Advertisement

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

      Abstract

      Demirbag D, Unlu E, Ozdemir F, Genchellac H, Temizoz O, Ozdemir H, Demir MK. 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.

      Objectives

      To investigate the differences in findings from magnetic resonance imaging (MRI) in the neutral and provocative positions, and to examine the relationship between these differences and the results of physical examination tests in patients with thoracic outlet syndrome (TOS).

      Design

      Prospective.

      Setting

      University physical medicine and rehabilitation outpatient and radiology clinics.

      Participants

      Twenty-nine patients and 12 healthy controls. All of the patients had positive bilateral TOS stress tests; control group participants were symptom free and had negative TOS stress tests bilaterally.

      Interventions

      Not applicable.

      Main Outcome Measures

      All participants underwent Adson’s test, the Halsted maneuver, and a hyperabduction test. All were evaluated with MRI while in 2 positions: the neutral position (upper extremities adducted) and in a provocative position. Measurements were obtained at the interscalene triangle, at the costoclavicular space, and at the retropectoralis minor space.

      Results

      There was a significant difference in MRI findings between the neutral and provocative position in the patients (P<.05), but there were no significant differences in the control group. There was a significant difference in the positional change values in MRI between the patients and the control subjects (P<.05). The difference was found in the minimum costoclavicular distance between patients with a positive Halsted maneuver and a negative Halsted maneuver (P<.05).

      Conclusions

      Our findings indicate that MRI findings in patients in a provocative position are more valuable in the diagnosis of TOS, and these findings are in accord with findings from the physical evaluation tests.

      Key Words

      THORACIC OUTLET SYNDROME (TOS) involves the compression of the brachial plexus or subclavian vessels in their course from the cervical area toward the axilla and proximal arm, either at the interscalene triangle, the costoclavicular triangle, or the subcoracoid space. Its clinical presentation varies from severe compression with permanent neural and/or vascular lesions to intermittent postural symptoms without any organic damage.
      • Roos D.B.
      • Owens J.C.
      Thoracic outlet syndrome.
      • Novak C.B.
      • Mackinnon S.E.
      Thoracic outlet syndrome.
      • Demondion X.
      • Bacqueville E.
      • Paul C.
      • Duquesnoy B.
      • Hachulla E.
      • Cotten A.
      Thoracic outlet: assessment with MR imaging in asymptomatic and symptomatic populations.
      Symptoms of vessel abnormalities and neurologic signs may coexist in the same patient because the nerves and vessels are in close proximity in the thoracic outlet.
      • Huang J.H.
      • Zager E.L.
      Thoracic outlet syndrome.
      There is no generally accepted protocol for the diagnosis of TOS.
      • Huang J.H.
      • Zager E.L.
      Thoracic outlet syndrome.
      • Davidovic L.B.
      • Kostic D.M.
      • Jakovljevic N.S.
      • Kuzmanovic I.L.
      • Simic T.M.
      Vascular thoracic outlet syndrome.
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      • Matsumura J.S.
      • Rilling W.S.
      • Pearce W.H.
      • Nemcek Jr, A.A.
      • Vogelzang R.L.
      • Yao J.S.
      Helical computed tomography of the normal thoracic outlet.
      • Panegyres P.K.
      • Moore N.
      • Gibson R.
      • Rushworth G.
      • Donaghy M.
      Thoracic outlet syndromes and magnetic resonance imaging.
      Usual diagnostic criteria include clinical symptoms such as regular or intermittent neurovascular compression in the upper arm, positive results with dynamic provocative tests, and exclusion of other diseases such as cervical vertebral disorder and peripheral nerve diseases.
      • Tateishi A.
      Diagnosis and management of thoracic outlet syndrome.
      • Ohkawa Y.
      • Isoda H.
      • Hasegawa S.
      • Furuya Y.
      • Takahashi M.
      • Kaneko M.
      MR angiography of thoracic outlet syndrome.
      Four specific provocative maneuvers have been defined for TOS. In the Adson’s maneuver (or test), the patient’s arm hangs at the side, and the head is turned toward the affected side. The patient is instructed to breathe deeply. The radial pulse is monitored and the test is considered positive if the radial pulse disappears. This test is often modified by rotating the patient’s head to the unaffected side. In the Halsted maneuver (exaggerated military position), the patient assumes a military posture, with the shoulders backward and in a downward direction so to narrow the costoclavicular space. The radial pulse is monitored and the test is positive if the pulse disappears. Wright described a test with the shoulder hyperabducted to 180° and the elbow flexed. The radial pulse is monitored; this maneuver is also considered positive if the radial pulse disappears. The Roos overhead exercise test involves arm elevation for 3 minutes, with 90° of shoulder abduction and external rotation and with the elbow flexed at 90°; the patient is then asked to open and close the hand rapidly. The Roos test is considered positive if the patient’s symptoms are reproduced.
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      A diagnosis of TOS is usually made through a combination of physical examination (history, provocative tests) and diagnostic modalities (radiography, electrodiagnostic tests, brachial plexus neurography, color Doppler sonography, computed tomography [CT], magnetic resonance imaging [MRI], digital subtraction angiography).
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      • Panegyres P.K.
      • Moore N.
      • Gibson R.
      • Rushworth G.
      • Donaghy M.
      Thoracic outlet syndromes and magnetic resonance imaging.
      • Tateishi A.
      Diagnosis and management of thoracic outlet syndrome.
      • Ohkawa Y.
      • Isoda H.
      • Hasegawa S.
      • Furuya Y.
      • Takahashi M.
      • Kaneko M.
      MR angiography of thoracic outlet syndrome.
      • Seror P.
      Frequency of neurogenic thoracic outlet syndrome in patients with definite carpal tunnel syndrome: an electrophysiological evaluation in 100 women.
      There is, however, considerable disagreement among examiners about the best methods for diagnosing TOS. The disagreement is mainly related to the lack of specific objective diagnostic tests for early and accurate recognition of the neurovascular structures that are compressed.
      • Huang J.H.
      • Zager E.L.
      Thoracic outlet syndrome.
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      • Remy-Jardin M.
      • Doyen J.
      • Remy J.
      • Artaud D.
      • Fribourg M.
      • Duhamel A.
      Functional anatomy of the thoracic outlet: evaluation with spiral CT.
      Furthermore, little has been written about evaluating the thoracic outlet and neurovascular structures by using MRI to image the result of provocative maneuvers. To our knowledge, there has been no study into the relationship between the qualitative and quantitative findings of MRI and the clinical examination maneuvers when a control group and a patient group are compared. Therefore, we investigated (1) the differences in MRI findings with subjects in the neutral and provocative positions, and (2) the differences between the participants’ MRI findings and the results of the physical examination tests.

      Methods

      We used MRI to evaluate 29 consecutive patients and 12 healthy volunteer controls from September 2004 to July 2005. None of the patients had undergone surgery of the upper-thoracic region before the MRI studies. All participants provided their informed consent before the study.
      A physical medicine and rehabilitation (PM&R) specialist evaluated every participant. Patient complaints included waking up with numbness in both upper extremities, swollen and tense hands in the morning, pain and weakness in the upper extremities during overhead activities, and shoulder and arm pain, especially when tired or during stressful times. None of the patients had a history of radiculopathy, entrapment neuropathy, or any other neurologic diseases. Subjects had no motor and/or sensory loss and their deep tendon reflexes were within the referent range. There was no atrophy in any muscle group, cyanosis, objective edema, or ischemia in the upper extremities. A TOS stress test was performed on all patients. To standardize the patient group, we included only patients with bilateral positive findings on the Roos test and patients with symptoms in both arms. The PM&R specialist also evaluated the control subjects. In selecting our subjects, we ensured that those in the control group had no history of pain or numbness in the upper extremities, neck, or back areas. The control subjects also had bilateral negative TOS stress tests and normative findings on physical examination of the upper extremities.
      Another PM&R specialist examined participants in a blind design in which the specialist did not know to which group any participant belonged. In this examination, all participants were administered provocative tests including Adson’s test, hyperabduction test, and the Halsted maneuver for TOS. Test findings were recorded as either positive or negative.
      All subjects had normal electromyographic findings. Anteroposterior and lateral cervical spine radiographs were obtained routinely to detect cervical rib apophysomegaly of the C7 vertebra and other possible skeletal pathologies. MRI protocol and qualitative-quantitative evaluation criteria were adapted from a study by Demondion et al.
      • Demondion X.
      • Bacqueville E.
      • Paul C.
      • Duquesnoy B.
      • Hachulla E.
      • Cotten A.
      Thoracic outlet: assessment with MR imaging in asymptomatic and symptomatic populations.
      MRI was done with the 1-T system
      Magnetom Expert; Siemens Medical Solutions, Henkestrasse 127, D-91052 Erlangen, Germany.
      with 20mT/m maximum gradient strength. All examinations were performed with a standard body coil. After scout images, coronal and sagittal T1-weighted spin-echo sequences (repetition time, 510; echo time, 14) were acquired with both arms in adduction (the neutral position). The spin-echo sequence was then repeated with the participant’s arms elevated above the head (≈130° of abduction of the arms, ≈130° of flexion of the elbows) for the provocative maneuver. For the sagittal sequences, we obtained 16 contiguous slices for both sides. The average total imaging time was 20 minutes.
      Four radiologists who were unaware of the subjects’ physical examination findings evaluated the MRI scans both quantitatively and qualitatively. For the quantitative analysis, we measured the interscalene triangle (angle between the anterior scalene muscle and the middle-posterior scalene muscle), maximum thickness of the anterior scalene muscle, minimum costoclavicular distance, maximum thickness of the subclavius muscle, the angle between first rib and horizontal axis, and the retropectoralis minor space distance (Fig 1, Fig 2, Fig 3).
      Figure thumbnail gr1
      Fig 1Sagittal-weighted image of the sagittal T1-weighted MR image of the interscalene angle and thickness of anterior scalene muscle (in millimeters). Abbreviations: a, subclavian artery; AS, thickness of anterior scalene muscle (in millimeters); p, brachial plexus; v, subclavian vein. *The interscalene angle.
      Figure thumbnail gr2
      Fig 2Sagittal T1-weighted MR image of the thickness of the subclavius muscle (in millimeters), minimum costoclavicular distance (in millimeters), and the angle between the first rib and horizontal axis. Abbreviations: a, subclavian artery; c, clavicle; CCD, minimum costoclavicular distance (in millimeters); p, brachial plexus; SC, thickness of subclavius muscle (in millimeters); v, subclavian vein. **The angle between first rib and horizontal axis.
      Figure thumbnail gr3
      Fig 3Sagittal T1-weighted MRI image of retropectoralis minor distance. Abbreviations: cp, coracoid process; RMD, retropectoralis minor distance (in millimeters).
      For qualitative analysis, we assessed the interscalene triangle-prescalene space, costoclavicular space, and retropectoralis minor space for side subclavian arteries, subclavian veins, and brachial plexus compression. A reduction in the diameter of more than 30% for the arteries and of more than 50% for the veins, plus disappearance of the perineural fat for the neural structure, were accepted as positive for compression.
      We compared the MRI findings of the neutral and provocative tests for all subjects. They were evaluated for: (1) MRI findings in the neutral position (patients, controls), (2) MRI findings in the provocative position (patients, controls), (3) MRI findings in the neutral and provocative position (patients only), and (4) MRI findings in the neutral and provocative position (controls only). In addition, to measure the changes that occur when subjects were in a provocative position, for all parameters, we subtracted the neutral position measures from the provocative position measures; we called the numeric difference “the positional change value” (provocative change value). All subjects were compared on the basis of this difference. Also, we investigated the relationship between MRI positional difference and the physical evaluation test results (positive or negative) separately for both arms. We compared the patients’ physical examination findings with those of the control subjects.
      All patients received conservative treatment that included modifying behaviors by avoiding provocative activities and arm positions, and individually tailored physical therapy programs that strengthen the muscles of the pectoral girdle and help restore normal posture. Initially, treatment was started with range of motion (ROM) exercises and stretching of short, tight muscles that were determined by physical examination, including strike palpation of those muscles, such as the upper trapezius, levator scapulae, scalenes, sternocleidomastoid, and pectoralis muscles. The stretching and ROM exercises were continued with strike strengthening exercises (ie, shoulder girdle, neck and scapular stabilizer muscles). Pain control was provided by nonsteroidal anti-inflamatory drugs and hot packs, which were applied on the posterior cervical region for 30 minutes a day. The conservative therapy was applied as a home program. Subjects were advised to do 10 repetitions twice a day.
      This was the exercise program:
      • 1
        Sit in an upright position, clasp both hands behind the back, lower the left shoulder, tilt the head toward the right (15−30s), and then return to the starting position. Lower the right shoulder and tilt the head toward the left until feeling a stretch (15−30s) (scalene stretch).
      • 2
        Pull the chin back while keeping the eyes level.
      • 3
        Stand in a doorway or corner with both arms on the wall and slightly above the head. Lean forward slowly until a stretch is felt in the front of the shoulders (15−30s) (pectoralis stretch).
      • 4
        Lie on the back, or sit in a chair, or stand. Bring both hands behind the head. Try to bring the elbows up and back.
      • 5
        Stand tall with the arms at one’s side. Shrug the shoulders forward and up. Relax. Shrug the shoulders backward and up. Relax. Shrug the shoulders straight up. Relax.
      • 6
        Stand tall with the arms straight out from the side at shoulder level. Raise the arms to the side until they meet over the head, keeping the elbows straight and the palms down.
      • 7
        While sitting or standing with the arms at the sides, squeeze the shoulder blades together (5s) (scapular squeezes).
      • 8
        Lie face down or stand. Grab the hands behind the back. While lifting the head and chest off the floor as high as possible, breathe in and squeeze the shoulder blades together, and hold the chin in (3s). Return to the starting position and breathe out.
      • 9
        Lie down on one’s back. Put a rolled towel between the shoulder blades. Start with the arms at the sides, raise them up in the front and over the head while breathing in. Return to the starting position and breathe out.

      Statistical Analysis

      We used the Kolmogorov-Smirnov test to determine whether the data were distributed normally. We analyzed the difference between the neutral and the provocative position with a paired-samples t test for the group that followed a normal distribution; we used the Wilcoxon test for the group whose data were not normally distributed. We used the Mann-Whitney U test to find whether there was a significant difference between the patient and the control groups in terms of neutral-provocative positioned MRI findings, and to test whether there was a difference between patients with positive and negative physical examination findings. We used the chi-square test to compare the categorical data.

      Results

      There were 41 participants in the study. The patient group included 23 women and 6 men; the control subjects were 10 women and 2 men. There were no significant sex distribution differences between the 2 groups (P>.05). The patients’ mean age was 41.34±8.27 years; the mean age of the controls was 46.16±7.69 years. There were no significant differences for age between these groups (P>.05).
      We detected no cervical rib or apophysomegaly of the C7 vertebra on the radiographs of any of the patients.
      We evaluated the patients’ and controls’ quantitative MRI findings separately. In the patient group, there was a difference between the neutral position values and provocative position values except for the retropectoralis minor distance. In the control group, however, there was no significant difference between the 2 measures (P>.05), except for the left anterior scalene muscle thickness and the thickness of the right subclavius muscle (P<.05) (figs 4A, 4B).
      Figure thumbnail gr4
      Fig 4Comparison of patient and control groups’ MRIs of the (A) left- and (B) right-side measures. Abbreviations: ABFRH, angle between first rib and s horizontal; CG, control group; IA, interscalene angle; MCD, minimum costoclavicular distance (mm); PG, patient group; RMD, retropectoralis minor distance (in millimeters); TASM, thickness of anterior scalene muscle (in millimeters); TSM, thickness of subclavius muscle (in millimeters).
      Table 1 shows the neutral and provocative position comparisons between the patients and the control subjects. In particular, the measures in the provocative position differed significantly between the 2 groups (P<.05).
      Table 1Comparison of the MRI Findings in the Neutral and Provocative Positions, and the Positional Change Value Between the Patient Group and Control Group
      MRI MeasurementsNeutral PositionProvocative PositionComparison of Provocative Change Value
      Patient GroupControl GroupPPatient GroupControl GroupPPatient GroupControl GroupP
      RIA34.8433.07.40630.4533.04.2074.390.33.000
      LIA34.5033.25.75230.5232.80.1293.980.44.007
      TRASM10.799.76.01411.3010.00.0060.500.23.112
      TLASM10.209.75.35810.8610.79.9770.661.04.870
      ABRFRH34.9737.08.09126.9436.00.0008.021.08.000
      ABLFRH35.6836.79.44727.0536.50.0008.630.29.000
      TRSM7.577.75.8606.188.01.0141.390.26.000
      TLSM7.558.00.1346.237.54.0101.320.46.036
      MRCD23.8223.50.6989.2023.62.00014.620.12.000
      MLCD22.6727.33.0099.4927.29.00013.170.04.000
      RRMD24.0321.20.00824.0021.18.0050.120.02.280
      LRMD24.5421.81.03924.4821.71.0310.060.10.781
      Abbreviations: ABLFRH, angle between left first rib and horizontal; ABRFRH, angle between right first rib and horizontal; LIA, left interscalene angle; LRMD, left retropectoralis minor distance (in millimeters); MLCD, minimum left costoclavicular distance (in millimeters); MRCD, minimum right costoclavicular distance (in millimeters); RIA, right interscalene angle; RRMD, right retropectoralis minor distance (in millimeters); TLASM, thickness of left anterior scalene muscle (in millimeters); TLSM, thickness of left subclavius muscle thickness (in millimeters); TRASM, thickness of right anterior scalene muscle (in millimeters); TRSM, thickness of right subclavius muscle (in millimeters).
      Table 1 also shows the positional change values that were the difference between the neutral and provocative MRI measures. We compared provocative change values of the patients and the controls to see whether there were any differences. There were no significant differences in anterior scalene muscle thickness and retropectoral distance between the 2 groups (P>.05); however, there was a significant difference for the other measures (P<.05) (see table 1). In short, the patients had a more significant change in the provocative position MRI.
      The patient group was divided into 2 groups according to the results (positive or negative) of the physical examination provocative tests. The 2 groups were compared for the positional change value detected on MRI. We found that the provocative change value in the minimum costoclavicular distance varied significantly between the group with a positive result on the Halsted maneuver and the group with a negative result. For the right arm, the patients with a positive Halsted maneuver registered a change value of 19.33mm, while the provocative change value in the group with a negative Halsted maneuver was 9.57mm; there was a significant difference between the 2 groups (P=.000). For the left arm, the group with a positive Halsted maneuver registered a change value of 16.3mm, while the provocative change value in the group with a negative test was 10.25mm; there was also a significant difference between the 2 groups (P=.001). In short, in the patients with a positive Halsted maneuver, the minimum costoclavicular distance showed a more significant contraction on the provocative MRI. There was no significant difference for the provocative change values between the patients with positive Adson’s tests and hyperabduction tests and those who had negative Adson’s and hyperabduction tests (P>.05).
      We compared the ratio of the positive Adson’s test, Halsted maneuver test, and hyperabduction test for both patients and controls (table 2). There was no significant difference found for Adson’s test between the 2 groups (P>.05); however, there was a significant difference for the Halsted maneuver and hyperabduction test (P<.05).
      Table 2The Ratio of Positive Findings in the Provocative Physical Examination Tests and the Difference Between the 2 Groups in This Ratio
      Physical Examination Test and Arm SidePositive Rate (%)P Values for Difference Between Patients and Control Subjects
      Patient GroupControl Group
      Adson’s test of right arm75.966.7.545
      Adson’s test of left arm79.366.7.391
      Exaggerated military position of right arm51.78.3.010
      Exaggerated military position of left arm48.38.3.016
      Hyperabduction test of right arm62.116.7.008
      Hyperabduction test of left arm62.125.0.031
      When the qualitative findings of the neural MRIs for both patients and control subjects were reviewed, there were no compression findings in the upper extremities in the vascular or neural structures. In the provocative maneuver, the symptomatic group had the following compression findings in the interscalene triangle-prescalene area: in the right extremity, 8 arterial, 21 venous, and 3 neural compressions; in the left extremity, 2 arterial and 17 venous compressions. The control group, on the other hand, yielded the following compression findings: in the right extremity, 6 venous compressions and in the left extremity, 7 venous compressions. In the same position, the symptomatic group had the following compressions in costoclavicular distances: in the right extremity, 15 arterial, 19 venous, and 8 neural; and in the left extremity, 14 arterial, 17 venous, and 10 neural compressions. In the control group there were 6 venous compressions in the right extremity and 1 arterial and 4 venous compressions in the left extremity. Evaluation of the retropectoralis minor distance in the symptomatic group showed compressions of 1 arterial and 9 venous structures in the right extremity and compressions of 8 venous structures in the left extremity; the control group showed compressions of 4 venous structures in the right extremity and 3 in the left extremity. When the bilateral critical distances were added together for the groups, the patient group showed an arterial compression in 40 segments (22.99%), venous compression in 91 segments (52.29%), and a nervous compression in 21 segments (12.06%), out of 174 total segments (figs 5A, 5B). The control group showed an arterial compression in 1 segment (1.39%) and venous compression in 30 segments (41.7%) out of 174 total segments. No neural compression was detected in the control group. In addition, according to our qualitative evaluation of the results, a unilateral fibrous band was detected in 10 (34.48%) of the 29 patients (Fig 6, Fig 7). No fibrous bands were detected in the control subjects.
      Figure thumbnail gr5
      Fig 5(A) Normal view of the subclavian vascular structure and brachial plexus in costoclavicular space in the neutral position. (B) Compressions of these structures in the provocative position in sagittal T1-weighted MRI in the same patient. Abbreviations: a, subclavian artery; p, brachial plexus; v, subclavian vein.
      Figure thumbnail gr6
      Fig 6Coronal T1-weighted image showing the left subclavian vascular structures and vertical oblique hypointense thick fibrous band that caused the distortion of brachial plexus (arrows).
      Figure thumbnail gr7
      Fig 7In another patient, the coronal T1-weighted image showing the hypointense thin fibrous band arising from the transverse process of C7 and crossing the left subclavian artery (arrows).

      Discussion

      Wilbourn
      • Wilbourn A.J.
      The thoracic outlet syndrome is overdiagnosed.
      subdivided TOS into 3 groups: true neurogenic, vascular (artery and/or vein), and disputed or nonspecific TOS. The latter classification was defined as chronic pain syndrome with subtle features suggestive of brachial plexus involvement, and because its etiology was obscure, it was hard to diagnose. According to the literature, in 98% of all patients with TOS, the symptoms were attributed to entrapment of the brachial plexus at the thoracic outlet, and in only 2% was vascular compression the cause.
      • Davidovic L.B.
      • Kostic D.M.
      • Jakovljevic N.S.
      • Kuzmanovic I.L.
      • Simic T.M.
      Vascular thoracic outlet syndrome.
      • Roos D.B.
      The thoracic outlet syndrome is underrated.
      TOS is more commonly observed in women, and the onset of symptoms usually occurs between the ages of 20 and 50.
      • Huang J.H.
      • Zager E.L.
      Thoracic outlet syndrome.
      • Wilbourn A.J.
      The thoracic outlet syndrome is overdiagnosed.
      Of the group of subjects with bilateral complaints who were categorized as disputed TOS, 23 (79.3%) were women. The thoracic outlet may constrict because of bone structure anomalies such as a cervical rib, an abnormal first rib, a prominent transverse process of the C7 vertebra, or soft tissue abnormalities such as congenital bands and ligaments and scalene muscle hypertrophies. Usual symptoms include shoulder stiffness or weakness, arm-back pain, arm numbness, tingling, coldness, and swelling of the extremity.
      • Roos D.B.
      • Owens J.C.
      Thoracic outlet syndrome.
      • Novak C.B.
      • Mackinnon S.E.
      Thoracic outlet syndrome.
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      Electrodiagnostic tests do not play an important diagnostic role; however, they are used for differential diagnosis.
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      • Seror P.
      Frequency of neurogenic thoracic outlet syndrome in patients with definite carpal tunnel syndrome: an electrophysiological evaluation in 100 women.
      The patient group in our study had normal electromyographic findings.
      Many diagnostic radiologic procedures have been described to confirm neurovascular compression or to differentiate the syndrome from other diseases. Plain radiography helps detect only the osseous pathologies,
      • Novak C.B.
      • Mackinnon S.E.
      Thoracic outlet syndrome.
      • Huang J.H.
      • Zager E.L.
      Thoracic outlet syndrome.
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      but a very important advantage of CT is that it can detect both the osseous structures and the vascular structures. There is, however, a critical disadvantage in the use of CT—the presence of ionizing radiation and the use of potentially nephrotoxic and allergic iodinated contrast material.
      • Remy-Jardin M.
      • Doyen J.
      • Remy J.
      • Artaud D.
      • Fribourg M.
      • Duhamel A.
      Functional anatomy of the thoracic outlet: evaluation with spiral CT.
      • Remy-Jardin M.
      • Remy J.
      • Masson P.
      • et al.
      CT angiography of thoracic outlet syndrome: evaluation of imaging protocols for the detection of arterial stenosis.
      Also, although the catheter angiography is the criterion standard in the diagnosis of vascular TOS, local and systematic complications (eg, hematoma, emboli, occlusion) may also develop. And because it is an invasive method, it is not possible to repeat it frequently.
      • Lang E.K.
      Arteriography of thoracic outlet syndrome.
      • Kadir S.
      Arteriography of the upper extremities.
      In recent years, contrast-enhanced magnetic resonance angiography (MRA) has circumvented some of these disadvantages, and there are reports that it is now more efficient in the diagnosis of vascular TOS.
      • Charon J.P.
      • Milne W.
      • Sheppard D.G.
      • Houston J.G.
      Evaluation of MR angiographic technique in the assessment of thoracic outlet syndrome.
      • Cosottini M.
      • Zampa V.
      • Petruzzi P.
      • Ortori S.
      • Cioni R.
      • Bartolozzi C.
      Contrast-enhanced three-dimensional MR angiography in the assessment of subclavian artery diseases.
      Sonography, another radiologic method, is usually able to detect arterial and venous pathologies, but obesity and surrounding osseous structures may sometimes prevent an accurate diagnosis. Also, it is usually not possible with sonography to see the compression of the neural structures and fibrous bands.
      There have been few reports on the potential value of MRI in the diagnosis of the vascular compression forms of TOS.
      • Panegyres P.K.
      • Moore N.
      • Gibson R.
      • Rushworth G.
      • Donaghy M.
      Thoracic outlet syndromes and magnetic resonance imaging.
      The benefits of MRI in the diagnosis of TOS are well known. The pioneering study of Panegyres et al
      • Panegyres P.K.
      • Moore N.
      • Gibson R.
      • Rushworth G.
      • Donaghy M.
      Thoracic outlet syndromes and magnetic resonance imaging.
      showed very high sensitivity (79%) and specificity rates (87.5%), and that MRI was able to detect brachial plexus compressions. In the same study, Panegyres noted that MRI could also detect the etiology of TOS (eg, cervical rib, band-like structure, posttraumatic callus, hypertrophied muscle). In recent years, Demondion et al
      • Demondion X.
      • Boutry N.
      • Drizenko A.
      • Paul C.
      • Francke J.P.
      • Cotten A.
      Thoracic outlet: anatomic correlation with MR imaging.
      investigated the critical anatomic localizations of postpostural maneuver conditions in TOS by using MRI in both cadavers and volunteers. More recently, Demondion et al
      • Demondion X.
      • Bacqueville E.
      • Paul C.
      • Duquesnoy B.
      • Hachulla E.
      • Cotten A.
      Thoracic outlet: assessment with MR imaging in asymptomatic and symptomatic populations.
      compared postpostural maneuver MRI evaluations of the upper extremities of patient and control groups; the results showed that MRI helps detect arterial or neural compression and the cause of the compression.
      There are some methodologic differences between our study and the above-mentioned study.
      • Demondion X.
      • Bacqueville E.
      • Paul C.
      • Duquesnoy B.
      • Hachulla E.
      • Cotten A.
      Thoracic outlet: assessment with MR imaging in asymptomatic and symptomatic populations.
      One difference is that in our study, the patient group consisted only of patients with bilateral complaints and patients who were categorized as having clinically disputed TOS, and bilateral MRIs were conducted. A second difference is that none of our patients had a cervical rib or elongated transverse process of the C7 vertebrae. Also, in the Demondion study
      • Demondion X.
      • Bacqueville E.
      • Paul C.
      • Duquesnoy B.
      • Hachulla E.
      • Cotten A.
      Thoracic outlet: assessment with MR imaging in asymptomatic and symptomatic populations.
      the MRI findings and the clinical examination data did not correlate; however, in our study the significance of the provocative tests was evaluated on the basis of the MRI findings.
      The quantitative analysis of the neutral- and provocative-positioned MRI findings showed that the patient group had a significant difference in all measures (P<.05) except for the bilateral retropectoralis minor distance. The only significant differences between the neutral- and the provocative-positioned MRI in the control group, however, were the left anterior scalene muscle thickness and the thickness of the right subclavius muscle (P<.05).
      When we compared the MRI findings for the patient and control groups, we found no differences in the neutral MRI measurement and significant differences in the provocative MRI measurements. In the Demondion study,
      • Demondion X.
      • Bacqueville E.
      • Paul C.
      • Duquesnoy B.
      • Hachulla E.
      • Cotten A.
      Thoracic outlet: assessment with MR imaging in asymptomatic and symptomatic populations.
      the difference between neutral and provocative positions was significant on more parameters; however, our study showed a significant difference on only 2 parameters. This difference may result from 2 things: (1) we equalized the patient and control groups for age and sex, while it is not clear whether Demondion and colleagues did this because they did not report age and sex equalization, and (2) the posture of the Turkish population may contribute to this difference. We could not, however, find any other studies on this subject with a Turkish population.
      Furthermore, when the positional change values were compared between the patient and control groups, all parameters except for the bilateral anterior scalene muscle thickness and left retropectoralis minor distance measurements, showed significant differences. This means that the provocative-positioned MRI findings on the change value in TOS patients were higher than in the control group. This finding resembles the clinical symptom that is evident in the TOS stress test, which is an important diagnostic criterion in determining TOS. In addition, our control group had no clinical complaints about the TOS stress test positions. In patients with positive TOS stress test, provocative MRI examination may be of value in diagnosing TOS.
      In our qualitative analysis of the MRI findings, we found that none of our control subjects had neurovascular compression, as shown in the neutral-positioned MRI. In the provocative position, the patient group showed 22.99% arterial, 52.29% venous, and 12.06% nervous compression, and the control group showed 1.39% arterial and 41.7% venous compression. There was no neural compression in the control group. The higher percentage of venous compression in the control group is not surprising; the findings are within the expected ranges and other studies show high false-positive rates in healthy subjects.
      • Demondion X.
      • Bacqueville E.
      • Paul C.
      • Duquesnoy B.
      • Hachulla E.
      • Cotten A.
      Thoracic outlet: assessment with MR imaging in asymptomatic and symptomatic populations.
      • Huang J.H.
      • Zager E.L.
      Thoracic outlet syndrome.
      • Davidovic L.B.
      • Kostic D.M.
      • Jakovljevic N.S.
      • Kuzmanovic I.L.
      • Simic T.M.
      Vascular thoracic outlet syndrome.
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      • Longley D.G.
      • Yedlicka J.W.
      • Molina E.J.
      • Schwabacher S.
      • Hunter D.W.
      • Letourneau J.G.
      Thoracic outlet syndrome: evaluation of the subclavian vessels by color duplex sonography.
      In another study, Demondion et al
      • 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.
      attempted to evaluate the feasibility and potential usefulness of power Doppler ultrasonography in the assessment of changes in the arterial cross-sectional area of the thoracic outlet during upper-limb elevation. Demondion’s study demonstrated that: (1) the subclavian artery presents a progressive narrowing during upper-limb abduction in the costoclavicular space, even in an asymptomatic population; (2) asymptomatic subjects may present with a significant narrowing of their subclavian artery after complete upper-limb elevation (170°); and (3) MRI examination may demonstrate false-negative findings because of the subject’s supine position during the MRI examination.
      • 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.
      Some authors
      • Novak C.B.
      • Mackinnon S.E.
      Thoracic outlet syndrome.
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      • Ide J.
      • Kataoka Y.
      • Yamaga M.
      • Kitamura T.
      • Takagi K.
      Compression and stretching of the brachial plexus in thoracic outlet syndrome: correlation between neuroradiographic findings and symptoms and signs produced by provocation manoeuvres.
      have also reported high false-positive rates for compression of the subclavian artery during diagnostic maneuvers. It has therefore been reported that the clinical complaints by TOS patients while undergoing provocative physical examinations are more important to note in the diagnosis.
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      Because of this, we used a TOS stress test based on the increase in patients’ complaints. Different studies show that the TOS stress test is the best tool for diagnosing TOS.
      • Huang J.H.
      • Zager E.L.
      Thoracic outlet syndrome.
      Although subjects in the patient group were categorized as having disputed TOS, the fact that there was vascular compression along with neurogenic compression may imply that these 2 structures’ courses are very much in parallel. In the patient group, 16.7% of arterial compressions, 20.68% of venous compressions, and 10.34% of neural compressions were in the costoclavicular distance. The positive results of the Halsted maneuver are of great importance in the evaluation of patients with TOS because of the abnormalities in the costoclavicular distance. In our study, the contraction in the provocative MRI position and minimum costoclavicular distance were more evident in patients with a positive exaggerated military test. This finding supports the relationship between MRI and physical examination. Venous compression was most frequent in the prescalene space (21.84%). In both the patient and the control groups, the least-frequent vascular compression localization was in the retropectoralis minor space; there was no neural structure compression in this area. There was no difference between the provocative-positioned and neutral-positioned MRI measures recorded through a retropectoral distance in the patient group. Although it may seem contradictory, the patient and the control groups showed significant differences between their neutral and provocative MRI measures of the retropectoral distances. Our main purpose in this study, however, was to determine the amount of change between the neutral-positioned and provocative-positioned MRI findings. Hence, the difference between the first and second measures of control and patient comparisons does not contradict the study findings.
      There was no significant difference in the retropectoralis minor distance of patients with negative and positive hyperabduction test results. As is well known, in the clinical hyperabduction maneuver, 180° of shoulder abduction is applied. It is also known that the retropectoral distance shows a significant contraction in the hyperabduction position. Because, however, the MRI scanner we used in the study was a closed tunnel magnet that permits only 130° of shoulder abduction, it was not possible to force enough abduction to contract the retropectoral distance. Smedby et al
      • Smedby Ö.
      • Rostad H.
      • Klaastad Ø.
      • Lilleås F.
      • Tillung T.
      • Fosse E.
      Functional imaging of the thoracic outlet syndrome in an open MR scanner.
      claim that an open MRI scanner is more functional for imaging of TOS patients, but the number of patients in their study was far smaller than in our study. Our finding that there was a significant difference between the 2 groups’ hyperabduction test results runs parallel to the finding that there is a difference between the neutral and the provocative MRI measures.
      Roos described 9 types of congenital bands and ligaments in the thoracic outlet area that may compress the neurovascular structures.
      • Mackinnon S.E.
      • Novak C.B.
      Thoracic outlet syndrome.
      In an MRI study,
      • Panegyres P.K.
      • Moore N.
      • Gibson R.
      • Rushworth G.
      • Donaghy M.
      Thoracic outlet syndromes and magnetic resonance imaging.
      75% of patients with TOS symptoms had a band-like structure. In our study, approximately 35% of the patients showed fibrous bands unilaterally. In our study, fibrous bands were detected in 10 (34.48%) of the 29 patients but were detected in our healthy subjects. TOS symptoms were present in all patients’ fibrous bands, but none were detected in asymptomatic patients. For this reason, we think that the patients’ symptoms were caused by fibrous bands.
      TOS research that uses different radiologic methods (eg, ultrasound, CT, CT angiography, catheter angiography, MRI, MRA) also uses provocative maneuvers to obtain more objective data that are similar to the clinical examination method. Gillard et al
      • Gillard J.
      • Perez-Cousin M.
      • Hachulla E.
      • et al.
      Diagnosing thoracic outlet syndrome: contribution of provocative tests, ultrasonography, electrophysiology, and helical-computed tomography in 48 patients.
      evaluated the diagnostic usefulness of provocative tests, Doppler ultrasonography, electrophysiologic investigations, and helical CT angiography in TOS. They found that Adson’s test was more useful than other provocative tests in diagnosing TOS. In our study, we conducted a provocative position imaging and compared the measures with the clinical examination tests. There was no significant difference in the positive and negative results of the Adson’s test between the patients and control subjects; however, this finding contradicted our qualitative MRI measures because the patient group showed a 22.99% ratio of subclavian artery pressure while the control group showed only a 1.39% ratio of arterial pressure. A possible reason for this paradox is that because the patients were evaluated using bilateral MRI, just as in the Adson’s test, no contralateral head rotation or deep inspiration was conducted. Another possible reason may be that Adson’s test yields a high ratio of false positives in healthy subjects.

      Conclusions

      On the basis of our findings, we conclude the following. First, neutral- and provocative-position MRI may be beneficial in the diagnosis of TOS. We believe that to develop a standard, studies involving large samples should be conducted. Criteria such as body mass index should be included in the evaluation for better standardization. Second, the positional change value between patient and control groups showed a difference in the quantitative analysis and this finding is in accord with the clinical examination tests. Third, for a qualitative analysis, provocative-position imaging alone is sufficient, because the findings did not show any compressions in the neutral position. Fourth, MRI can provide information about vascular or neural structure pressure in different compartments. It can also detect the fibrous band-like structures that cause the pressure. Because healthy subjects show a high ratio of venous compression, however, this finding should be evaluated carefully. Fifth, it is possible to use modified versions of the tests in a clinical evaluation using the MRI machine; this can provide an opportunity to obtain more objective results.
      Supplier
      aMagnetom Expert; Siemens Medical Solutions, Henkestrasse 127, D-91052 Erlangen, Germany.

      References

        • Roos D.B.
        • Owens J.C.
        Thoracic outlet syndrome.
        Arch Surg. 1966; 93: 71-74
        • Novak C.B.
        • Mackinnon S.E.
        Thoracic outlet syndrome.
        Orthop Clin North Am. 1996; 27: 747-762
        • 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
        • Huang J.H.
        • Zager E.L.
        Thoracic outlet syndrome.
        Neurosurgery. 2004; 55: 897-903
        • Davidovic L.B.
        • Kostic D.M.
        • Jakovljevic N.S.
        • Kuzmanovic I.L.
        • Simic T.M.
        Vascular thoracic outlet syndrome.
        World J Surg. 2003; 27: 545-550
        • Mackinnon S.E.
        • Novak C.B.
        Thoracic outlet syndrome.
        Curr Probl Surg. 2002; 39: 1070-1145
        • Matsumura J.S.
        • Rilling W.S.
        • Pearce W.H.
        • Nemcek Jr, A.A.
        • Vogelzang R.L.
        • Yao J.S.
        Helical computed tomography of the normal thoracic outlet.
        J Vasc Surg. 1997; 26: 776-783
        • Panegyres P.K.
        • Moore N.
        • Gibson R.
        • Rushworth G.
        • Donaghy M.
        Thoracic outlet syndromes and magnetic resonance imaging.
        Brain. 1993; 116: 823-841
        • Tateishi A.
        Diagnosis and management of thoracic outlet syndrome.
        Nippon Seikeigeka Gakkai Zasshi. 1980; 54: 817-827
        • Ohkawa Y.
        • Isoda H.
        • Hasegawa S.
        • Furuya Y.
        • Takahashi M.
        • Kaneko M.
        MR angiography of thoracic outlet syndrome.
        J Comput Assist Tomogr. 1992; 16: 475-477
        • Seror P.
        Frequency of neurogenic thoracic outlet syndrome in patients with definite carpal tunnel syndrome: an electrophysiological evaluation in 100 women.
        Clin Neurophysiol. 2005; 116: 259-263
        • Remy-Jardin M.
        • Doyen J.
        • Remy J.
        • Artaud D.
        • Fribourg M.
        • Duhamel A.
        Functional anatomy of the thoracic outlet: evaluation with spiral CT.
        Radiology. 1997; 205: 843-851
        • Wilbourn A.J.
        The thoracic outlet syndrome is overdiagnosed.
        Arch Neurol. 1990; 47: 328-330
        • Roos D.B.
        The thoracic outlet syndrome is underrated.
        Arch Neurol. 1990; 47: 327-330
        • Remy-Jardin M.
        • Remy J.
        • Masson P.
        • et al.
        CT angiography of thoracic outlet syndrome: evaluation of imaging protocols for the detection of arterial stenosis.
        J Comput Assist Tomogr. 2000; 24: 349-361
        • Lang E.K.
        Arteriography of thoracic outlet syndrome.
        in: Baum S. Abrams angiography: vascular and interventional radiology. Little, Brown Medical, Boston1997: 995-1010
        • Kadir S.
        Arteriography of the upper extremities.
        in: Kadir S. Diagnostic angiography. WB Saunders, Philadelphia1986: 172-206
        • Charon J.P.
        • Milne W.
        • Sheppard D.G.
        • Houston J.G.
        Evaluation of MR angiographic technique in the assessment of thoracic outlet syndrome.
        Clin Radiol. 2004; 59: 588-595
        • Cosottini M.
        • Zampa V.
        • Petruzzi P.
        • Ortori S.
        • Cioni R.
        • Bartolozzi C.
        Contrast-enhanced three-dimensional MR angiography in the assessment of subclavian artery diseases.
        Eur Radiol. 2000; 10: 1737-1744
        • Demondion X.
        • Boutry N.
        • Drizenko A.
        • Paul C.
        • Francke J.P.
        • Cotten A.
        Thoracic outlet: anatomic correlation with MR imaging.
        AJR Am J Roentgenol. 2000; 175: 417-422
        • Longley D.G.
        • Yedlicka J.W.
        • Molina E.J.
        • Schwabacher S.
        • Hunter D.W.
        • Letourneau J.G.
        Thoracic outlet syndrome: evaluation of the subclavian vessels by color duplex sonography.
        AJR Am J Roentgenol. 1992; 158: 623-630
        • 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
        • Ide J.
        • Kataoka Y.
        • Yamaga M.
        • Kitamura T.
        • Takagi K.
        Compression and stretching of the brachial plexus in thoracic outlet syndrome: correlation between neuroradiographic findings and symptoms and signs produced by provocation manoeuvres.
        J Hand Surg [Br]. 2003; 28: 218-223
        • Smedby Ö.
        • Rostad H.
        • Klaastad Ø.
        • Lilleås F.
        • Tillung T.
        • Fosse E.
        Functional imaging of the thoracic outlet syndrome in an open MR scanner.
        Eur Radiol. 2000; 10: 597-600
        • Gillard J.
        • Perez-Cousin M.
        • Hachulla E.
        • et al.
        Diagnosing thoracic outlet syndrome: contribution of provocative tests, ultrasonography, electrophysiology, and helical-computed tomography in 48 patients.
        Joint Bone Spine. 2001; 68: 416-424