazards models were
used to calculate the relative hazards for each outcome. Multivariable models
were constructed to include confounders which changed the main effects by >10%
(e.g., hospitalization, cancer, hypertension, hyperlipidemia, liver disease,
oral glucocorticoid use, NSAID use, body mass index, smoking, and alcohol use).  An interaction with disease modifying
anti-rheumatic drug (DMARD) was hypothesized a priori and was significant. Adjusted
hazard ratios (aHR) stratified by DMARD are presented.
Severe psoriasis was defined as a prescription for a DMARD specific to
psoriasis or phototherapy.

Results:
Patients with PsA (N=12,447), RA (N=51,949), psoriasis (N=194,639) and
unexposed controls (N=1,247,933) were identified. Patients with RA (with and
without a DMARD) and patients with psoriasis without a DMARD prescription had a
significantly elevated risk of VTE (See Table: aHR 1.50,
1.40, and 1.19 respectively).  This risk
was elevated but not statistically significant among patients with severe psoriasis
and PsA with a DMARD (aHR 1.16 and 1.22 respectively)
and not elevated in patients with PsA without a DMARD.  While the findings were similar for DVT,
the aHR were different for PE. Patients with RA (with
and without DMARD) and patients with PsA prescribed a DMARD had a significantly
higher risk of PE than controls. Adjusted HR for patients with psoriasis were not statistically significant. These results were
robust to several sensitivity analyses.

Conclusion:   Patients with RA had a significantly
increased risk of VTE compared with the general population even after
accounting for risk factors for VTE.  VTE and DVT were not significantly
elevated in PsA. 
However, patients with PsA on a DMARD had a substantially increased risk
of PE.  Further investigation of the
increased risk for PE in patients with PsA on a DMARD is needed.