What MRI Should You Order for Thoracic Outlet Syndrome?

The right MRI for thoracic outlet syndrome depends on which type you suspect, because TOS is not one condition but several, and each compresses different structures in the space between the collarbone and first rib. For suspected neurogenic TOS, you want MR neurography focused on the brachial plexus. For suspected vascular TOS, you want contrast-enhanced MR angiography or venography. In either case, the scan should include images taken with the arm raised into a provocative position, not just resting at the side. That dynamic element is arguably more important than the specific sequences, and skipping it is one of the most common reasons a TOS-focused MRI comes back unhelpful.

Why the Type of TOS Dictates the Protocol

TOS breaks down into three main categories. Neurogenic TOS, where the brachial plexus nerves are compressed, accounts for the vast majority of cases. Arterial TOS involves compression of the subclavian artery, and venous TOS (sometimes called Paget-Schroetter syndrome) involves the subclavian vein. Some patients have overlap. A study that used a combined MRI protocol including MR angiography and nerve-focused sequences on 27 patients found a mix: some had purely neurogenic findings, others had arterial-neurogenic or arterial-venous-neurogenic combinations, and a subset had post-traumatic changes like brachial plexus edema or rupture.¹ A generic “MRI of the chest” or even a standard “MRI of the shoulder” will not answer the clinical question. You need to communicate the suspected type to the radiologist so they can build the right protocol.

MR Neurography for Neurogenic TOS

Neurogenic TOS is the hardest subtype to confirm with imaging because the nerve compression can be subtle. The go-to study is MR neurography, which uses high-resolution sequences tuned to visualize peripheral nerves. In a 3T MR neurography study of 30 patients with suspected neurogenic TOS, researchers identified clear structural causes in seven, including fibrous bands and rib anomalies. The compressed portions of the brachial plexus showed increased signal on T2-weighted images, which helped pinpoint the exact level of compression. All seven findings were confirmed during surgery.²

That hit rate of about one in four may sound modest, but neurogenic TOS has always been a diagnosis that leans heavily on clinical history and physical examination. The value of MR neurography is that when it does show something, the finding tends to be surgically actionable. Researchers have concluded that high-resolution MR neurography and diffusion-based tractography can confirm a neurogenic TOS diagnosis and localize the compression, allowing surgeons to plan a more targeted decompression.³

One emerging technique adds zero echo time (ZTE) imaging to the neurography protocol. ZTE sequences excel at depicting bone and are especially good at catching subtle rib anomalies like cervical ribs or elongated transverse processes that standard nerve-focused sequences can miss. In a recent study, adding ZTE to MR neurography improved agreement between readers for identifying anomaly-related nerve impingement, bumping inter-reader agreement from fair to substantial.⁴ If your facility has this capability, it is worth requesting, particularly when a cervical rib or bony variant is suspected but plain X-rays are equivocal.

MR Angiography and Venography for Vascular TOS

When the concern is arterial compression, the study to order is contrast-enhanced 3D MR angiography. This involves injecting gadolinium contrast and acquiring images while the patient holds their arms in different positions. Researchers have shown that this approach provides excellent visualization of the arteries and veins on both sides and offers a noninvasive alternative to traditional catheter-based angiography for suspected vascular TOS.⁵ The contrast-enhanced protocol can capture both arterial and venous phases in the same session, so you do not necessarily need to order separate studies for arteries and veins.

For venous TOS specifically, the MRI needs to demonstrate whether the subclavian vein is being compressed or thrombosed. A case report using ferumoxytol-enhanced MR venography illustrated this well: with the patient’s arm relaxed at the side, the right subclavian vein was moderately narrowed at the thoracic inlet. When the arm was raised into the abducted position, the vein was completely occluded, and chronic clot was visible as filling defects. The opposite side also showed dynamic occlusion with arm abduction but was clot-free at rest.⁶ That kind of side-to-side comparison and position-dependent change is exactly what clinicians need to see.

Why Provocative Arm Positioning Matters So Much

A standard MRI with the arms resting at the sides will miss the compression that defines TOS, because the thoracic outlet opens up when the arm is down. Early work showed that hyperabduction maneuvers during MRI could reveal the location of compression that was invisible in the neutral position.⁷ A study comparing symptomatic TOS patients with healthy volunteers confirmed this: except for venous thrombosis, compressions of the nerves and vessels showed up only with the arm elevated. Patients with TOS also had a significantly smaller costoclavicular distance after the postural maneuver compared to volunteers.⁸

The practical implication is clear: if you are ordering an MRI for TOS and the order says nothing about arm positioning, there is a real chance the scan will be read as normal even in a patient with genuine compression. When speaking with the radiology department, specify that images should be obtained with the arms in both the neutral (adducted) and the provocative (abducted, elevated) positions. Some centers refer to this as a “dynamic TOS protocol.” If the facility does not routinely perform this, it may be worth referring the patient to one that does.

One case series focused on surgical planning took this a step further, looking for loss of the normal fat signal around the brachial plexus nerves when the arm was in the provocative position. When that fat signal disappeared, it meant the nerves were being squeezed directly between the clavicle and the first rib. The combination of this provocative-position MRI finding with the right clinical picture strongly predicted which patients would respond well to first rib resection surgery.⁹

The Trap of Vascular Compression in Normal People

Here is where ordering clinicians need to be careful. Vascular compression at the thoracic outlet is surprisingly common in people who have no symptoms at all. A recent MRI study measured subclavian artery and vein compression in patients with one-sided TOS, using the asymptomatic side as an internal control. On the asymptomatic side, median venous compression in the stress position was about 64%, and three out of four of these symptom-free sides showed more than 50% narrowing of the vein.¹⁰

The researchers also found that the costoclavicular distance dropped significantly in the stress position and that this reduction was greatest in patients with higher body mass. Their conclusion was blunt: imaging evidence of vascular narrowing alone should not be the sole criterion for diagnosing TOS.¹⁰ This is a critical point for anyone reading TOS MRIs. Seeing a squeezed subclavian vein on a provocative image does not, by itself, mean the patient has TOS. You have to match the imaging findings to the clinical syndrome.

3T Versus 1.5T Scanners

If you have the choice, a 3-Tesla MRI scanner is preferable for TOS imaging. A study comparing brachial plexus imaging at 1.5T and 3T found that the higher-field scanner produced significantly better signal and contrast ratios across all sequences, and nerve visibility was significantly improved.¹¹ For neurogenic TOS in particular, where you are trying to see subtle changes in small nerves and tiny fibrous bands, that extra resolution matters. The 3T MR neurography study that identified fibrous bands causing brachial plexus lesions was specifically designed to exploit the advantages of high-field imaging.²

That said, 3T scanners are not universally available, and a well-designed 1.5T protocol with proper coils and provocative positioning will still outperform a 3T scan ordered without the right sequences or arm positions. The contrast-enhanced MR angiography protocol for vascular TOS has been validated on both 1.5T and 3T scanners.⁵ Protocol design trumps magnet strength.

What to Tell the Radiologist

Because TOS imaging is not a one-size-fits-all order, the referring clinician’s communication with the radiologist meaningfully affects what comes back. A systematic scoping review of MRI for suspected neurogenic TOS noted the lack of standardized examination protocols and called for establishing clear guidelines to optimize MRI’s ability to assess nerves and soft tissues in these patients.¹² Until that standardization happens, the ordering provider essentially needs to build the protocol request themselves, or at least point the radiologist in the right direction.

At minimum, the order or clinical note should specify:

Suspected type: neurogenic, arterial, venous, or uncertain. If uncertain, a combined protocol covering both nerves and vessels is reasonable.
Dynamic imaging: images in both neutral and provocative (arm abducted/elevated) positions.
Contrast: gadolinium-enhanced MR angiography if vascular TOS is on the differential. MR neurography for neurogenic TOS does not always require contrast, though some centers use it.
Side: bilateral imaging is ideal for comparison, since the asymptomatic side provides an internal baseline.

Simply ordering “MRI chest” or “MRI shoulder” without these specifications is a recipe for a normal-looking scan that does not answer the clinical question. Some institutions have a dedicated “TOS protocol” built into their MRI menu; if yours does, use it and confirm it includes the provocative positioning component.

How MRI Catches Anatomical Culprits

One of MRI’s strengths over other imaging is its ability to identify the specific structure causing compression. An early study showed that MRI detected all seven cervical ribs that were visible on plain X-rays, confirming it could replace plain films for bony anatomy.¹³ But MRI goes further than X-ray because it also reveals soft-tissue causes that bones-only imaging misses entirely. Fibrous bands, an enlarged subclavius muscle, anomalous arteries crossing the plexus, tumors, and post-traumatic scarring are all identifiable on MRI. The combined-protocol study mentioned earlier found causes ranging from fibrous bands and elongated transverse processes to schwannomas and lymph node metastases compressing the brachial plexus.¹

This matters for surgical planning. A surgeon who knows before the operation that a fibrous band is compressing the C8 nerve at a specific level can plan a more targeted approach than one operating on clinical suspicion alone. It also matters for catching diagnoses you were not expecting. Not every case of arm pain and numbness turns out to be TOS, and MRI occasionally reveals an alternative explanation like a tumor or vascular anomaly that changes management entirely.

TOS Imaging in Children and Adolescents

TOS is overwhelmingly a condition of young adults, but it shows up in older children and teenagers too, particularly in athletes. A dedicated review of MRI for pediatric TOS recommended that a complete examination should include routine anatomical sequences, MR angiography with the arms at the sides, and a dynamic assessment of vessel compression with the arms abducted.¹⁴ The approach mirrors adult protocols, but pediatric radiologists may need to adjust coil placement and sequence timing for smaller body habitus. If you are referring a teenager for TOS imaging, flagging the patient’s age on the order helps the MRI team plan accordingly.

When MRI Finds Nothing

A negative MRI does not rule out TOS, particularly the neurogenic type. In several of the studies reviewed here, a meaningful percentage of patients with clinically confirmed TOS had normal-appearing MRIs. In the combined-protocol study, seven out of 27 patients had no detectable abnormality on MRI.¹ This does not mean the MRI was wasted. A negative scan with provocative positioning narrows the differential and may steer the workup toward nerve conduction studies, scalene block injections, or other diagnostic steps. It also provides a useful baseline if the patient later undergoes surgery and post-operative imaging is needed to assess whether decompression was achieved.

The absence of a universally standardized MRI protocol for TOS means that a “negative” result at one center might become a positive finding when re-imaged at a center with more experience. If the clinical suspicion remains high despite a normal MRI, repeating the study at a facility with a dedicated TOS imaging program is a reasonable next step rather than abandoning the diagnosis.