PB1016 - Case-control study on dynamic upper limb ultrasound and biomarkers in idiopathic upper extremity deep vein thrombosis

Background: Idiopathic upper extremity deep vein thrombosis (IUEDVT) is postulated to arise from venous thoracic outlet syndrome (VTOS), where the subclavian vein is compressed intermittently upon arm abduction; resultant changes to vascular rheology may contribute to enhanced platelet activation. VTOS theory underpins the rationale for surgical decompression for IUEDVT. However, VTOS diagnostic criteria are lacking, and often subclavian artery (SCA) compression as demonstrated on dynamic upper limb ultrasonography (DULUS) peak systolic velocity ratio (PSVR, abduction PSV expressed over neutral PSV) and change in SCA dimensions with abduction are used as surrogate markers for venous compression.

Aims: • To establish a standardised protocol and reference intervals for DULUS
• To compare DULUS results in cases with controls, specifically SCA PSVR at 90° and 130° abduction and describe SCA change in dimension with 170° abduction compared to neutral(∆change)
• To compare D-dimer (fibrinolysis) and soluble glycoprotein VI (sGPVI; platelet activation) levels in cases and controls

Methods: Case-control study to establish DULUS reference ranges and compare results between 23 cases with antecedent IUEDVT treated with anticoagulation alone and 30 age- and sex-matched controls. Secondary outcomes included comparing D-dimer and sGPVI levels between groups.

Results: Our 23 cases (median age 33.9 years, 39% male, Table 1) were 26 months from IUEDVT diagnosis and received 7.1 months anticoagulation (range, 3 months-ongoing). Our controls were age- and sex-matched but lighter than cases. DULUS results could not differentiate between cases and controls (SCA PSVR, Figure 1, ∆change SCA diameter erect, case -7.7%[16.5], control -6.8%[17.8]) and neither could D-dimer (0.08vs0.08mg/l, p=0.86) or sGPVI (8.2vs13.5ng/ml, p=0.1).

Conclusion(s): DULUS measurements including SCA PSVR at 90° and 130° abduction and change in SCA dimensions, D-dimer and sGPVI could not differentiate between IUEDVT and controls. Our findings challenge the utility of DULUS in VTOS diagnosis and the role of VTOS in the pathogenesis of IUEDVT.