New model may help identify which AAV patients are at risk of blood clots
Study links disease activity, cardiac involvement, immune cells to VTE
Written by |
A new prediction tool may help identify which people with ANCA-associated vasculitis (AAV) are at risk of developing venous thromboembolism (VTE), a condition in which blood clots block blood flow through veins, a study showed.
Using real-world patient data, researchers developed the AAV-VTE model based on four factors that independently predicted VTE in AAV patients: higher disease activity, cardiac involvement, and elevated blood levels of two immune cell types.
“The AAV-VTE model might help to facilitate individualized risk prediction and support clinical decision-making in patients with AAV,” the researchers wrote.
Model uses clinical, immune factors to estimate VTE risk in AAV
The study, “Immunological and clinical predicators for venous thromboembolism in ANCA-associated vasculitis patients: elevated natural killer cells and cardiac involvement,” was published in Arthritis Research & Therapy.
AAV is a group of autoimmune diseases in which the immune system mistakenly produces self-reactive antibodies, called ANCAs, that cause inflammation and damage to small blood vessels.
People with AAV are at increased risk of VTE, which includes two related conditions: deep vein thrombosis, in which blood clots form in deep veins, most often in the legs, and pulmonary embolism, in which a clot breaks off and travels to the lungs.
To identify factors that predict VTE in people with AAV, a team of researchers reviewed demographic and clinical data from 138 AAV patients treated at a single hospital in China.
Participants (80 men and 58 women) had an average age of 59.66 years. Among AAV types, 77 had microscopic polyangiitis (MPA), 29 had granulomatosis with polyangiitis (GPA), and 32 had eosinophilic granulomatosis with polyangiitis (EGPA).
A total of 26 patients (11 with MPA, 10 with EGPA, and five with GPA) experienced VTE. These included 24 deep vein thrombosis events and four pulmonary embolism events, with two patients having both.
AAV patients who experienced VTE were significantly more likely to have EGPA, higher disease activity, as measured by the Birmingham Vasculitis Activity Score (BVAS), and involvement of the kidneys, lungs, and heart.
Blood, immune markers differ in AAV patients who develop VTE
Blood tests also showed that VTE patients had significantly higher levels of inflammation markers and immune cells involved in inflammation, including neutrophils and eosinophils. The VTE group also had significantly lower numbers of regulatory T-cells, which help dampen inflammation and immune responses.
A separate group of 45 AAV patients, with characteristics similar to those of the larger group and follow-up data for up to 18 months (about 1.5 years), served as a validation group. Over a median follow-up of nine months, nine patients (20%) developed VTE: deep vein thrombosis in seven and pulmonary embolism in two.
In the validation group, the VTE rate increased over time: 2.22% at three months, 6.67% at six months, 11.11% at nine months, 15.56% at 12 months, 17.78% at 15 months, and 20% at 18 months. Rates were highest during the first several months of follow-up.
Statistical analysis identified four factors independently associated with an increased risk of VTE: higher disease activity, cardiac involvement, and elevated blood levels of two immune cell types, eosinophils and natural killer cells (which typically kill infected and diseased cells).
In particular, each one-point increase in the BVAS disease activity score was significantly associated with a 19.1% increase in the odds of VTE, and patients with cardiac involvement had about 25 times higher odds of VTE than those without.
Each one-unit increase in eosinophil counts was associated with about a 2.5-fold increase in the odds of VTE, while each one-unit increase in natural killer cell counts was associated with about a 0.4% increase in the odds.
Model built from four key factors shows strong discrimination
Using these four predictive factors, researchers built an AAV-VTE prediction model that showed a strong ability to distinguish patients who developed VTE from those who did not.
The model could discriminate between these two groups of patients, with scores of 0.887 in the initial group of 138 patients and 0.949 in the validation group.
To assess the model’s clinical utility, the team divided patients into three risk groups based on predicted probability: low risk (below 3%), intermediate risk (3% to 13.9%), and high risk (above 13.9%).
The team found that the model’s predictions closely matched the actual observed VTE rates in the 138 AAV patients in the low-risk group (1.7% vs. 2.2%), the intermediate-risk group (6.9% vs. 10.9%), and the high-risk group (47.9% vs. 43.5%). Similar patterns were also observed in the validation group.
“Our study highlighted the immunological and clinical potential predictors for VTE in AAV patients, especially the elevated [natural killer] cells and [heart] involvement, which should be taken seriously,” the researchers wrote.
“AAV patients might benefit from the AAV-VTE prediction model and the appropriate interventions as well as management guided by risk stratification to reduce the disease burden of patients with AAV and improve prognosis in clinical practice,” they concluded.
