False-negative upper extremity ultrasound in the initial evaluation of patients with suspected subclavian vein thrombosis due to thoracic outlet syndrome (Paget-Schroetter syndrome).

OBJECTIVE

To assess the utilization and consequences of upper extremity Duplex ultrasound in the initial diagnostic evaluation of patients with suspected subclavian vein (SCV) thrombosis and venous thoracic outlet syndrome (VTOS).

METHODS

A retrospective single-center review was conducted for patients that underwent primary surgical treatment for VTOS between 2008 and 2017, in whom an upper extremity ultrasound had been performed as the initial diagnostic test (n = 214). Clinical and treatment characteristics were compared between patients with positive and false-negative ultrasound studies.

RESULTS

There were 122 men (57%) and 92 women (43%) that had presented with spontaneous idiopathic arm swelling, including 28 (13%) with proven pulmonary embolism, at a mean age of 30.7 ± 0.8 years (range 14-69). Upper extremity ultrasound had been performed 23.8 ± 12.2 days after the onset of symptoms, with confirmation of axillary-SCV thrombosis in 169 patients (79%) and negative results in 45 (21%). Of the false-negative ultrasound study reports, only 8 (18%) acknowledged limitations in visualizing the central SCV. Definitive diagnostic imaging (DDI) had been obtained by upper extremity venography in 175 (82%), computed tomography angiography in 24 (11%), and magnetic resonance angiography in 15 (7%), with 142 (66%) undergoing catheter-directed axillary-SCV thrombolysis. The mean interval between initial ultrasound and DDI was 48.9 ± 14.2 days with no significant difference between groups, but patients with a positive ultrasound were more likely to have DDI within 48 hours than those with a false-negative ultrasound (44% vs 24%; P = .02). At the time of surgical treatment, the SCV was widely patent following paraclavicular decompression and external venolysis alone in 74 patients (35%). Patch angioplasty was performed for focal SCV stenosis in 76 (36%) and bypass graft reconstruction for long-segment axillary-SCV occlusion in 63 (29%). Patients with false-negative initial ultrasound studies were significantly more likely to require SCV bypass reconstruction than those with a positive ultrasound (44% vs 25%; P = .02).

CONCLUSIONS

Duplex ultrasound has significant limitations in the initial evaluation of patients with suspected SCV thrombosis, with false-negative results in 21% of patients with proven VTOS. This is rarely acknowledged in ultrasound reports, but false-negative ultrasound studies have the potential to delay definitive imaging, thrombolysis, and further treatment for VTOS. Initial false-negative ultrasound results are associated with progressive thrombus extension and a more frequent need for SCV bypass reconstruction at the time of surgical treatment.