Va s c u l a r a n d I n t e r ve n t i o n a l R a d i o l o g y • O r i g i n a l R e s e a r c h Ersoy et al. 3D MR Angiography of Vascular Thoracic Outlet Syndrome

Downloaded from www.ajronline.org by 37.44.207.158 on 01/16/17 from IP address 37.44.207.158. Copyright ARRS. For personal use only; all rights reserved

Vascular and Interventional Radiology Original Research

Hale Ersoy 1,2 Michael L. Steigner 1 Karl B. Coyner 1 Marie D. Gerhard-Herman 3 Frank J. Rybicki1 Raphael Bueno 4 Louis L. Nguyen 5 Ersoy H, Steigner ML, Coyner KB, et al.

Keywords: MR angiography, MRI, thoracic outlet syndrome DOI:10.2214/AJR.11.6417 Received January 2, 2011; accepted after revision October 4, 2011. 1

Department of Radiology, Brigham and Women’s Hospital, Boston, MA. 2

Present affiliation: Department of Radiology, Ohio State University Medical Center, 395 W 12th Ave, 4th Fl, Ste 426, Columbus, OH 43210. Address correspondence to H. Ersoy ([email protected]).

3 Department of Vascular Medicine, Brigham and Women’s Hospital, Boston, MA. 4 Department of Thoracic Surgery, Brigham and Women’s Hospital, Boston, MA. 5 Department of Vascular Surgery, Brigham and Women’s Hospital, Boston, MA.

AJR 2012; 198:1180–1187 0361–803X/12/1985–1180 © American Roentgen Ray Society

1180

Vascular Thoracic Outlet Syndrome: Protocol Design and Diagnostic Value of ContrastEnhanced 3D MR Angiography and Equilibrium Phase Imaging on 1.5and 3-T MRI Scanners OBJECTIVE. The purpose of this article is to evaluate the efficiency and reproducibility of a contrast-enhanced 3D MR angiography (MRA) protocol, using the provocative arm position on 1.5- and 3-T MRI scanners, and to determine the frequency and distribution of vascular compression and vascular complications in the thoracic outlet. MATERIALS AND METHODS. Seventy-eight consecutive patients with clinically suspected thoracic outlet syndrome (TOS) were included in the study. Two radiologists independently analyzed all eligible vessel segments, and interobserver agreement was determined using kappa statistics. The distribution of vascular compression with regard to the clinical presentation at referral was also analyzed. RESULTS. A venous component, which presented with mainly venous symptoms and findings, was confirmed in 85% of the subjects. An arterial component, which presented with clinical symptoms and findings of vascular TOS syndrome, was seen in 82% of the subjects. The vascular component of TOS, which presented with mainly neurogenic or indeterminate symptoms or findings, was excluded in 61% of the subjects. CONCLUSION. Contrast-enhanced 3D MRA using provocative arm positioning allows excellent imaging of the arteries and veins on both sides and thus provides a noninvasive imaging alternative to digital subtraction angiography in patients with suspected vascular TOS. Contrast-enhanced 3D MRA is also an ideal imaging modality for postsurgical follow-up for identifying restenosis or residual vascular compression. However, all imaging studies, including the contrast-enhanced 3D MRA protocol described here, should be treated as complementary tests for the diagnosis of TOS.

T

horacic outlet syndrome (TOS) is caused by the impingement of the brachial plexus nerves, subclavian artery (SCA), and subclavian vein (SCV) in the thoracic outlet, the area just above the first rib and behind the clavicle. More than 90% of cases present with neurogenic symptoms, whereas fewer than 10% of patients have vascular only or combined neurogenic and vascular symptoms [1–6]. Most vascular compressions occur in the costoclavicular space, whereas the retro pectoralis minor space is rarely the site of compression [7]. When vascular involvement is a concern, further investigation is warranted because early recognition and treatment of the vascular compression and associated complications are essential to prevent rare but devastating clinical outcomes. These include pulmonary embolism and venous gangrene of the hand associated with venous TOS (primary effort thrombosis, also

known as Paget–von Schrötter syndrome) [2, 8, 9], or digital ischemia and stroke as a result of emboli originating from the injured SCA [1, 2, 5, 6, 10]. The diagnosis of TOS is usually done with careful history, physical examination (provocative tests), radiography, electrodiagnostic tests, and brachial plexus neurography. However, in patients with a suspected vascular component, the diagnosis may be aided by imaging modalities. Most clinicians consider conventional digital subtraction angiography (DSA) the standard of care in the evaluation of the vascular anatomy in the thoracic outlet. However, DSA carries the risk of potential nephrotoxicity from the iodinated contrast agents. Ionizing radiation can be substantial when combined with the therapeutic interventions. In addition, arterial puncture site complications and stroke are part of the procedural risk. Therefore, DSA is usually reserved for transcatheter inter-

AJR:198, May 2012

3D MR Angiography of Vascular Thoracic Outlet Syndrome TABLE 1:  Thoracic Outlet Contrast-Enhanced 3D MR Angiography (MRA) Protocols for 1.5- and 3-T MRI Scanners Scanner, Pulse Sequence

Slice Thickness (mm)

TR/TE

Matrix

Flip Angle (°)

No. of Excitations

Bandwidth (Hz/pixel)

5

1008/91

256 × 160

90

0.55

488

1.5 T

Downloaded from www.ajronline.org by 37.44.207.158 on 01/16/17 from IP address 37.44.207.158. Copyright ARRS. For personal use only; all rights reserved

Single-shot fast spin-echo

2.6 (1.3)

4/1.4

288 × 192

35

0.75

244

Equilibrium fast acquisition with multiphase 3D elliptical centric fast gradient-echo (coronal)

2.6

3.3/1.2

256 × 192

12

0.75

244

Equilibrium liver acquisition with volume acquisition (axial)

5

4.5/2.1

320 × 192

12

0.75

244

HASTE

5.5

1200/101

320 × 224

150

1

780

3D contrast-enhanced MRA

3D contrast-enhanced MRA

3T 1.6

2.8/1.1

384 × 362

24

1

650

Equilibrium volume interpolated gradient-echo (coronal)

4

3.3/1.3

320 × 256

10

1

505

Equilibrium volume interpolated gradient-echo (axial)

4

4.4/1.8

320 × 168

10

1

505

ventions and preoperative evaluation of the vessels to determine surgical approach and bypass graft possibilities. With the purpose of establishing a reliable noninvasive imaging service for patients with suspected TOS, we have developed a contrastenhanced 3D MR angiography (MRA) protocol that involves imaging during arm abduction and rest. We assessed the efficiency of this protocol for identifying the vascular compression and associated vascular complications in the thoracic outlet. MRA results were correlated with the clinical symptoms and findings to determine the frequency and distribution of vascular compression with regard to the clinical presentation and the usefulness of this MRA protocol in patient management. Materials and Methods Patients The institutional review board approved the study and waived the requirement for informed patient consent because of the retrospective nature of using an anonymous-subject database. The patients’ medical records were obtained from the cardiovascular imaging section. Seventy-eight consecutive patients (55 women; mean age, 38 years; age range, 15–65 years) with clinically suspected TOS were included in the study. Eleven patients had undergone previous unilateral partial or total first rib resection and decompression surgery for TOS.

Contrast-Enhanced 3D MRA Protocol MRA examinations were performed on a 1.5T scanner (Signa HDx, GE Healthcare) for 37

patients and on a 3-T scanner (Magnetom Trio, Tim System, Siemens Healthcare) for 41 patients. Multichannel phased-array coils were used for signal reception. Gadobenate dimeglumine (0.5 mol/L; MultiHance, Bracco Diagnostics) or gadopentetate dimeglumine (0.5 mol/L; Magnevist Bayer HealthCare) was used as contrast agent. An automated injector was used for contrast agent and saline chaser administration. The pulse sequences and their imaging parameters for this protocol are provided in Table 1. First, T2-weighted imaging (single-shot fast spinecho on the 1.5-T and HASTE on the 3-T scanner) was performed. This is followed by breathhold arterial and venous phase contrast-enhanced 3D MRA and equilibrium phase imaging using a 3D gradient-echo pulse sequence with fat suppression. The first set of MRA and equilibrium phase images was acquired during 150–160° of bilateral arm abduction with the head and neck in the neutral position. A coronal oblique 3D slab of the MRA was prescribed to cover the bilateral subclavian and axillary vessels. Unenhanced mask imaging was followed by multiphase contrast-­ enhanced dynamic acquisition using the identical 3D slab and imaging parameters with a mask. Bolus timing was established using fluoroscopic triggering. The patients were instructed to hold their breath during the acquisitions. Contrast-enhanced images were obtained with the IV administration of 20 mL of gadolinium-based contrast agent and 20 mL of saline flush at a rate of 2 mL/s. All pulse sequences were repeated with the arm at rest next to the torso with the administration of 15 mL of gadolinium-based contrast agent and 20 mL of saline flush at a rate of 2 mL/s.

Image Interpretation and Data Analysis Image interpretation was performed on a 3D workstation using source images to obtain multiplanar reformatted, maximum intensity projection, and 3D volume-rendered images. For statistical analysis and clinical correlation purposes, each arm side was considered as a separate subject. Two radiologists independently analyzed all eligible vessel segments; observer 1 had 6 years of experience, and observer 2 had 2 years of experience in noninvasive cardiovascular imaging. Both observers were blinded to the clinical and other imaging data. Image quality was assessed by observer 1 for adequate positioning, motion artifacts, and venous contamination. The degree of compression was determined as the percentage of vessel diameter reduction (mild, < 50%; moderate, 50–75%, and severe, > 75%) during the arm abduction when compared with arm rest (Fig. 1). Any degree of SCA compression and greater than 50% SCV compression during the arm abduction were considered to be significant (Fig. 2). Thrombosis, occlusion, and persistent stenoses (present during both rest and arm abduction) with or without associated poststenotic dilatation or aneurysm were considered as supportive findings for the diagnosis of a vascular component of TOS in the presence of clinically suspected TOS. Interobserver agreement was determined by using the kappa statistic with quadratic weights. A few weeks later, the studies with discordant results were reanalyzed by both observers in consensus for the purpose of clinical correlation. Clinical presentation at referral and physical examination findings, including provocative test results, were used for clinical correlation.

AJR:198, May 2012 1181

Downloaded from www.ajronline.org by 37.44.207.158 on 01/16/17 from IP address 37.44.207.158. Copyright ARRS. For personal use only; all rights reserved

Ersoy et al.

A

B Fig. 1—18-year-old woman with left arm fatigue and neurogenic pain. Her medical history was remarkable for previous right Paget–von Schrötter syndrome that was successfully treated with first rib resection, anterior scalenotomy, and subclavian vein (SCV) angioplasty. A, Venous phase MR angiography (MRA) coronal maximum-intensity-projection (MIP) image shows significant compression of left SCV (arrow) at costoclavicular region during arm abduction. B, Sagittal oblique MIP image through left thoracic outlet space confirms significant SCV compression (arrow). C, Venous phase MRA coronal MIP image shows resolution of left SCV compression during arm rest.

C

A

B

Fig. 2—16-year-old girl with vasospastic symptoms in right hand. Adson test showed loss of radial pulse with arm abduction and contralateral turning of head. A, Coronal maximum-intensity-projection (MIP) image of arterial phase MR angiography (MRA) during arm abduction shows severe stenosis (arrow) of right subclavian artery. B, Coronal MIP image of arterial phase MRA during arm rest shows resolution of stenosis.

Subjects were classified as having predominantly arterial (e.g., Raynaud phenomenon, coldness, or reduced pulse or blood pressures during provocative maneuvers), predominantly venous (e.g.,

1182

color change, edema, or visible varicose veins), predominantly neurogenic (e.g., sensory-motor disturbance or muscular atrophy), or indeterminate (e.g., mixed presentation or nonspecific pain

and paresthesia) TOS. MRA findings were also compared with the DSA findings whenever DSA was performed within 15 days of the contrast-­ enhanced 3D MRA study date.

AJR:198, May 2012

Downloaded from www.ajronline.org by 37.44.207.158 on 01/16/17 from IP address 37.44.207.158. Copyright ARRS. For personal use only; all rights reserved

3D MR Angiography of Vascular Thoracic Outlet Syndrome

A

B

C

Fig. 3—34-year-old man with suspected thoracic outlet syndrome on right. A, Linear compression (arrow) over left subclavian artery (SCA) is seen. B, Source image of MR angiography (MRA) shows anterior scalenus muscle tendon, fibrous band, or aberrant ligament compression (arrow) over left SCA. C, Coronal maximum-intensity-projection image of venous phase MRA during arm rest shows resolution of linear compression of left SCA (arrow).

TABLE 2:  Distribution of MR Angiography Findings With Respect to the Predominant Clinical Presentation Predominant Clinical Presentation Findings

Asymptomatic

Arterial

Venous

Neurogenic

Indeterminate

Normal to mild

30

Moderate to severe

26

0

3

2

34

3

17

3

18

Occlusion

0

0

3

0

0

0

0

0

0

0

Both

6

5

4

0

2

Total

62

8

27

5

54

Subclavian vein only

Subclavian artery only

Results All patients tolerated the MRA study well, except one patient who developed facial rush, nausea, and nasal stuffiness after administration of contrast agent (MultiHance, Bracco Diagnostics). Some degree of venous contamination in the central thoracic and neck veins was unavoidable, because the contrast injection was done via the peripheral arm vein, but venous contamination did not obscure the arteries. Magnetic susceptibility artifact from high contrast agent concentrations in the axillary or SCVs on the same side of the IV injection was seen at arterial phase MRA in almost all of the patients. However, the artifact resolved on the following phases, and we were able to assess vessel lumens in all cases. Venous contamination from the first contrast agent administration was never an issue on the second phase (rest) of MRA, and we were able to evaluate arteries and veins in all of the patients. A suboptimal venous phase signal-to-noise ratio was common, but

it was not a problem because adequate venous enhancement on the equilibrium phase images was present in all subjects. In summary, all studies were eligible for image analysis. Interobserver agreement was very good (weighted κ = 0.83; 95% CI, 0.70–0.96; standard error [SE], 0.064) in detecting and grading of the arterial compression and very good (weighted κ = 0.83; 95% CI, 0.79–0.88; SE, 0.023) in detecting and grading of the venous compression and occlusion. Almost all significant SCV compressions were identified at the costoclavicular space. All arterial compressions were slightly distal in location. Arterial compression was always accompanied by significant venous compression during arm abduction. We have seen linear compression of the SCA during arm abduction, which resolved at rest in two subjects. We were not able to show the structure causing linear compression and concluded that the compression could be due to a tendon, a fibrous band, or an aberrant ligament (Fig. 3).

The distribution of vascular compression with respect to the predominant clinical presentation at referral is provided in Table 2. There was no significant vascular compression in 36 of 59 subjects (61%) who presented with mainly neurogenic or indeterminate symptoms or findings, and thus vascular TOS was excluded. Arterial compression was seen in 12 patients. Five of them (42%) had bilateral arterial compression. Eleven of 17 subjects (65%) with arterial compression had clinical symptoms or findings on the same side that were suggestive of vascular TOS in nine (82%) and indeterminate TOS in two (18%) subjects. Arterial compression was seen in six of 62 asymptomatic subjects (10%). There was no thrombosis, persistent arterial stenosis, occlusion, poststenotic dilatation, or aneurysm in our study population. Because previous publications reported an association between the arterial compression and bone abnormalities in 90% of cases, we also reviewed the

AJR:198, May 2012 1183

Downloaded from www.ajronline.org by 37.44.207.158 on 01/16/17 from IP address 37.44.207.158. Copyright ARRS. For personal use only; all rights reserved

Ersoy et al.

A

B

Fig. 4—34-year-old man with right upper extremity swelling and mild tenderness to palpation over right subclavian vessels. A, Equilibrium phase MR angiography image shows complete thrombosis (arrow) of right subclavian vein (SCV). B, Conventional venography confirmed right SCV occlusion (arrow).

chest radiographs to determine the frequency of such pathologic abnormalities. Chest radiographs were available for comparison in 11 of 17 subjects but did not show any bone abnormality or cervical rib that could be the cause of arterial compression. Significant venous compression was seen in 48 patients. Thirty-four of 48 patients (71%) had bilateral significant venous compression, but only seven of those 34 patients (21%) had bilateral clinical symptoms or findings suggestive of TOS. Significant venous compression was seen in 32 of 62 (52%) asymptomatic subjects. There was a strong clinical suspicion of venous TOS in 27 subjects. Medical records revealed a history of SCV thrombosis in eight of 27 subjects. Eleven of 27 subjects presented with acute symptoms and findings of SCV thrombosis. Complete SCV thrombosis was identified independently by both observers and was confirmed on DSA in three subjects (Fig. 4). Partial thrombus in the SCV was identified independently by both observers in four subjects. DSA was available for comparison and confirmed partial SCV thrombosis in three of them. Persistent SCV stenosis (two severe cases and one mild case) was seen in three subjects. MRA revealed severe SCV compression during arm abduction in 23 of 27 (85%) patients, and TOS was established as a contributing cause of clinical symptoms and findings. MRA did not show significant SCV compression during arm abduction in three subjects, and the vascular component of TOS was excluded. In one subject, an enhancing superior anterior mediastinal mass was the cause of SCV stenosis or compression.

1184

Eleven patients had prior unilateral first rib resection (three right and eight left). Six of 11 patients were symptomatic on the surgery side. Three had mainly vascular and three had indeterminate symptoms. MRA showed persistent severe stenosis of the SCV in one patient (Fig. 5) and residual severe compression of the SCV in two patients but failed to reveal the cause of extrinsic compression. Three patients with indeterminate symptoms had no or mild SCV compression. Four patients had a history of thrombolysis with or without balloon angioplasty for SCV thrombosis. MRA revealed residual partial thrombus in one patient (Fig. 6) and significant venous compression in three patients. Discussion Contrast-enhanced MRA combined with provocative maneuvers has been described in a few case reports [11–13]. The accuracy of the technique was investigated by Charon et al. [14] in 28 patients. In the present study, we introduced a modified contrastenhanced 3D MRA protocol in which pulse sequence parameters are optimized for acquisition time and imaging FOV and added a contrast-enhanced equilibrium phase acquisition to improve reader confidence in diagnosing venous complications and identifying extraanatomical causes of vascular compression in the thoracic outlet. We also added a T2-weighted pulse sequence for better characterization of the abnormal soft tissues or masses causing neurovascular compression in the thoracic outlet.

The described contrast-enhanced 3D MRA protocol consistently yields arterial and venous phase images after the first (during arm abduction) and second (during arm rest) contrast agent injections, whereas DSA requires separate procedures for arteries and veins. The imaging FOV of the MRA is large enough to provide vessel coverage of the aortic root through the bilateral distal axillary arteries and of the superior vena cava through the bilateral axillary veins. This enables assessment of the bilateral thoracic outlet vessels for compression and reversible or irreversible vascular damage. The nearly isotropic voxel diameter of contrast-enhanced 3D MRA data eliminates stairstep artifacts and allows reformations in any desired plane, including 2D projection images similar to those for the DSA projections. Contrast-enhanced equilibrium phase imaging enables better delineation of the veins and more confident diagnosis of a venous stenosis or thrombosis, particularly when venous enhancement was not adequate on the venous phase of the MRA. Equilibrium phase acquisition is also helpful for identifying other thoracic pathologic abnormalities that could explain the patient’s clinical symptoms and findings, such as extrinsic masses compressing the neurovascular structures. The superior soft-tissue imaging capability of MRI allows demonstration of radiographically invisible nonosseous causes of TOS. T2-weighted images through the chest are helpful in visualization of the brachial plexus, as well as softtissue masses within the thoracic inlet.

AJR:198, May 2012

Downloaded from www.ajronline.org by 37.44.207.158 on 01/16/17 from IP address 37.44.207.158. Copyright ARRS. For personal use only; all rights reserved

3D MR Angiography of Vascular Thoracic Outlet Syndrome

A

C

B

D

Fig. 5—40-year-old woman with previous left first rib resection for thoracic outlet syndrome who presented with new soreness in left arm and visible superficial veins on left anterior chest wall. A, Venous phase maximum-intensity-projection (MIP) image shows bilateral significant compression (arrows) of subclavian veins (SCVs) and absence of enhancement in right SCV (arrowhead), which was suspicious for SCV thrombosis. B, Venous phase MIP image shows resolution of compression and patent right SCV (arrowhead). Lack of enhancement in SCV during arm abduction was due to late enhancement secondary to severe compression of right SCV. C, Rest MR angiography coronal MIP image shows short segment severe stenosis (arrow) of left SCV. D, Digital subtraction angiography confirms severe stenosis (arrow) of left SCV and collateral venous circulation.

Contrast-enhanced 3D MRA has some advantages over other noninvasive imaging modalities, such as duplex ultrasound, CT angiography (CTA), and time-of-flight MRA, which have been explored and proven to be helpful for the diagnosis of vascular TOS. However, these imaging techniques have some disadvantages. The diagnostic accuracy of duplex ultrasound strongly depends on the experience of the operator, the patient’s ability to keep the provocative arm

position for long scan durations, the patient’s body habitus, and attenuation from the osseus structures limiting an adequate imaging window. Large collateral veins can be mistaken for the SCV in patients with thrombosed SCV. DSA is usually required to confirm the abnormalities identified on duplex ultrasound [15–17]. CTA is more reproducible compared with the duplex ultrasound because it is independent from the operator experience and can reveal the vascular struc-

tures of the thoracic outlet and their relationship to the osseus structures [7, 18, 19]. Similar to MRA, CTA allows image postprocessing using maximum intensity projection and multiplanar reformatted techniques and, thus, a more comprehensive evaluation of the vessels. However, CTA also uses iodinated contrast agents and ionizing radiation, which can be substantial with multiphase acquisition during the arm abduction and arm rest. In addition, streak artifacts

AJR:198, May 2012 1185

Downloaded from www.ajronline.org by 37.44.207.158 on 01/16/17 from IP address 37.44.207.158. Copyright ARRS. For personal use only; all rights reserved

Ersoy et al.

A

B

Fig. 6—17-year-old boy with spontaneous development of right subclavian vein (SCV) thrombosis. A, Equilibrium phase image shows complete thrombosis (arrow) of right SCV. B, Follow-up MR angiography study after right SCV thrombolysis and angioplasty shows residual thrombus (arrow) in right SCV.

from the high concentration of the iodinated contrast agent within the central veins may obscure thoracic outlet vessels on the arterial phase and thus can result in suboptimal image quality for assessing the thoracic outlet vessels. Arterial enhancement may not be optimal during the venous phase. Time-offlight MRA requires long acquisition times when combined with postural maneuvers, may be unacceptable for patients with severe clinical symptoms, and suffers from flow artifacts that could lead to misdiagnosis of stenosis or thrombosis [7, 20–22]. In our study population, arterial compression was always accompanied by significant venous compression during arm abduction, even in asymptomatic subjects. SCV compression was almost always seen in the costoclavicular space, whereas arterial compressions usually occurred slightly distal in location. According to our results, significant vascular compression tends to occur bilaterally. The frequency of bilateral significant venous compression with or without accompanied arterial compression was 71%, whereas the frequency of bilateral symptoms and findings of TOS was 21%. This may be explained with the concept of the dominant upper extremity, which is used more frequently than the nondominant side and is more likely to be used during strenuous activity. Significant venous compression was seen in 52% of subjects, and arterial compression was seen in 11% of the asymptomatic subjects in our study population. Previous studies have also reported that venous compression is frequently observed in asymptomatic

1186

individuals in all the compartments of the thoracic outlet after arm elevation [16, 19, 20, 23]. Therefore, contrast-enhanced 3D MRA alone cannot establish a confident diagnosis of vascular TOS, and treatment decisions based on false-positive outcomes have serious implications for mistreatment, such as inappropriate surgical intervention. Our study has some limitations. First, DSA results were available for only nine patients for comparison. However, DSA comparison in all cases would be impractical because the number of studies would be twice the number of symptomatic subjects (i.e., separate studies for arteriography and venography). Second, the T2-weighted imaging included in our protocol does not have adequate spatial resolution to identify anomalous ligaments or fibrous bands causing the neurovascular compression. We have seen linear SCA compression during arm abduction in two subjects but could not identify the structure causing linear compression. In these cases, focused MRI with multiple weightings and higher spatial resolution is required. Third, our protocol requires patient repositioning and two separate injections for arm abduction and rest positions. The double injection could be avoided by using blood pool MRA contrast agents such as gadofosveset. Blood pool MRA contrast agents provide a longer intravascular contrast phase, allowing repositioning of the arms while the contrast agent is still in the intravascular space. Finally, our MRA protocol is not optimized for the assessment of distal arm

and hand arteries. In patients with clinically suspected distal emboli, DSA should be the choice of imaging modality because it would allow diagnosis and therapeutic interventions at the same time. In summary, contrast-enhanced 3D MRA is an excellent noninvasive alternative to DSA in patients with suspected TOS because it allows imaging of the arteries and veins on both sides. Contrast-enhanced 3D MRA using provocative arm positions is helpful to determine the presence and degree of vascular compression and associated complications in the thoracic outlet. MRA successfully identifies the existence and extension of complete or partial venous thrombosis and persistent stenosis of the thoracic outlet vessels in patients with residual or recurrent vascular symptoms after interventional or surgical treatments. MRA is also an ideal imaging modality for postsurgical follow-up to detect restenosis or residual vascular compressions. However, all imaging studies, including the contrast-enhanced 3D MRA protocol described in this article, should be treated as complementary tests in diagnosing vascular TOS. References 1. Cooke RA. Thoracic outlet syndrome: aspects of diagnosis in the differential diagnosis of handarm vibration syndrome. Occup Med (Lond) 2003; 53:331–336 2. Davidović LB, Koncar IB, Pejkic SD, Kuzmanovic IB. Arterial complications of thoracic outlet syndrome. Am Surg 2009; 75:235–239 3. Hood DB, Kuehne J, Yellin AE, Weaver FA. Vas-

AJR:198, May 2012

Downloaded from www.ajronline.org by 37.44.207.158 on 01/16/17 from IP address 37.44.207.158. Copyright ARRS. For personal use only; all rights reserved

3D MR Angiography of Vascular Thoracic Outlet Syndrome cular complications of thoracic outlet syndrome. Am Surg 1997; 63:913–917 4. Sanders RJ, Hammond SL, Rao NM. Diagnosis of thoracic outlet syndrome. J Vasc Surg 2007; 46:601–604 5. Sanders RJ, Haug C. Review of arterial thoracic outlet syndrome with a report of five new instances. Surg Gynecol Obstet 1991; 173:415–425 6. Sheth RN, Belzberg AJ. Diagnosis and treatment of thoracic outlet syndrome. Neurosurg Clin N Am 2001; 12:295–309 7. Demondion X, Herbinet P, Van Sint Jan S, Boutry N, Chantelot C, Cotten A. Imaging assessment of thoracic outlet syndrome. RadioGraphics 2006; 26:1735–1750 8. Azakie A, McElhinney DB, Thompson RW, Raven RB, Messina LM, Stoney RJ. Surgical management of subclavian-vein effort thrombosis as a result of thoracic outlet compression. J Vasc Surg 1998; 28:777–786 9. Angle N, Gelabert HA, Farooq MM, et al. Safety and efficacy of early surgical decompression of the thoracic outlet for Paget-Schroetter syndrome. Ann Vasc Surg 2001; 15:37–42 10. Dorazio RA, Ezzet F. Arterial complications of the thoracic outlet syndrome. Am J Surg 1979; 138:246–250 11. Dymarkowski S, Bosmans H, Marchal G, Bogaert J. Three-dimensional MR angiography in the

evaluation of thoracic outlet syndrome. AJR 1999; 173:1005–1008 12. Hagspiel KD, Spinosa DJ, Angle JF, Matsumoto AH. Diagnosis of vascular compression at the thoracic outlet using gadolinium-enhanced high-resolution ultrafast MR angiography in abduction and adduction. Cardiovasc Intervent Radiol 2000; 23:152–154 13. Reid JR, Morrison SC, DiFiore JW. Thoracic outlet syndrome with subclavian aneurysm in a very young child: the complementary value of MRA and 3D-CT in diagnosis. Pediatr Radiol 2002; 32:22–24 14. Charon JP, Milne W, Sheppard DG, Houston JG. Evaluation of MR angiographic technique in the assessment of thoracic outlet syndrome. Clin Radiol 2004; 59:588–595 15. Demondion X, Vidal C, Herbinet P, Gautier C, Duquesnoy B, Cotten A. Ultrasonographic assessment of arterial cross-sectional area in the thoracic outlet on postural maneuvers measured with power Doppler ultrasonography in both asymptomatic and symptomatic populations. J Ultrasound Med 2006; 25:217–224 16. Longley DG, Yedlicka JW, Molina EJ, Schwabacher S, Hunter DW, Letourneau JG. Thoracic outlet syndrome: evaluation of the subclavian vessels by color duplex sonography. AJR 1992; 158:623–630

17. Wadhwani R, Chaubal N, Sukthankar R, Shroff M, Agarwala S. Color Doppler and duplex sonography in 5 patients with thoracic outlet syndrome. J Ultrasound Med 2001; 20:795–801 18. Remy-Jardin M, Remy J, Masson P, et al. Helical CT angiography of thoracic outlet syndrome: functional anatomy. AJR 2000; 174:1667–1674 19. Matsumura JS, Rilling WS, Pearce WH, Nemcek AA Jr, Vogelzang RL, Yao JS. Helical computed tomography of the normal thoracic outlet. J Vasc Surg 1997; 26:776–783 20. Demondion X, Bacqueville E, Paul C, Duquesnoy B, Hachulla E, Cotten A. Thoracic outlet: assessment with MR imaging in asymptomatic and symptomatic populations. Radiology 2003; 227:461–468 21. Demirbag D, Unlu E, Ozdemir F, et al. The relationship between magnetic resonance imaging findings and postural maneuver and physical examination tests in patients with thoracic outlet syndrome: results of a double-blind, controlled study. Arch Phys Med Rehabil 2007; 88:844–851 22. Smedby O, Rostad H, Klaastad O, Lilleas F, Tillung T, Fosse E. Functional imaging of the thoracic outlet syndrome in an open MR scanner. Eur Radiol 2000; 10:597–600 23. Stapleton C, Herrington L, George K. Sonographic evaluation of the subclavian artery during thoracic outlet syndrome shoulder manoeuvres. Man Ther 2009; 14:19–27

AJR:198, May 2012 1187