WG = Wegener's granulomatosis; anti-PR3 = anti–proteinase 3 antibodies. Anti-PR3 were measured at the time of thrombotic event using standard assay in the clinical laboratory of University of California at San Francisco. Normal level of anti-PR3 is <7 U/ml.
Thrombosis and pediatric Wegener's granulomatosis: Acquired and genetic risk factors for hypercoagulability
Article first published online: 8 DEC 2003
Copyright © 2003 by the American College of Rheumatology
Arthritis Care & Research
Volume 49, Issue 6, pages 862–865, 15 December 2003
How to Cite
Von Scheven, E., Lu, T. T., Emery, H. M., Elder, M. E. and Wara, D. W. (2003), Thrombosis and pediatric Wegener's granulomatosis: Acquired and genetic risk factors for hypercoagulability. Arthritis & Rheumatism, 49: 862–865. doi: 10.1002/art.11454
- Issue published online: 8 DEC 2003
- Article first published online: 8 DEC 2003
- Manuscript Accepted: 19 OCT 2002
- Manuscript Received: 29 JUN 2002
- NIH. Grant Number: M01-RR-01271
- Pediatric Clinical Research Center
Wegener's granulomatosis (WG) is not frequently associated with thrombosis, although intracranial thromboses of large and small vessels have been described in both adults and children. Historically, these events have been attributed to either extension of granulomatous lesions from the nasal cavity or to intracranial vasculitis (1, 2). More recently, isolated reports of thromboses at other sites have been described in several adults and 1 child (3, 4). Systematic investigation of factors contributing to hypercoagulability in these patients with WG was not reported.
We describe 5 pediatric patients who developed deep venous thromboses shortly after their presentation with WG. Two individuals simultaneously developed arterial thromboses. These thrombotic events were associated with significant morbidity and emphasize the need to identify individuals at risk in order to provide timely therapy. We evaluated these patients for genetic and acquired risk factors for hypercoagulability, and identified the presence of antiphospholipid antibodies (aPL), factor V Leiden mutation, and nephrotic range proteinuria. The potential contribution of these risk factors in the setting of WG is discussed.
This 8-year-old girl was diagnosed with WG after presenting with pulmonary hemorrhage, sinusitis, nephritis, and anti–proteinase 3 (anti-PR3) antibodies (880 U/ml, normal <7 U/ml). Treatment included intravenous methylprednisolone (30 mg/kg/day) for 3 days followed by daily oral prednisone (2 mg/kg/day) and cyclophosphamide (2 mg/kg/day). Pulmonary and renal vasculitis improved, however antibodies to anti-PR3 remained detectable. Two weeks after diagnosis, she developed a painful swollen left thigh, and doppler ultrasound demonstrated a deep venous thrombosis extending from the left popliteal vein to the left external iliac vein (Table 1). Evaluation for hypercoagulability revealed aPL by a false-positive rapid plasma reagin test (RPR) result on a single occasion with negative antitreponemal antibodies (Table 2). Other assays for aPL, including partial thromboplastin time (PTT), dilute Russell's viper venom time (dRVVT), and enzyme-linked immunoabsorbent assay for anticardiolipin antibodies (aCL), were all negative. In addition, antithrombin III activity, protein C activity, and free protein S levels were normal, and a factor V Leiden mutation was not identified. Treatment included intravenous heparin followed by oral coumadin with resultant recanalization of affected vessels. She subsequently developed multiple episodes of thromboses at the venous tip of her dialysis catheter despite ongoing anticoagulation therapy. Two years after presentation, she received a kidney transplant that was managed perioperatively with intravenous heparin therapy. She has remained on oral coumadin to prevent renal graft thrombosis and has done well during the 2.5 years since renal transplantation.
|Patient||Time from WG diagnosis to thrombotic event||Anti-PR3, U/ml||Location of thrombosis||Therapy||Recurrence of thrombosis|
|1||2 weeks||23||Left external iliac to popliteal vein||Heparin Coumadin||Venous dialysis catheter (multiple episodes)|
|2||1 week||21 (at diagnosis)||Portal vein||Heparin||None|
|Left pulmonary artery||Coumadin|
|3||2 months||25||Left internal jugular to axillary vein||Enoxaparin||None|
|4||5 months||<7||Left external iliac vein||Heparin||None|
|Right pulmonary artery||Coumadin|
|5||9 days||47||Right posterior tibial to lesser sapheneous vein||Enoxaparin||None|
|Left superficial femoral to lesser saphenous vein|
|Patient||aPL: false-positive RPR||aPL: PTT||aPL: dRVVT||aPL: aCL||aPL: β2GPI||Protein C activity||Free protein S level||ATIII activity||Factor V Leiden mutation||MTHFR C677T mutation||Prothrombin 20210A mutation||Proteinuria (mg/day) at time of thrombosis|
A 10-year-old girl presented with sinusitis, evolving saddle-nose deformity, pulmonary hemorrhage, nephritis with nephrotic range proteinuria (4 gm/day), and anti-PR3 antibodies (21 U/ml, normal <2 U/ml). One week after diagnosis and initiation of methylprednisolone (3 mg/kg/day) and oral cyclophosphamide (2 mg/kg/day), she developed abdominal pain and worsening nephritis. Doppler ultrasound revealed a portal vein thrombosis (Table 1). Anticoagulation was not started immediately due to the risk of worsening pulmonary hemorrhage. One week later, the patient developed upper back pain and magnetic resonance angiography (MRA) and doppler ultrasound demonstrated a large left descending pulmonary artery thrombus as well as extension of the portal vein thrombus into the inferior vena cava. A confirmatory dRVVT demonstrated the presence of a lupus anticoagulant on 2 separate occasions, and evaluation for antibodies to the phospholipid-binding protein β2 glycoprotein I (β2GPI) revealed low levels of IgM antibodies (22 SMU, normal <20 SMU). PTT, protein S level, and protein C and antithrombin III activities were normal. RPR and aCL were negative. She did not have factor V Leiden, methylenetetrahydrofolate reductase, or prothrombin mutations (Table 2). With progression of the thrombus, anticoagulation with intravenous heparin followed by oral coumadin and aspirin (81 mg/day) was instituted. Coumadin was discontinued 1 year later when repeat MRA demonstrated recanalization of the portal vein thrombus and reduction of the size of the pulmonary artery thrombus. She has done well on low-dose aspirin without thrombotic recurrence during the 1.5 years since discontinuation of coumadin.
A 13-year-old boy presented to an outside institution with sinusitis, glomerulonephritis requiring hemodialysis, and anti-PR3 antibodies (1,264 U/ml, normal <7 U/ml). He initially responded well to plasmapheresis, intravenous cyclophosphamide (750 mg/m2), and intravenous methylprednisolone (1 gm) followed by oral prednisone (1 mg/kg/day). However, 2 weeks later he re-presented with massive pulmonary hemorrhage and recurrent renal failure. Treatment included additional doses of intravenous cyclophosphamide and methylprednisolone. A left internal jugular central venous catheter was placed for intravenous access for 4 days. Two weeks later, while taking prednisone (2 mg/kg/day) and oral cyclophosphamide (3 mg/kg/day), he developed a swollen left arm. A thrombus extending from the left internal jugular vein to the axillary vein was documented by doppler ultrasound. At that time, anti-PR3 antibodies were decreased but still detectable (Table 1). He was anticoagulated with subcutaneous heparin but showed progression of his clot. He died of overwhelming fungal sepsis 1 month later. Hypercoagulability workup revealed that he was heterozygous for a factor V Leiden mutation. Protein C activity, antithrombin III activity, and free protein S level were normal; aPL were not detected (Table 2).
A 15-year-old boy was diagnosed with WG after presenting with pulmonary hemorrhage, sinusitis, nephritis, and anti-PR3 antibodies (168 U/ml, normal <7 U/ml). Five months after diagnosis, during treatment with prednisone (0.5 mg/kg/day) and oral cyclophosphamide (3 mg/kg/day), he presented with acute chest pain, mild worsening of nephritis, and an increased erythrocyte sedimentation rate to 62 mm/hour. His previous sedimentation rate had been normal, and his anti-PR3 antibodies were undetectable. A large right main pulmonary artery thrombus with extension to the descending pulmonary artery and a left common iliac vein thrombus were documented by computed tomography (CT) (Table 1). Evaluation revealed no detectable aPL and normal antithrombin III activity, protein C activity, free protein S level, and factor V genotyping (Table 3). Anticoagulant therapy included intravenous heparin followed by oral coumadin, which was discontinued 6 months later when recanalization was demonstrated. He continues taking aspirin (325 mg/day) with no recurrence of thrombosis 3 years later.
A 15-year-old boy was diagnosed with WG after presenting with gross hematuria, fever, 15-lb weight loss, and cough with a single episode of hemoptysis. Laboratory testing revealed anti-PR3 antibodies (1:80, normal <1:40), and renal biopsy demonstrated crescentic segmental necrotizing glomerulonephritis. Sinus CT demonstrated minimal mucosal thickening and chest radiograph revealed patchy infiltrates. Initial treatment included intravenous methylprednisolone followed by oral prednisone (0.7 mg/kg/day) and cyclophosphamide (2.5 mg/kg/day). Renal failure ensued and hemodialysis was initiated. Nine days later, he developed hemoptysis and bilateral leg pain. Doppler ultrasound revealed bilateral lower extremity thrombi extending from the posterior tibial vein to the lesser saphenous vein on the right and from the superficial femoral vein to the lesser saphenous vein on the left. Evaluation for hypercoagulable risk factors revealed the presence of a lupus anticoagulant by confirmatory dRVVT and by activated PTT with confirmation by the hexagonal phospholipid neutralization procedure. RPR, aCL, and β2GPI antibodies were negative. Factor V genotyping, antithrombin III and protein C activities, and free protein S level were not evaluated. Anticoagulation with enoxaparin was instituted and he has had no recurrence of thrombosis in the interval year.
These cases illustrate that significant thromboses can occur in large vessels at sites distal to the primarily affected areas of the sinuses, lungs, and kidneys in childhood Wegener's granulomatosis. All thrombotic events occurred soon after disease presentation and before remission had been achieved. With the exception of patient 1, who developed recurrent dialysis catheter thromboses, no recurrent thrombotic events have occurred in the followup periods, which range from 1 to 5.5 years. Antiphospholipid antibodies, acquired or genetic predisposition for hypercoagulability, and other unidentified factors may have contributed to the prothrombotic states.
Three patients (1, 2, and 5) demonstrated the presence of aPL at the time of their thrombotic events, although only patients 2 and 5 had detectable aPL on repeated testing. A retrospective chart review of 36 adult WG patients reported a 17% prevalence for aPL, however the presence of aPL was not correlated with a greater risk for thrombotic events in these patients (4). Another report (5) of 83 adult patients with WG reported an 11.5–23% prevalence for aPL, depending on the method used; however, these patients similarly did not demonstrate an increased prevalence of aPL-related clinical manifestations. Although strongly associated with thrombosis in the context of systemic lupus erythematosus, aPL are not necessarily associated with thrombotic events in other systemic vasculitidies, such as Behcet's disease (6), rheumatoid arthritis (7), or giant cell arteritis (8, 9). However, the prevalence of these antibodies and their relationship to the risk of thrombosis in childhood WG remain to be examined.
In addition to the presence of aPL, patient 2 also had nephrotic-range proteinuria. Nephrotic syndrome is associated with an increased risk for both venous and arterial thrombosis. Renal vein thrombosis is the most common manifestation, although deep vein thrombosis of the extremities and portal vein thrombosis have also been reported (10, 11). The hypercoagulable state is attributed to increased urinary excretion of anticoagulant proteins, such as antithrombin III and protein S; as in our patient; however, normal serum levels may be found. Disturbances of the fibrinolytic system have also been implicated (10). The combined effect of aPL and nephrotic range proteinuria may have contributed to hypercoagulability for patient 2.
Patient 3 was heterozygous for factor V Leiden mutation. This is associated with a 7-fold increased risk of venous thrombosis with most thrombotic events occurring in association with additional risk factors, such as use of oral contraceptives (12). In patient 3, it is possible that damaged endothelium resulting from active vasculitis of WG may have served as the additional risk factor.
Patient 4 had no identifiable risk factors for hypercoagulability, suggesting the potential role for other unmeasured risk factors, such as active vasculitis. Although WG primarily affects the small vessels of the skin, lungs, and kidneys, the thrombotic events in our patients involved large vessels and suggests that the endothelial dysfunction associated with Wegener's may be more widespread than is traditionally understood. Because proinflammatory cytokines are elevated in the serum of Wegener's patients, circulating tumor necrosis factor (TNF) could potentially mediate activation of endothelial beds in both large and small vessels (13). Notably, TNF can increase gene expression of procoagulant proteins, such as tissue factor, in large vessel endothelial cells (14). Similarly, in the same type of cells, anti-PR3 antibodies can also induce increased expression of tissue factor and thus potentially promote a prothrombotic state in large vessels (15). Cytokine-mediated thrombophilia may explain why these subjects developed thromboses during the initial stage of disease when inflammation was not yet controlled. Alternatively, thrombotic potential during this phase of disease may reflect treatment, such as high-dose steroids, or physical inactivity resulting from severe illness.
This report adds to the small but growing number of reports describing thrombotic events in the setting of WG, and is the first to address the potential contribution of known genetic and acquired risk factors for hypercoagulability in children with these complications. The potential for thrombosis in WG should be anticipated, especially in recently diagnosed patients whose disease is not yet under control, because these thrombotic events are frequently associated with significant morbidity and potential loss of transplanted allografts. The incidence of thrombosis in WG, the exact contribution of known hypercoagulable risk factors (including aPL, Factor V Leiden mutation, and nephrotic-range proteinuria), and the possible role of other factors awaits systematic prospective study. Ultimately, routine screening for hypercoagulable risk factors in these patients may be of use for identifying individuals at increased risk for thrombosis, so that timely and appropriate prophylactic and therapeutic anticoagulation can be initiated. However, the development of reliable treatment guidelines will require prospective study through placebo-controlled trials.